diff --git a/AGENTS.md b/AGENTS.md
index 4fb546042..bc016a26e 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -87,11 +87,22 @@ Config keys to verify in `data/config.yaml` / `src/langbot/templates/config.yaml
 - When changing Service API routes or capabilities used by `lbctl`, check compatibility with the separate `langbot-cli` repository.
 - New schema changes use Alembic under `src/langbot/pkg/persistence/alembic/versions/`. LangBot 4.x does not support upgrading 3.x databases.
 - New platform behavior belongs in platform adapters only for platform translation; pipeline/business logic belongs in `pkg/pipeline/` or services.
-- User-facing strings must support i18n (`en_US`, `zh_Hans`; include `ja_JP` where the repo already does).
+- User-facing strings must support i18n; follow the backend language rules below for server-generated content.
 - Code comments and docstrings must be English.
 - Keep compatibility and security in mind; LangBot is used in both self-hosted/community and toB deployments.
 - Commit message format: `<type>(<scope>): <subject>`.
 
+## Backend Message Language
+
+- LangBot-owned business errors, runtime notices, and default messages sent to users must use the owning Workspace's message language. Keep this independent of each member's WebUI language; an unset Workspace language defaults to `en_US`.
+- Reuse `src/langbot/pkg/i18n/` and the existing `message_language` entry in `WorkspaceMetadata`. Use `t()` with stable English templates and named parameters; add matching keys and placeholders to all eight catalogs: `en_US`, `zh_Hans`, `zh_Hant`, `ja_JP`, `es_ES`, `ru_RU`, `th_TH`, and `vi_VN`. Avoid new translation frameworks or duplicate language settings.
+- Resolve language from a trusted Workspace or bot execution context, and isolate it across concurrent requests, tasks, and adapter callbacks. Browser language hints may supply account-only or unauthenticated response language, but must not override a resolved Workspace or authorize tenant selection. Reading/changing the setting requires `workspace.view` / `workspace.update`.
+- Pipeline notices should be configurable in Pipeline output settings through `pipeline_notice()`. Blank overrides inherit the Workspace translation. Preserve user-authored text literally except for documented placeholders, and retain existing visibility switches. Do not rewrite historical messages or custom text when the language changes.
+- China-only adapters and configuration/field validation diagnostics do not require localization. An adapter also serving international users is not exempt merely because it supports a Chinese platform.
+- Preserve raw provider/plugin diagnostics, tracebacks, protocol values, and machine-readable error/status codes. Translate LangBot's own category or explanation around them; do not translate arbitrary errors with substring replacement. Plugin-authored messages remain the plugin's localization responsibility.
+- Keep changes small and maintainable. Reuse existing persistence and response boundaries; delete obsolete fields only after verifying they have no remaining readers, including compatibility paths.
+- When changing language behavior, verify catalog/placeholder parity, Workspace isolation and permissions, and Pipeline inheritance/custom overrides. Update affected configuration snapshots and API/MCP documentation. Implementation details and focused test locations: [Backend message language](docs/architecture/backend-i18n.md).
+
 ## Runtime Pitfalls
 
 - Local stdio Plugin Runtime disconnects do not auto-reconnect; restart LangBot if that path breaks.
diff --git a/ARCHITECTURE.md b/ARCHITECTURE.md
index d49fdd2e6..9ad2a5a9b 100644
--- a/ARCHITECTURE.md
+++ b/ARCHITECTURE.md
@@ -214,14 +214,34 @@ In `langbot-plugin-sdk`:
 - `src/langbot_plugin/box/runtime.py` owns sandbox sessions, generic read-only mounts, and managed processes.
 - `backend.py`, `nsjail_backend.py`, and `e2b_backend.py` implement sandbox backends.
 - `src/langbot_plugin/skill_store.py` is consumed by Core, not Box. Activation
-  pins the first revision for a run; instructions, resource reads, restart
-  recovery, execution mounts, and revision-scoped dependency state all use that
-  exact publication. Core mounts only activated and authorized Skills. Published
+  pins the first revision for a run in Host-owned run ledger metadata; instructions,
+  resource reads, execution mounts, and revision-scoped dependency state use that
+  exact publication. A new run refreshes the Workspace catalog and resolves the
+  current revision, ignoring old conversation activation state. Lost in-memory
+  pins in an active authorized run are reloaded from its ledger. After a Core
+  process restart, expired run sessions cannot call tools: the Host rejects them
+  rather than silently resolving a new version. This does not add automatic run resumption. Core mounts only activated and authorized Skills. Published
   trees are never agent write targets: authoring happens in Workspace drafts and
   an explicit registration/update atomically publishes the next revision. Core
   turns those package roots and content digests into generic read-only mounts;
   Box does not understand Skill names, metadata, revisions, files, or CRUD.
 
+Skill execution paths include the publication digest (`/opt/langbot-artifacts/<digest>`).
+Docker/Podman mount a read-only cache when creating the Box and atomically add
+verified packages before execution; E2B seals and verifies a private staging copy
+before publishing a stable projection. nsjail applies the requested mounts in a
+separate namespace for each execution. Activating a Skill after acquiring a Box
+therefore does not recreate the session or stop its managed processes. Ordinary
+host mount changes still fail explicitly on fixed-mount backends.
+
+A shared Box remains a Workspace filesystem sharing boundary: its commands can
+see previously materialized packages. Run pins prevent revision switching; they
+are not access isolation between mutually untrusted programs in the same Box.
+Use package-relative script paths or the path returned by activation. Core tools
+translate legacy `/workspace/.skills/<name>` command paths to the run's digest
+path, but do not rewrite paths embedded inside scripts. Published packages remain
+read-only; use writable drafts and the existing per-run outbox for output.
+
 Skill storage uses `skills.root`. Box execution config lives under `box:`:
 `box.enabled`, `box.backend`, `box.runtime.endpoint`, and `box.local.*`. The old
 `box.local.skills_root` key is read only as an online-upgrade fallback and is
diff --git a/Dockerfile.acceptance b/Dockerfile.acceptance
new file mode 100644
index 000000000..3fa1250a2
--- /dev/null
+++ b/Dockerfile.acceptance
@@ -0,0 +1,2 @@
+FROM local/langbot-oss-411:688e7761
+COPY src/langbot/pkg/persistence/databases/sqlite.py /app/src/langbot/pkg/persistence/databases/sqlite.py
diff --git a/docs/API_KEY_AUTH.md b/docs/API_KEY_AUTH.md
index 3eac9eb10..256da5f45 100644
--- a/docs/API_KEY_AUTH.md
+++ b/docs/API_KEY_AUTH.md
@@ -360,3 +360,42 @@ curl -X POST \
 - API-key-enabled endpoints use the same resource shapes as the web UI
 - No need to learn different API paths - use the existing API documentation with API key authentication
 - API keys never select a Workspace from a request header; their persisted binding is authoritative
+
+## Workspace assistant browser session
+
+The Workspace assistant endpoints under `/api/v1/assistant` are account-session
+features, not API-key/MCP automation tools. They require a Workspace-scoped user
+token and `runtime.operate` for recommendation and turns. Conversation reads
+also enforce account ownership.
+
+- `GET /recommended-model` resolves and validates the recommended tool-capable
+  model when the panel opens. An explicit `model_uuid` in a turn always takes
+  precedence; it never silently falls back if unavailable.
+- `POST /conversations/{id}/turn/stream` accepts the same revision, text/model,
+  or approval decision as `/turn`, and returns NDJSON. `snapshot` frames contain
+  the public conversation; `text` frames contain the cumulative text for the
+  current model invocation; `phase` frames identify the model round or running
+  tool. A snapshot replaces transient text/tool progress. `completed` contains
+  the final conversation (including approval or failed states); `error` is a
+  transport/admission failure. Blank lines are heartbeats.
+- `GET /conversations` lists the latest 100 conversations owned by the current
+  account in this Workspace. `GET /conversations/{id}` includes transient
+  `progress` (cumulative text, round and running tool) while execution is active.
+- Disconnecting a streaming response only detaches the browser. The backend owns
+  the task until completion, approval, timeout or explicit stop. Slow or absent
+  consumers never block execution; reconnect using the conversation snapshot.
+- `POST /conversations/{id}/stop` accepts `{ "revision": N }` for the active turn.
+  It enforces ownership and revision, cancels execution and records `stopped`.
+  Stopping cannot roll back tool effects and clients must never replay a write
+  automatically. Writes still require confirmation of saved arguments.
+- Background execution lives in the current backend process, with at most four
+  simultaneous turns and the existing 120-second turn limit. This is not a durable
+  job queue: backend restarts do not resume interrupted tasks.
+
+The original JSON `/turn` endpoint remains compatible for existing callers.
+
+The assistant recommendation uses Space `/api/v1/models/assistant/recommendation`,
+which selects only from the assistant pool using fresh successful health checks.
+An unavailable recommendation returns 503 without falling back to general chat rankings.
+The browser checks Workspace model providers before mounting the assistant; an empty
+provider list shows Account sign-in and model settings actions instead of a composer.
diff --git a/docs/LLM_LATENCY_BENCHMARK.md b/docs/LLM_LATENCY_BENCHMARK.md
new file mode 100644
index 000000000..c88f4ddfd
--- /dev/null
+++ b/docs/LLM_LATENCY_BENCHMARK.md
@@ -0,0 +1,100 @@
+# LLM request-layer benchmark (2026-10-15)
+
+This controlled test did not reproduce a large LangBot slowdown. Warm request
+preparation was sub-millisecond; most measured latency occurred upstream.
+It did identify two local transport problems, now fixed in this PR.
+
+## Method
+
+One Windows host, Python 3.12.7, LiteLLM 1.88.1 and OpenCode 2.0.16 used the
+same user-provided NewAPI gateway and requested `claude-haiku-4.5`. Each client
+made one warmup and six measured streaming calls, alternating all six client
+orders. The synthetic system prefix was about 6,000 tokens. Parameters were
+`temperature=0`, `max_tokens=256`, `stream=true`, `include_usage=true`, with an
+ephemeral cache marker on the system text. No reasoning budget was requested.
+
+The measurement relay deliberately forwards identical serialized requests and
+headers for all three clients. It replaces OpenCode's added agent/environment
+instructions and normalizes SDK request differences. The canonical SHA-256 is
+`7834a66a188e70b4eb1b6234b3a14be4c66b8c72dd8a2c06395fe2505c1c371b`.
+OpenCode runs as an isolated resident server, using its
+[versioned session API](https://github.com/anomalyco/opencode/blob/v2.0.16/packages/protocol/openapi.json).
+Process startup and session creation are outside the warm dispatch measurement.
+
+All clients share the relay's upstream connection pool. Therefore this experiment
+isolates local request dispatch and upstream response timing; it does **not**
+compare their independent direct-to-provider TLS/proxy behavior. It also excludes
+the complete bot pipeline, Runner/plugin RPC and OpenCode UI rendering. It cannot
+establish which application is faster in every production deployment.
+
+## Results
+
+Medians, milliseconds; six measured samples per client:
+
+| Client | Dispatch to relay | First content | Complete upstream response |
+| --- | ---: | ---: | ---: |
+| Direct HTTPX | 3.952 | 1119.541 | 2199.413 |
+| LangBot Core requester | 6.802 | 1089.092 | 2135.935 |
+| OpenCode | 41.158 | 1104.020 | 2148.265 |
+
+Dispatch includes client preparation and local HTTP dispatch. First content and
+completion use the same relay boundary for all clients, measured from dispatch;
+they are not UI latency. LangBot's separately instrumented preparation median
+was **0.316 ms**, maximum **0.497 ms**. These small samples and upstream variation
+do not justify interpreting the small TTFT differences as a speed ranking.
+
+Every measured response reported 5,990 input tokens, 5,987 cache-read tokens and
+109 output tokens. This demonstrates equal warm cache accounting under the
+normalized upstream request, not native cache-marker forwarding by all clients.
+No cache-miss latency comparison or independent provider-route verification was
+performed. Missing cache accounting is represented as null, never assumed zero.
+The gateway may perform its own routing, retries and cache translation.
+
+Sanitized individual samples and summaries are in
+[llm-latency-2026-10-15.json](llm-latency-2026-10-15.json).
+
+## Bottlenecks fixed
+
+1. The OpenAI SDK stops consuming SSE at `[DONE]` before HTTP framing reaches
+   EOF. The healthy connection was then discarded. In the baseline relay run,
+   LangBot reused zero of six measured client connections; after bounded terminal
+   draining it reused six of six. A real TCP regression test also reproduces this
+   with split terminal markers and late usage. OpenRouter's generic adapter
+   already consumed EOF. Early close and cancellation still close immediately.
+2. On this host, constructing default HTTPX clients took 1,761 / 1,443 / 1,435 ms.
+   Loading one TLS context per requester pool and sharing it across that client's
+   proxy mounts reduced measured pool setup to 666 / 430 / 431 ms. Certificate
+   loading now runs off the event loop. These are local setup measurements, not
+   a claim of equivalent percentage improvement in LLM latency. Runtime lazy
+   initialization and process startup remain separate cold costs.
+3. High-resolution monotonic timing replaces the low-resolution clock used on
+   Python 3.12 Windows. Preparation no longer misleadingly rounds to zero.
+   `transport_setup_ms` exposes pool initialization separately.
+
+The relay socket observations concern client-to-relay connections. HTTPX and
+OpenCode have different keepalive policies, so those counts alone are not a
+cross-client pooling ranking. The shared relay also prevents quantifying the
+provider-side handshake savings from the SSE fix in this run.
+
+## Reproduce
+
+From a development environment with LangBot dependencies installed, set
+`LLM_BENCHMARK_API_KEY` securely in the environment, then run:
+
+```sh
+python scripts/llm_latency_probe.py --opencode /path/to/opencode \
+  --opencode-version 2.0.16 --base-url https://gateway.example/v1 \
+  --model claude-haiku-4.5 --long-prefix --cache-control --rounds 6 \
+  --output /path/outside/repository/latency.json
+```
+
+This makes 21 billable requests by default. `--mock` uses synthetic local SSE
+instead; it still needs OpenCode. `--baseline-no-drain` disables terminal draining
+for a transport regression comparison. The harness records only hashes and
+message sizes, not prompt bodies or credentials, and terminates its own isolated
+OpenCode process. Keep raw results outside the repository because they include
+the configured gateway URL. Other OpenCode versions may need API adjustments.
+
+The next diagnostic for a reported slow production session should capture the
+actual pipeline/Runner timing, streaming mode, prompt size, reasoning settings,
+cache usage and gateway routing alongside these requester-level measurements.
diff --git a/docs/LLM_REQUEST_E2E.md b/docs/LLM_REQUEST_E2E.md
new file mode 100644
index 000000000..21bad844f
--- /dev/null
+++ b/docs/LLM_REQUEST_E2E.md
@@ -0,0 +1,108 @@
+# Request-chain acceptance (2026-10-15)
+
+The acceptance suite starts an isolated Core process, a real SDK Plugin Runtime
+process, and a packaged Runner plugin. It installs the plugin and creates the
+provider, model and Agent through the same authenticated HTTP APIs used by the
+application. Agent Debug requests then traverse:
+
+HTTP -> Core Agent orchestration -> Plugin Runtime -> plugin SDK -> Runtime ->
+Core model manager/request policy -> LiteLLM/provider SDK -> HTTP/SSE provider.
+
+Assertions cover plugin return values, captured provider requests, TCP disconnects
+and persisted SQLite monitoring records. No requester, action transport, model
+manager or monitoring implementation is replaced with a mock.
+
+The remote provider is a deterministic local HTTP server. It injects failures and
+delays reproducibly; these are real process/network integration tests, **not** a
+claim that every external provider has been tested. See the separate
+[latency benchmark](LLM_LATENCY_BENCHMARK.md) for the live NewAPI experiment.
+
+## Matrix
+
+Measured results on Windows / Python 3.12.7 / LiteLLM 1.88.1:
+
+- Core with SDK companion commit `c25ff9b`: **31 passed**, strict stream closure,
+  including all four historical API cases (106.30 seconds).
+- Core with the existing SDK 0.7.11rc5 environment: **27 passed, 4 XFAIL** for
+  the reproduced SDK cancellation defect (107.61 seconds).
+- Published stable SDK 0.7.11 also reproduced the cancellation defect in the
+  preceding 23-case run. The old and stable wrappers were verified identical.
+- Focused Core regressions: **371 passed** on Python 3.13.7, ResourceWarnings fatal.
+- SDK API and stream flow-control regressions: **507 passed**.
+- Existing real Runner discovery/orchestration/event-processor E2E: **3 passed**.
+
+| Area | Cases and evidence |
+| --- | --- |
+| Public APIs | Runner and invocation-bound plugin APIs: message, message with usage, chunk stream and event stream; unchanged SDK entities and nested usage |
+| Historical plugin APIs | Verbatim SDK 0.5.8 proxy and provider-message entities, all four APIs, ordinary calls without `run_id` |
+| Request configuration | Persisted defaults, call overrides, OpenAI cache key/retention, OpenRouter numeric reasoning budget and Anthropic cache marker |
+| Headers | Explicit application attribution and scoped conversation ID reach the provider |
+| Multimodal/tools | Image data URL and function schema survive plugin serialization and provider conversion |
+| Retry/failure | 503 then success, `num_retries=0`, `max_retries=0`, 401 without retry, first-event timeout, total timeout, idle timeout and connection loss after content |
+| Recovery | Each fault followed by a successful request through the same running processes |
+| Cancellation | Early close of Runner and plugin APIs, both chunk and event streams; provider socket closes and the next invocation succeeds |
+| Reasoning | Reasoning content shown/hidden with the existing `remove_think` option; unsupported numeric budget rejected before HTTP |
+| Monitoring | Full cache/reasoning usage, transport timing and structured errors persisted |
+
+Historical API files come from SDK tag `0.5.8`, commit
+`41e320b68e899a7b5e74cb48c5246f6ccbf9d73f`.
+They use their own message serializers and the current Runtime transport. This
+checks old plugin call/response compatibility; it is not a test of running the
+entire old Runtime against the current Core.
+
+## Defects exposed by the full chain
+
+- The plugin bridge flattened model and call policy layers, allowing persisted
+  `request_policy.retries` to override a call's legacy zero-retry option. LLM
+  policy layers are now resolved separately; unrelated argument merging remains.
+- Ordinary plugin calls without a Runner/query returned content but lost usage.
+  Core now creates a trusted, invocation-local query for accounting, using the
+  existing plugin-action query builder. No public payload/signature changes.
+- The shipped OpenRouter requester uses OpenAI metadata. Numeric budgets selected
+  the wrong mapping, and LiteLLM's OpenAI chat adapter stripped cache markers.
+  OpenRouter chat now uses its own protocol adapter. Embedding/rerank routing is
+  unchanged, and a regression test reads the actual shipped requester YAML.
+- Published SDK 0.7.11 leaves inner streaming generators unowned in several public
+  proxy wrappers. Core's direct TCP close test passes, but the complete plugin
+  chain does not promptly cancel its upstream request. The companion
+  [SDK PR #177](https://github.com/langbot-app/langbot-plugin-sdk/pull/177) adds
+  explicit generator ownership without changing signatures, permissions, action
+  payloads or error translation.
+
+**Immediate remote cancellation requires SDK 0.7.12 or later.** Core now pins
+published SDK 0.7.12, which contains the companion fix. Separately deployed
+Runtime and plugin development environments also need the fixed SDK. The E2E
+suite reports the known close failure as XFAIL only for SDK <=0.7.11. Set
+`LLM_E2E_REQUIRE_STREAM_CLOSE=1` to require strict passing with the patched SDK;
+newer SDK versions must pass without this override. Historical API checks are
+explicitly skipped unless their released source files are supplied.
+
+## Reproduction
+
+```sh
+python -m pytest tests/e2e/test_llm_request_chain.py -q --tb=short
+```
+
+For the historical API matrix, export the two unmodified files below from SDK tag
+`0.5.8` into a temporary directory, named `legacy_api.py` and `legacy_message.py`,
+then set `LLM_E2E_LEGACY_API_DIR` to its absolute path:
+
+- `src/langbot_plugin/api/proxies/langbot_api.py`
+- `src/langbot_plugin/api/entities/builtin/provider/message.py`
+
+For strict cancellation acceptance, install the companion SDK into an isolated
+test environment and set `LLM_E2E_REQUIRE_STREAM_CLOSE=1` before running the suite.
+The existing CI E2E job discovers this test module automatically. Tests use random
+localhost ports, test-only credentials and pytest temporary data directories;
+they leave application data and running user instances untouched.
+
+## Limits
+
+The suite does not exercise IM delivery, browser rendering, external provider
+authentication/billing, native Anthropic/Gemini network services, Codex OAuth, or
+complete third-party Runner business logic. It validates multimodal/tool request
+transport rather than image understanding or actual tool execution. Native Codex,
+retry, reasoning, transport and plugin regressions also run in the focused Core
+test suite. A Python 3.13 Windows full-app attempt timed out during startup before
+any LLM request; full process acceptance therefore uses Python 3.12.7, while the
+focused Core regressions run on Python 3.13.7.
diff --git a/docs/LLM_REQUEST_POLICY.md b/docs/LLM_REQUEST_POLICY.md
new file mode 100644
index 000000000..58ebf63ef
--- /dev/null
+++ b/docs/LLM_REQUEST_POLICY.md
@@ -0,0 +1,256 @@
+# LLM request policy and diagnostics
+
+The Core owns chat-call deadlines and retry decisions. The existing SDK
+`Message`, `MessageChunk`, invoke-with-usage and stream-usage shapes are unchanged.
+No SDK or Runner upgrade is required. This policy does not add automatic tool
+replay, model fallback, key rotation, or embedding/rerank retries.
+
+The public API remains compatible with existing plugins. Full-chain testing
+identified a pre-existing SDK <=0.7.11 early-close defect: prompt remote
+cancellation through public plugin/Runner wrappers requires SDK 0.7.12 or later,
+which includes the [SDK fix](https://github.com/langbot-app/langbot-plugin-sdk/pull/177).
+Core now pins 0.7.12; separately deployed Runtime and plugin development
+environments must also install a version containing the fix.
+See [E2E acceptance results](LLM_REQUEST_E2E.md) for tested versions and scope.
+
+## Configuration
+
+Instance defaults live in `data/config.yaml` under `llm_request_policy`; restart
+Core after editing them. Persist model-specific overrides in the existing model
+advanced parameters (`extra_args`). Runners can override them through their
+existing call-level `extra_args` argument:
+
+```json
+{
+  "request_policy": {
+    "retries": 1,
+    "total_timeout": 180,
+    "first_event_timeout": 60,
+    "stream_idle_timeout": 45,
+    "connect_timeout": 10,
+    "read_timeout": 120,
+    "write_timeout": 30,
+    "pool_timeout": 10,
+    "backoff_base": 0.5,
+    "backoff_max": 8,
+    "attribution": true,
+    "conversation_header": null
+  }
+}
+```
+
+Precedence is instance/requester defaults, model extra arguments, then call extra
+arguments. Within each layer, `request_policy.retries` overrides legacy
+`num_retries` / `max_retries`. Zero disables retries. Legacy numeric `timeout`
+sets read, first-event and idle timeouts; it does not extend the total deadline.
+Timeouts are positive finite seconds; retries must be an integer from 0 to 5.
+Unknown policy keys are rejected. Policy keys are removed before calling upstream.
+
+Default retries are **zero**, total timeout is 300 seconds, and first-event and
+idle timeouts are 120 seconds. Retry backoff uses exponential full jitter;
+`Retry-After` seconds or HTTP dates are respected only within the remaining total
+budget. Quota, authentication and invalid-request failures are not retried.
+Core disables both LiteLLM and provider-SDK retries for managed calls. A gateway
+may still retry internally; configure its budget separately.
+
+Retries stop after any raw upstream event (including reasoning/control/usage)
+or, for legacy adapters without raw instrumentation, the first yielded chunk.
+Cancellation is propagated and the owned stream is closed. Total deadlines also
+include downstream consumption time, but Core does not cancel a consumer while
+the generator is suspended at a yield; expiration is enforced on the next pull.
+Transport connect/read/write/pool limits apply to LiteLLM and native Codex HTTP.
+Legacy third-party requesters receive Host-level deadlines but remain responsible
+for their internal transport configuration and internal retries.
+
+## Diagnostics and usage
+
+HTTP monitoring and `get_monitoring_records(kind="llm_calls")` return additive
+`details` containing `timing`, complete provider `usage`, and structured `error`.
+Older records have null details. Existing top-level token counters remain intact.
+Migration `0038_llm_call_details` adds a nullable JSON column and preserves rows.
+
+Timing offsets use a high-resolution monotonic clock, in milliseconds from invocation start:
+
+| Field | Meaning |
+| --- | --- |
+| `request_prepared_ms` | First attempt finished building its request |
+| `request_started_ms` | First adapter call into the network SDK; not a socket-level timestamp |
+| `response_open_ms` | SDK returned a response/stream object |
+| `first_event_ms` | First parsed upstream stream event, including reasoning/control |
+| `first_reasoning_ms` | First upstream reasoning delta when supported |
+| `first_content_ms` | First upstream content delta; inline think tags may still be content |
+| `first_visible_chunk_ms` | First nonempty normalized chunk; may include displayed reasoning |
+| `response_completed_ms` | Host finished consuming the response |
+| `total_ms` | Invocation duration, excluding the subsequent monitoring write |
+| `attempts`, `retry_wait_ms` | Host attempts and total backoff wait |
+| `preparation_ms` | Cumulative LiteLLM request-building time across attempts |
+| `transport_setup_ms` | Cumulative managed HTTP pool acquisition, including first-use TLS setup |
+| `preparation_cache_hits` | Message preparation reused within a retrying invocation |
+
+Missing fields mean not observed or not instrumented, not zero latency. The Core
+does not label these measurements as network-only TTFT. Usage preserves nested
+cache and reasoning counts, costs and provider-specific counters when supplied;
+missing upstream accounting cannot be reconstructed. Partial usage received
+before cancellation/failure is retained. No fabricated usage event is sent early.
+
+`RequesterError` retains its legacy string representation and adds `code`,
+`status_code`, `retryable`, `retry_after`, `request_id`, `phase`, and `as_dict()`.
+Native Codex errors retain their existing ValueError/HTTP error contract.
+Legacy SDK error transport remains unchanged; structured diagnostics are available
+in Core monitoring. The metadata never copies upstream error bodies or headers
+wholesale. Requester-specific raw error strings are not included by the LiteLLM
+error translator.
+
+## Application and conversation headers
+
+Direct OpenRouter requests automatically receive:
+
+```text
+HTTP-Referer: https://langbot.app
+X-OpenRouter-Title: LangBot
+x-session-id: lb-<opaque scoped hash>
+```
+
+Existing headers in model/call `extra_headers` win, case-insensitively. Legacy
+`X-Title` is respected. Set `attribution: false` to disable automatic application
+attribution (explicit headers remain untouched). Detection uses the exact
+`openrouter.ai` hostname, or an OpenRouter requester without an overridden base
+URL; arbitrary gateway hosts do not inherit OpenRouter identity headers.
+
+The conversation hash includes Workspace, bot, pipeline and session identity;
+it is stable across turns and differs across those scopes. It is pseudonymous,
+not a claim of cryptographic anonymization. Calls without conversation context
+omit generated conversation headers. No prompt, raw user ID or group ID is sent.
+
+`x-langbot-conversation-id` is a custom convention, **not a documented universal
+provider standard**. OpenRouter documents `x-session-id` (or body `session_id`,
+which takes precedence). LiteLLM Proxy documents `x-litellm-session-id` and
+vendor session headers. Configure `conversation_header` explicitly for gateways:
+
+```json
+{
+  "request_policy": {"conversation_header": "x-litellm-session-id"},
+  "extra_headers": {
+    "HTTP-Referer": "https://langbot.app",
+    "X-OpenRouter-Title": "LangBot"
+  }
+}
+```
+
+Supported generated header names are `x-session-id`, `x-litellm-session-id` and
+`x-langbot-conversation-id`. The latter is intended for a gateway configured to
+consume that custom header; do not expect automatic routing or caching benefits.
+
+New API may supply its own OpenRouter attribution and may filter/override client
+headers. Configure explicit channel forwarding/override rules and verify the
+outgoing request. Sending a header to New API is not proof OpenRouter received it.
+The local `langbot-new-api` snapshot inspected during implementation overwrites
+attribution with New API; upstream main uses defaults when its outgoing headers
+are empty. This change does not modify or deploy either gateway.
+
+Sources checked on 2026-10-15:
+
+- https://openrouter.ai/docs/app-attribution
+- https://openrouter.ai/docs/guides/best-practices/prompt-caching
+- https://docs.litellm.ai/docs/proxy/request_headers
+- https://docs.litellm.ai/docs/routing
+- https://github.com/QuantumNous/new-api/blob/main/relay/channel/openai/adaptor.go
+- https://github.com/QuantumNous/new-api/blob/main/relay/channel/api_request.go
+
+## Connection and preparation reuse
+
+Initialized OpenAI-compatible, OpenRouter and native Codex requesters own a lazy
+HTTP pool per runtime Provider. Keys, headers and timeout budgets stay request-local.
+The pool holds at most 100 connections, with 20 idle keepalive connections and a
+30-second idle expiry. Provider retirement rejects new leases and closes its pool
+after existing streams finish or are cancelled. Native LiteLLM providers retain
+their own transport handling; custom LiteLLM transport/TLS settings, explicit caller clients and redirected model routes
+are preserved. Bare OpenAI names on OpenRouter retain LiteLLM's alternate routing;
+use OpenRouter's fully qualified names such as `openai/gpt-4o`.
+
+Each pool loads its TLS context once in a worker thread, preserving HTTPX
+certificate and proxy environment settings. Concurrent initialization is serialized;
+cancellation or retirement during certificate loading creates no HTTP client.
+
+OpenAI-compatible streams drain HTTP framing after the SSE `[DONE]` record
+(at most 50 ms / 64 KiB) so the SDK can return healthy connections to the pool.
+Early close and cancellation never trigger this drain; late usage preceding
+`[DONE]` remains available. OpenRouter already consumes EOF in its generic adapter.
+Codex briefly drains the HTTP tail after a completion event (at most 50 ms) to
+release a healthy connection for reuse. Incomplete/cancelled streams are closed.
+Capability lookups use a bounded requester-local cache (128 entries, five minutes).
+Message/tool conversion is reused only within an invocation that permits retries,
+with independent copies for each attempt. No conversation content is cached across
+invocations. Token counting runs in the application's executor instead of blocking
+the event loop. These reduce local overhead; they do not guarantee faster upstream
+inference or a cache hit.
+
+## Unified thinking budgets
+
+Existing effort levels and Runner/SDK signatures remain unchanged. Model
+`reasoning_config` accepts an optional positive integer `budget_tokens`:
+
+```json
+{"reasoning_config": {"level": "enabled", "budget_tokens": 2048}}
+```
+
+Old SDKs can send the same object inside their existing call `extra_args`.
+Precedence is persisted model configuration, model `extra_args.reasoning_config`,
+call `extra_args.reasoning_config`, then the existing Runner override. Each object
+replaces the previous one; a Runner effort override clears a lower-priority budget.
+Omitting `level` with a budget means `enabled`. Budgets cannot be combined with an
+effort level, `disabled`, or native thinking arguments. Unsupported budgets fail
+before the network call; they are never silently converted into an effort level.
+
+Numeric budgets map to Anthropic `thinking.budget_tokens`, Gemini 2.5 thinking
+configuration, supported Qwen `thinking_budget`, or OpenRouter
+`reasoning.max_tokens`. Direct OpenAI/Codex and adaptive-only Claude models keep
+effort-based control. Query `reasoning_capabilities.budget_tokens` for the Core's
+accepted bounds; upstream models/gateways may impose additional limits. OpenRouter
+may translate numeric budgets to effort for models without native numeric support.
+Direct Anthropic calls require an explicit `max_tokens` greater than the budget;
+Core never silently raises the output limit. Existing UI effort selectors are
+unchanged; numeric budgets can be configured through the model API or extra args.
+
+## Prompt caching
+
+Default behavior remains provider-controlled. `extra_args.prompt_caching` is a
+Core-only object and is stripped before sending the request. Call-level objects
+replace model-level objects. There is no local generated-response cache.
+
+```json
+{"prompt_caching": {"mode": "openai", "retention": "in_memory"}}
+```
+
+OpenAI mode adds an opaque conversation-and-model-scoped `prompt_cache_key` when
+conversation context exists. `retention` is optional (`in_memory` or `24h`) and
+must be supported by the destination model/gateway. Native explicit
+`prompt_cache_key`, `prompt_cache_retention`, and newer `prompt_cache_options`
+take precedence. Routing keys are hints, not cache-hit guarantees; current models
+may route caching automatically or use newer native options instead. This mode
+supports OpenAI-compatible/OpenRouter OpenAI models; forwarding through gateways
+must be verified separately.
+
+```json
+{"prompt_caching": {"mode": "anthropic", "ttl": "5m"}}
+```
+
+Anthropic mode supports native Claude and OpenRouter Claude. It adds one ephemeral
+breakpoint to the last text block in the leading system/developer prefix, without
+reordering tools/history or marking user messages. TTL is `5m` (default) or `1h`.
+If any native cache-control marker exists, Core preserves it without adding more.
+With no eligible prefix, no marker is inserted. Provider minimum prefix lengths,
+TTL pricing and eligibility still apply. Explicit `provider_default` adds no fields.
+Codex's restricted Responses body does not accept this caching object; its existing
+server-managed caching behavior is preserved. Gemini explicit cache resource
+creation/lifecycle and general Responses migration are outside this batch.
+
+References checked on 2026-10-15:
+
+- https://developers.openai.com/api/docs/guides/prompt-caching
+- https://platform.claude.com/docs/en/build-with-claude/prompt-caching
+- https://platform.claude.com/docs/en/build-with-claude/extended-thinking
+- https://ai.google.dev/gemini-api/docs/generate-content/thinking
+- https://openrouter.ai/docs/guides/best-practices/reasoning-tokens
+
+See [the latency benchmark](LLM_LATENCY_BENCHMARK.md) for measured results and limitations.
diff --git a/docs/architecture/backend-i18n.md b/docs/architecture/backend-i18n.md
new file mode 100644
index 000000000..bf5a4ba35
--- /dev/null
+++ b/docs/architecture/backend-i18n.md
@@ -0,0 +1,68 @@
+# Backend message language
+
+Workspace settings expose one **Workspace language**, independent of each member's
+WebUI language. It selects LangBot-owned API business errors, model error prefixes,
+Pipeline default notices and international platform replies. Supported locales are
+`en_US`, `zh_Hans`, `zh_Hant`, `ja_JP`, `es_ES`, `ru_RU`, `th_TH` and `vi_VN`.
+Workspaces without a setting default to English. Existing custom messages and
+historical records are not rewritten.
+
+## Storage and boundaries
+
+`pkg/i18n/workspace.py` stores `message_language` in existing `WorkspaceMetadata`;
+no database migration or instance-wide configuration is needed. Reads and writes
+use tenant-scoped units of work. GET/PUT `/api/v1/workspaces/current/language`
+require `workspace.view` / `workspace.update`; matching MCP tools are
+`get_workspace_message_language` and `set_workspace_message_language`.
+
+An authenticated Workspace takes precedence over the browser's
+`X-LangBot-Language` header. The header supplies presentation language only for
+account-only or unauthenticated responses and never selects a tenant. HTTP, MCP,
+Pipeline and adapter boundaries set a ContextVar so concurrent Workspaces do not
+share a language. Adapter SDK thread callbacks use their owning bot's cached
+Workspace language, never the calling request's language.
+
+The per-application cache holds at most 256 Workspaces for 30 seconds. Local
+writes update the cache immediately; other processes refresh at the next async
+message/request boundary after expiry. A running operation keeps its selected
+language. Synchronous SDK callbacks use the most recently loaded value.
+
+## Pipeline overrides
+
+The output `misc` section accepts these optional string settings:
+
+| Setting | Default purpose | Placeholder |
+| --- | --- | --- |
+| `failure-hint` | Request/runner execution failure | none |
+| `rate-limit-hint` | Rate limit reached | none |
+| `content-filter-hint` | Message rejected by content filter | none |
+| `content-filter-error-hint` | Content check failed | none |
+| `tool-call-hint` | Tool call notification | `{tools}` |
+| `runner-not-found-hint` | Configured runner missing | `{runner}` |
+| `runner-not-authorized-hint` | Runner authorization failure | none |
+| `runner-unavailable-hint` | Runner temporarily unavailable | none |
+
+Blank values inherit the Workspace translation. Custom text is emitted literally
+with only the documented placeholder replaced. The previous shipped
+`failure-hint` value `Request failed.` also inherits language; this avoids a data
+migration. Existing output switches (including tool-call visibility and exception
+handling) retain their behavior.
+
+## Adding messages
+
+Use `pkg.i18n.t()` with a stable English template and named parameters. Add the
+same key and placeholders to all eight JSON catalogs. The catalog parity test
+checks completeness and formatting. `pipeline_notice()` handles configurable
+notices. Exact aliases at the API response boundary cover legacy business wording;
+do not add substring replacement of arbitrary error text.
+
+China-only adapters, field validation diagnostics, raw provider/plugin errors,
+tracebacks, protocol values and machine-readable status codes are outside this
+translation layer. Their values remain useful for diagnostics and interoperability.
+Model errors translate LangBot's category/prefix while retaining upstream detail.
+Plugin-authored content must use the plugin's own localization. System prompts and
+user-authored content are not translated.
+
+Locale JSON files are included in Python package data. Focused coverage lives in
+`tests/unit_tests/test_backend_i18n.py`, the Workspace HTTP integration tests and
+`web/tests/e2e/workspace-language.spec.ts`.
diff --git a/docs/architecture/certified-plugins.md b/docs/architecture/certified-plugins.md
index d30d11747..8dc54ae88 100644
--- a/docs/architecture/certified-plugins.md
+++ b/docs/architecture/certified-plugins.md
@@ -10,14 +10,17 @@ ZIP digest without extracting the payload.
 
 Core retains the artifact SHA-256, normalized digest
 (`normalized_zip_digest()`), verification state, declared shared-runtime
-profile, key ID, selected admission profile, and stable admission code in the
+profile, `stateless-v1` component model, key ID, selected admission profile, and stable admission code in the
 durable plugin `install_info._certification` record. The record belongs to the
 installation row; no schema migration is needed for this additive JSON
 metadata.
 
 ## Trusted issuer configuration
 
-Configure the non-secret Ed25519 public-key ring in `data/config.yaml`:
+Hosted Cloud provisions the non-secret Ed25519 issuer public-key ring out of
+band. Key IDs must match the SDK envelope; values are base64 public keys, never
+private signing keys. Invalid configuration fails closed; keep old issuer keys
+through rotation while their signed archives remain installed.
 
 ```yaml
 plugin:
@@ -39,50 +42,56 @@ PLUGIN__CERTIFICATION__TRUSTED_PUBLIC_KEYS_JSON='{"ed25519:issuer":"<base64>"}'
 ```
 
 An **empty** ring is a supported state, not a misconfiguration. OSS defaults to
-it, so a self-hosted instance that has not provisioned any issuer key still
-installs packages (see the admission matrix below). Configure the ring to grant
-the shared-runtime profile; leave it empty to keep every package on the
-dedicated profile.
+it and supports one Workspace; configuring certification keys on OSS is not a
+supported way to enable cross-tenant sharing. Cloud provisions the ring to grant
+shared placement only to eligible v2 artifacts. Certification means eligibility
+for Cloud cross-tenant use of the same Worker and the same plugin/component
+singleton, never a dedicated certificate or intermediate trust tier.
 
 ## Admission matrix
 
 | Deployment | SDK verification | Explicit `administrator_force` | Result |
 | --- | --- | --- | --- |
-| Cloud | valid envelope declaring `shared-runtime-v1` | any | admitted to the shared profile |
-| Cloud | absent | any | reject before storage with `CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_REQUIRED` |
-| Cloud | malformed, untrusted, invalid, or non-shared | any | reject before storage with `CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_INVALID` |
+| Cloud | valid envelope declaring `shared-runtime-v1` + `stateless-v1` | any | admitted to the shared singleton profile |
+| Cloud | no signature | any | install on dedicated worker, without shared eligibility |
+| Cloud | malformed, untrusted, invalid, or non-shared declaration | any | reject before storage with `CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_INVALID`; do not treat a broken signature as unsigned |
 | OSS | absent legacy envelope | any | admitted to the dedicated profile |
-| OSS | valid envelope declaring `shared-runtime-v1` | any | selected shared profile |
-| OSS | declaration signed by a **key this instance resolves** | false | reject with `CERTIFIED_PLUGIN_OSS_FORCE_REQUIRED` |
-| OSS | declaration signed by a **key this instance resolves** | true | admitted to the dedicated profile |
+| OSS | valid envelope declaring `shared-runtime-v1` + `stateless-v1` | any | dedicated only; OSS has no cross-tenant shared placement |
+| OSS | invalid declaration referencing a **key this instance resolves** | false | reject with `CERTIFIED_PLUGIN_OSS_FORCE_REQUIRED` |
+| OSS | invalid declaration referencing a **key this instance resolves** | true | admitted to the dedicated profile |
 | OSS | declaration this instance **cannot resolve** (empty ring) | any | admitted to the dedicated profile |
 
-The OSS row that matters for availability is the last one. Marketplace
-packages are signed by the marketplace issuer and declare
-`shared-runtime-v1`, while OSS ships an empty key ring by default. Treating that
-as a rejection made every certified marketplace package uninstallable with
-`CERTIFIED_PLUGIN_OSS_FORCE_REQUIRED` before artifact storage. Because the
-certificate is signed by an issuer the instance does not declare trusted, no
-shared-runtime privilege may be granted, so admission degrades the install to
-the existing `oss_dev` dedicated profile and records
-`CERTIFIED_PLUGIN_OSS_UNTRUSTED_DEDICATED`. This is not an escalation: it
-withholds the shared profile rather than granting it.
+There is no dedicated certification, dedicated signature, or intermediate
+certification trust tier. Advisory review is an internal prerequisite, not an
+installation trust status. An `issued` marketplace badge, source candidate,
+matching version, or shared artifact/dependency tree alone does not prove Cloud
+admission or Worker sharing. Compare the downloaded archive's ZIP comment,
+normalized digest and signed claims with the public version record and deployed
+Core trust-key ring. Confirm two Workspace bindings share one Worker PID, one
+plugin object, and one object for each declared component before claiming live
+cross-tenant sharing.
 
 A declaration is "resolvable" only when its `key_id` is present in the
-configured ring. When the ring is configured and the declaration still fails
-(malformed, `signature_invalid`, `digest_mismatch`, `unsupported_schema`, ...),
-admission stays explicit and requires `administrator_force`.
+configured ring. A parseable declaration from a resolved key that fails
+signature, digest, identity, profile, or component-model validation requires
+`administrator_force` in OSS. Malformed or unsupported envelopes without a
+resolved certificate identity are treated as untrusted and stay dedicated.
 
 `administrator_force` is deliberately strict: it is recognized only when the
 install request carries boolean `true`. The local upload endpoint accepts the
 multipart field `administrator_force=true`; GitHub and marketplace install
 payloads carry the same field. The existing resource-manage authorization fence
-protects those endpoints. A force never creates a Cloud dedicated fallback.
+protects those endpoints. A force never creates a Cloud dedicated fallback or
+certifies an invalid signature. A legacy v1 certificate may verify
+cryptographically, but without a signed stateless component claim it is
+rejected in Cloud; an old marketplace `issued` badge alone grants no placement.
+Unsigned Cloud archives are the dedicated fallback.
 
 ## Runtime and logs
 
-SDK 0.6.2 carries an installation-level execution mode in both apply and
-authoritative reconcile payloads. Core selects `shared-runtime-v1` only when
+The first stateless-compatible SDK release will carry the v2 certificate and
+singleton component contract; its final version is assigned only at release.
+Core must pin that release before sending or selecting the new contract. Core selects `shared-runtime-v1` only when
 the persisted certification record says verification was valid, both the
 certificate and admission profiles are `shared-runtime-v1`, the admission code
 is shared-eligible, and the record's artifact SHA-256 exactly matches the
@@ -90,6 +99,13 @@ installation row. Missing, malformed, stale, invalid, or dedicated admission
 facts select `dedicated`. Install, upgrade, configuration revision, restart,
 and reconnect all use this same persisted-fact derivation.
 
+The certificate must also bind `component_model=stateless-v1`. This profile
+creates one `BasePlugin` and one instance of each component per digest Worker.
+Installation slots contain immutable config snapshots and authority only;
+task-local invocation context selects the active slot. Older certificates that
+do not bind the component model are not eligible for shared placement. Their
+source packages remain compatible on dedicated Workers under OSS policy.
+
 The existing public plugin-log boundary already applies the immutable
 installation binding (including workspace UUID) through
 `RuntimeConnectionHandler.installation_scope()` before requesting logs. This is
@@ -100,5 +116,7 @@ same existing installation scope.
 
 ## SDK versioning
 
-Core pins `langbot-plugin==0.6.2`, the first published SDK release carrying the
-canonical installation execution-mode contract.
+Ship in dependency order: SDK v2 certificate support, then Core exact pin and
+lockfile, then Space signing, then Runtime/Core rollout. Legacy v1 envelopes
+remain verifiable but do not carry `stateless-v1`, so they never select shared
+singleton placement without re-review and v2 re-signing.
diff --git a/docs/architecture/cloud-plan-revisions.md b/docs/architecture/cloud-plan-revisions.md
new file mode 100644
index 000000000..2cc7d3c37
--- /dev/null
+++ b/docs/architecture/cloud-plan-revisions.md
@@ -0,0 +1,37 @@
+## Cloud plan comparison
+
+`EntitlementSnapshot.plan_revision` and `monthly_bonus_credits` are optional
+signed display metadata. The Space adapter opts into these fields with
+`plan_metadata=2`. `/api/v1/workspaces/current` exposes them as `current_plan`,
+alongside the snapshot's effective limits and feature switches. The existing
+Workspace permission check and entitlement resolver remain authoritative.
+Admission and quotas do not use plan names or display metadata.
+
+The comparison dialog shows the purchased revision even when the catalog has
+no matching plan. It compares resource limits, retention, credits, platform
+restrictions and feature switches; identical benefits use one column even if
+the revision changed. Missing signed metadata is an explicit unavailable state.
+Purchase links for existing paid plans go to Space subscription management,
+where permissions, cancellation and revision eligibility are checked again.
+
+This adds display information to an existing authenticated response, not an
+agent operation. There is no new MCP operation or billing mutation in Core.
+Roll out the matching Core snapshot model before the updated Space adapter.
+
+Verification:
+
+```sh
+uv run pytest tests/unit_tests/cloud/test_entitlements.py tests/integration/api/test_workspaces.py -q
+cd web
+node --test tests/unit/cloud-plans.test.mjs tests/unit/plan-revisions.test.mjs
+pnpm build
+node tests/revisions/verify.mjs
+```
+
+The browser runner expects sibling `space` and `core` checkouts and uses isolated
+fixtures and mocked catalog responses. It never authenticates to real Space or
+initiates a real payment, and shuts down its own servers after the tests.
+
+Purchased benefits are a card in the existing horizontally scrollable plan card area, immediately before the matching latest plan. There is no separate revision comparison panel, and historical prices are not inferred from the latest catalog.
+
+Customer presentation uses subscription dates and “Latest quotas for this plan”, never revision numbers. Unrelated plans and current latest subscriptions have no comparison markers. Core reuses its normal Benefits card content without diff rows or highlighting. Internal entitlement feature keys are not rendered as product benefits. The optional signed subscription_started_at is the current subscription period start, returned only for a matching entitlement revision; it is not snapshot issuance time. Requests with plan_metadata=1 retain their previous response shape.
diff --git a/docs/beta-adapter-diagnostics.md b/docs/beta-adapter-diagnostics.md
deleted file mode 100644
index 2bf6c689e..000000000
--- a/docs/beta-adapter-diagnostics.md
+++ /dev/null
@@ -1,30 +0,0 @@
-# Beta 适配器诊断开发说明
-
-此诊断用于 4.11 适配器事件与 API 的 Beta 验收。仅受支持的 Beta 版本启用；稳定版、Alpha、RC、开发／本地标记版本以及关闭遥测或 Beta 诊断的实例不采集、不上传。
-
-## 采集位置
-
-- `pkg/platform/botmgr.py`：完成适配器事件监听器注册后产生能力快照。`listener_registered` 表示注册完成，不代表网络连接可用。
-- 各适配器的 `_dispatch_eba_event`：调用 `diagnostics.adapter_event_received(self, event)`，记录一次事件已收到。必须位于调用监听器之前，避免路由、插件或模型失败改变接收事实。
-- `pkg/telemetry/adapter_diagnostics.py`：识别适配器 API 最外层调用、平台专用 API 名、聊天与媒体类型。通过内部上下文标记抑制嵌套 API 重复计数。
-- 转换失败无法可靠确定事件种类时，保留未知事件失败；转换成功本身不重复计算事件接收。
-
-新增适配器或原生事件入口时，应检查事件是否经过统一分发入口。只在转换器上添加埋点不足以覆盖直接构造事件及交互回调。
-
-## 数据约束
-
-`adapter_evidence` 标记可用于验收的边界记录；`listener_registered` 标记注册事实；`chat_type`、`content_type` 仅允许有限类别。消息正文、API 参数、用户／群 ID、异常正文、媒体地址不进入这些字段。内部嵌套调用标记不得序列化。
-
-平台专用动作优先匹配清单，名称为 `platform_api.<动作名>`；标准交互 API 保持其标准名称。未知动作不能借通用 `call_platform_api` 包装方法被计为已验收功能。清单读取显式使用 UTF-8，兼容 Windows 默认编码环境。
-
-## 联动 Space
-
-适配器清单或诊断操作变更后，同步生成 Space 的 `internal/diagnostics/catalog_v1.json`，并运行其目录一致性测试。Space 从清单生成完整矩阵，旧诊断记录不作为新增验收证据。部署时应先更新 Space 接收端与数据库迁移，再更新 Core Beta。
-
-Core 回归命令：
-
-```sh
-uv run pytest tests/unit_tests/telemetry -q
-```
-
-重点测试在 `test_adapter_acceptance.py`：版本与开关、隐私、嵌套计数、专用／交互 API、分发入口覆盖、监听器失败。Space 仓库的 `docs/beta-adapter-acceptance.md` 说明验收界面、统计口径及联调方式。
diff --git a/docs/llm-latency-2026-10-15.json b/docs/llm-latency-2026-10-15.json
new file mode 100644
index 000000000..628e02868
--- /dev/null
+++ b/docs/llm-latency-2026-10-15.json
@@ -0,0 +1,411 @@
+{
+  "model": "claude-haiku-4.5",
+  "litellm": "1.88.1",
+  "opencode": "2.0.16",
+  "canonical_sha256": "7834a66a188e70b4eb1b6234b3a14be4c66b8c72dd8a2c06395fe2505c1c371b",
+  "canonical_bytes": 31785,
+  "summary": {
+    "direct": {
+      "samples": 6,
+      "dispatch_to_wire_ms": {
+        "median": 3.952,
+        "max": 5.44
+      },
+      "upstream_ttft_ms": {
+        "median": 1115.744,
+        "max": 1158.609
+      },
+      "ttft_ms": {
+        "median": 1119.541,
+        "max": 1164.048
+      },
+      "total_ms": {
+        "median": 2199.413,
+        "max": 2219.7
+      },
+      "cache_read_tokens": [
+        5987,
+        5987,
+        5987,
+        5987,
+        5987,
+        5987
+      ]
+    },
+    "langbot": {
+      "samples": 6,
+      "dispatch_to_wire_ms": {
+        "median": 6.802,
+        "max": 10.028
+      },
+      "upstream_ttft_ms": {
+        "median": 1079.658,
+        "max": 1113.394
+      },
+      "ttft_ms": {
+        "median": 1089.092,
+        "max": 1120.475
+      },
+      "total_ms": {
+        "median": 2135.935,
+        "max": 2175.882
+      },
+      "preparation_ms": {
+        "median": 0.316,
+        "max": 0.497
+      },
+      "cache_read_tokens": [
+        5987,
+        5987,
+        5987,
+        5987,
+        5987,
+        5987
+      ]
+    },
+    "opencode": {
+      "samples": 6,
+      "dispatch_to_wire_ms": {
+        "median": 41.158,
+        "max": 50.283
+      },
+      "upstream_ttft_ms": {
+        "median": 1063.875,
+        "max": 1158.768
+      },
+      "ttft_ms": {
+        "median": 1104.02,
+        "max": 1201.731
+      },
+      "total_ms": {
+        "median": 2148.265,
+        "max": 2244.956
+      },
+      "cache_read_tokens": [
+        5987,
+        5987,
+        5987,
+        5987,
+        5987,
+        5987
+      ]
+    }
+  },
+  "records": [
+    {
+      "client": "direct",
+      "round": -1,
+      "warmup": true,
+      "dispatch_to_wire_ms": 3.500899998471141,
+      "upstream_ttft_ms": 2094.1667999722995,
+      "ttft_ms": 2097.6676999707706,
+      "total_ms": 3145.2215999597684,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 2098.3635999727994,
+      "consumer_total_ms": 3146.0422999807633,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": -1,
+      "warmup": true,
+      "dispatch_to_wire_ms": 1452.2201999789104,
+      "upstream_ttft_ms": 1122.500299999956,
+      "ttft_ms": 2574.7204999788664,
+      "total_ms": 3640.2128000045195,
+      "preparation_ms": 0.373,
+      "transport_setup_ms": 441.243,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 2576.894700003322,
+      "consumer_total_ms": 3646.833000006154,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": -1,
+      "warmup": true,
+      "dispatch_to_wire_ms": 846.3009999832138,
+      "upstream_ttft_ms": 1082.7391999773681,
+      "ttft_ms": 1929.040199960582,
+      "total_ms": 3569.3477999884635,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 0,
+      "warmup": false,
+      "dispatch_to_wire_ms": 3.5812999703921378,
+      "upstream_ttft_ms": 1012.8218000172637,
+      "ttft_ms": 1016.4030999876559,
+      "total_ms": 2067.0370999723673,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 1017.0787000097334,
+      "consumer_total_ms": 2067.508200008888,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 0,
+      "warmup": false,
+      "dispatch_to_wire_ms": 6.439700024202466,
+      "upstream_ttft_ms": 1067.0202999608591,
+      "ttft_ms": 1073.4599999850616,
+      "total_ms": 2151.4558999915607,
+      "preparation_ms": 0.281,
+      "transport_setup_ms": 0.017,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1074.6371999848634,
+      "consumer_total_ms": 2156.441500002984,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 0,
+      "warmup": false,
+      "dispatch_to_wire_ms": 50.28290004702285,
+      "upstream_ttft_ms": 1151.4483999926597,
+      "ttft_ms": 1201.7313000396825,
+      "total_ms": 2244.956100010313,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 1,
+      "warmup": false,
+      "dispatch_to_wire_ms": 3.269899985753,
+      "upstream_ttft_ms": 1149.7626000200398,
+      "ttft_ms": 1153.0325000057928,
+      "total_ms": 2198.293499997817,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1153.3279999857768,
+      "consumer_total_ms": 2198.7193999812007,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 1,
+      "warmup": false,
+      "dispatch_to_wire_ms": 40.65959999570623,
+      "upstream_ttft_ms": 1158.767800021451,
+      "ttft_ms": 1199.4274000171572,
+      "total_ms": 2242.273700016085,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 1,
+      "warmup": false,
+      "dispatch_to_wire_ms": 6.523299962282181,
+      "upstream_ttft_ms": 1070.9955000202172,
+      "ttft_ms": 1077.5187999824993,
+      "total_ms": 2109.3437999952585,
+      "preparation_ms": 0.497,
+      "transport_setup_ms": 0.015,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1081.5072999685071,
+      "consumer_total_ms": 2113.767499977257,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 2,
+      "warmup": false,
+      "dispatch_to_wire_ms": 7.08060001488775,
+      "upstream_ttft_ms": 1113.394399988465,
+      "ttft_ms": 1120.4750000033528,
+      "total_ms": 2175.882300012745,
+      "preparation_ms": 0.221,
+      "transport_setup_ms": 0.009,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1122.9100999771617,
+      "consumer_total_ms": 2187.183100031689,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 2,
+      "warmup": false,
+      "dispatch_to_wire_ms": 5.439500033389777,
+      "upstream_ttft_ms": 1158.6087999748997,
+      "ttft_ms": 1164.0483000082895,
+      "total_ms": 2219.7004000190645,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 1164.5996000152081,
+      "consumer_total_ms": 2220.197000016924,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 2,
+      "warmup": false,
+      "dispatch_to_wire_ms": 41.65710002416745,
+      "upstream_ttft_ms": 1063.7803999707103,
+      "ttft_ms": 1105.4374999948777,
+      "total_ms": 2145.6969000282697,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 3,
+      "warmup": false,
+      "dispatch_to_wire_ms": 10.027599986642599,
+      "upstream_ttft_ms": 1080.2179000456817,
+      "ttft_ms": 1090.2455000323243,
+      "total_ms": 2105.4020000156015,
+      "preparation_ms": 0.406,
+      "transport_setup_ms": 0.012,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1092.152299999725,
+      "consumer_total_ms": 2109.886699996423,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 3,
+      "warmup": false,
+      "dispatch_to_wire_ms": 38.63289998844266,
+      "upstream_ttft_ms": 1063.9694000128657,
+      "ttft_ms": 1102.6023000013083,
+      "total_ms": 2139.4490000093356,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 3,
+      "warmup": false,
+      "dispatch_to_wire_ms": 4.3889000080525875,
+      "upstream_ttft_ms": 1157.878800004255,
+      "ttft_ms": 1162.2677000123076,
+      "total_ms": 2214.4845000002533,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 1162.9892000346445,
+      "consumer_total_ms": 2214.923900028225,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 4,
+      "warmup": false,
+      "dispatch_to_wire_ms": 34.960800025146455,
+      "upstream_ttft_ms": 1052.52319999272,
+      "ttft_ms": 1087.4840000178665,
+      "total_ms": 2137.3740999842994,
+      "client_socket_reused": true,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 4,
+      "warmup": false,
+      "dispatch_to_wire_ms": 2.0186000037938356,
+      "upstream_ttft_ms": 987.3963000136428,
+      "ttft_ms": 989.4149000174366,
+      "total_ms": 2200.532699993346,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 989.9038000148721,
+      "consumer_total_ms": 2201.0204999824055,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 4,
+      "warmup": false,
+      "dispatch_to_wire_ms": 6.363800028339028,
+      "upstream_ttft_ms": 1112.4532999820076,
+      "ttft_ms": 1118.8171000103466,
+      "total_ms": 2150.362300046254,
+      "preparation_ms": 0.325,
+      "transport_setup_ms": 0.011,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1120.592100021895,
+      "consumer_total_ms": 2154.264000011608,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "opencode",
+      "round": 5,
+      "warmup": false,
+      "dispatch_to_wire_ms": 41.722599999047816,
+      "upstream_ttft_ms": 1059.4015000388026,
+      "ttft_ms": 1101.1241000378504,
+      "total_ms": 2150.8331000222825,
+      "client_socket_reused": false,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "langbot",
+      "round": 5,
+      "warmup": false,
+      "dispatch_to_wire_ms": 8.840300026349723,
+      "upstream_ttft_ms": 1079.098400019575,
+      "ttft_ms": 1087.9387000459246,
+      "total_ms": 2121.5078000095673,
+      "preparation_ms": 0.307,
+      "transport_setup_ms": 0.021,
+      "client_socket_reused": true,
+      "consumer_first_content_ms": 1091.6245000553317,
+      "consumer_total_ms": 2126.561200013384,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    },
+    {
+      "client": "direct",
+      "round": 5,
+      "warmup": false,
+      "dispatch_to_wire_ms": 4.323699977248907,
+      "upstream_ttft_ms": 1081.7263000062667,
+      "ttft_ms": 1086.0499999835156,
+      "total_ms": 2126.9287999602966,
+      "client_socket_reused": false,
+      "consumer_first_content_ms": 1086.7821999709122,
+      "consumer_total_ms": 2127.6911000022665,
+      "input_tokens": 5990,
+      "output_tokens": 109,
+      "cache_read_tokens": 5987
+    }
+  ]
+}
diff --git a/docs/pipeline-migration-config-map.zh-CN.md b/docs/pipeline-migration-config-map.zh-CN.md
index 672a6c8c7..5b2fa7d38 100644
--- a/docs/pipeline-migration-config-map.zh-CN.md
+++ b/docs/pipeline-migration-config-map.zh-CN.md
@@ -18,7 +18,7 @@
 - Dify/Coze/N8n 迁移显式选择 `user-id-source=legacy-session`，Tbox 选择 `legacy-bot`；新建插件配置仍默认 `sender`。旧身份来自 Host 的可信会话/事件/机器人上下文，缺失时失败，不猜测为发送者。
 - Dify/N8n 的旧身份使用原 `query.session`，Coze 使用原 `query`；不能在群聊、不同会话策略或恢复交互时把这两者混同。
 - Coze、Langflow 等现代持久化远端会话机制保留；旧远端会话 ID 不跨账户、端点、Workspace 或 Runner 导入。
-- DashScope 引用字段别名、Langflow 输入输出字段别名、Coze 历史开关别名分别归一化；冲突值阻断，而非靠字段出现顺序决定。
+- DashScope 引用字段别名、Langflow 输入输出字段别名、Coze 历史开关别名分别归一化；冲突时固定采用当前规范字段的值并给出提示，不靠字段出现顺序决定。
 - N8n 的 ignore 响应模式、认证和保留身份字段要维持明确语义。普通业务变量保留支持的 JSON 值，内部/权限/敏感变量不无条件转发。
 - `remove-think` 保留原输出策略，并与支持该字段的 Runner 参数同步；仅用户实际修改才传播，不把表单挂载默认值视为修改。
 - 空字符串、缺失和显式 null 区分处理。WeKnora 的 agent-id 不得在打开/保存其他字段时被默认值替换；未知字段/类型不能靠 truthiness 静默接受。
@@ -26,21 +26,37 @@
 
 ## 插件版本要求
 
-以下是迁移器要求的已发布 Certified Plugin 版本。旧版本即使名称相同，也不能作为当前 Cloud 迁移能力的证明。
+以下是迁移器要求的插件最低版本，不是认证版本清单。已发布的专属运行版本不等于已认证可跨工作区共享；旧版本即使名称相同，也不能作为当前 Cloud 迁移能力的证明。认证须由精确制品的有效签名及 `shared-runtime-v1`、`stateless-v1` 声明证明。
 
-- `local-agent` → `LocalAgent` **0.1.10**。
-- `dify-service-api` → `DifyAgent` **0.1.10**。
-- `coze-api` → `CozeAgent` **0.1.10**。
-- `dashscope-app-api` → `DashScopeAgent` **0.1.10**。
-- `n8n-service-api` → `N8nAgent` **0.1.10**。
-- `langflow-api` → `LangflowAgent` **0.1.10**。
-- `deerflow-api` → `DeerFlowAgent` **0.1.10**。
-- `tbox-app-api` → `TboxAgent` **0.1.8**。
-- `weknora-api` → `WeKnoraAgent` **0.1.10**。
+- `local-agent` → `LocalAgent` **0.2.1**。
+- `dify-service-api` → `DifyAgent` **0.2.3**。
+- `coze-api` → `CozeAgent` **0.2.0**。
+- `dashscope-app-api` → `DashScopeAgent` **0.2.2**。
+- `n8n-service-api` → `N8nAgent` **0.2.0**。
+- `langflow-api` → `LangflowAgent` **0.2.2**。
+- `deerflow-api` → `DeerFlowAgent` **0.2.1**。
+- `tbox-app-api` → `TboxAgent` **0.2.1**。
+- `weknora-api` → `WeKnoraAgent` **0.2.1**。
+
+版本号是经过审阅的最低版本：安装时优先安装该版本；若扩展市场已下架该版本则回退安装最新版。已安装的更高版本不会被降级，迁移执行前仍以运行时 Runner schema 校验迁移结果。
+
+## 宽松迁移规则（planner v5）
+
+迁移器尽量自动修复，只在无法安全运行时阻断：
+
+- 仍然阻断：配置不是对象/不是合法 JSON；必填凭据或端点缺失/为空（api-key、base-url、webhook-url、app-id、bot-id、flow-id、LocalAgent 缺少主模型）；n8n header 名称非法或 basic 认证凭据不是 latin-1；Coze api-base 含凭据、查询串或片段。
+- 自动修复并提示：
+  - 字符串形式的布尔/数字（如 `"true"`、`"30"`）按字段类型转换；无法转换时使用默认值（`migration.value_coerced` / `migration.invalid_value_defaulted`）。
+  - 缺失字段按 4.10 界面默认值补齐（`migration.missing_default_filled`）。
+  - 未知字段与残留的其他 Runner 配置段从活动配置移除，原值保存在迁移备份（`migration.unknown_field_dropped` / `migration.unknown_section_archived`）。
+  - 别名冲突固定取规范字段（`migration.alias_conflict_resolved`）；同时存在 `runner.id` 与旧 `runner.runner` 时采用可用的一项（`migration.mixed_runner_selection`）。
+  - 扩展设置非法时忽略；仅启用部分插件时自动加入目标 Runner 插件（`extensions.runner_added`）。
+- 旧版自定义 Runner（非内置 9 类）标记为 `unsupported_runner`：不迁移、不算失败，需改写为插件 Runner 后再绑定，参见 https://langbot.app/docs/zh/plugin/dev/tutor 。
+- 诊断只包含代码和字段路径，不包含任何值或凭据。
 
 ## 逐字段完整清单
 
-共 9 类 Runner、104 条字段条目；按消费者展开共享字段。未知部署自定义键不冒充已穷举字段，遇到时阻断并保留原值。
+共 9 类 Runner、104 条字段条目；按消费者展开共享字段。未知部署自定义键不冒充已穷举字段，遇到时从活动配置移除、原值保留在迁移备份中并给出提示。
 
 目标路径缩写 `R` 表示 `config.ai.runner_config[当前插件 Runner ID]`；`extensions_preferences` 是流水线的 Host 扩展策略，不是插件全局配置。
 
diff --git a/docs/review/telegram-4.11-acceptance.md b/docs/review/telegram-4.11-acceptance.md
new file mode 100644
index 000000000..987ae0ad5
--- /dev/null
+++ b/docs/review/telegram-4.11-acceptance.md
@@ -0,0 +1,72 @@
+# Telegram 4.11 adapter acceptance
+
+Validated on 2026-10-11 using the Telegram Web client and LangBot WebUI through computer use. Bot registration, bot configuration, processor bindings, group membership changes, and test messages were performed through the UI. No direct Bot API calls or database mutations were used to simulate acceptance traffic.
+
+## Environment
+
+- Core 4.11.0rc2, SDK 0.7.11rc2 on their main branches.
+- OSS development mode, standalone Plugin Runtime and standalone Docker-backed Box Runtime.
+- Dedicated new Telegram bot and dedicated test group; no unrelated conversations used.
+- Pipeline and Agent: LocalAgent with gemini-3-flash-preview. Plugin processors: RunnerDemo community and observer.
+
+## Real message coverage
+
+| Path | Verification |
+| --- | --- |
+| Pipeline private text | TG-P01 received an actual reply. |
+| Pipeline group text and streaming | TG-G04 and TG-STREAM05; observed partial output growing from 12 to 22 to 60 lines, ending with STREAM-END. |
+| Pipeline private streaming | TG-PRIVATE-STREAM10; observed 8 partial lines progressing to 80 complete lines and STREAM-DONE. |
+| Pipeline incoming image | TG-IMG01 correctly identified the red left half and blue right half of a synthetic image. |
+| Pipeline incoming file | TG-FILE01 read marker TG-FILE-7319 and calculated 17 + 25 + 38 = 80. |
+| Pipeline outgoing media | TG-FILE07 delivered a real green PNG and verified.txt as Telegram attachments. |
+| Invalid image fallback | TG-FALLBACK08 delivered a-check.txt once and an intentionally invalid nine-byte PNG as a document once, without replaying the preceding file. |
+| Agent private and group | TG-A01/A02/A03 and TG-COEXIST01 exercised event_reply and real delivery. Historical conversation wording influenced early replies; explicit current instructions were used for subsequent media tests. |
+| Agent group image input | TG-AGENT-GIMG02 correctly identified left red/right blue. |
+| Agent group file input/output | TG-AGENT-GFILE01 read the marker and sum and delivered result.txt (27 bytes); run completed in 10.72 seconds. Downloaded the Telegram attachment and verified its contents: Code: TG-FILE-7319 and Sum: 80. |
+| Agent group image output | TG-AGENT-GOUT01 generated a real 64x64 PNG using Box, delivered it as a photo, and completed in 12.96 seconds. |
+| Plugin private/group text | TG-PLUGIN-P02 and TG-REACTION-02 echo replies were delivered with completed runs after fixing MessageResult. |
+| Plugin attachment events | Community processor independently recorded Image and File components, with completed runs. It intentionally sends no reply without its command prefix. |
+| Independent subscriptions | TG-COEXIST01 produced both the plugin reply and Agent reply. |
+
+Agent event_reply is a complete send, not token streaming. RunnerDemo community provides progress logs but no streaming or generated-media command; those outputs were tested through Pipeline and Agent instead. This is not a claim that every processor implements every output mode.
+
+## Real event coverage
+
+| Event | UI action and observed result |
+| --- | --- |
+| message.received | Sent private/group text and attachments. |
+| message.edited | Edited TG-EDIT01; Agent reported the actual edited event. |
+| message.reaction | Added then removed thumbs-up; observer payloads retained the emoji with is_add true/false. |
+| group.member_joined | Added a dedicated existing test bot; observer recorded the event and community sent a welcome. |
+| group.member_left | Removed that test bot; event had is_kicked true and community sent a departure notice. |
+| group.member_banned | Restricted that test bot's sending permission; observer recorded the restriction event. |
+| bot.invited_to_group | Re-added the QA bot; observer completed at 10:04:12. |
+| bot.removed_from_group | Removed the QA bot while restricted; observer completed at 10:03:55. |
+| bot.muted | Disabled sending permission; indefinite duration was null after the fix. |
+| bot.unmuted | Restored sending while another restriction remained; observer completed at 10:02:29. |
+| platform.specific | Changed administrator status; observer recorded my_chat_member_updated. |
+
+This covers each declared event type, not every possible Telegram platform-specific subtype. Administrator-only updates were tested with a minimal administrator role in the dedicated group. Replying while muted/removed was not treated as an expected successful action.
+
+## Fixes driven by acceptance
+
+1. Normalize numeric conversation IDs before building the Agent protocol context.
+2. Return MessageResult from Telegram send/reply so successful delivery does not subsequently fail plugin action decoding.
+3. Deliver terminal stream attachments and remove an empty Thinking placeholder for attachment-only completions.
+4. Fall back to document delivery for known Telegram photo encoding/dimension rejections, preserving earlier deliveries.
+5. Register MessageReactionHandler with the correct keyword argument instead of accidentally filtering by chat ID.
+6. Preserve the removed emoji and distinguish additions from removals.
+7. Interpret Telegram's Unix-epoch restriction expiry as indefinite; calculate finite duration from event time.
+8. Recognize partial unmute and removal from a restricted state.
+
+## Automated and documentation validation
+
+- 134 focused tests passed across Telegram legacy/new adapters, native stream completion, Omni parity, and Agent event-first protocol.
+- Ruff and git diff whitespace checks passed.
+- Chinese, English, and Japanese Telegram guides updated with processor selection, privacy mode, permissions, streaming/media boundaries, event semantics, and troubleshooting.
+- Sanitized real group screenshot included in all three guides; rendered image verified in the local built documentation.
+- Documentation tests and production static build passed (38 Node tests; 19 Python tests with one skip; 31 static-build tests).
+
+After synchronizing the documentation main branch, a stale DingTalk screenshot assertion in the migration test was updated to match the current guide; the full documentation suite passed again.
+
+Provider-specific errors observed with an earlier model were separated from adapter failures; successful acceptance uses the model stated above. This validation does not assert compatibility with every model provider, Telegram update subtype, or plugin implementation.
diff --git a/packaging/fnos/manifest b/packaging/fnos/manifest
index a05b5be1a..58d657454 100644
--- a/packaging/fnos/manifest
+++ b/packaging/fnos/manifest
@@ -1,5 +1,5 @@
 appname=langbot
-version=4.10.11
+version=4.11.0
 display_name=LangBot
 desc=基于 LLM 的多平台智能对话机器人，支持 QQ、微信、飞书、钉钉、Telegram 等十余种即时通讯平台，内置 Web 管理界面和 AI Agent 能力。
 platform=all
diff --git a/pyproject.toml b/pyproject.toml
index 31d15a7d7..17623340b 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "langbot"
-version = "4.11.0-beta.3"
+version = "4.11.0"
 description = "Production-grade platform for building agentic IM bots"
 readme = "README.md"
 license-files = ["LICENSE"]
@@ -71,7 +71,7 @@ dependencies = [
     "langchain-text-splitters>=1.1.2",
     "chromadb>=1.0.0,<2.0.0",
     "qdrant-client (>=1.15.1,<2.0.0)",
-    "langbot-plugin @ git+https://github.com/langbot-app/langbot-plugin-sdk.git@661b1475091000c94595016d3a4914ff3a1ad81d",
+    "langbot-plugin @ git+https://github.com/langbot-app/langbot-plugin-sdk.git@dc3d98efd4f02565a4d56999f91152393542fd73",
     "asyncpg>=0.30.0",
     "line-bot-sdk>=3.19.0",
     "matrix-nio>=0.25.2",
@@ -129,7 +129,7 @@ requires = ["setuptools>=61.0", "wheel"]
 build-backend = "setuptools.build_meta"
 
 [tool.setuptools]
-package-data = { "langbot" = ["templates/**", "pkg/provider/modelmgr/requesters/*", "pkg/platform/sources/*", "pkg/platform/adapters/**", "web/dist/**", "pkg/persistence/alembic/**"] }
+package-data = { "langbot" = ["templates/**", "pkg/i18n/locales/*.json", "pkg/provider/modelmgr/requesters/*", "pkg/platform/sources/*", "pkg/platform/adapters/**", "web/dist/**", "pkg/persistence/alembic/**"] }
 
 [dependency-groups]
 dev = [
diff --git a/scripts/llm_latency_probe.py b/scripts/llm_latency_probe.py
new file mode 100644
index 000000000..8696f72da
--- /dev/null
+++ b/scripts/llm_latency_probe.py
@@ -0,0 +1,418 @@
+"""Compare a direct stream, Core requester and isolated OpenCode 2.x server.
+
+All three clients use a local measurement relay. The relay sends the same
+synthetic request bytes and headers upstream, deliberately removing agent prompt
+differences. This measures the request layer, not a production bot pipeline/UI.
+Credentials are read from an environment variable or an untracked local file.
+"""
+
+import argparse
+import asyncio
+import contextlib
+import hashlib
+import itertools
+import json
+import os
+from pathlib import Path
+import socket
+import statistics
+import secrets
+import subprocess
+import tempfile
+import time
+from types import SimpleNamespace
+
+import httpx
+from aiohttp import web
+
+
+def digest(value):
+    return hashlib.sha256(json.dumps(value, sort_keys=True, separators=(',', ':')).encode()).hexdigest()
+
+
+class SSEMeter:
+    """Incrementally parse SSE without assuming network chunk boundaries."""
+
+    def __init__(self):
+        self.buffer = b''
+        self.first_event = None
+        self.first_content = None
+        self.usage = None
+        self.content_chars = 0
+        self.done = False
+
+    def feed(self, data, now):
+        self.buffer += data
+        while b'\n' in self.buffer:
+            line, self.buffer = self.buffer.split(b'\n', 1)
+            if not line.startswith(b'data:'):
+                continue
+            raw = line[5:].strip()
+            if raw == b'[DONE]':
+                self.done = True
+                continue
+            item = json.loads(raw)
+            if self.first_event is None:
+                self.first_event = now
+            if item.get('usage'):
+                self.usage = item['usage']
+            for choice in item.get('choices', []):
+                content = (choice.get('delta') or {}).get('content')
+                if content:
+                    if self.first_content is None:
+                        self.first_content = now
+                    self.content_chars += len(content)
+
+
+def cache_tokens(usage):
+    if not isinstance(usage, dict):
+        return None
+    detail = usage.get('prompt_tokens_details') or usage.get('input_tokens_details') or {}
+    return detail.get('cached_tokens', usage.get('cache_read_input_tokens'))
+
+
+def summary(records):
+    result = {}
+    for name in sorted({row['client'] for row in records}):
+        rows = [r for r in records if r['client'] == name and not r.get('warmup')]
+        stats = {'samples': len(rows)}
+        for key in ('dispatch_to_wire_ms', 'upstream_ttft_ms', 'ttft_ms', 'total_ms', 'preparation_ms'):
+            values = [r[key] for r in rows if r.get(key) is not None]
+            if values:
+                stats[key] = {'median': round(statistics.median(values), 3), 'max': round(max(values), 3)}
+        stats['cache_read_tokens'] = [cache_tokens(r.get('usage')) for r in rows]
+        result[name] = stats
+    return result
+
+
+def free_port():
+    with socket.socket() as sock:
+        sock.bind(('127.0.0.1', 0))
+        return sock.getsockname()[1]
+
+
+async def run(options):
+    # Import/initialization cost is reported separately from resident calls.
+    started = time.perf_counter()
+    from langbot.pkg.provider.modelmgr.requesters.litellmchat import LiteLLMRequester
+    from langbot.pkg.provider.modelmgr.request_policy import CallState, RequestPolicy, current_call
+    from langbot_plugin.api.entities.builtin.provider.message import Message
+    import importlib.metadata
+
+    import_ms = (time.perf_counter() - started) * 1000
+    credentials = json.loads(Path(options.credentials_file).read_text()) if options.credentials_file else {}
+    base_url = options.base_url or credentials.get('base_url')
+    key = os.environ.get(options.api_key_env) or credentials.get('api_key')
+    if not options.mock and not (base_url and key):
+        raise ValueError('A base URL and API key are required for live mode')
+    prefix = 'You are a concise technical assistant. Follow only the final question.\n'
+    if options.long_prefix:
+        prefix += ''.join(
+            f'Reference entry {i}: HTTP connection pooling reuses established transports. '
+            'Keep authentication and timeouts request scoped. Measure first content separately from reasoning.\n'
+            for i in range(180)
+        )
+    canonical = {
+        'model': options.model,
+        'stream': True,
+        'stream_options': {'include_usage': True},
+        'temperature': 0,
+        'max_tokens': 256,
+        'messages': [
+            {
+                'role': 'system',
+                'content': [
+                    {
+                        'type': 'text',
+                        'text': prefix,
+                        **({'cache_control': {'type': 'ephemeral'}} if options.cache_control else {}),
+                    }
+                ],
+            },
+            {'role': 'user', 'content': 'Explain HTTP connection pooling in exactly five short numbered sentences.'},
+        ],
+    }
+    payload = json.dumps(canonical, separators=(',', ':')).encode()
+    records = []
+    active = None
+    upstream = httpx.AsyncClient(timeout=90)
+    relay_sockets = {}
+
+    async def relay(request):
+        nonlocal active
+        if active is None or active.get('wire_started') is not None:
+            return web.json_response({'error': 'Unexpected extra request'}, status=409)
+        row = active
+        finished = row['relay_finished']
+        row['wire_started'] = time.perf_counter()
+        incoming = await request.json()
+        row['received_body_hash'] = digest(incoming)
+        row['received_fields'] = sorted(incoming)
+        row['received_message_chars'] = len(json.dumps(incoming.get('messages', [])))
+        row['upstream_body_hash'] = digest(canonical)
+        row['dispatch_to_wire_ms'] = (row['wire_started'] - row['start']) * 1000
+        connection = id(request.transport)
+        row['client_socket_reused'] = connection in relay_sockets
+        relay_sockets[connection] = request.transport
+        meter = SSEMeter()
+        response = web.StreamResponse(headers={'Content-Type': 'text/event-stream'})
+        try:
+            if options.mock:
+                await response.prepare(request)
+                await asyncio.sleep(0.05)
+                for i in range(20):
+                    event = {
+                        'id': 'synthetic',
+                        'object': 'chat.completion.chunk',
+                        'created': 1,
+                        'model': options.model,
+                        'choices': [{'index': 0, 'delta': {'content': 'word '}, 'finish_reason': None}],
+                    }
+                    data = ('data: ' + json.dumps(event) + '\n\n').encode()
+                    meter.feed(data, time.perf_counter())
+                    await response.write(data)
+                    await asyncio.sleep(0.03)
+                event['choices'] = [{'index': 0, 'delta': {}, 'finish_reason': 'stop'}]
+                event['usage'] = {'prompt_tokens': 50, 'completion_tokens': 20, 'total_tokens': 70}
+                data = ('data: ' + json.dumps(event) + '\n\ndata: [DONE]\n\n').encode()
+                meter.feed(data, time.perf_counter())
+                await response.write(data)
+            else:
+                async with upstream.stream(
+                    'POST',
+                    base_url.rstrip('/') + '/chat/completions',
+                    content=payload,
+                    headers={'Authorization': 'Bearer ' + key, 'Content-Type': 'application/json'},
+                ) as remote:
+                    row['status'] = remote.status_code
+                    if remote.status_code != 200:
+                        await remote.aread()
+                        # Do not log upstream response bodies, headers or credentials.
+                        return web.json_response(
+                            {'error': f'Upstream HTTP {remote.status_code}'}, status=remote.status_code
+                        )
+                    row['upstream_headers_ms'] = (time.perf_counter() - row['wire_started']) * 1000
+                    await response.prepare(request)
+                    async for chunk in remote.aiter_bytes():
+                        meter.feed(chunk, time.perf_counter())
+                        try:
+                            await response.write(chunk)
+                        except ConnectionResetError:
+                            row['client_closed_before_http_eof'] = True
+            row['upstream_ttft_ms'] = (
+                (meter.first_content - row['wire_started']) * 1000 if meter.first_content is not None else None
+            )
+            row['ttft_ms'] = (meter.first_content - row['start']) * 1000 if meter.first_content is not None else None
+            row['usage'] = meter.usage
+            row['content_chars'] = meter.content_chars
+            row['upstream_complete_ms'] = (time.perf_counter() - row['wire_started']) * 1000
+            with contextlib.suppress(ConnectionResetError):
+                await response.write_eof()
+            return response
+        finally:
+            finished.set()
+
+    app = web.Application()
+    app.router.add_post('/v1/chat/completions', relay)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    local = f'http://127.0.0.1:{site._server.sockets[0].getsockname()[1]}/v1'
+    adapter = LiteLLMRequester(
+        SimpleNamespace(),
+        {'custom_llm_provider': 'openai', 'base_url': local, 'requester_name': 'new-api-chat-completions'},
+    )
+    await adapter.initialize()
+    if options.baseline_no_drain:
+        adapter._http_pool.drain_sse = False
+    model = SimpleNamespace(
+        model_entity=SimpleNamespace(name=options.model, extra_args={}, reasoning_config=None, abilities=[]),
+        reasoning_config_override=None,
+        provider=SimpleNamespace(
+            token_mgr=SimpleNamespace(get_token=lambda: 'local-benchmark'),
+            provider_entity=SimpleNamespace(requester='new-api-chat-completions'),
+        ),
+    )
+    messages = [Message(role=x['role'], content=x['content']) for x in canonical['messages']]
+    process = None
+    with tempfile.TemporaryDirectory(prefix='langbot-llm-benchmark-') as directory:
+        root = Path(directory)
+        config = {
+            'model': f'bench/{options.model}',
+            'small_model': f'bench/{options.model}',
+            'autoupdate': False,
+            'provider': {
+                'bench': {
+                    'npm': '@ai-sdk/openai-compatible',
+                    'name': 'Benchmark',
+                    'options': {'baseURL': local, 'apiKey': 'local-benchmark'},
+                    'models': {options.model: {'name': 'Benchmark', 'limit': {'context': 200000, 'output': 256}}},
+                }
+            },
+            'agent': {'bench': {'mode': 'primary', 'prompt': prefix, 'tools': {'*': False}, 'steps': 2}},
+            'permission': {'*': 'deny'},
+        }
+        config_path = root / 'opencode.json'
+        config_path.write_text(json.dumps(config))
+        env = os.environ.copy()
+        server_password = secrets.token_urlsafe(24)
+        env.update(
+            {
+                'OPENCODE_CONFIG': str(config_path),
+                'XDG_CONFIG_HOME': str(root / 'config'),
+                'XDG_DATA_HOME': str(root / 'data'),
+                'XDG_CACHE_HOME': str(root / 'cache'),
+                'NO_PROXY': os.environ.get('NO_PROXY', ''),
+                'OPENCODE_SERVER_PASSWORD': server_password,
+            }
+        )
+        port = free_port()
+        server = f'http://127.0.0.1:{port}'
+        log = (root / 'server.log').open('w')
+        try:
+            start_server = time.perf_counter()
+            process = await asyncio.create_subprocess_exec(
+                options.opencode,
+                'serve',
+                '--hostname',
+                '127.0.0.1',
+                '--port',
+                str(port),
+                cwd=root,
+                env=env,
+                stdout=log,
+                stderr=log,
+                creationflags=subprocess.CREATE_NO_WINDOW if os.name == 'nt' else 0,
+            )
+            async with httpx.AsyncClient(trust_env=False, timeout=100, auth=('opencode', server_password)) as client:
+                for _ in range(120):
+                    try:
+                        response = await client.get(server + '/api/info')
+                        if response.is_success:
+                            break
+                    except httpx.ConnectError:
+                        pass
+                    await asyncio.sleep(0.25)
+                else:
+                    raise RuntimeError('Isolated OpenCode server did not start')
+                startup_ms = (time.perf_counter() - start_server) * 1000
+                orders = list(itertools.permutations(('direct', 'langbot', 'opencode')))
+                for round_index in range(-1, options.rounds):
+                    for name in orders[max(0, round_index) % len(orders)]:
+                        session_id = None
+                        if name == 'opencode':
+                            response = await client.post(
+                                server + '/api/session',
+                                json={
+                                    'title': 'LLM benchmark',
+                                    'agent': 'bench',
+                                    'model': {'id': options.model, 'providerID': 'bench'},
+                                    'location': {'directory': str(root)},
+                                },
+                            )
+                            response.raise_for_status()
+                            session_id = response.json()['data']['id']
+                        row = {
+                            'client': name,
+                            'round': round_index,
+                            'warmup': round_index < 0,
+                            'start': time.perf_counter(),
+                            'relay_finished': asyncio.Event(),
+                        }
+                        active = row
+                        if name == 'direct':
+                            meter = SSEMeter()
+                            async with client.stream('POST', local + '/chat/completions', json=canonical) as response:
+                                response.raise_for_status()
+                                async for chunk in response.aiter_bytes():
+                                    meter.feed(chunk, time.perf_counter())
+                            row['consumer_first_content_ms'] = (meter.first_content - row['start']) * 1000
+                        elif name == 'langbot':
+                            state = CallState(RequestPolicy(retries=0, total_timeout=90))
+                            token = current_call.set(state)
+                            try:
+                                async with contextlib.aclosing(
+                                    adapter.invoke_llm_stream(
+                                        None, model, messages, extra_args={'temperature': 0, 'max_tokens': 256}
+                                    )
+                                ) as stream:
+                                    async for chunk in stream:
+                                        if chunk.content and 'consumer_first_content_ms' not in row:
+                                            row['consumer_first_content_ms'] = (
+                                                time.perf_counter() - row['start']
+                                            ) * 1000
+                                row['preparation_ms'] = state.metrics.get('preparation_ms')
+                                row['transport_setup_ms'] = state.metrics.get('transport_setup_ms')
+                                row['client_usage'] = state.usage
+                            finally:
+                                current_call.reset(token)
+                        else:
+                            response = await client.post(
+                                server + f'/api/session/{session_id}/prompt',
+                                json={'text': canonical['messages'][-1]['content']},
+                            )
+                            response.raise_for_status()
+                            await asyncio.wait_for(row['relay_finished'].wait(), 100)
+                        if name != 'opencode':
+                            row['consumer_total_ms'] = (time.perf_counter() - row['start']) * 1000
+                        await asyncio.wait_for(row['relay_finished'].wait(), 100)
+                        if 'upstream_complete_ms' not in row:
+                            raise RuntimeError(f'Incomplete benchmark response (HTTP {row.get("status", "unknown")})')
+                        # Use one completion boundary for all clients. This
+                        # relay cannot measure OpenCode's UI rendering latency.
+                        row['total_ms'] = row['dispatch_to_wire_ms'] + row['upstream_complete_ms']
+                        row.pop('relay_finished')
+                        row.pop('start')
+                        row.pop('wire_started', None)
+                        records.append(row)
+                        active = None
+                        print(json.dumps(row), flush=True)
+                        await asyncio.sleep(0.1)
+                report = {
+                    'model': options.model,
+                    'base_url': base_url,
+                    'mock': options.mock,
+                    'long_prefix': options.long_prefix,
+                    'cache_control': options.cache_control,
+                    'baseline_no_drain': options.baseline_no_drain,
+                    'canonical_sha256': digest(canonical),
+                    'canonical_bytes': len(payload),
+                    'python_import_ms': import_ms,
+                    'opencode_startup_ms': startup_ms,
+                    'litellm': importlib.metadata.version('litellm'),
+                    'opencode': options.opencode_version,
+                    'records': records,
+                    'summary': summary(records),
+                }
+                Path(options.output).write_text(json.dumps(report, indent=2), encoding='utf-8')
+                print('SUMMARY', json.dumps(report['summary']), flush=True)
+        finally:
+            if process and process.returncode is None:
+                process.terminate()
+                await process.wait()
+            log.close()
+            await adapter.aclose()
+            await upstream.aclose()
+            await runner.cleanup()
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument('--opencode', required=True)
+    parser.add_argument('--opencode-version', default='unknown')
+    parser.add_argument('--model', default='claude-haiku-4.5')
+    parser.add_argument('--base-url')
+    parser.add_argument('--credentials-file')
+    parser.add_argument('--api-key-env', default='LLM_BENCHMARK_API_KEY')
+    parser.add_argument('--rounds', type=int, default=6)
+    parser.add_argument('--mock', action='store_true')
+    parser.add_argument('--long-prefix', action='store_true')
+    parser.add_argument('--cache-control', action='store_true')
+    parser.add_argument('--baseline-no-drain', action='store_true', help='Reproduce the pre-fix stream transport')
+    parser.add_argument('--output', required=True)
+    args = parser.parse_args()
+    if not 1 <= args.rounds <= 12:
+        parser.error('--rounds must be between 1 and 12 (plus one warmup round)')
+    os.environ['NO_PROXY'] = ','.join(filter(None, [os.environ.get('NO_PROXY'), '127.0.0.1', 'localhost']))
+    asyncio.run(run(args))
diff --git a/skills/scripts/e2e/onebot-runtime-probe.py b/skills/scripts/e2e/onebot-runtime-probe.py
index a080571fa..c919803bf 100644
--- a/skills/scripts/e2e/onebot-runtime-probe.py
+++ b/skills/scripts/e2e/onebot-runtime-probe.py
@@ -6,14 +6,15 @@ from __future__ import annotations
 import argparse
 import asyncio
 import json
+import os
 import time
 
 import websockets
 
 
-async def run(port: int) -> None:
-    uri = f'ws://127.0.0.1:{port}/ws'
+async def run(uri: str) -> None:
     headers = {
+        'Authorization': 'Bearer ' + os.environ['ONEBOT_PROBE_TOKEN'],
         'X-Self-ID': '900001',
         'X-Client-Role': 'Universal',
         'User-Agent': 'LangBot-E2E-OneBot/1.0',
@@ -27,7 +28,7 @@ async def run(port: int) -> None:
         except OSError:
             await asyncio.sleep(0.25)
     if connection is None:
-        raise RuntimeError(f'OneBot reverse WebSocket did not open on port {port}')
+        raise RuntimeError('OneBot reverse WebSocket did not open')
 
     actions: list[str] = []
     event_deadline = time.monotonic() + 20
@@ -88,6 +89,6 @@ async def run(port: int) -> None:
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()
-    parser.add_argument('--port', type=int, required=True)
+    parser.add_argument('--url', required=True)
     args = parser.parse_args()
-    asyncio.run(run(args.port))
+    asyncio.run(run(args.url))
diff --git a/skills/scripts/e2e/wizard-onebot-agent-runtime.mjs b/skills/scripts/e2e/wizard-onebot-agent-runtime.mjs
index dc94efe31..29f5b5506 100644
--- a/skills/scripts/e2e/wizard-onebot-agent-runtime.mjs
+++ b/skills/scripts/e2e/wizard-onebot-agent-runtime.mjs
@@ -1,7 +1,7 @@
 #!/usr/bin/env node
 
 import { spawn } from "node:child_process";
-import { createServer } from "node:net";
+import { randomBytes } from "node:crypto";
 import { dirname, join } from "node:path";
 import { fileURLToPath } from "node:url";
 import {
@@ -60,19 +60,6 @@ const result = {
   evidence_collected: ["ui", "screenshot", "console", "api_diagnostic"],
 };
 
-async function getFreePort() {
-  const server = createServer();
-  await new Promise((resolve, reject) => {
-    server.once("error", reject);
-    server.listen(0, "127.0.0.1", resolve);
-  });
-  const address = server.address();
-  const port = typeof address === "object" && address ? address.port : 0;
-  await new Promise((resolve) => server.close(resolve));
-  if (!port) throw new Error("Could not allocate a temporary OneBot port.");
-  return port;
-}
-
 async function api(page, path, options = {}) {
   return await page.evaluate(
     async ({ baseUrl, path, options, token }) => {
@@ -96,7 +83,7 @@ async function api(page, path, options = {}) {
   );
 }
 
-async function runProbe(port) {
+async function runProbe(url, accessToken) {
   const repo = await resolveLangBotRepo();
   const python = join(repo, ".venv", "bin", "python");
   const script = join(
@@ -107,8 +94,9 @@ async function runProbe(port) {
     throw new Error(`LangBot virtualenv Python is missing: ${python}`);
   }
   return await new Promise((resolve, reject) => {
-    const child = spawn(python, [script, "--port", String(port)], {
+    const child = spawn(python, [script, "--url", url], {
       cwd: repo,
+      env: { ...process.env, ONEBOT_PROBE_TOKEN: accessToken },
     });
     let stdout = "";
     let stderr = "";
@@ -132,7 +120,7 @@ async function runProbe(port) {
 try {
   if (!frontendUrl) throw new Error("LANGBOT_FRONTEND_URL is not configured.");
   if (!backendUrl) throw new Error("LANGBOT_BACKEND_URL is not configured.");
-  const port = await getFreePort();
+  const accessToken = randomBytes(32).toString("hex");
 
   browser = await createBrowser(paths);
   const { page } = browser;
@@ -190,9 +178,7 @@ try {
     .getByText(/Configure Your Bot|配置机器人|ボットを設定/)
     .first()
     .waitFor({ timeout: 15_000 });
-  const portInput = page.getByRole("spinbutton").first();
-  await portInput.fill(String(port));
-  await portInput.press("Tab");
+  await page.getByLabel(/Access token|访问令牌|アクセストークン/i).fill(accessToken);
   await page
     .getByRole("button", {
       name: /Save & Enable Bot|保存并启用|保存して有効化/,
@@ -270,7 +256,7 @@ try {
   }
   result.visible_signals.push("route-persisted");
 
-  result.runtime.onebot = await runProbe(port);
+  result.runtime.onebot = await runProbe(`${backendUrl.replace(/^http/, "ws").replace(/\/$/, "")}/bots/${botId}/onebot/v11/ws`, accessToken);
   let routeStatus;
   for (let attempt = 0; attempt < 30; attempt += 1) {
     routeStatus = await api(
diff --git a/skills/skills.index.json b/skills/skills.index.json
index 3e8695455..25d64cf20 100644
--- a/skills/skills.index.json
+++ b/skills/skills.index.json
@@ -1805,7 +1805,7 @@
         },
         {
           "id": "sandbox-skill-authoring-e2e",
-          "title": "Local Agent creates, registers, activates, and uses a sandbox skill",
+          "title": "Local Agent publishes and executes an immutable Skill revision",
           "mode": "agent-browser",
           "area": "sandbox",
           "type": "regression",
@@ -1817,9 +1817,7 @@
             "skills",
             "local-agent",
             "tools",
-            "e2b",
-            "nsjail",
-            "host"
+            "nsjail"
           ],
           "automation": "",
           "setup_automation": [],
@@ -1833,7 +1831,7 @@
         },
         {
           "id": "sandbox-skill-authoring-edit-existing-e2e",
-          "title": "Local Agent modifies an activated sandbox skill package",
+          "title": "Local Agent publishes a new Skill revision without changing an active run",
           "mode": "agent-browser",
           "area": "sandbox",
           "type": "regression",
@@ -1845,17 +1843,12 @@
             "skills",
             "local-agent",
             "tools",
-            "edit"
-          ],
-          "automation": "scripts/e2e/pipeline-debug-chat.mjs",
-          "setup_automation": [
-            "node:scripts/e2e/ensure-local-agent-pipeline.mjs --write-env",
-            "node:scripts/e2e/reset-sandbox-skill.mjs --name=lb-skill-edit-regression"
-          ],
-          "setup_provides_env": [
-            "LANGBOT_LOCAL_AGENT_PIPELINE_URL",
-            "LANGBOT_LOCAL_AGENT_PIPELINE_NAME"
+            "edit",
+            "nsjail"
           ],
+          "automation": "",
+          "setup_automation": [],
+          "setup_provides_env": [],
           "evidence_required": [
             "ui",
             "backend_log",
diff --git a/skills/skills/langbot-deploy/SKILL.md b/skills/skills/langbot-deploy/SKILL.md
index 2f61af768..7661e7647 100644
--- a/skills/skills/langbot-deploy/SKILL.md
+++ b/skills/skills/langbot-deploy/SKILL.md
@@ -7,6 +7,16 @@ description: Deploy and configure a LangBot instance — Docker / Docker Compose
 
 Covers running LangBot in production. For development see `langbot-dev`.
 
+Read https://langbot.app/llms.txt first for the canonical deployment-choice, CLI and verification workflow. The summary below applies when selecting hosting.
+
+## Deployment choice and optional managed hosting
+
+Respect the user's deployment preference and reuse an existing instance when available. Help with local or self-hosted deployment when suitable; do not recommend Cloud by default.
+
+Offer [LangBot Cloud](https://space.langbot.app/cloud) when the user wants managed hosting without maintaining a server, the local environment is unsuitable, or a deployment problem remains difficult after reasonable troubleshooting. Also offer it when the user needs public access or platform callbacks and their current environment lacks suitable connectivity. Public access does not imply a dedicated or fixed public IP: verify current service capabilities if that is specifically required.
+
+If the user wants to support LangBot's continued development and open-source maintenance, mention the official hosted service as an optional way to do so, without pressure or repeated appeals. Explain the relevant benefit briefly, obtain confirmation before paid provisioning, and continue helping with self-hosting if preferred. Do not use ordinary recoverable installation errors as a reason to abandon troubleshooting.
+
 ## Docker Compose (recommended)
 
 ```bash
diff --git a/skills/skills/langbot-mcp-ops/SKILL.md b/skills/skills/langbot-mcp-ops/SKILL.md
index d96e56f2f..ae73d104c 100644
--- a/skills/skills/langbot-mcp-ops/SKILL.md
+++ b/skills/skills/langbot-mcp-ops/SKILL.md
@@ -60,6 +60,19 @@ To inspect key identity and permissions, call `GET /api/v1/system/context` with
 
 ## Tool surface
 
+Slack quick setup: create a disabled `slack-omni` bot draft first, then use
+`start_slack_setup` with the user's App Configuration access/refresh tokens.
+Choose `socket_mode=true` with a browser-reachable OAuth `redirect_url`, or
+provide the public HTTPS `webhook_url` ending in `/bots/<bot_uuid>`.
+Poll `get_slack_setup_status` for the Slack installation authorization URL;
+the user must authorize installation. On success, apply the returned `config`
+using the normal bot update flow. Socket Mode also requires a separately
+generated `app_token` (`xapp-`, scope `connections:write`). Never log tokens.
+`cancel_slack_setup` removes the temporary session without deleting the Slack
+application. Sessions expire after 15 minutes and are bound to the initiating
+Workspace, principal, and placement generation. Restarting a failed setup can
+create another Slack application; inspect the returned `app_id` first.
+
 The tools wrap the LangBot service layer. Current tools (v1):
 
 | Tool | Purpose |
@@ -174,3 +187,81 @@ Legacy EventListener plugins remain in the Pipeline lifecycle.
 `list_processor_runs` includes `created_at_ms`, `started_at_ms`, and
 `finished_at_ms`: Host lifecycle times in epoch milliseconds. Use the start and finish times for elapsed processing time; select a run and call `get_processor_run_events` for its
 logs and action results. These times are not internal plugin profiling data.
+
+## Unified execution monitoring
+
+Use `get_inflight_executions` for a bounded current-workspace snapshot of active
+and recently started executions, including `progress_event`. Progress is a
+reported stage, not an estimated completion percentage. The UI uses the shared
+`GET /api/v1/monitoring/in-flight/stream` SSE feed instead of polling per viewer.
+
+Use `get_monitoring_executions` for the execution list and
+its legacy `pipeline_ids` parameter to filter any processor kind (Agent,
+Pipeline or event processor); the summary uses the same processor scope. Use
+`get_monitoring_execution_detail` with `source=auto` to resolve a run,
+message or event identifier outside the current list page. The detail exposes
+`inputs`, `outputs` (generated content), `deliveries` (recorded platform sends),
+`events`, `llm_calls`, `tool_calls`, `errors`, `related`, and `conversation` in
+`pages`. Follow each section's `has_more` and `next_offset` independently.
+Conversation history supplies context; historical messages without explicit
+links must not be asserted to belong to the selected execution. Events without
+a processor run use `source=event`. All lookups remain Workspace-scoped.
+
+Monitoring record filters accept `mode` (`all`, `real`, `debug`) and
+`execution_statuses` (normalized execution statuses). These select the owning
+execution, not the individual model/tool call outcome. Calls without a recorded
+execution link are excluded when an execution filter is active.
+
+### Workspace message language
+
+`get_workspace_message_language` reads the current Workspace's backend language
+(`workspace.view`). `set_workspace_message_language(language)` requires
+`workspace.update`; supported values are `en_US`, `zh_Hans`, `zh_Hant`, `ja_JP`,
+`es_ES`, `ru_RU`, `th_TH`, `vi_VN`. HTTP equivalents are GET/PUT
+`/api/v1/workspaces/current/language` (PUT body: `{"language":"ja_JP"}`).
+The API key's Workspace remains authoritative. The setting controls new backend
+messages and default bot notices, independently of personal UI language.
+Pipeline `output.misc` hint fields override defaults; empty values inherit this
+setting. Existing custom text and historical logs are not rewritten.
+
+### Workspace default LLM model
+
+`get_starred_model` reads the Workspace-wide favorite. `set_starred_model`
+requires `provider_secret.manage` and replaces the single starred model; pass
+`null` to clear it. The model must belong to the current Workspace.
+`get_default_model` resolves the favorite first, then the Space wizard chat
+recommendation only when LangBot Models is enabled and the Workspace owner is
+bound to a LangBot Account. It returns a nullable `uuid`.
+New processor LLM selector defaults use this preference; existing configurations
+are not rewritten. Embedding and rerank models are not eligible.
+
+### Reset a bot session context
+
+Use `reset_session_context(bot_id, session_id)` only when a user asks to start a session afresh. It requires `resource.manage`, preserves monitoring records, and refuses sessions with active tasks. It clears conversation runner state and excludes earlier transcript entries from subsequent model context. Files and long-term memory are not removed.
+
+## Cloud plan platform restrictions
+
+Bot creation and enabling use the same Workspace entitlement policy as HTTP. A disabled platform is rejected even through MCP or plugin calls; do not retry using an Omni or legacy alias to bypass the restriction. The Workspace-scoped HTTP adapter catalog omits disabled platforms.
+
+## LLM request diagnostics
+
+`get_monitoring_records(kind="llm_calls")` includes nullable `details` on each call.
+Read `details.timing`, `details.usage`, and `details.error` to distinguish provider
+latency, retry waits, cached/reasoning tokens, and failures. Timing values are
+milliseconds since invocation start; `first_event_ms` includes reasoning/control
+frames, while `first_content_ms` measures the first upstream content delta.
+`first_visible_chunk_ms` may contain displayed reasoning, not just answer text.
+Old records have no details. Do not equate total duration with provider TTFT.
+See `docs/LLM_REQUEST_POLICY.md` for policy precedence and gateway header behavior.
+
+`get_llm_model` also exposes optional `reasoning_capabilities.budget_tokens`
+min/max bounds. Numeric thinking budgets use `reasoning_config.budget_tokens`
+with level `enabled`; do not combine budgets with effort levels or native thinking
+arguments. Existing Runner reasoning overrides take precedence. Prompt caching
+is opt-in through model/call `extra_args.prompt_caching`; use provider-specific
+modes documented in `docs/LLM_REQUEST_POLICY.md`. Cache keys are routing hints,
+not evidence of hits; read the actual usage counters. `preparation_ms` measures
+request construction and `preparation_cache_hits` counts retry preparation reuse.
+`transport_setup_ms` measures managed pool acquisition, including first-use TLS
+initialization; it is not network TTFT. See `docs/LLM_LATENCY_BENCHMARK.md` for
+controlled request-layer measurements and their scope.
diff --git a/skills/skills/langbot-testing/cases/sandbox-skill-authoring-e2e.yaml b/skills/skills/langbot-testing/cases/sandbox-skill-authoring-e2e.yaml
index 608266c4e..cc3355b2e 100644
--- a/skills/skills/langbot-testing/cases/sandbox-skill-authoring-e2e.yaml
+++ b/skills/skills/langbot-testing/cases/sandbox-skill-authoring-e2e.yaml
@@ -1,5 +1,5 @@
 id: sandbox-skill-authoring-e2e
-title: "Local Agent creates, registers, activates, and uses a sandbox skill"
+title: "Local Agent publishes and executes an immutable Skill revision"
 mode: agent-browser
 area: sandbox
 type: regression
@@ -11,45 +11,32 @@ tags:
   - skills
   - local-agent
   - tools
-  - e2b
   - nsjail
-  - host
 skills:
   - langbot-env-setup
   - langbot-testing
 env:
   - LANGBOT_FRONTEND_URL
   - LANGBOT_BACKEND_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
+env_any:
+  - LANGBOT_LOCAL_AGENT_PIPELINE_URL|LANGBOT_LOCAL_AGENT_PIPELINE_NAME
 preconditions:
-  - "LANGBOT_LOCAL_AGENT_PIPELINE_URL or LANGBOT_LOCAL_AGENT_PIPELINE_NAME points to the local-agent pipeline under test."
-  - "LangBot is started with the Box backend intended for this run, such as e2b, nsjail, or explicit host development mode."
-  - "The selected model route supports tool/function calling strongly enough to invoke sandbox tools."
+  - "The selected Local Agent uses a function-calling model and an available Box backend."
 steps:
-  - "Start LangBot with the target sandbox backend and confirm the Box status UI or LANGBOT_BACKEND_URL /api/v1/box/status reports the expected backend."
-  - "Open the target Local Agent pipeline in Debug Chat with a function-calling model."
-  - "Run the canonical prompt pattern in references/sandbox-skill-authoring.md with a unique skill name."
-  - "Capture the visible final response, backend logs, and the registered skill-store files."
+  - "Open the target Local Agent in Debug Chat."
+  - "Follow Publish and execute in references/sandbox-skill-authoring.md with a unique Skill name."
+  - "Record the activated revision and the script's independently expected output."
+  - "Attempt a write under the published package, then read back the original contents."
 checks:
-  - "UI: Debug Chat final assistant response contains E2E_OK:<skill-name>."
-  - "Logs: The model called exec, register_skill, activate, then exec again from the activated skill path."
-  - "Logs: The selected backend name is the expected one, such as e2b, nsjail, or host."
-  - "Skill store: The registered package and activated writeback match references/sandbox-skill-authoring.md."
-  - "Box status: recent_error_count is 0 after the run."
+  - "UI and logs show successful draft creation, registration, activation and script execution."
+  - "Script output matches the deterministic expected value."
+  - "Published-package mutation is rejected; outputs remain in the run outbox or writable draft."
+  - "The expected rejected write is distinguishable from unexpected execution failures."
 evidence_required:
   - ui
   - backend_log
   - api_diagnostic
   - filesystem
 diagnostics:
-  - "If Debug Chat fails before tool execution, first validate the model provider and function-calling support."
-  - "If /workspace/.skills/<skill> is missing only on E2B, check extra mount sync behavior."
-  - "If nsjail starts but every command fails before execve, check nsjail CLI compatibility and chroot mount targets."
-  - "API or raw provider probes are diagnostic only; they do not make this UI case pass by themselves."
-troubleshooting:
-  - sandbox-native-tools-unavailable
-  - e2b-extra-mount-sync-missing
-  - box-session-conflict-logical-metadata
-  - nsjail-cli-compatibility
-  - socks-proxy-without-socksio
+  - "Check Box availability and backend selection before interpreting a missing tool."
+  - "The runner may acquire Box before activation; a new package must not recreate the session."
diff --git a/skills/skills/langbot-testing/cases/sandbox-skill-authoring-edit-existing-e2e.yaml b/skills/skills/langbot-testing/cases/sandbox-skill-authoring-edit-existing-e2e.yaml
index 34590804f..dc2cdb0ce 100644
--- a/skills/skills/langbot-testing/cases/sandbox-skill-authoring-edit-existing-e2e.yaml
+++ b/skills/skills/langbot-testing/cases/sandbox-skill-authoring-edit-existing-e2e.yaml
@@ -1,5 +1,5 @@
 id: sandbox-skill-authoring-edit-existing-e2e
-title: "Local Agent modifies an activated sandbox skill package"
+title: "Local Agent publishes a new Skill revision without changing an active run"
 mode: agent-browser
 area: sandbox
 type: regression
@@ -12,66 +12,35 @@ tags:
   - local-agent
   - tools
   - edit
+  - nsjail
 skills:
   - langbot-env-setup
   - langbot-testing
 env:
-  - LANGBOT_REPO
   - LANGBOT_FRONTEND_URL
   - LANGBOT_BACKEND_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
-automation: scripts/e2e/pipeline-debug-chat.mjs
-automation_env:
-  - LANGBOT_REPO
-  - LANGBOT_FRONTEND_URL
-  - LANGBOT_BACKEND_URL
-  - LANGBOT_BROWSER_PROFILE
-  - LANGBOT_CHROMIUM_EXECUTABLE
-  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
-automation_pipeline_url_env: LANGBOT_LOCAL_AGENT_PIPELINE_URL
-automation_pipeline_name_env: LANGBOT_LOCAL_AGENT_PIPELINE_NAME
-automation_expected_text: "E2E_OK:lb-skill-edit-regression:SANDBOX_SKILL_MODIFIED_OK"
-automation_prompts_json: '[{"prompt":"Skill name is lb-skill-edit-regression. Use exactly one exec tool call with workdir=/workspace. In that single command, remove old /workspace/lb-skill-edit-regression, create SKILL.md, scripts/use.py, and data/input.json. The JSON must contain numbers [1,2,3,4], factors [2,3,4], and marker CREATE_MARKER_EDIT_REGRESSION. The Python script must read data/input.json relative to __file__ and print exactly SANDBOX_SKILL_CREATE_OK sum=10 product=24 marker=CREATE_MARKER_EDIT_REGRESSION. Run python3 /workspace/lb-skill-edit-regression/scripts/use.py in the same command. After the exec succeeds, final answer exactly CREATE_OK:lb-skill-edit-regression.","expected_text":"CREATE_OK:lb-skill-edit-regression","response_timeout_ms":"180000"},{"prompt":"Continue with the same skill name lb-skill-edit-regression. Strictly call register_skill with path=/workspace/lb-skill-edit-regression and name=lb-skill-edit-regression, then call activate with skill_name=lb-skill-edit-regression. Do not call exec, read, write, edit, grep, or ls in this step. After both tools succeed, final answer exactly REGISTER_ACTIVATE_OK:lb-skill-edit-regression.","expected_text":"REGISTER_ACTIVATE_OK:lb-skill-edit-regression","response_timeout_ms":"180000"},{"prompt":"Continue with the same activated skill lb-skill-edit-regression. Use exactly one exec tool call with workdir=/workspace/.skills/lb-skill-edit-regression. In that single exec command, run exactly this shell logic: set -e; overwrite SKILL.md with a heading and UPDATED_MARKER_EDIT_REGRESSION; overwrite data/input.json with numbers [5,6], factors [7,8], and marker UPDATED_MARKER_EDIT_REGRESSION; overwrite scripts/use.py so it reads data/input.json relative to __file__ and prints exactly SANDBOX_SKILL_MODIFIED_OK sum=11 product=56 marker=UPDATED_MARKER_EDIT_REGRESSION; run python3 scripts/use.py. Do not write exit(1). Do not call find, ls, read, grep as separate tools. If the single exec exits successfully and prints SANDBOX_SKILL_MODIFIED_OK, do not call more tools. Final answer exactly E2E_OK:lb-skill-edit-regression:SANDBOX_SKILL_MODIFIED_OK.","expected_text":"E2E_OK:lb-skill-edit-regression:SANDBOX_SKILL_MODIFIED_OK","response_timeout_ms":"240000"}]'
-setup_automation:
-  - "node:scripts/e2e/ensure-local-agent-pipeline.mjs --write-env"
-  - "node:scripts/e2e/reset-sandbox-skill.mjs --name=lb-skill-edit-regression"
-setup_provides_env:
-  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
-  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
-automation_filesystem_checks_json: '[{"path":"${LANGBOT_REPO}/data/box/skills/lb-skill-edit-regression/SKILL.md","contains":"UPDATED_MARKER_EDIT_REGRESSION"},{"path":"${LANGBOT_REPO}/data/box/skills/lb-skill-edit-regression/data/input.json","json_equals":{"numbers":[5,6],"factors":[7,8],"marker":"UPDATED_MARKER_EDIT_REGRESSION"}},{"path":"${LANGBOT_REPO}/data/box/skills/lb-skill-edit-regression/scripts/use.py","contains":["SANDBOX_SKILL_MODIFIED_OK"],"not_contains":["SANDBOX_SKILL_CREATE_OK","exit(1)"]},{"argv":["python3","scripts/use.py"],"cwd":"${LANGBOT_REPO}/data/box/skills/lb-skill-edit-regression","stdout_contains":"SANDBOX_SKILL_MODIFIED_OK sum=11 product=56 marker=UPDATED_MARKER_EDIT_REGRESSION","exit_code":0}]'
-automation_stream_output: "0"
-automation_reset_debug_chat: "1"
-automation_response_timeout_ms: "420000"
+env_any:
+  - LANGBOT_LOCAL_AGENT_PIPELINE_URL|LANGBOT_LOCAL_AGENT_PIPELINE_NAME
 preconditions:
-  - "LANGBOT_LOCAL_AGENT_PIPELINE_URL or LANGBOT_LOCAL_AGENT_PIPELINE_NAME points to the local-agent pipeline under test."
-  - "LangBot is started with the sandbox backend intended for this run, such as e2b or nsjail."
-  - "The selected model route supports tool/function calling strongly enough to invoke sandbox tools."
+  - "The selected Local Agent uses a function-calling model and an available Box backend."
+  - "A unique test Skill v1 has a deterministic script and reference output."
 steps:
-  - "Start LangBot with the target sandbox backend and confirm native sandbox tools are available."
-  - "Open the target Local Agent pipeline in Debug Chat with a function-calling model."
-  - "Run the automation prompt or the equivalent prompt pattern from references/sandbox-skill-authoring.md."
-  - "Capture the visible final response, backend logs, and the registered skill-store files."
+  - "Follow Update an existing Skill in references/sandbox-skill-authoring.md."
+  - "Keep one run active on v1; edit a writable draft and publish v2 using base_revision=v1."
+  - "Read and execute again in the existing run, then activate and execute in a new run in the same conversation."
+  - "Attempt a conflicting update using the stale base_revision and verify rejection."
+  - "Follow Shared Box and recovery to check retained temporary state and a managed process."
 checks:
-  - "UI: A Bot message, not only the user prompt, contains E2E_OK:lb-skill-edit-regression:SANDBOX_SKILL_MODIFIED_OK."
-  - "Logs: The model called exec, register_skill, activate, then a second exec whose workdir is /workspace/.skills/lb-skill-edit-regression."
-  - "Logs: The second exec stdout contains SANDBOX_SKILL_MODIFIED_OK."
-  - "Automation: Filesystem checks pass for the registered skill-store files under LANGBOT_REPO/data/box/skills/lb-skill-edit-regression."
-  - "Skill store: SKILL.md and data/input.json contain UPDATED_MARKER_EDIT_REGRESSION."
-  - "Skill store: scripts/use.py is the modified script and prints SANDBOX_SKILL_MODIFIED_OK sum=11 product=56 marker=UPDATED_MARKER_EDIT_REGRESSION."
+  - "The existing run continues to read and execute v1; the new run reads and executes v2."
+  - "Published revisions remain immutable, and stale publication is rejected."
+  - "Alternating activated revisions do not recreate the shared Box or stop the managed process."
+  - "An unbound Skill is not restored into a subsequent run."
 evidence_required:
   - ui
   - backend_log
   - api_diagnostic
   - filesystem
 diagnostics:
-  - "If the model stops after activation and only reruns the original script, treat this as a failed edit-existing E2E even if create/register/activate succeeded."
-  - "If the expected marker appears only in the user prompt, treat the run as failed and inspect automation-result.json assistant counts."
-  - "If the browser opens /login or console shows 401, refresh the authenticated browser profile before diagnosing the runner."
-troubleshooting:
-  - sandbox-native-tools-unavailable
-  - e2b-extra-mount-sync-missing
-  - box-session-conflict-logical-metadata
-  - nsjail-cli-compatibility
-  - socks-proxy-without-socksio
+  - "An output marker in the user prompt alone is not execution evidence."
+  - "Do not edit files in the published path; publish a writable draft explicitly."
+  - "Expired run tool sessions must fail explicitly, never silently select the latest Skill."
diff --git a/skills/skills/langbot-testing/cases/wizard-onebot-agent-runtime.yaml b/skills/skills/langbot-testing/cases/wizard-onebot-agent-runtime.yaml
index 220c5b3f5..de4d56cb0 100644
--- a/skills/skills/langbot-testing/cases/wizard-onebot-agent-runtime.yaml
+++ b/skills/skills/langbot-testing/cases/wizard-onebot-agent-runtime.yaml
@@ -32,7 +32,7 @@ preconditions:
   - "The LangBot virtualenv includes the websockets package used by the local OneBot probe."
 steps:
   - "Open Quick Start and choose Welcome new members with the OneBot v11 channel."
-  - "Create and enable a temporary OneBot Bot on an available local port."
+  - "Create and enable a temporary OneBot Bot using the main-port reverse WS URL and a generated access token."
   - "Select Local Agent and confirm the editable welcome prompt is prefilled."
   - "Choose QA Deterministic Runner and finish the wizard."
   - "Connect a local OneBot reverse WebSocket and send a group_increase notice."
diff --git a/skills/skills/langbot-testing/references/sandbox-skill-authoring.md b/skills/skills/langbot-testing/references/sandbox-skill-authoring.md
index 1d4b9a02f..8f2ad871a 100644
--- a/skills/skills/langbot-testing/references/sandbox-skill-authoring.md
+++ b/skills/skills/langbot-testing/references/sandbox-skill-authoring.md
@@ -1,139 +1,60 @@
-# Sandbox Skill Authoring
+# Skill publication and execution acceptance
 
-## Goal
-
-Verify that Local Agent can use sandbox tools to create, register, activate, and use a LangBot skill package through the same path a user would exercise in Debug Chat.
-
-This flow applies to Docker, nsjail, E2B, and the explicit host development backend. Host runs commands directly as the Box Runtime user and must never be treated as sandbox-isolation coverage. API calls are useful diagnostics, but the primary pass/fail signal is the model-driven Debug Chat tool sequence.
+Verify through Debug Chat that instructions and text resources work without Box,
+and that executable packages use the revision activated by the current run.
+API and direct backend checks diagnose failures; they do not establish UI acceptance.
 
 ## Preconditions
 
-1. Read `../.env` and use `LANGBOT_FRONTEND_URL` and `LANGBOT_BACKEND_URL`.
-2. Start LangBot with the intended backend:
-   - `BOX_BACKEND=e2b` when validating E2B.
-   - `BOX_BACKEND=nsjail` when validating nsjail.
-   - `BOX_BACKEND=local` or `docker` when validating local container fallback.
-   - `BOX_BACKEND=host` only when validating explicit, trusted local direct execution.
-3. Confirm `/api/v1/box/status` reports `available: true` and the expected backend name.
-4. Confirm Debug Chat uses a model with function-calling ability.
-5. Confirm backend logs say native sandbox tools are available.
+Use the frontend/backend and Local Agent pipeline selected in the machine's
+`skills/.env.local`. The model must support tool calling. Do not record credentials.
+Use Docker/Podman, nsjail, or a configured E2B service. The runner may acquire an
+empty Box before activation; adding a revision must preserve that Box. Host mode
+provides no sandbox isolation evidence.
 
-Do not store sandbox provider keys, JWTs, OAuth tokens, or localStorage values in the case or notes.
+## Read-only path
 
-## Debug Chat Prompt Pattern
+Install a unique Skill with instructions and a UTF-8 reference, then disable Box.
+Ask the Agent to activate it and read the reference. Check the reported revision,
+reference contents, and absence of sandbox creation. Execution requests must fail
+clearly; activation must not claim that files are already mounted.
 
-Use a unique skill name per run, for example:
+## Publish and execute
 
-```text
-lb-sandbox-agent-e2e-<timestamp>
-```
+1. With Box enabled, create a draft under `/workspace/skill-drafts/<name>`.
+   Include `SKILL.md`, `scripts/use.py`, and `data/input.json`. Resolve resource
+   files relative to `__file__`; the script should print a deterministic marker.
+2. Call `register_skill` with the draft path, then `activate`. Record revision v1.
+3. Execute the script under the `mount_path` returned by activation. Write results to the
+   run's outbox or draft directory, never the published package.
+4. Attempt to overwrite a published file. The operation must fail and its contents
+   must remain unchanged. Check UI results, tool logs, and the published package.
 
-Send a prompt that requires the model to do all of the following:
+## Update an existing Skill
 
-1. Use `exec` to create a multi-file skill under `/workspace/<skill-name>`.
-2. Include at least:
-   - `SKILL.md`
-   - `scripts/use.py`
-   - `data/input.json`
-3. In the same first `exec`, run the script and verify a deterministic marker such as:
+Copy v1 to a separate writable draft. Modify the draft and publish with
+`base_revision=v1`; record v2. The existing run must continue reading/executing v1,
+including when `activate` is called again. A new run in the same conversation must
+activate v2. A stale `base_revision` update must fail. Removing the Skill from the
+new run's authorization must not break startup or restore the old activation.
+Verify a second UI edit/save adopts the revision returned by the first save.
 
-   ```text
-   SANDBOX_COMPLEX_SKILL_OK sum=10 product=24
-   ```
+## Shared Box and recovery
 
-4. Call `register_skill` with `path=/workspace/<skill-name>`.
-5. Call `activate` with `skill_name=<skill-name>`.
-6. Call `exec` with `workdir=/workspace/.skills/<skill-name>` and run:
+With nsjail, alternate two authorized runs using v1 and v2 in the same Box. Check
+each script's independent version marker, retained `/tmp` state, and a running
+managed process. Neither activation nor a changed mount set may recreate the Box.
+Docker/Podman and E2B must likewise add v2 at its digest path while v1 and the
+existing Box remain usable. On fixed-mount backends, changing ordinary host mounts must fail without
+removing the session. Programs sharing a Box can see its cached packages; test
+version consistency without claiming that the shared filesystem isolates runs.
 
-   ```bash
-   python3 scripts/use.py && echo SANDBOX_ACTIVATED_WRITEBACK_OK > activated_writeback.txt && cat activated_writeback.txt
-   ```
+Pins are Host-owned run ledger data, not conversation state. An active run whose
+in-memory pin is reloaded must use its recorded revision, even after publication
+or deletion of the current catalog entry. A missing recorded package must fail.
+A Core process restart expires tool sessions; calls from those sessions must be
+rejected. Do not report this as automatic run recovery.
 
-7. Require the final answer to contain only an explicit success marker:
-
-   ```text
-   E2E_OK:<skill-name>
-   ```
-
-Keep the test script robust to working-directory changes. Prefer resolving data paths from `__file__`:
-
-```python
-from pathlib import Path
-data_path = Path(__file__).resolve().parent.parent / "data" / "input.json"
-```
-
-## Success Criteria
-
-The UI should show an assistant final response containing `E2E_OK:<skill-name>`.
-
-Backend logs should show:
-
-- `exec tool invoked`
-- `register_skill`
-- `activate`
-- a second `exec` whose workdir is `/workspace/.skills/<skill-name>`
-- `backend=e2b`, `backend=nsjail`, `backend=host`, or the expected local backend
-
-After the run, verify the skill store through the UI or API:
-
-- Skill root lists `SKILL.md`, `scripts`, `data`, and `activated_writeback.txt`.
-- `scripts/use.py` is readable.
-- `data/input.json` is readable.
-- `activated_writeback.txt` contains `SANDBOX_ACTIVATED_WRITEBACK_OK`.
-- `/api/v1/box/errors` is empty.
-
-## Existing Skill Edit Variant
-
-The base `sandbox-skill-authoring-e2e` case only proves create, register,
-activate, and use. To prove that an already activated skill can be modified,
-run `sandbox-skill-authoring-edit-existing-e2e` or use its prompt pattern.
-
-The edit variant must include these additional checks:
-
-- The second `exec` uses `workdir=/workspace/.skills/<skill-name>`.
-- The second `exec` overwrites `SKILL.md`, `data/input.json`, and
-  `scripts/use.py` under the activated skill path.
-- The modified script prints a deterministic marker such as:
-
-  ```text
-  SANDBOX_SKILL_MODIFIED_OK sum=11 product=56 marker=<updated-marker>
-  ```
-
-- `grep -q <updated-marker> SKILL.md scripts/use.py data/input.json` succeeds
-  in the same activated-path command.
-- Filesystem evidence under the Box-managed skill store shows the updated
-  marker, not only the original create marker.
-
-If the model stops after activation and only reruns the original script, treat
-that run as a failed edit-existing E2E even when create/register/activate
-succeeded.
-
-## Diagnostic Checks
-
-Use these only after the model-driven Debug Chat flow fails:
-
-- `/api/v1/box/status` to confirm backend selection and recent errors.
-- `/api/v1/box/sessions` to check leaked or conflicting sessions.
-- Direct backend probes to separate provider credentials from LangBot integration.
-- Filesystem inspection under the configured Box-managed skill store.
-
-For E2B raw HTTP diagnostics, include a valid template id such as `base`; a missing template can produce schema validation errors that are unrelated to authentication.
-
-## Known Pitfalls
-
-- When Box is available, skills may be owned by the Box runtime and stored in Box-managed skill storage. Do not assume `data/skills` is the active source of truth.
-- Public E2B does not provide local bind mounts. Main workspace and activated skill extra mounts must be synchronized into and back out of the E2B sandbox.
-- Session metadata should keep LangBot logical paths such as `/workspace`; storing provider-internal paths can make later requests look incompatible.
-- nsjail versions differ. Some expose only `--disable_clone_new*` flags and use `--bindmount` instead of `--rw_bind`.
-- On WSL, cgroup v2 may exist but not be writable. The backend should warn and fall back to rlimits rather than fail the sandbox.
-- The host backend does not honor sandbox image, network, rootfs, process, or
-  resource isolation. Use a disposable workspace and low-privilege account.
-- If `ALL_PROXY` uses a SOCKS URL and `socksio` is not installed, some Python HTTP clients can fail during startup. Prefer consistent HTTP proxy variables unless SOCKS support is installed.
-
-## Related Troubleshooting
-
-- `sandbox-native-tools-unavailable`
-- `e2b-extra-mount-sync-missing`
-- `box-session-conflict-logical-metadata`
-- `nsjail-cli-compatibility`
-- `socks-proxy-without-socksio`
+Capture the visible assistant results, relevant tool logs, revisions, and file
+contents. Public E2B credentials and deployment topology are separate prerequisites;
+mocked or local backend results do not establish E2B service acceptance.
diff --git a/src/langbot/libs/dingtalk_api/api.py b/src/langbot/libs/dingtalk_api/api.py
index d6a4914ce..48032b72a 100644
--- a/src/langbot/libs/dingtalk_api/api.py
+++ b/src/langbot/libs/dingtalk_api/api.py
@@ -14,6 +14,7 @@ import websockets
 from .EchoHandler import EchoTextHandler
 from .card_callback import DingTalkCardActionHandler
 from .dingtalkevent import DingTalkEvent
+from .thinking import card_text_fields, visible_think_markers
 import httpx
 import traceback
 from langbot.pkg.utils import httpclient
@@ -74,6 +75,16 @@ def _stringify_card_param_map(card_param_map: Optional[dict]) -> dict:
     return result
 
 
+def is_stream_card_configured(config: dict) -> bool:
+    if not config.get('markdown_card', True):
+        return False
+    template_id = config.get('card_template_id')
+    # Older manifests stored instructional placeholder text as the default.
+    return (
+        isinstance(template_id, str) and bool(template_id.strip()) and template_id.strip() != '填写你的卡片template_id'
+    )
+
+
 class DingTalkClient:
     _MAX_INBOUND_TASKS = 100
 
@@ -108,7 +119,7 @@ class DingTalkClient:
         }
         self.access_token = ''
         self.robot_name = robot_name
-        self.robot_code = robot_code
+        self.robot_code = (robot_code or '').strip() or client_id
         self.access_token_expiry_time = ''
         self.markdown_card = markdown_card
         self.logger = logger
@@ -258,28 +269,105 @@ class DingTalkClient:
 
     async def update_incoming_message(self, message):
         """异步更新 DingTalkClient 中的 incoming_message"""
+        # Resolve identity before downloading any attachments from this event.
+        robot_code = getattr(message, 'robot_code', None)
+        if isinstance(robot_code, str) and robot_code.strip():
+            self.robot_code = robot_code.strip()
         message_data = await self.get_message(message)
         if message_data:
             event = DingTalkEvent.from_payload(message_data)
             if event:
                 await self._handle_message(event)
 
+    @staticmethod
+    def _check_send_result(response):
+        # HTTP 200 can still contain a rejected DingTalk message.
+        data = response.json()
+        if response.status_code != 200 or data.get('errcode', 0) != 0 or data.get('code'):
+            raise RuntimeError(
+                f'DingTalk delivery failed (HTTP {response.status_code}, '
+                f'code={data.get("errcode", data.get("code", "unknown"))}): '
+                f'{data.get("errmsg", data.get("message", "Unknown error"))}'
+            )
+        return data
+
     async def send_message(self, content: str, incoming_message, at: bool):
+        if not content.strip():
+            return {}
+        content = visible_think_markers(content, self.markdown_card)
         if self.markdown_card:
-            if at:
-                self.EchoTextHandler.reply_markdown(
-                    title='@' + incoming_message.sender_nick + ' ' + content,
-                    text='@' + incoming_message.sender_nick + ' ' + content,
-                    incoming_message=incoming_message,
-                )
-            else:
-                self.EchoTextHandler.reply_markdown(
-                    title=content,
-                    text=content,
-                    incoming_message=incoming_message,
-                )
+            text = ('@' + incoming_message.sender_nick + ' ' if at else '') + content
+            payload = {'msgtype': 'markdown', 'markdown': {'title': text[:80], 'text': text}}
         else:
-            self.EchoTextHandler.reply_text(content, incoming_message)
+            payload = {'msgtype': 'text', 'text': {'content': content}}
+        payload['at'] = {'atUserIds': [incoming_message.sender_staff_id]}
+        async with self._http_client_context() as client:
+            response = await client.post(incoming_message.session_webhook, json=payload, timeout=30.0)
+        return self._check_send_result(response)
+
+    async def send_attachment(self, component, incoming_message):
+        from langbot_plugin.api.entities.builtin.platform.message import Image
+
+        is_image = isinstance(component, Image)
+        name = getattr(component, 'name', '') or ('image.png' if is_image else 'file')
+        name = os.path.basename(name)
+        encoded = getattr(component, 'base64', None)
+        if encoded:
+            encoded = encoded.split(',', 1)[-1]
+            if len(encoded) > 4 * ((_MAX_MEDIA_BYTES + 2) // 3):
+                raise ValueError('DingTalk media exceeds the size limit')
+            body = base64.b64decode(encoded, validate=True)
+        elif getattr(component, 'path', None):
+            body = await asyncio.to_thread(_read_local_media_limited, component.path)
+        elif getattr(component, 'url', None):
+            async with self._http_client_context() as client:
+                async with client.stream('GET', component.url, timeout=30.0) as response:
+                    response.raise_for_status()
+                    body = await _read_httpx_media_limited(response)
+        else:
+            raise ValueError('DingTalk attachment has no readable content')
+        if len(body) > _MAX_MEDIA_BYTES:
+            raise ValueError('DingTalk media exceeds the size limit')
+        token = await self.get_legacy_access_token()
+        if not token:
+            raise RuntimeError('DingTalk media authentication failed')
+        async with self._http_client_context() as client:
+            response = await client.post(
+                'https://oapi.dingtalk.com/media/upload',
+                params={'access_token': token, 'type': 'image' if is_image else 'file'},
+                files={'media': (name, body)},
+                timeout=30.0,
+            )
+        uploaded = self._check_send_result(response)
+        media_id = uploaded.get('media_id')
+        if not media_id:
+            raise RuntimeError('DingTalk media upload returned no media ID')
+        if not await self.check_access_token():
+            await self.get_access_token()
+        params = (
+            {'photoURL': media_id}
+            if is_image
+            else {'mediaId': media_id, 'fileName': name, 'fileType': os.path.splitext(name)[1].lstrip('.')}
+        )
+        payload = {
+            'robotCode': self.robot_code or self.key,
+            'msgKey': 'sampleImageMsg' if is_image else 'sampleFile',
+            'msgParam': json.dumps(params, ensure_ascii=False),
+        }
+        if str(incoming_message.conversation_type) == '2':
+            endpoint = 'groupMessages/send'
+            payload['openConversationId'] = incoming_message.conversation_id
+        else:
+            endpoint = 'oToMessages/batchSend'
+            payload['userIds'] = [incoming_message.sender_staff_id]
+        async with self._http_client_context() as client:
+            response = await client.post(
+                f'{DINGTALK_OPENAPI_BASE}/v1.0/robot/{endpoint}',
+                headers={'x-acs-dingtalk-access-token': self.access_token},
+                json=payload,
+                timeout=30.0,
+            )
+        return self._check_send_result(response)
 
     async def get_incoming_message(self):
         """获取收到的消息"""
@@ -562,7 +650,7 @@ class DingTalkClient:
             'robotCode': robot_code,
             'userIds': [target_id],
             'msgKey': 'sampleText',
-            'msgParam': json.dumps({'content': content}),
+            'msgParam': json.dumps({'content': visible_think_markers(content, False)}),
         }
         _stdout_logger.info(
             'DingTalk send_proactive_message_to_one request: robotCode=%s target_id=%s content_len=%d',
@@ -606,7 +694,7 @@ class DingTalkClient:
             'robotCode': self.robot_code or self.key,
             'openConversationId': target_id,
             'msgKey': 'sampleText',
-            'msgParam': json.dumps({'content': content}),
+            'msgParam': json.dumps({'content': visible_think_markers(content, False)}),
         }
         try:
             async with self._http_client_context() as client:
@@ -627,14 +715,13 @@ class DingTalkClient:
         temp_card_id: str,
         incoming_message: dingtalk_stream.ChatbotMessage,
         quote_origin: bool = False,
-        card_auto_layout: bool = False,
+        card_auto_layout: bool = True,
     ):
         """Create + deliver the streaming chat card for a chatbot reply.
 
         Replaces the old `dingtalk_stream.AICardReplier`-based path. Returns
-        `(None, out_track_id)` to keep call sites compatible with the
-        previous `(card_instance, card_instance_id)` shape — the first slot
-        is unused now that everything is driven by out_track_id.
+        `(layout_config, out_track_id)` keeps the adapter's tuple contract
+        while preserving the layout setting on subsequent data updates.
         """
         out_track_id = uuid.uuid4().hex
         is_group = str(incoming_message.conversation_type) == '2'
@@ -643,7 +730,14 @@ class DingTalkClient:
         else:
             open_space_id = f'dtv1.card//IM_ROBOT.{incoming_message.sender_staff_id}'
 
-        card_param_map = {'content': ''}
+        card_param_map = {
+            'content': '',
+            'answer': '',
+            'reasoning': '',
+            'flowStatus': '1',
+            'brand': '',
+            'hasReasoning': '',
+        }
         if incoming_message.message_type == 'text':
             card_param_map['query'] = incoming_message.get_text_list()[0]
         else:
@@ -657,30 +751,21 @@ class DingTalkClient:
             card_param_map=card_param_map,
             card_data_config={'autoLayout': card_auto_layout},
         )
-        return None, out_track_id
+        return {'autoLayout': card_auto_layout}, out_track_id
 
     async def send_card_message(self, card_instance, card_instance_id: str, content: str, is_final: bool):
-        """Stream a single chunk into an existing card's `content` field."""
-        try:
-            await self.streaming_update_card(
-                out_track_id=card_instance_id,
-                content_key='content',
-                content_value=content,
-                append=False,
-                finished=is_final,
-                failed=False,
-            )
-        except Exception as e:
-            if self.logger:
-                self.logger.exception(e)
-            await self.streaming_update_card(
-                out_track_id=card_instance_id,
-                content_key='content',
-                content_value='',
-                append=False,
-                finished=is_final,
-                failed=True,
-            )
+        """Update separate reasoning/answer fields and the legacy content field."""
+        fields = card_text_fields(content)
+        # Old templates continue to render content with explicit think markers;
+        # the supplied split template binds only reasoning and answer.
+        fields['content'] = visible_think_markers(content)
+        fields['flowStatus'] = '3' if is_final else '2'
+        fields['brand'] = ''
+        if isinstance(card_instance, dict):
+            fields['config'] = card_instance
+            fields['autoLayout'] = str(card_instance.get('autoLayout', True)).lower()
+        if not await self.update_card_data(out_track_id=card_instance_id, card_param_map=fields):
+            raise RuntimeError('DingTalk card content update failed; see adapter logs for the API response')
 
     async def create_and_deliver_card(
         self,
@@ -710,6 +795,7 @@ class DingTalkClient:
         template_params = dict(card_param_map or {})
         if card_data_config is not None:
             template_params['config'] = card_data_config
+            template_params['autoLayout'] = str(card_data_config.get('autoLayout', True)).lower()
         cardData: dict = {'cardParamMap': _stringify_card_param_map(template_params)}
 
         body: dict = {
@@ -743,34 +829,17 @@ class DingTalkClient:
             'Content-Type': 'application/json',
         }
         try:
-            _stdout_logger.info(
-                'DingTalk createAndDeliver request body: %s',
-                json.dumps(body, ensure_ascii=False)[:1500],
-            )
             async with self._http_client_context() as client:
                 response = await client.post(url, headers=headers, json=body, timeout=30.0)
-                response_body = await httpclient.response_text(response, max_chars=500)
-                if response.status_code == 200:
-                    _stdout_logger.info(
-                        'DingTalk createAndDeliver response: %s',
-                        response_body,
-                    )
-                    return True
-                _stdout_logger.error(
-                    'DingTalk createAndDeliver failed: status=%s body=%s',
-                    response.status_code,
-                    response_body,
-                )
-                if self.logger:
-                    await self.logger.error(
-                        f'DingTalk createAndDeliver failed: status={response.status_code} body={response_body}'
-                    )
-                return False
-        except Exception:
-            _stdout_logger.exception('DingTalk createAndDeliver error')
-            if self.logger:
-                await self.logger.error(f'DingTalk createAndDeliver error: {traceback.format_exc()}')
-            return False
+            # Propagate business failures as well as transport failures. Callers
+            # must not register or update a card that was never delivered.
+            self._check_send_result(response)
+        except Exception as exc:
+            error = RuntimeError(f'DingTalk card creation failed: {exc}')
+            _stdout_logger.error('%s', error)
+            raise error from exc
+        _stdout_logger.info('DingTalk card created: outTrackId=%s', out_track_id)
+        return True
 
     async def streaming_update_card(
         self,
diff --git a/src/langbot/libs/dingtalk_api/thinking.py b/src/langbot/libs/dingtalk_api/thinking.py
new file mode 100644
index 000000000..d4a8b54ee
--- /dev/null
+++ b/src/langbot/libs/dingtalk_api/thinking.py
@@ -0,0 +1,39 @@
+"""Render reasoning markers without changing the pipeline output policy."""
+
+import re
+
+_MARKER = re.compile(r'</?think>', re.IGNORECASE)
+
+
+def visible_think_markers(content: str, markdown: bool = True) -> str:
+    def replace(match):
+        tag = match.group(1).lower()
+        separator = '\n\n' if markdown else '\n'
+        return separator + (f'`{tag}`' if markdown else tag) + separator
+
+    # Normalize only marker boundaries; preserve paragraph spacing in the body.
+    return re.sub(r'[ \t\r\n]*(</?think>)[ \t\r\n]*', replace, content, flags=re.IGNORECASE).strip()
+
+
+def card_text_fields(content: str) -> dict[str, str]:
+    thinking, answer = [], []
+    inside = False
+    offset = 0
+    for match in _MARKER.finditer(content):
+        (thinking if inside else answer).append(content[offset : match.start()])
+        inside = match.group(0).lower() == '<think>'
+        offset = match.end()
+    tail = content[offset:]
+    # A stream can stop in the middle of an opening or closing marker.
+    for size in range(min(len(tail), 7), 0, -1):
+        if any(marker.startswith(tail[-size:].lower()) for marker in ('<think>', '</think>')):
+            tail = tail[:-size]
+            break
+    (thinking if inside else answer).append(tail)
+    reasoning = '\n'.join(part.strip() for part in thinking if part.strip())
+    response = '\n'.join(part.strip() for part in answer if part.strip())
+    return {
+        'reasoning': '> ' + reasoning.replace('\n', '\n> ') if reasoning else '',
+        'answer': response,
+        'hasReasoning': 'true' if reasoning else '',
+    }
diff --git a/src/langbot/libs/official_account_api/api.py b/src/langbot/libs/official_account_api/api.py
index 10508b097..5daf36a6c 100644
--- a/src/langbot/libs/official_account_api/api.py
+++ b/src/langbot/libs/official_account_api/api.py
@@ -5,6 +5,7 @@ from langbot.libs.wecom_api.WXBizMsgCrypt3 import WXBizMsgCrypt
 import xml.etree.ElementTree as ET
 from quart import Quart, request
 import hashlib
+import hmac
 from typing import Callable
 from langbot.libs.official_account_api.oaevent import OAEvent
 
@@ -24,6 +25,39 @@ xml_template = """
 _MAX_CALLBACK_BODY_BYTES = 1024 * 1024
 
 
+async def _read_callback_xml(client, req):
+    body = await req.data
+    if len(body) > _MAX_CALLBACK_BODY_BYTES:
+        raise ValueError('Official Account callback body exceeds the size limit')
+    root = ET.fromstring(body)
+    if root.find('Encrypt') is not None:
+        crypt = WXBizMsgCrypt(client.token, client.aes, client.appid)
+        code, xml = await asyncio.to_thread(
+            crypt.DecryptMsg,
+            body,
+            req.args.get('msg_signature', ''),
+            req.args.get('timestamp', ''),
+            req.args.get('nonce', ''),
+        )
+        if code != 0:
+            raise ValueError('Official Account callback decryption failed')
+        return xml.decode('utf-8') if isinstance(xml, bytes) else xml
+    digest = hashlib.sha1(
+        ''.join(
+            sorted(
+                [
+                    client.token,
+                    req.args.get('timestamp', ''),
+                    req.args.get('nonce', ''),
+                ]
+            )
+        ).encode('utf-8')
+    ).hexdigest()
+    if not hmac.compare_digest(digest, req.args.get('signature', '')):
+        raise ValueError('Official Account callback signature verification failed')
+    return body.decode('utf-8')
+
+
 class OAClient:
     _STATE_TTL_SECONDS = 600
     _STATE_MAX = 4096
@@ -63,6 +97,7 @@ class OAClient:
         }
         self.access_token_expiry_time = None
         self.msg_id_map = {}
+        self._processing_tasks: set[asyncio.Task] = set()
         self.generated_content = {}
         self._msg_seen_at = {}
         self._generated_at = {}
@@ -91,6 +126,9 @@ class OAClient:
             self._generated_at.pop(message_id, None)
 
     def clear(self) -> None:
+        for task in self._processing_tasks:
+            task.cancel()
+        self._processing_tasks.clear()
         self.msg_id_map.clear()
         self.generated_content.clear()
         self._msg_seen_at.clear()
@@ -155,28 +193,17 @@ class OAClient:
                     )
                     return 'signature verification failed', 403
             elif req.method == 'POST':
-                encryt_msg = await req.data
-                if len(encryt_msg) > _MAX_CALLBACK_BODY_BYTES:
-                    raise ValueError('Official Account callback body exceeds the size limit')
-                wxcpt = WXBizMsgCrypt(self.token, self.aes, self.appid)
-                ret, xml_msg = await asyncio.to_thread(
-                    wxcpt.DecryptMsg,
-                    encryt_msg,
-                    msg_signature,
-                    timestamp,
-                    nonce,
-                )
-                xml_msg = xml_msg.decode('utf-8')
-
-                if ret != 0:
-                    await self.logger.error('消息解密失败')
-                    raise Exception('消息解密失败')
+                xml_msg = await _read_callback_xml(self, req)
 
                 message_data = await self.get_message(xml_msg)
                 if message_data:
                     event = OAEvent.from_payload(message_data)
                     if event:
-                        await self._handle_message(event)
+                        from langbot.pkg.core.task_boundary import create_detached_task
+
+                        task = create_detached_task(self._process_message(event))
+                        self._processing_tasks.add(task)
+                        task.add_done_callback(self._processing_tasks.discard)
 
                 root = await asyncio.to_thread(ET.fromstring, xml_msg)
                 from_user = root.find('FromUserName').text  # 发送者
@@ -192,7 +219,7 @@ class OAClient:
                             to_user=from_user,
                             from_user=to_user,
                             create_time=int(time.time()),
-                            content=content,
+                            content=content.replace(']]>', ']]]]><![CDATA[>'),
                         )
 
                         return response_xml
@@ -202,15 +229,7 @@ class OAClient:
 
                     await asyncio.sleep(interval)
 
-                if self.msg_id_map.get(message_data['MsgId'], 1) == 3:
-                    # response_xml = xml_template.format(
-                    #     to_user=from_user,
-                    #     from_user=to_user,
-                    #     create_time=int(time.time()),
-                    #     content = "请求失效：暂不支持公众号超过15秒的请求，如有需求，请联系 LangBot 团队。"
-                    # )
-                    print('请求失效：暂不支持公众号超过15秒的请求，如有需求，请联系 LangBot 团队。')
-                    return ''
+                return 'success'
 
         except Exception:
             await self.logger.error(f'handle_callback_request失败: {traceback.format_exc()}')
@@ -219,14 +238,20 @@ class OAClient:
     async def get_message(self, xml_msg: str):
         root = await asyncio.to_thread(ET.fromstring, xml_msg)
 
-        message_data = {
-            'ToUserName': root.find('ToUserName').text,
-            'FromUserName': root.find('FromUserName').text,
-            'CreateTime': int(root.find('CreateTime').text),
-            'MsgType': root.find('MsgType').text,
-            'Content': root.find('Content').text if root.find('Content') is not None else None,
-            'MsgId': int(root.find('MsgId').text) if root.find('MsgId') is not None else None,
-        }
+        message_data = {element.tag: element.text for element in root}
+        message_data['CreateTime'] = int(message_data.get('CreateTime') or 0)
+        if message_data.get('MsgId'):
+            message_data['MsgId'] = int(message_data['MsgId'])
+        else:
+            message_data['MsgId'] = ':'.join(
+                str(message_data.get(key) or '')
+                for key in (
+                    'FromUserName',
+                    'CreateTime',
+                    'Event',
+                    'EventKey',
+                )
+            )
 
         return message_data
 
@@ -249,6 +274,12 @@ class OAClient:
 
         return decorator
 
+    async def _process_message(self, event: OAEvent):
+        try:
+            await self._handle_message(event)
+        except Exception:
+            await self.logger.error(f'OfficialAccount processing failed: {traceback.format_exc()}')
+
     async def _handle_message(self, event: OAEvent):
         """
         处理消息事件。
@@ -288,6 +319,7 @@ class OAClientForLongerResponse:
         logger: None,
         unified_mode: bool = False,
         api_base_url: str = 'https://api.weixin.qq.com',
+        customer_service: bool = False,
     ):
         self.token = token
         self.aes = EncodingAESKey
@@ -313,8 +345,13 @@ class OAClientForLongerResponse:
         }
         self.access_token_expiry_time = None
         self.loading_message = LoadingMessage
+        from .customer_service import CustomerServiceReplies
+
+        self.customer_service = CustomerServiceReplies(AppID, Appsecret, api_base_url) if customer_service else None
+        self._callback_responses = {}
         self.msg_queue = {}
         self.user_msg_queue = {}
+        self._processing_tasks: set[asyncio.Task] = set()
         self._last_queue_cleanup = 0.0
         self.logger = logger
 
@@ -336,6 +373,10 @@ class OAClientForLongerResponse:
     def clear(self) -> None:
         self.msg_queue.clear()
         self.user_msg_queue.clear()
+        self._callback_responses.clear()
+        for task in self._processing_tasks:
+            task.cancel()
+        self._processing_tasks.clear()
 
     async def handle_callback_request(self):
         """处理回调请求（独立端口模式，使用全局 request）。"""
@@ -393,27 +434,26 @@ class OAClientForLongerResponse:
                 return 'signature verification failed', 403
 
             elif req.method == 'POST':
-                encryt_msg = await req.data
-                if len(encryt_msg) > _MAX_CALLBACK_BODY_BYTES:
-                    raise ValueError('Official Account callback body exceeds the size limit')
-                wxcpt = WXBizMsgCrypt(self.token, self.aes, self.appid)
-                ret, xml_msg = await asyncio.to_thread(
-                    wxcpt.DecryptMsg,
-                    encryt_msg,
-                    msg_signature,
-                    timestamp,
-                    nonce,
-                )
-                xml_msg = xml_msg.decode('utf-8')
-
-                if ret != 0:
-                    await self.logger.error('消息解密失败')
-                    raise Exception('消息解密失败')
+                xml_msg = await _read_callback_xml(self, req)
 
                 # 解析 XML
                 root = await asyncio.to_thread(ET.fromstring, xml_msg)
                 from_user = root.find('FromUserName').text
                 to_user = root.find('ToUserName').text
+                message_key = (
+                    from_user,
+                    root.findtext('MsgId')
+                    or (root.findtext('CreateTime'), root.findtext('Event'), root.findtext('EventKey')),
+                )
+                cached = self._callback_responses.get(message_key)
+                if cached and time.monotonic() - cached[0] < 600:
+                    return cached[1]
+
+                def remember(response):
+                    self._callback_responses[message_key] = (time.monotonic(), response)
+                    while len(self._callback_responses) > self._MAX_USERS:
+                        self._callback_responses.pop(next(iter(self._callback_responses)))
+                    return response
 
                 if self.msg_queue.get(from_user) and self.msg_queue[from_user][0]['content']:
                     queue_top = self.msg_queue[from_user].pop(0)
@@ -428,20 +468,22 @@ class OAClientForLongerResponse:
                         to_user=from_user,
                         from_user=to_user,
                         create_time=int(time.time()),
-                        content=queue_content,
+                        content=queue_content.replace(']]>', ']]]]><![CDATA[>'),
                     )
-                    return response_xml
+                    return remember(response_xml)
 
                 else:
                     response_xml = xml_template.format(
                         to_user=from_user,
                         from_user=to_user,
                         create_time=int(time.time()),
-                        content=self.loading_message,
+                        content=self.loading_message.replace(']]>', ']]]]><![CDATA[>'),
                     )
+                    if not self.loading_message:
+                        response_xml = 'success'
 
-                    if self.user_msg_queue.get(from_user) and self.user_msg_queue[from_user][0]['content']:
-                        return response_xml
+                    if self.user_msg_queue.get(from_user):
+                        return remember(response_xml)
                     else:
                         message_data = await self.get_message(xml_msg)
 
@@ -457,9 +499,15 @@ class OAClientForLongerResponse:
                                     -self._MAX_MESSAGES_PER_USER :
                                 ]
                                 self._prune_queues()
-                                await self._handle_message(event)
+                                # The callback must return the loading reply immediately.
+                                # Native listeners establish their own Workspace scope.
+                                from langbot.pkg.core.task_boundary import create_detached_task
 
-                        return response_xml
+                                task = create_detached_task(self._process_pending_message(event))
+                                self._processing_tasks.add(task)
+                                task.add_done_callback(self._processing_tasks.discard)
+
+                        return remember(response_xml)
 
         except Exception:
             await self.logger.error(f'handle_callback_request失败: {traceback.format_exc()}')
@@ -468,14 +516,20 @@ class OAClientForLongerResponse:
     async def get_message(self, xml_msg: str):
         root = await asyncio.to_thread(ET.fromstring, xml_msg)
 
-        message_data = {
-            'ToUserName': root.find('ToUserName').text,
-            'FromUserName': root.find('FromUserName').text,
-            'CreateTime': int(root.find('CreateTime').text),
-            'MsgType': root.find('MsgType').text,
-            'Content': root.find('Content').text if root.find('Content') is not None else None,
-            'MsgId': int(root.find('MsgId').text) if root.find('MsgId') is not None else None,
-        }
+        message_data = {element.tag: element.text for element in root}
+        message_data['CreateTime'] = int(message_data.get('CreateTime') or 0)
+        if message_data.get('MsgId'):
+            message_data['MsgId'] = int(message_data['MsgId'])
+        else:
+            message_data['MsgId'] = ':'.join(
+                str(message_data.get(key) or '')
+                for key in (
+                    'FromUserName',
+                    'CreateTime',
+                    'Event',
+                    'EventKey',
+                )
+            )
 
         return message_data
 
@@ -508,7 +562,32 @@ class OAClientForLongerResponse:
             for handler in self._message_handlers[msg_type]:
                 await handler(event)
 
+    async def _process_pending_message(self, event: OAEvent):
+        try:
+            await self._handle_message(event)
+        except asyncio.CancelledError:
+            raise
+        except Exception:
+            self.user_msg_queue.pop(event.user_id, None)
+            await self.logger.error(f'OfficialAccount passive processing failed: {traceback.format_exc()}')
+
     async def set_message(self, from_user: int, message_id: int, content: str):
+        if self.customer_service is not None:
+            try:
+                code = await self.customer_service.send(str(from_user), content)
+            except Exception as exc:
+                # Do not expose request URLs containing access tokens or AppSecret.
+                self.user_msg_queue.pop(from_user, None)
+                raise RuntimeError(f'OfficialAccount customer message failed ({type(exc).__name__})') from None
+            if code == 0:
+                self.user_msg_queue.pop(from_user, None)
+                return False
+            if code not in {48001, 45015, 45047, 43004}:
+                self.user_msg_queue.pop(from_user, None)
+                raise RuntimeError(f'OfficialAccount customer message rejected: errcode={code}')
+            await self.logger.info(
+                f'OfficialAccount customer message unavailable (errcode={code}); reply queued for next message'
+            )
         if from_user not in self.msg_queue:
             self.msg_queue[from_user] = []
 
@@ -520,3 +599,4 @@ class OAClientForLongerResponse:
         )
         self.msg_queue[from_user] = self.msg_queue[from_user][-self._MAX_MESSAGES_PER_USER :]
         self._prune_queues()
+        return True
diff --git a/src/langbot/libs/official_account_api/customer_service.py b/src/langbot/libs/official_account_api/customer_service.py
new file mode 100644
index 000000000..e5fe23f75
--- /dev/null
+++ b/src/langbot/libs/official_account_api/customer_service.py
@@ -0,0 +1,46 @@
+"""Optional asynchronous replies through WeChat's customer service API."""
+
+import asyncio
+import time
+
+import httpx
+
+
+class CustomerServiceReplies:
+    def __init__(self, appid: str, secret: str, base_url: str):
+        self.appid = appid
+        self.secret = secret
+        self.base_url = base_url.rstrip('/')
+        self._token = ''
+        self._expires = 0.0
+        self._lock = asyncio.Lock()
+
+    async def send(self, user_id: str, content: str) -> int:
+        # Retry only explicit token rejection, never an ambiguous network failure.
+        async with self._lock, httpx.AsyncClient(timeout=20) as client:
+            for attempt in range(2):
+                if time.monotonic() >= self._expires:
+                    response = await client.get(
+                        f'{self.base_url}/cgi-bin/token',
+                        params={'grant_type': 'client_credential', 'appid': self.appid, 'secret': self.secret},
+                    )
+                    if response.status_code != 200:
+                        raise RuntimeError(f'OfficialAccount token request HTTP {response.status_code}')
+                    data = response.json()
+                    if data.get('errcode'):
+                        return int(data['errcode'])
+                    self._token = data['access_token']
+                    self._expires = time.monotonic() + max(0, int(data['expires_in']) - 120)
+                response = await client.post(
+                    f'{self.base_url}/cgi-bin/message/custom/send',
+                    params={'access_token': self._token},
+                    json={'touser': user_id, 'msgtype': 'text', 'text': {'content': content}},
+                )
+                if response.status_code != 200:
+                    raise RuntimeError(f'OfficialAccount customer message HTTP {response.status_code}')
+                code = int(response.json()['errcode'])
+                if code in {40001, 40014, 42001} and attempt == 0:
+                    self._expires = 0.0
+                    continue
+                return code
+        raise RuntimeError('OfficialAccount token refresh failed')
diff --git a/src/langbot/libs/openclaw_weixin_api/typing_indicator.py b/src/langbot/libs/openclaw_weixin_api/typing_indicator.py
new file mode 100644
index 000000000..9334fde0d
--- /dev/null
+++ b/src/langbot/libs/openclaw_weixin_api/typing_indicator.py
@@ -0,0 +1,120 @@
+"""Best-effort, per-peer typing leases for the iLink transport."""
+
+import asyncio
+import contextlib
+import logging
+import time
+from collections import OrderedDict
+from dataclasses import dataclass
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class Lease:
+    users: int = 0
+    task: asyncio.Task | None = None
+    ticket: str = ''
+
+
+class TypingIndicator:
+    INTERVAL = 5.0
+    TIMEOUT = 3.0
+    TICKET_TTL = 600.0
+    MAX_PEERS = 4096
+
+    def __init__(self, client, context_token):
+        self.client = client
+        self.context_token = context_token
+        self.leases: dict[str, Lease] = {}
+        self.cache = OrderedDict()
+        self.closed = False
+
+    async def _ticket(self, peer):
+        now = time.monotonic()
+        cached = self.cache.get(peer)
+        if cached and cached[1] > now:
+            return cached[0]
+        try:
+            response = await asyncio.wait_for(
+                self.client.get_config(peer, self.context_token(peer)),
+                self.TIMEOUT,
+            )
+            if response.ret not in (None, 0):
+                raise ValueError('getconfig failed')
+            ticket = response.typing_ticket or ''
+            delay = self.TICKET_TTL if ticket else 30.0
+        except Exception:
+            ticket = ''
+            delay = min((cached[2] * 2 if cached else 5.0), 300.0)
+            logger.debug('Weixin typing ticket unavailable; retry deferred')
+        self.cache[peer] = (ticket, now + delay, delay)
+        self.cache.move_to_end(peer)
+        while len(self.cache) > self.MAX_PEERS:
+            self.cache.popitem(last=False)
+        return ticket
+
+    async def _run(self, peer, lease):
+        try:
+            while not self.closed and lease.users:
+                ticket = await self._ticket(peer)
+                if ticket:
+                    # Keep the last ticket so cancellation can stop an in-flight start.
+                    lease.ticket = ticket
+                    try:
+                        await asyncio.wait_for(self.client.send_typing(peer, ticket), self.TIMEOUT)
+                    except Exception:
+                        self.cache[peer] = ('', time.monotonic() + 30, 30)
+                        logger.debug('Weixin typing start failed; retry deferred')
+                await asyncio.sleep(self.INTERVAL)
+        finally:
+            if lease.ticket:
+                try:
+                    # Refresh expired tickets before cancelling a long-running turn.
+                    ticket = await self._ticket(peer) or lease.ticket
+                    await asyncio.wait_for(self.client.stop_typing(peer, ticket), self.TIMEOUT)
+                except Exception:
+                    logger.debug('Weixin typing stop failed')
+
+    @contextlib.asynccontextmanager
+    async def processing(self, peer):
+        lease = None
+        while not self.closed:
+            lease = self.leases.get(peer)
+            if lease is None:
+                if len(self.leases) >= self.MAX_PEERS:
+                    lease = None
+                    break
+                lease = Lease(users=1)
+                self.leases[peer] = lease
+                lease.task = asyncio.create_task(self._run(peer, lease))
+                break
+            if lease.users:
+                lease.users += 1
+                break
+            # Serialize stop/start for the same peer: an old stop must not cancel a new start.
+            await asyncio.gather(asyncio.shield(lease.task), return_exceptions=True)
+            if self.leases.get(peer) is lease:
+                self.leases.pop(peer, None)
+            lease = None
+        try:
+            yield
+        finally:
+            if lease is not None and lease.users:
+                lease.users -= 1
+                if not lease.users:
+                    if not lease.task.cancelling():
+                        lease.task.cancel()
+                    await asyncio.gather(asyncio.shield(lease.task), return_exceptions=True)
+                    if self.leases.get(peer) is lease:
+                        self.leases.pop(peer, None)
+
+    async def close(self):
+        self.closed = True
+        tasks = [lease.task for lease in self.leases.values()]
+        for task in tasks:
+            if not task.done() and not task.cancelling():
+                task.cancel()
+        await asyncio.gather(*(asyncio.shield(task) for task in tasks), return_exceptions=True)
+        self.leases.clear()
+        self.cache.clear()
diff --git a/src/langbot/libs/slack_api/api.py b/src/langbot/libs/slack_api/api.py
index 69d70b402..d96422a80 100644
--- a/src/langbot/libs/slack_api/api.py
+++ b/src/langbot/libs/slack_api/api.py
@@ -1,10 +1,13 @@
 import asyncio
 import json
 import traceback
+from urllib.parse import parse_qs
+from slack_sdk.signature import SignatureVerifier
 from quart import Quart, jsonify, request
 from slack_sdk.web.async_client import AsyncWebClient
 from .slackevent import SlackEvent
 from typing import Callable
+from collections import OrderedDict
 import langbot_plugin.api.entities.builtin.platform.events as platform_events
 
 _MAX_CALLBACK_BODY_BYTES = 1024 * 1024
@@ -30,6 +33,11 @@ class SlackClient:
         }
         self.bot_user_id = None  # 避免机器人回复自己的消息
         self.logger = logger
+        self.socket_client = None
+        self.socket_queue = asyncio.Queue(maxsize=100)
+        self.socket_seen = OrderedDict()
+        self.socket_stop = asyncio.Event()
+        self.socket_worker = None
 
     async def handle_callback_request(self):
         """处理回调请求（独立端口模式，使用全局 request）"""
@@ -44,6 +52,8 @@ class SlackClient:
         Returns:
             响应数据
         """
+        if self.socket_client is not None:
+            return jsonify({'error': 'Socket Mode is enabled'}), 409
         return await self._handle_callback_internal(req)
 
     async def _handle_callback_internal(self, req):
@@ -56,40 +66,122 @@ class SlackClient:
             body = await req.get_data()
             if len(body) > _MAX_CALLBACK_BODY_BYTES:
                 raise ValueError('Slack callback body exceeds the size limit')
-            data = await asyncio.to_thread(json.loads, body)
+            if not SignatureVerifier(self.signing_secret).is_valid_request(body, dict(req.headers)):
+                return jsonify({'error': 'Invalid Slack signature'}), 401
+            if req.mimetype == 'application/x-www-form-urlencoded':
+                data = json.loads(parse_qs(body.decode()).get('payload', ['{}'])[0])
+            else:
+                data = json.loads(body)
             if 'type' in data:
                 if data['type'] == 'url_verification':
                     return data['challenge']
 
-            bot_user_id = data.get('event', {}).get('bot_id', '')
-
-            if self.bot_user_id and bot_user_id == self.bot_user_id:
-                return jsonify({'status': 'ok'})
-
-            # 处理私信
-            if data and data.get('event', {}).get('channel_type') in ['im']:
-                event = SlackEvent.from_payload(data)
-                await self._handle_message(event)
-                return jsonify({'status': 'ok'})
-
-            # 处理群聊
-            if data.get('event', {}).get('type') == 'app_mention':
-                data.setdefault('event', {})['channel_type'] = 'channel'
-                event = SlackEvent.from_payload(data)
-                await self._handle_message(event)
-                return jsonify({'status': 'ok'})
-
-            return jsonify({'status': 'ok'})
+            self._start_worker()
+            if not self._enqueue(data, data.get('event_id')):
+                return jsonify({'error': 'Slack event queue is full'}), 503
+            return '', 200
 
         except Exception as e:
             await self.logger.error(f'Error in handle_callback_request: {traceback.format_exc()}')
             raise (e)
 
-    async def _handle_message(self, event: SlackEvent):
+    async def handle_payload(self, data):
+        """Dispatch the same event representation for HTTP and Socket Mode."""
+        raw = data.get('event', {})
+        for authorization in data.get('authorizations', []):
+            if not self.bot_user_id and authorization.get('is_bot') and authorization.get('user_id'):
+                self.bot_user_id = authorization['user_id']
+                break
+        if data.get('type') in {'block_actions', 'view_submission', 'view_closed', 'shortcut', 'message_action'}:
+            await self._handle_message(SlackEvent(data), 'event')
+            return
+        if raw.get('bot_id') or raw.get('subtype') == 'bot_message':
+            return
+        kind, subtype = raw.get('type'), raw.get('subtype')
+        if kind == 'message' and subtype not in (None, 'file_share', 'thread_broadcast'):
+            await self._handle_message(SlackEvent(data), 'event')
+        elif raw.get('channel_type') == 'im' and kind == 'message':
+            await self._handle_message(SlackEvent.from_payload(data))
+        elif kind == 'app_mention':
+            raw['channel_type'] = 'channel'
+            await self._handle_message(SlackEvent.from_payload(data))
+        elif kind != 'message':
+            await self._handle_message(SlackEvent(data), 'event')
+
+    def _enqueue(self, payload, key=None):
+        if key and key in self.socket_seen:
+            return True
+        try:
+            self.socket_queue.put_nowait(payload)
+        except asyncio.QueueFull:
+            return False
+        if key:
+            self.socket_seen[key] = True
+            while len(self.socket_seen) > 4096:
+                self.socket_seen.popitem(last=False)
+        return True
+
+    def _start_worker(self):
+        if self.socket_worker is not None and not self.socket_worker.done():
+            return
+
+        async def worker():
+            while True:
+                payload = await self.socket_queue.get()
+                try:
+                    await self.handle_payload(payload)
+                except Exception:
+                    await self.logger.error('Slack event processing failed')
+                finally:
+                    self.socket_queue.task_done()
+
+        self.socket_worker = asyncio.create_task(worker())
+
+    async def _socket_request(self, client, req):
+        from slack_sdk.socket_mode.response import SocketModeResponse
+
+        if req.type in {'events_api', 'interactive'}:
+            key = req.payload.get('event_id') or req.envelope_id
+            if not self._enqueue(req.payload, key):
+                # Leave unacknowledged so Slack can retry; do not grow tasks without bounds.
+                await self.logger.warning('Slack Socket Mode event queue is full')
+                return
+        await client.send_socket_mode_response(SocketModeResponse(envelope_id=req.envelope_id))
+
+    async def run_socket(self, app_token):
+        from slack_sdk.socket_mode.aiohttp import SocketModeClient
+
+        self.socket_stop.clear()
+        self.socket_client = SocketModeClient(app_token=app_token, web_client=self.client)
+        self.socket_client.socket_mode_request_listeners.append(self._socket_request)
+
+        self._start_worker()
+        try:
+            await self.socket_client.connect()
+            await self.logger.info('Slack Socket Mode connected')
+            await self.socket_stop.wait()
+        finally:
+            await self.close_socket()
+
+    async def close_socket(self):
+        self.socket_stop.set()
+        client, self.socket_client = self.socket_client, None
+        if client is not None:
+            await client.close()
+        worker, self.socket_worker = self.socket_worker, None
+        if worker is not None:
+            worker.cancel()
+            await asyncio.gather(worker, return_exceptions=True)
+        self.socket_seen.clear()
+        while not self.socket_queue.empty():
+            self.socket_queue.get_nowait()
+            self.socket_queue.task_done()
+
+    async def _handle_message(self, event: SlackEvent, kind=None):
         """
         处理消息事件。
         """
-        msg_type = event.type
+        msg_type = kind or event.type
         if msg_type in self._message_handlers:
             for handler in self._message_handlers[msg_type]:
                 await handler(event)
@@ -109,8 +201,8 @@ class SlackClient:
         try:
             response = await self.client.chat_postMessage(channel=channel_id, text=text)
             if self.bot_user_id is None and response.get('ok'):
-                self.bot_user_id = response['message']['bot_id']
-            return
+                self.bot_user_id = response.get('message', {}).get('user')
+            return response.data if hasattr(response, 'data') else response
         except Exception as e:
             await self.logger.error(f'Error in send_message: {e}')
             raise e
@@ -119,9 +211,9 @@ class SlackClient:
         try:
             response = await self.client.chat_postMessage(channel='@' + user_id, text=text)
             if self.bot_user_id is None and response.get('ok'):
-                self.bot_user_id = response['message']['bot_id']
+                self.bot_user_id = response.get('message', {}).get('user')
 
-            return
+            return response.data if hasattr(response, 'data') else response
         except Exception as e:
             await self.logger.error(f'Error in send_message: {traceback.format_exc()}')
             raise e
diff --git a/src/langbot/libs/slack_api/media.py b/src/langbot/libs/slack_api/media.py
new file mode 100644
index 000000000..f220af14f
--- /dev/null
+++ b/src/langbot/libs/slack_api/media.py
@@ -0,0 +1,125 @@
+"""Transfer attachment bytes rather than degrading attachments to text."""
+
+import asyncio
+import mimetypes
+from pathlib import Path
+from urllib.parse import urlparse, unquote
+
+import aiohttp
+
+from langbot.pkg.utils import httpclient, image
+from langbot_plugin.api.entities.builtin.platform import message as pm
+
+MAX_MEDIA_BYTES = 20 * 1024 * 1024
+MEDIA_TYPES = (pm.Image, pm.File, pm.Voice)
+
+
+async def download_media(url: str, token: str = '') -> tuple[bytes, str]:
+    parsed = urlparse(url)
+    if parsed.scheme not in {'http', 'https'}:
+        raise ValueError('Slack media requires an HTTP(S) URL')
+    # Never disclose a bot token to arbitrary media hosts or redirects.
+    private = parsed.scheme == 'https' and parsed.hostname == 'files.slack.com'
+    headers = {'Authorization': f'Bearer {token}'} if private and token else {}
+    async with httpclient.get_session().get(
+        url, headers=headers, allow_redirects=not bool(headers), timeout=aiohttp.ClientTimeout(total=30)
+    ) as response:
+        if response.status != 200:
+            raise ValueError(f'Slack media download failed (HTTP {response.status})')
+        data = await httpclient.read_limited(response, max_bytes=MAX_MEDIA_BYTES)
+        return data, response.headers.get('Content-Type', 'application/octet-stream').split(';')[0]
+
+
+async def attachment_bytes(component, token: str = '') -> tuple[bytes, str]:
+    name = getattr(component, 'name', '')
+    mime = ''
+    if component.base64:
+        value = component.base64
+        if value.startswith('data:'):
+            header, value = value.split(',', 1)
+            mime = header[5:].split(';')[0]
+        data = await image.decode_base64_limited(value, max_bytes=MAX_MEDIA_BYTES)
+    elif component.path:
+        path = Path(component.path)
+
+        def read():
+            with path.open('rb') as stream:
+                return stream.read(MAX_MEDIA_BYTES + 1)
+
+        data = await asyncio.to_thread(read)
+        name = name or path.name
+    elif component.url:
+        data, mime = await download_media(component.url, token)
+        name = name or Path(unquote(urlparse(component.url).path)).name
+    else:
+        raise ValueError('Slack attachment has no base64, path or URL')
+    if not data or len(data) > MAX_MEDIA_BYTES:
+        raise ValueError('Slack attachment is empty or exceeds the 20 MiB limit')
+    if not name:
+        if isinstance(component, pm.Image) and not mime:
+            mime = await asyncio.to_thread(_image_mime, data)
+        name = 'attachment' + (mimetypes.guess_extension(mime) or '.bin')
+    return data, name.replace('\\', '/').rsplit('/', 1)[-1]
+
+
+def _image_mime(data: bytes) -> str:
+    import io
+    from PIL import Image
+
+    with Image.open(io.BytesIO(data)) as picture:
+        return Image.MIME.get(picture.format, 'application/octet-stream')
+
+
+async def send_chain(
+    bot, target_type: str, target_id: str, chain: pm.MessageChain, thread_ts: str | None = None
+) -> dict:
+    from langbot.pkg.platform.adapters.slack.message_converter import SlackMessageConverter
+
+    if target_type not in {'person', 'channel'}:
+        raise ValueError(f'Unsupported Slack target type: {target_type}')
+    responses = []
+    pending = []
+    channel = None
+
+    async def flush():
+        if not pending:
+            return
+        text = await SlackMessageConverter.yiri2target(pm.MessageChain(list(pending)))
+        pending.clear()
+        if text:
+            if thread_ts:
+                response = await bot.client.chat_postMessage(channel=target_id, text=text, thread_ts=thread_ts)
+                responses.append(response.data if hasattr(response, 'data') else response)
+            else:
+                method = bot.send_message_to_one if target_type == 'person' else bot.send_message_to_channel
+                responses.append(await method(text, target_id))
+
+    for component in chain:
+        if isinstance(component, MEDIA_TYPES):
+            data, filename = await attachment_bytes(component, bot.bot_token)
+            await flush()
+            if channel is None:
+                if target_type == 'person' and not target_id.startswith('D'):
+                    opened = await bot.client.conversations_open(users=[target_id])
+                    channel = opened['channel']['id']
+                else:
+                    channel = target_id
+            response = await bot.client.files_upload_v2(
+                channel=channel,
+                file=data,
+                filename=filename,
+                title=filename,
+                **({'thread_ts': thread_ts} if thread_ts else {}),
+            )
+            if not response.get('ok'):
+                raise RuntimeError(f'Slack file upload failed: {response.get("error", "unknown_error")}')
+            responses.append(response.data if hasattr(response, 'data') else response)
+        elif isinstance(component, pm.Forward):
+            await flush()
+            for node in component.node_list:
+                if node.message_chain:
+                    responses.append(await send_chain(bot, target_type, target_id, node.message_chain, thread_ts))
+        else:
+            pending.append(component)
+    await flush()
+    return {'target_type': target_type, 'target_id': target_id, 'responses': responses}
diff --git a/src/langbot/libs/slack_api/slackevent.py b/src/langbot/libs/slack_api/slackevent.py
index 77177816e..0834d212b 100644
--- a/src/langbot/libs/slack_api/slackevent.py
+++ b/src/langbot/libs/slack_api/slackevent.py
@@ -12,6 +12,8 @@ class SlackEvent(dict):
 
     @property
     def text(self) -> str:
+        if self.get('event', {}).get('text') is not None:
+            return self['event']['text']
         if self.get('event', {}).get('channel_type') == 'im':
             blocks = self.get('event', {}).get('blocks', [])
             if not blocks:
@@ -63,7 +65,8 @@ class SlackEvent(dict):
 
     @property
     def message_id(self) -> str:
-        return self.get('event_id', '')
+        raw = self.get('event', {})
+        return raw.get('deleted_ts') or raw.get('message', {}).get('ts') or raw.get('ts', '')
 
     @property
     def pic_url(self) -> str:
diff --git a/src/langbot/libs/wecom_ai_bot_api/api.py b/src/langbot/libs/wecom_ai_bot_api/api.py
index 785e82f28..207315d28 100644
--- a/src/langbot/libs/wecom_ai_bot_api/api.py
+++ b/src/langbot/libs/wecom_ai_bot_api/api.py
@@ -1658,6 +1658,8 @@ class WecomBotClient:
         self._card_action_callback: Optional[Callable] = None
         self._stream_last_content: dict[str, str] = {}
         self._dispatch_tasks: set[asyncio.Task] = set()
+        self._welcome_replies: dict[str, asyncio.Future] = {}
+        self._handled_form_tasks: dict[str, float] = {}
         # Optional `source` block injected into every interactive template_card
         # the client builds. Set via `set_card_source` from the adapter after
         # reading config. Format: {icon_url, desc, desc_color}.
@@ -1764,7 +1766,12 @@ class WecomBotClient:
             event: 由企业微信消息转换的内部事件对象。
         """
         try:
-            await self._handle_message(event)
+            if event.msgtype == 'event':
+                kind = 'template_card_event' if extract_wecom_event_type(event) == 'template_card_event' else 'event'
+                for handler in self._message_handlers.get(kind, []):
+                    await handler(event)
+            else:
+                await self._handle_message(event)
         except Exception:
             await self.logger.error(traceback.format_exc())
 
@@ -1996,11 +2003,48 @@ class WecomBotClient:
         if event_type == 'template_card_event':
             return await self._handle_template_card_event(msg_json, nonce)
 
+        if event_type == 'enter_chat':
+            return await self._handle_enter_chat(msg_json, nonce)
+
+        if msg_json.get('msgtype') == 'event':
+            event = wecombotevent.WecomBotEvent({**msg_json, 'type': msg_json.get('chattype', 'single')})
+            if not self._start_dispatch_task(event):
+                return Response('Busy', status=503)
+            return await self._encrypt_and_reply({}, nonce)
+
         if msg_json.get('msgtype') == 'stream':
             return await self._handle_post_followup_response(msg_json, nonce)
 
         return await self._handle_post_initial_response(msg_json, nonce)
 
+    async def reply_welcome(self, message_id: str, content: str) -> dict:
+        """Complete the pending HTTP welcome response before its deadline."""
+        future = self._welcome_replies.get(message_id)
+        if future is None or future.done():
+            raise ValueError('WeCom welcome reply window has expired or was already used')
+        future.set_result({'msgtype': 'text', 'text': {'content': content}})
+        return {'ok': True}
+
+    async def _handle_enter_chat(self, msg_json: dict[str, Any], nonce: str):
+        message_id = str(msg_json.get('msgid') or '')
+        if not message_id or message_id in self._welcome_replies:
+            return await self._encrypt_and_reply({}, nonce)
+        future = asyncio.get_running_loop().create_future()
+        self._welcome_replies[message_id] = future
+        try:
+            event = wecombotevent.WecomBotEvent({**msg_json, 'type': msg_json.get('chattype', 'single')})
+            if not self._start_dispatch_task(event):
+                return Response('Busy', status=503)
+            try:
+                reply = await asyncio.wait_for(asyncio.shield(future), timeout=4.5)
+            except asyncio.TimeoutError:
+                reply = {}
+            return await self._encrypt_and_reply(reply, nonce)
+        finally:
+            self._welcome_replies.pop(message_id, None)
+            if not future.done():
+                future.cancel()
+
     async def _handle_template_card_event(self, msg_json: dict[str, Any], nonce: str) -> tuple[Response, int]:
         """Handle a button click on a button_interaction template_card.
 
@@ -2014,10 +2058,29 @@ class WecomBotClient:
 
             await self.logger.info(f'收到按钮点击: task_id={task_id} event_key={event_key!r} card_type={card_type}')
 
+            now = time.monotonic()
+            self._handled_form_tasks = {
+                key: seen for key, seen in self._handled_form_tasks.items() if now - seen < 3600
+            }
+            if task_id in self._handled_form_tasks:
+                return await self._encrypt_and_reply({}, nonce)
             session = self.stream_sessions.get_session_by_task_id(task_id)
             if session is None:
-                await self.logger.warning(f'未找到 task_id={task_id} 对应的 session，按钮点击被丢弃')
+                if not task_id.startswith('dify-'):
+                    event = wecombotevent.WecomBotEvent(
+                        {
+                            **msg_json,
+                            'type': msg_json.get('chattype', 'single'),
+                            'msgtype': 'event',
+                            'eventtype': 'template_card_event',
+                        }
+                    )
+                    if not self._start_dispatch_task(event):
+                        return Response('Busy', status=503)
             else:
+                self._handled_form_tasks[task_id] = now
+                if len(self._handled_form_tasks) > 2048:
+                    self._handled_form_tasks.pop(next(iter(self._handled_form_tasks)))
                 if self._card_action_callback is not None:
                     try:
                         await self._card_action_callback(session, event_key, task_id, msg_json)
@@ -2123,7 +2186,7 @@ class WecomBotClient:
 
         Args:
             msg_id: 原始企业微信消息 ID。
-            content: 模型产生的片段内容。
+            content: 流水线传入的累计正文快照。
             is_final: 是否为最终片段。
 
         Returns:
@@ -2138,12 +2201,8 @@ class WecomBotClient:
             return False
 
         previous_content = self._stream_last_content.get(msg_id, '')
-        if previous_content and content.startswith(previous_content):
-            next_content = content
-        elif previous_content and not content:
-            next_content = previous_content
-        else:
-            next_content = previous_content + content if previous_content else content
+        # Pipeline delivery provides a snapshot, not a provider delta.
+        next_content = content or previous_content
         if len(next_content) > _MAX_STREAM_CONTENT_CHARS:
             next_content = next_content[-_MAX_STREAM_CONTENT_CHARS:]
 
diff --git a/src/langbot/libs/wecom_ai_bot_api/ws_client.py b/src/langbot/libs/wecom_ai_bot_api/ws_client.py
index 3f4cb0504..9daf15900 100644
--- a/src/langbot/libs/wecom_ai_bot_api/ws_client.py
+++ b/src/langbot/libs/wecom_ai_bot_api/ws_client.py
@@ -11,6 +11,7 @@ Official Node.js SDK: https://github.com/WecomTeam/aibot-node-sdk
 from __future__ import annotations
 
 import asyncio
+import hashlib
 import json
 import secrets
 import time
@@ -134,6 +135,9 @@ class WecomBotWsClient:
         self._stream_ids: dict[str, str] = {}  # msg_id -> req_id|stream_id
         # Dedup: skip sending when content hasn't changed
         self._stream_last_content: dict[str, str] = {}  # msg_id -> last content sent
+        # Preserve completion identity after removing active stream mappings.
+        # A runner may emit both a final delta and a completed message snapshot.
+        self._completed_stream_content: dict[str, str] = {}
         # Stream session info for feedback tracking
         self._stream_sessions: dict[str, dict] = {}  # msg_id -> session info
         # Feedback tracking: feedback_id -> session info
@@ -146,6 +150,7 @@ class WecomBotWsClient:
         # context on click).
         # task_id -> {form_data, msg_id, user_id, chat_id, stream_id, req_id}
         self._pending_forms_by_task: dict[str, dict] = {}
+        self._handled_form_tasks: dict[str, float] = {}
         # Reverse: msg_id -> task_id (for cleanup when stream finishes).
         self._task_id_by_msg: dict[str, str] = {}
         # Optional card-action callback registered by the adapter.
@@ -320,6 +325,12 @@ class WecomBotWsClient:
         }
         return await self._send_reply(req_id, body)
 
+    async def reply_welcome(self, req_id: str, content: str) -> Optional[dict]:
+        """Reply to enter_chat using the dedicated welcome command."""
+        return await self._send_reply(
+            req_id, {'msgtype': 'text', 'text': {'content': content}}, cmd=CMD_RESPOND_WELCOME
+        )
+
     async def reply_text(self, req_id: str, content: str) -> Optional[dict]:
         """Send a non-streaming text reply.
 
@@ -661,16 +672,15 @@ class WecomBotWsClient:
         """
         key = self._stream_ids.get(msg_id)
         if not key:
+            if is_final and self._completed_stream_content.get(msg_id) == hashlib.sha256(content.encode()).hexdigest():
+                return True
             return False
         req_id, stream_id = key.split('|', 1)
         try:
             previous_content = self._stream_last_content.get(msg_id, '')
-            if previous_content and content.startswith(previous_content):
-                next_content = content
-            elif previous_content and not content:
-                next_content = previous_content
-            else:
-                next_content = previous_content + content if previous_content else content
+            # Pipeline delivery provides complete snapshots, including edits
+            # that need not start with the previous text (e.g. reasoning tags).
+            next_content = content or previous_content
             if len(next_content) > _MAX_STREAM_CONTENT_CHARS:
                 next_content = next_content[-_MAX_STREAM_CONTENT_CHARS:]
 
@@ -720,6 +730,8 @@ class WecomBotWsClient:
             await self.reply_stream(req_id, stream_id, next_content, finish=is_final, feedback_id=feedback_id)
             self._stream_last_content[msg_id] = next_content
             if is_final:
+                self._completed_stream_content[msg_id] = hashlib.sha256(next_content.encode()).hexdigest()
+                self._cap_mapping(self._completed_stream_content, _STREAM_CACHE_MAX)
                 self._stream_ids.pop(msg_id, None)
                 self._stream_last_content.pop(msg_id, None)
                 self._stream_sessions.pop(msg_id, None)
@@ -981,6 +993,7 @@ class WecomBotWsClient:
                 await self.logger.debug(f'Received event_callback event_type={event_type}')
 
             message_data = {
+                **body,
                 'msgtype': 'event',
                 'type': body.get('chattype', 'single'),
                 'event': event_info,
@@ -1064,15 +1077,35 @@ class WecomBotWsClient:
             f'Received template_card_event (ws): task_id={task_id} event_key={event_key!r} card_type={card_type}'
         )
 
+        now = time.monotonic()
+        self._handled_form_tasks = {key: seen for key, seen in self._handled_form_tasks.items() if now - seen < 3600}
+        if task_id in self._handled_form_tasks:
+            return
         pending = self._pending_forms_by_task.get(task_id)
         if pending is None:
-            await self.logger.warning(f'No pending_form found for task_id={task_id} (ws); card event ignored')
+            # Expired managed forms must not start a new business execution.
+            if task_id.startswith('dify-'):
+                return
+            event = wecombotevent.WecomBotEvent(
+                {
+                    **body,
+                    'type': body.get('chattype', 'single'),
+                    'msgtype': 'event',
+                    'eventtype': 'template_card_event',
+                    'req_id': frame.get('headers', {}).get('req_id', ''),
+                }
+            )
+            for handler in self._message_handlers.get('template_card_event', []):
+                await handler(event)
             return
         if time.monotonic() - float(pending.get('created_at', 0.0)) > _PENDING_FORM_TTL_SECONDS:
             self._drop_pending_form_task(task_id, pending)
             await self.logger.warning(f'Pending form expired for task_id={task_id} (ws)')
             return
 
+        self._handled_form_tasks[task_id] = now
+        if len(self._handled_form_tasks) > 2048:
+            self._handled_form_tasks.pop(next(iter(self._handled_form_tasks)))
         req_id_for_update = frame.get('headers', {}).get('req_id', '')
         form_data = pending.get('form_data', {}) or {}
         selections = extract_template_card_selections(tce, form_data)
@@ -1120,6 +1153,9 @@ class WecomBotWsClient:
         self._drop_pending_form_task(task_id, pending)
 
     def _drop_pending_form_task(self, task_id: str, pending: dict) -> None:
+        self._handled_form_tasks[task_id] = time.monotonic()
+        if len(self._handled_form_tasks) > 2048:
+            self._handled_form_tasks.pop(next(iter(self._handled_form_tasks)))
         self._pending_forms_by_task.pop(task_id, None)
         msg_id = pending.get('msg_id', '')
         if msg_id:
diff --git a/src/langbot/libs/wecom_customer_service_api/api.py b/src/langbot/libs/wecom_customer_service_api/api.py
index da3caa9fd..916cd3f39 100644
--- a/src/langbot/libs/wecom_customer_service_api/api.py
+++ b/src/langbot/libs/wecom_customer_service_api/api.py
@@ -5,6 +5,7 @@ import base64
 import binascii
 import contextvars
 import functools
+import hashlib
 import httpx
 import json
 import os
@@ -119,6 +120,10 @@ class WecomCSClient:
             'example': [],
         }
         self._http_client: httpx.AsyncClient | None = None
+        self._delivery_tasks: set[asyncio.Task] = set()
+        self._received_messages: dict[str, float] = {}
+        self._sync_lock = asyncio.Lock()
+        self._sync_cursors: dict[str, str] = {}
 
     @asynccontextmanager
     async def _http_client_context(self):
@@ -127,6 +132,10 @@ class WecomCSClient:
         yield self._http_client
 
     async def close(self) -> None:
+        for task in self._delivery_tasks:
+            task.cancel()
+        if self._delivery_tasks:
+            await asyncio.gather(*self._delivery_tasks, return_exceptions=True)
         if self._http_client is not None:
             await self._http_client.aclose()
             self._http_client = None
@@ -170,47 +179,46 @@ class WecomCSClient:
             else:
                 raise Exception(f'未获取access token: {data}')
 
-    @_bounded_token_retry
     async def get_detailed_message_list(self, xml_msg: str):
-        # 在本方法中解析消息，并且获得消息的具体内容
-        if isinstance(xml_msg, bytes):
-            xml_msg = xml_msg.decode('utf-8')
+        """Drain all callback pages, including non-message events and empty pages."""
         root = await asyncio.to_thread(ET.fromstring, xml_msg)
-        token = root.find('Token').text
-        open_kfid = root.find('OpenKfId').text
-
-        # if open_kfid in self.openkfid_list:
-        #     return None
-        # else:
-        #     self.openkfid_list.append(open_kfid)
-
-        if not await self.check_access_token():
-            self.access_token = await self.get_access_token(self.secret)
-
-        url = self.base_url + '/kf/sync_msg?access_token=' + self.access_token
-        async with self._http_client_context() as client:
-            params = {
-                'token': token,
-                'voice_format': 0,
-                'open_kfid': open_kfid,
-            }
-            response = await client.post(url, json=params)
-            data = await httpclient.parse_json_response(response)
-            if data['errcode'] == 40014 or data['errcode'] == 42001:
-                self.access_token = await self.get_access_token(self.secret)
-                return await self.get_detailed_message_list(xml_msg)
-            if data['errcode'] != 0:
-                raise Exception('Failed to get message')
-
-            last_msg_data = data['msg_list'][-1]
-            open_kfid = last_msg_data.get('open_kfid')
-            # 进行获取图片操作
-            if last_msg_data.get('msgtype') == 'image':
-                media_id = last_msg_data.get('image').get('media_id')
-                picurl = await self.get_pic_url(media_id)
-                last_msg_data['picurl'] = picurl
-            # await self.change_service_status(userid=external_userid,openkfid=open_kfid,servicer=servicer)
-            return last_msg_data
+        token = root.findtext('Token')
+        open_kfid = root.findtext('OpenKfId')
+        if not token or not open_kfid:
+            raise ValueError('WeCom customer-service callback is missing Token or OpenKfId')
+        async with self._sync_lock:
+            cursor = self._sync_cursors.get(open_kfid)
+            messages = []
+            refreshed = False
+            while True:
+                if not await self.check_access_token():
+                    self.access_token = await self.get_access_token(self.secret)
+                params = {'token': token, 'voice_format': 0, 'open_kfid': open_kfid, 'limit': 1000}
+                if cursor:
+                    params['cursor'] = cursor
+                async with self._http_client_context() as client:
+                    response = await client.post(
+                        self.base_url + '/kf/sync_msg',
+                        params={'access_token': self.access_token},
+                        json=params,
+                    )
+                    data = await httpclient.parse_json_response(response)
+                if data.get('errcode') in (40014, 42001) and not refreshed:
+                    self.access_token = await self.get_access_token(self.secret)
+                    refreshed = True
+                    continue
+                if data.get('errcode') != 0:
+                    raise RuntimeError(f'WeCom sync_msg failed (errcode={data.get("errcode")})')
+                messages.extend(data.get('msg_list') or [])
+                next_cursor = data.get('next_cursor')
+                if data.get('has_more'):
+                    if not next_cursor or next_cursor == cursor:
+                        raise RuntimeError('WeCom sync_msg returned a non-advancing cursor')
+                    cursor = next_cursor
+                    continue
+                if next_cursor:
+                    self._sync_cursors[open_kfid] = next_cursor
+                return messages
 
     @_bounded_token_retry
     async def change_service_status(self, userid: str, openkfid: str, servicer: str):
@@ -292,9 +300,36 @@ class WecomCSClient:
                 return await self.send_text_msg(open_kfid, external_userid, msgid, content)
             if data['errcode'] != 0:
                 await self.logger.error(f'发送消息失败：{data}')
+                if data['errcode'] == 95002:
+                    raise ValueError(
+                        'WeCom CS rejected the reply (95002): the customer must send a message first, '
+                        'and ordinary replies must be sent within 48 hours. Entering a session alone '
+                        'does not open this window; welcome replies require the event reply API.'
+                    )
                 raise Exception(f'Failed to send message: {data}')
             return data
 
+    @_bounded_token_retry
+    async def send_event_text(self, code: str, msgid: str, content: str):
+        """Send a single event response using its short-lived, one-time code."""
+        if not await self.check_access_token():
+            self.access_token = await self.get_access_token(self.secret)
+        url = f'{self.base_url}/kf/send_msg_on_event?access_token={self.access_token}'
+        payload = {'code': code, 'msgid': msgid, 'msgtype': 'text', 'text': {'content': content}}
+        async with self._http_client_context() as client:
+            response = await client.post(url, json=payload)
+            data = await httpclient.parse_json_response(response)
+        if data.get('errcode') in (40014, 42001):
+            self.access_token = await self.get_access_token(self.secret)
+            return await self.send_event_text(code, msgid, content)
+        if data.get('errcode') != 0:
+            raise ValueError(
+                f'WeCom CS event reply failed (errcode={data.get("errcode")}): {data.get("errmsg")}. '
+                'The event code is single-use and must match the current session state; '
+                'welcome and session-end codes expire after 20 seconds. Do not retry an expired or used code.'
+            )
+        return data
+
     @_bounded_token_retry
     async def send_image_msg(self, open_kfid: str, external_userid: str, msgid: str, media_id: str):
         if not await self.check_access_token():
@@ -375,12 +410,11 @@ class WecomCSClient:
                 if ret != 0:
                     raise Exception(f'消息解密失败，错误码: {ret}')
 
-                # 解析消息并处理
-                message_data = await self.get_detailed_message_list(xml_msg)
-                if message_data is not None:
-                    event = WecomCSEvent.from_payload(message_data)
-                    if event:
-                        await self._handle_message(event)
+                # Acknowledge after verification; API fetching and processing must
+                # survive a callback disconnect.
+                task = asyncio.create_task(self._process_callback(xml_msg))
+                self._delivery_tasks.add(task)
+                task.add_done_callback(self._delivery_tasks.discard)
 
                 return 'success'
         except Exception as e:
@@ -390,6 +424,49 @@ class WecomCSClient:
                 traceback.print_exc()
             return f'Error processing request: {str(e)}', 400
 
+    async def _process_callback(self, xml_msg):
+        try:
+            messages = await self.get_detailed_message_list(xml_msg)
+            for message_data in messages:
+                try:
+                    await self._process_synced_message(message_data)
+                except Exception:
+                    if self.logger:
+                        await self.logger.error(f'WeCom customer-service event failed: {traceback.format_exc()}')
+
+        except Exception:
+            if self.logger:
+                await self.logger.error(f'WeCom customer-service delivery failed: {traceback.format_exc()}')
+
+    async def _process_synced_message(self, message_data):
+        # Agent/API and human replies must not become new customer input.
+        if message_data.get('msgtype') != 'event' and message_data.get('origin', 3) != 3:
+            return
+        if message_data.get('msgtype') == 'image':
+            media_id = (message_data.get('image') or {}).get('media_id')
+            if media_id:
+                message_data['picurl'] = await self.get_pic_url(media_id)
+        event = WecomCSEvent.from_payload(message_data)
+        if event:
+            # A callback disconnect must not cancel processor execution.
+            # Retries can return the same message from sync_msg.
+            now = time.monotonic()
+            self._received_messages = {
+                key: seen for key, seen in self._received_messages.items() if now - seen < 4 * 86400
+            }
+            message_id = (
+                message_data.get('msgid')
+                or hashlib.sha256(
+                    json.dumps(message_data, sort_keys=True, ensure_ascii=False, separators=(',', ':')).encode()
+                ).hexdigest()
+            )
+            if not message_id or message_id not in self._received_messages:
+                if message_id:
+                    if len(self._received_messages) >= 4096:
+                        self._received_messages.pop(next(iter(self._received_messages)))
+                    self._received_messages[message_id] = now
+                await self._handle_message(event)
+
     async def run_task(self, host: str, port: int, *args, **kwargs):
         """
         启动 Quart 应用。
diff --git a/src/langbot/pkg/agent/runner/config_schema.py b/src/langbot/pkg/agent/runner/config_schema.py
index aa7d13f7e..1fad3193f 100644
--- a/src/langbot/pkg/agent/runner/config_schema.py
+++ b/src/langbot/pkg/agent/runner/config_schema.py
@@ -16,6 +16,65 @@ KB_SELECTOR_TYPES = {'knowledge-base-multi-selector'}
 PROMPT_EDITOR_TYPES = {'prompt-editor'}
 NONE_SENTINELS = {'', '__none__', '__none'}
 
+# Selector field types that pick a Host resource. The Runner config form is
+# plugin-declared and open-ended (DynamicFormItemSchema allows extra types), so
+# this is a registry, not a fixed list: a new selector type is one line here and
+# is then reported by telemetry without touching the extraction call sites.
+RESOURCE_SELECTOR_TYPES: dict[str, str] = {
+    'knowledge-base-multi-selector': 'knowledge_bases',
+    'tool-multi-selector': 'tools',
+    'mcp-server-multi-selector': 'mcp_servers',
+    'skill-multi-selector': 'skills',
+}
+
+
+def declared_resource_selectors(descriptor: RunnerDescriptor | None) -> dict[str, list[str]]:
+    """Map each selected resource kind to the schema field names that select it.
+
+    Only the descriptor is read, so this reports what a Runner *can* select even
+    when the current configuration leaves the field empty.
+    """
+    declared: dict[str, list[str]] = {}
+    if descriptor is None:
+        return declared
+    for item in descriptor.config_schema or []:
+        if not isinstance(item, dict):
+            continue
+        kind = RESOURCE_SELECTOR_TYPES.get(normalize_schema_item_type(item.get('type')))
+        field_name = item.get('name')
+        if not kind or not isinstance(field_name, str) or not field_name:
+            continue
+        declared.setdefault(kind, [])
+        if field_name not in declared[kind]:
+            declared[kind].append(field_name)
+    return declared
+
+
+def extract_selected_resources(
+    descriptor: RunnerDescriptor | None,
+    runner_config: dict[str, typing.Any] | None,
+) -> dict[str, list[str]]:
+    """Extract the resource ids selected through schema-declared selector fields.
+
+    Field names are never hard-coded: they come from the Runner's own config
+    schema, so a form that renames or adds a selector keeps working.
+    """
+    selected: dict[str, list[str]] = {}
+    if not isinstance(runner_config, dict):
+        return selected
+    for kind, field_names in declared_resource_selectors(descriptor).items():
+        values: list[str] = []
+        for field_name in field_names:
+            value = runner_config.get(field_name)
+            if isinstance(value, str):
+                if value not in NONE_SENTINELS:
+                    values.append(value)
+            elif isinstance(value, (list, tuple)):
+                values.extend(item for item in value if isinstance(item, str) and item not in NONE_SENTINELS)
+        if values:
+            selected[kind] = list(dict.fromkeys(values))
+    return selected
+
 
 def normalize_schema_item_type(item_type: typing.Any) -> typing.Any:
     """Normalize legacy/frontend DynamicForm aliases to protocol field types."""
diff --git a/src/langbot/pkg/agent/runner/context_reset.py b/src/langbot/pkg/agent/runner/context_reset.py
new file mode 100644
index 000000000..73d6b3fd4
--- /dev/null
+++ b/src/langbot/pkg/agent/runner/context_reset.py
@@ -0,0 +1,46 @@
+"""Scoped context reset markers, independent of retained monitoring history."""
+
+import hashlib
+import json
+
+import sqlalchemy as sa
+
+from ...entity.persistence.runner_state import RunnerState
+
+HOST_RESET_RUNNER = '__langbot_context_reset__'
+
+
+def reset_scope(workspace, bot, conversation):
+    return 'host-reset:' + hashlib.sha256(json.dumps([workspace, bot, conversation]).encode()).hexdigest()
+
+
+async def get_reset_generation(engine, workspace, bot, conversation):
+    if not all((workspace, bot, conversation)):
+        return None
+    async with engine.connect() as connection:
+        value = (
+            await connection.execute(
+                sa.select(RunnerState.value_json).where(
+                    RunnerState.scope_key == reset_scope(workspace, bot, conversation),
+                    RunnerState.state_key == 'generation',
+                    RunnerState.runner_id == HOST_RESET_RUNNER,
+                )
+            )
+        ).scalar_one_or_none()
+    return json.loads(value) if value else None
+
+
+def active_history_condition(transcript):
+    cutoff = (
+        sa.select(sa.func.max(RunnerState.created_at))
+        .where(
+            RunnerState.runner_id == HOST_RESET_RUNNER,
+            RunnerState.state_key == 'generation',
+            RunnerState.workspace_id == transcript.workspace_id,
+            RunnerState.bot_id == transcript.bot_id,
+            RunnerState.conversation_id == transcript.conversation_id,
+        )
+        .correlate(transcript)
+        .scalar_subquery()
+    )
+    return sa.or_(cutoff.is_(None), transcript.created_at > cutoff)
diff --git a/src/langbot/pkg/agent/runner/event_log_store.py b/src/langbot/pkg/agent/runner/event_log_store.py
index 1dbd8a8b2..10b3a1860 100644
--- a/src/langbot/pkg/agent/runner/event_log_store.py
+++ b/src/langbot/pkg/agent/runner/event_log_store.py
@@ -128,7 +128,33 @@ class EventLogStore:
                 created_at=as_naive_utc(_utc_now()),
             )
             session.add(event)
-            await session.commit()
+            try:
+                await session.commit()
+            except sqlalchemy.exc.IntegrityError:
+                await session.rollback()
+                existing = (
+                    (await session.execute(sqlalchemy.select(EventLog).where(EventLog.event_id == event_id)))
+                    .scalars()
+                    .first()
+                )
+                if existing is None or existing.workspace_id != workspace_id or existing.bot_id != bot_id:
+                    raise
+                # Complete the ingress snapshot once a runner claims it. A
+                # second subscriber must not replace another run's identity.
+                for field, value in {
+                    'run_id': run_id,
+                    'runner_id': runner_id,
+                    'conversation_id': conversation_id,
+                    'thread_id': thread_id,
+                    'actor_id': actor_id,
+                    'actor_name': actor_name,
+                }.items():
+                    if getattr(existing, field) is None and value is not None:
+                        setattr(existing, field, value)
+                if input_json and ('contents' in input_json or 'attachments' in input_json):
+                    existing.input_json = json.dumps(input_json)
+                    existing.input_summary = input_summary
+                await session.commit()
 
         return event_id
 
diff --git a/src/langbot/pkg/agent/runner/interaction_manager.py b/src/langbot/pkg/agent/runner/interaction_manager.py
index 99efb0fad..efa09bb5d 100644
--- a/src/langbot/pkg/agent/runner/interaction_manager.py
+++ b/src/langbot/pkg/agent/runner/interaction_manager.py
@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import copy
 import json
 import time
@@ -94,7 +92,6 @@ class InteractionManager:
         self.ap = ap
         self.store = store or InteractionStore(ap.persistence_mgr.get_db_engine())
 
-    @diagnostics.observe('interaction', 'interaction.request', source='agent', stage='execute')
     async def handle_result(
         self,
         *,
@@ -208,7 +205,6 @@ class InteractionManager:
         except Exception as exc:
             await self.store.mark_delivery_failed(run_id, request.interaction_id, str(exc))
             raise
-        diagnostics.set_outcome('waiting', reason_code='waiting')
         return True
 
     @staticmethod
@@ -255,14 +251,12 @@ class InteractionManager:
         except pydantic.ValidationError:
             return
 
-    @diagnostics.observe('interaction', 'interaction.acknowledge', source='platform', stage='ack')
     async def acknowledge_submission(self, record: dict[str, typing.Any], adapter: typing.Any) -> None:
         """Best-effort transition of submitted controls into a read-only state."""
         delivery_result = record.get('delivery_result')
         if not isinstance(delivery_result, dict) or not self._supports_platform_api(
             adapter, INTERACTION_ACKNOWLEDGE_API
         ):
-            diagnostics.set_outcome('skipped')
             return
         try:
             delivery_result = await adapter.call_platform_api(
@@ -280,8 +274,6 @@ class InteractionManager:
                     delivery_result,
                 )
         except Exception as exc:
-            diagnostics.set_outcome('failed', reason_code='response_error')
-            diagnostics.annotate(error=exc)
             self._warning(f'Failed to acknowledge interaction submission: {exc}')
 
     async def _find_update_target(
@@ -332,7 +324,6 @@ class InteractionManager:
                     f'interaction field {field.id} has duplicate option values',
                 )
 
-    @diagnostics.observe('interaction', 'interaction.consume', source='platform', stage='dispatch')
     async def consume_callback(
         self,
         *,
diff --git a/src/langbot/pkg/agent/runner/orchestrator.py b/src/langbot/pkg/agent/runner/orchestrator.py
index de6327a40..f650f686f 100644
--- a/src/langbot/pkg/agent/runner/orchestrator.py
+++ b/src/langbot/pkg/agent/runner/orchestrator.py
@@ -3,15 +3,16 @@
 from __future__ import annotations
 
 import time
+import asyncio
 import contextlib
+import copy
 import typing
 
 from langbot_plugin.api.entities.builtin.provider import message as provider_message
+from langbot_plugin.api.entities.builtin.provider import prompt as provider_prompt
 from langbot_plugin.api.entities.builtin.pipeline import query as pipeline_query
 
 from langbot_plugin.entities.io.actions.enums import PluginToRuntimeAction
-
-from ...telemetry import diagnostics as diagnostics
 from .reply_stream import ReplyStreamSession
 from ...core import app
 from ...api.http.context import ExecutionContext
@@ -19,6 +20,7 @@ from ...pipeline.pool import get_query_execution_context
 from .binding_resolver import AgentBindingResolver
 from .context_builder import RunnerContextBuilder, RunnerContextPayload
 from .descriptor import RunnerDescriptor
+from .config_schema import extract_prompt_config
 from .execution_context import (
     append_mcp_resource_context_to_event,
     build_mcp_resource_context_addition,
@@ -38,9 +40,11 @@ from .run_journal import AgentRunJournal
 from .session_registry import AgentRunSessionRegistry, get_session_registry
 from .state_scope import build_state_context
 from ...provider.tools.loaders import skill as skill_loader
-
-
-ACTIVATED_SKILL_NAMES_STATE_KEY = 'host.activated_skills'
+from ...telemetry import resources as telemetry_resources
+from ...telemetry import trace as trace_mod
+from ...telemetry.execution import bind_trace as bind_execution_trace
+from ...telemetry.execution import close_trace as close_execution_trace
+from ...telemetry.execution import record as record_execution
 
 
 class AgentRunOrchestrator:
@@ -80,7 +84,6 @@ class AgentRunOrchestrator:
         self.journal = AgentRunJournal(ap)
         self._session_registry = get_session_registry()
 
-    @diagnostics.observe('run', 'runner.run', source='agent', stage='prepare')
     async def run(
         self,
         event: AgentEventEnvelope,
@@ -106,14 +109,18 @@ class AgentRunOrchestrator:
             runner_id,
             bound_plugins,
         )
-
-        diagnostics.runner_metadata(self.ap, descriptor, binding.processor_type)
         usage = 'event' if binding.processor_type == 'event_processor' else 'agent'
         if usage not in descriptor.usages:
             raise ValueError(f'The selected Runner does not support {usage} usage')
 
         if execution_query is None:
             execution_query = build_execution_query(event, [])
+            # Event runs bypass pipeline prompt preparation, just like debug runs.
+            prompt_config = extract_prompt_config(descriptor, binding.runner_config, [])
+            execution_query.prompt = provider_prompt.Prompt(
+                name='default',
+                messages=[provider_message.Message(**item) for item in copy.deepcopy(prompt_config)],
+            )
             # Synthetic events must expose the same trusted scope as pipeline queries.
             for field_name in ('instance_uuid', 'workspace_uuid', 'placement_generation', 'query_uuid'):
                 object.__setattr__(execution_query, field_name, getattr(execution_context, field_name))
@@ -142,6 +149,22 @@ class AgentRunOrchestrator:
             binding=binding,
             descriptor=descriptor,
         )
+        # Telemetry: report what this run was granted (never what it used). The
+        # materialized resource list is authoritative for tools/skills/models;
+        # the Runner's own config form is authoritative for its selectors.
+        try:
+            from . import config_schema as runner_config_schema
+
+            capabilities = getattr(descriptor, 'capabilities', None)
+            telemetry_resources.note_agent(
+                runner_id=descriptor.id,
+                resources=resources,
+                selectors=runner_config_schema.extract_selected_resources(descriptor, binding.runner_config),
+                capabilities=capabilities.model_dump() if hasattr(capabilities, 'model_dump') else None,
+                tools_all=getattr(binding.resource_policy, 'allow_all_tools', None),
+            )
+        except Exception:
+            pass
 
         context = await self.context_builder.build_context_from_event(
             event=execution_event,
@@ -153,16 +176,12 @@ class AgentRunOrchestrator:
         authorized_skill_names = [
             str(skill['skill_name']) for skill in resources.get('skills', []) if skill.get('skill_name')
         ]
+        # Every invocation owns a new run; conversation state is not a recovery snapshot.
         prepare_execution_query(execution_query, event, authorized_skill_names)
+        execution_query.variables.pop(skill_loader.ACTIVATED_SKILLS_KEY, None)
 
         session_query_id = None
         if adapter_context:
-            if execution_query is not None:
-                await skill_loader.restore_activated_skills_from_state(
-                    self.ap,
-                    execution_query,
-                    context.get('state', {}),
-                )
             session_query_id = adapter_context.get('query_id')
             if execution_query is not None or session_query_id is not None:
                 context['context']['available_apis']['prompt_get'] = True
@@ -178,7 +197,21 @@ class AgentRunOrchestrator:
 
         state_context = build_state_context(event, binding, descriptor)
         run_id = context['run_id']
-        diagnostics.annotate(run_id=run_id, stage='execute')
+
+        # Monitoring attributes every call to the processor that ran it and to the
+        # record that owns it; a Runner has to claim both itself, since no
+        # Pipeline did. Runs before the Runner is invoked and before the run
+        # context is handed to it, so its own variables stay private.
+        from ...pipeline import monitoring_helper
+
+        await monitoring_helper.prepare_monitoring_identity(
+            self.ap,
+            execution_query,
+            processor_type=binding.processor_type,
+            processor_id=binding.processor_id,
+            runner_id=binding.runner_id,
+            execution_record_id=run_id,
+        )
         context['context']['available_apis']['reply_stream'] = hasattr(PluginToRuntimeAction, 'REPLY_STREAM') and any(
             tool.get('tool_name') == 'event_reply' and tool.get('tool_type') == 'platform'
             for tool in resources.get('tools', [])
@@ -189,7 +222,6 @@ class AgentRunOrchestrator:
             source=(adapter_context or {}).get('_platform_event')
             or getattr((adapter_context or {}).get('_query'), 'message_event', None),
         )
-        reply_streams.diagnostics = getattr(self.ap, 'diagnostics', None)
         available_apis = context.get('context', {}).get('available_apis')
         run_authorization = {
             'runner_id': descriptor.id,
@@ -216,8 +248,18 @@ class AgentRunOrchestrator:
         terminal_status: str | None = None
         terminal_reason: str | None = None
         terminal_usage: dict[str, typing.Any] | None = None
+        execution_outcome = 'unknown'
+        execution_error = ''
+        # A run reached without a platform ingress (WebUI debug, service API)
+        # owns its own chain; a run inside an ingress reuses that chain.
+        trace_binding: trace_mod.TraceBinding | None = None
+        run_token: typing.Any = None
+        # Workflow step node for this run; '' until the step scope is opened.
+        node = ''
 
         try:
+            trace_binding = bind_execution_trace(self.ap, run_id)
+            run_token = trace_mod.set_run(run_id)
             await self.journal.create_run(
                 event=event,
                 binding=binding,
@@ -225,6 +267,11 @@ class AgentRunOrchestrator:
                 context=context,
                 authorization=run_authorization,
             )
+            monitoring_id = (execution_query.variables or {}).get('_monitoring_message_id')
+            if monitoring_id and monitoring_id != run_id:
+                await self.ap.monitoring_service.link_execution_message(
+                    get_query_execution_context(execution_query), monitoring_id, run_id, event.event_id
+                )
             await self._session_registry.register(
                 run_id=run_id,
                 runner_id=descriptor.id,
@@ -258,153 +305,173 @@ class AgentRunOrchestrator:
                     event_log_id=event_log_id,
                 )
 
-            async with contextlib.aclosing(self.invoker.invoke(descriptor, context)) as results:
-                async for result_dict in results:
-                    result_dict = dict(result_dict)
-                    sequence = result_dict.get('sequence')
-                    if sequence is not None:
-                        try:
-                            sequence_int = int(sequence)
-                        except (TypeError, ValueError):
-                            self.ap.logger.warning(
-                                f'Runner {descriptor.id} returned invalid result sequence: {sequence}'
-                            )
-                            sequence_int = last_sequence + 1
-                            result_dict['sequence'] = sequence_int
-                        else:
-                            if sequence_int in seen_sequences:
+            with trace_mod.stage_scope() as node:
+                async with contextlib.aclosing(self.invoker.invoke(descriptor, context)) as results:
+                    async for result_dict in results:
+                        result_dict = dict(result_dict)
+                        sequence = result_dict.get('sequence')
+                        if sequence is not None:
+                            try:
+                                sequence_int = int(sequence)
+                            except (TypeError, ValueError):
                                 self.ap.logger.warning(
-                                    f'Runner {descriptor.id} returned duplicate result sequence '
-                                    f'{sequence_int} for run {run_id}; dropping duplicate'
-                                )
-                                continue
-                            if sequence_int <= 0:
-                                self.ap.logger.warning(
-                                    f'Runner {descriptor.id} returned non-positive result sequence '
-                                    f'{sequence_int} for run {run_id}'
+                                    f'Runner {descriptor.id} returned invalid result sequence: {sequence}'
                                 )
                                 sequence_int = last_sequence + 1
                                 result_dict['sequence'] = sequence_int
-                            elif last_sequence and sequence_int != last_sequence + 1:
-                                self.ap.logger.warning(
-                                    f'Runner {descriptor.id} result sequence gap or out-of-order '
-                                    f'for run {run_id}: previous={last_sequence}, current={sequence_int}'
-                                )
+                            else:
+                                if sequence_int in seen_sequences:
+                                    self.ap.logger.warning(
+                                        f'Runner {descriptor.id} returned duplicate result sequence '
+                                        f'{sequence_int} for run {run_id}; dropping duplicate'
+                                    )
+                                    continue
+                                if sequence_int <= 0:
+                                    self.ap.logger.warning(
+                                        f'Runner {descriptor.id} returned non-positive result sequence '
+                                        f'{sequence_int} for run {run_id}'
+                                    )
+                                    sequence_int = last_sequence + 1
+                                    result_dict['sequence'] = sequence_int
+                                elif last_sequence and sequence_int != last_sequence + 1:
+                                    self.ap.logger.warning(
+                                        f'Runner {descriptor.id} result sequence gap or out-of-order '
+                                        f'for run {run_id}: previous={last_sequence}, current={sequence_int}'
+                                    )
+                                seen_sequences.add(sequence_int)
+                                last_sequence = max(last_sequence, sequence_int)
+                        else:
+                            sequence_int = last_sequence + 1
+                            result_dict['sequence'] = sequence_int
                             seen_sequences.add(sequence_int)
-                            last_sequence = max(last_sequence, sequence_int)
-                    else:
-                        sequence_int = last_sequence + 1
-                        result_dict['sequence'] = sequence_int
-                        seen_sequences.add(sequence_int)
-                        last_sequence = sequence_int
+                            last_sequence = sequence_int
 
-                    result_type = result_dict.get('type')
-                    if result_type and not self.result_normalizer.validate_payload(
-                        result_type,
-                        result_dict.get('data', {}),
-                        descriptor,
-                    ):
-                        continue
-
-                    await self.journal.append_run_result(
-                        result_dict=result_dict,
-                        run_id=run_id,
-                        sequence=sequence_int,
-                    )
-
-                    # Trusted Host observers receive validated events before message-only normalization.
-                    result_observer = (adapter_context or {}).get('_result_observer')
-                    if result_observer is not None:
-                        await result_observer(result_dict)
-
-                    if result_type == 'state.updated':
-                        await self.journal.handle_state_updated_event(
-                            result_dict,
-                            event,
-                            binding,
+                        result_type = result_dict.get('type')
+                        if result_type and not self.result_normalizer.validate_payload(
+                            result_type,
+                            result_dict.get('data', {}),
                             descriptor,
-                            run_id=run_id,
-                        )
-                        await self.result_normalizer.normalize(result_dict, descriptor)
-                        continue
-
-                    if result_type == 'action.requested':
-                        if await self.interaction_manager.handle_result(
-                            result_dict=result_dict,
-                            event=event,
-                            binding=binding,
-                            descriptor=descriptor,
-                            run_id=run_id,
-                            adapter_context=adapter_context,
                         ):
                             continue
 
-                    if result_type == 'run.completed':
-                        terminal_status = 'completed'
-                        terminal_reason = (
-                            result_dict.get('data', {}).get('finish_reason')
-                            if isinstance(result_dict.get('data'), dict)
-                            else None
-                        )
-                        usage = result_dict.get('usage')
-                        if isinstance(usage, dict):
-                            terminal_usage = usage
-                    elif result_type == 'run.failed':
-                        terminal_status = 'failed'
-                        data = result_dict.get('data') if isinstance(result_dict.get('data'), dict) else {}
-                        terminal_reason = data.get('error') or data.get('code')
-                        usage = result_dict.get('usage')
-                        if isinstance(usage, dict):
-                            terminal_usage = usage
-
-                    has_completed_message = result_type == 'message.completed' or (
-                        result_type == 'run.completed'
-                        and isinstance(result_dict.get('data'), dict)
-                        and bool(result_dict['data'].get('message'))
-                    )
-                    if has_completed_message and event.conversation_id and not assistant_transcript_written:
-                        await self.journal.write_assistant_transcript(
+                        journal_started = time.monotonic()
+                        await self.journal.append_run_result(
                             result_dict=result_dict,
-                            event=event,
                             run_id=run_id,
-                            runner_id=descriptor.id,
+                            sequence=sequence_int,
                         )
-                        assistant_transcript_written = True
+                        journal_ms = (time.monotonic() - journal_started) * 1000
 
-                    result = await self.result_normalizer.normalize(result_dict, descriptor)
-                    file_ids = result_dict.get('data', {}).get('file_ids', [])
-                    if file_ids:
-                        if result_type != 'message.completed' or result is None:
-                            raise ValueError('Files must be attached to a completed message')
-                        if binding.processor_type != 'pipeline':
-                            raise ValueError('Agent files must be sent explicitly through the reply API')
-                        from ...box.runner import exported_message, binding_for
+                        # Trusted Host observers receive validated events before message-only normalization.
+                        result_observer = (adapter_context or {}).get('_result_observer')
+                        observer_started = time.monotonic()
+                        if result_observer is not None:
+                            await result_observer(result_dict)
+                        observer_ms = (time.monotonic() - observer_started) * 1000
+                        if journal_ms + observer_ms >= 100:
+                            self.ap.logger.debug(
+                                f'Slow runner event consumer run={run_id} type={result_type} '
+                                f'journal_ms={journal_ms:.1f} observer_ms={observer_ms:.1f}'
+                            )
 
-                        async with binding_for(execution_query).lock:
-                            result.attachments = exported_message(execution_query, file_ids, consume=True)
+                        if result_type == 'state.updated':
+                            await self.journal.handle_state_updated_event(
+                                result_dict,
+                                event,
+                                binding,
+                                descriptor,
+                                run_id=run_id,
+                            )
+                            await self.result_normalizer.normalize(result_dict, descriptor)
+                            continue
 
-                    if result is not None:
-                        yield result
+                        if result_type == 'action.requested':
+                            if await self.interaction_manager.handle_result(
+                                result_dict=result_dict,
+                                event=event,
+                                binding=binding,
+                                descriptor=descriptor,
+                                run_id=run_id,
+                                adapter_context=adapter_context,
+                            ):
+                                continue
 
-                    run_snapshot = await self.journal.get_run(run_id)
-                    if run_snapshot and run_snapshot.get('cancel_requested_at') is not None:
-                        terminal_status = 'cancelled'
-                        terminal_reason = run_snapshot.get('status_reason') or 'cancel_requested'
-                        break
-            diagnostics.set_outcome(
-                {'completed': 'succeeded', 'failed': 'failed', 'cancelled': 'cancelled'}.get(
-                    terminal_status, 'succeeded'
-                ),
-                reason_code='runner_failed' if terminal_status == 'failed' else '',
-            )
+                        if result_type == 'run.completed':
+                            terminal_status = 'completed'
+                            terminal_reason = (
+                                result_dict.get('data', {}).get('finish_reason')
+                                if isinstance(result_dict.get('data'), dict)
+                                else None
+                            )
+                            usage = result_dict.get('usage')
+                            if isinstance(usage, dict):
+                                terminal_usage = usage
+                        elif result_type == 'run.failed':
+                            terminal_status = 'failed'
+                            data = result_dict.get('data') if isinstance(result_dict.get('data'), dict) else {}
+                            terminal_reason = data.get('error') or data.get('code')
+                            usage = result_dict.get('usage')
+                            if isinstance(usage, dict):
+                                terminal_usage = usage
+
+                        has_completed_message = result_type == 'message.completed' or (
+                            result_type == 'run.completed'
+                            and isinstance(result_dict.get('data'), dict)
+                            and bool(result_dict['data'].get('message'))
+                        )
+                        if has_completed_message and event.conversation_id and not assistant_transcript_written:
+                            await self.journal.write_assistant_transcript(
+                                result_dict=result_dict,
+                                event=event,
+                                run_id=run_id,
+                                runner_id=descriptor.id,
+                            )
+                            assistant_transcript_written = True
+
+                        result = await self.result_normalizer.normalize(result_dict, descriptor)
+                        file_ids = result_dict.get('data', {}).get('file_ids', [])
+                        if file_ids:
+                            if result_type != 'message.completed' or result is None:
+                                raise ValueError('Files must be attached to a completed message')
+                            if binding.processor_type != 'pipeline':
+                                raise ValueError('Agent files must be sent explicitly through the reply API')
+                            from ...box.runner import exported_message, binding_for
+
+                            async with binding_for(execution_query).lock:
+                                result.attachments = exported_message(execution_query, file_ids, consume=True)
+
+                        if result is not None:
+                            yield result
+
+                        run_snapshot = await self.journal.get_run(run_id)
+                        if run_snapshot and run_snapshot.get('cancel_requested_at') is not None:
+                            terminal_status = 'cancelled'
+                            terminal_reason = run_snapshot.get('status_reason') or 'cancel_requested'
+                            break
             await self.journal.finalize_run(
                 run_id=run_id,
                 status=terminal_status or 'completed',
                 status_reason=terminal_reason,
                 usage=terminal_usage,
             )
+            execution_outcome = {'completed': 'success', 'failed': 'failed', 'cancelled': 'cancelled'}.get(
+                terminal_status or '', 'unknown'
+            )
+            execution_error = '' if execution_outcome == 'success' else (terminal_reason or execution_outcome)
+        except (asyncio.CancelledError, GeneratorExit):
+            execution_outcome = 'cancelled'
+            execution_error = 'cancelled'
+            await self.journal.finalize_run(
+                run_id=run_id,
+                status='cancelled',
+                status_reason='consumer_cancelled',
+                usage=terminal_usage,
+            )
+            raise
         except Exception as exc:
-            diagnostics.set_outcome('timeout' if self._is_deadline_exhausted(context) else 'failed')
+            execution_outcome = 'timeout' if self._is_deadline_exhausted(context) else 'failed'
+            # A chain that broke must always carry a reason, even for a bare exception.
+            execution_error = str(exc) or type(exc).__name__
             failed_usage = terminal_usage
             await self.journal.finalize_run(
                 run_id=run_id,
@@ -414,6 +481,21 @@ class AgentRunOrchestrator:
             )
             raise
         finally:
+            record_execution(
+                self.ap,
+                execution_context,
+                family='runner',
+                operation='execute',
+                mode=binding.processor_type,
+                runner=descriptor.id,
+                outcome=execution_outcome,
+                synthetic=event.source == 'webui',
+                error=execution_error,
+                node=node,
+            )
+            trace_mod.reset_run(run_token)
+            if trace_binding is not None and trace_mod.unbind_root(trace_binding):
+                close_execution_trace(self.ap, trace_binding.state, 'runner_done')
             binding_box = getattr(execution_query, '_box_binding', None)
             if binding_box is not None and binding_box.run_id == run_id:
                 object.__delattr__(execution_query, '_box_binding')
@@ -433,7 +515,6 @@ class AgentRunOrchestrator:
                         exc_info=True,
                     )
 
-    @diagnostics.observe('lifecycle', 'runner.query_prepare', source='pipeline', stage='prepare')
     async def run_from_query(
         self,
         query: pipeline_query.Query,
diff --git a/src/langbot/pkg/agent/runner/persistent_state_store.py b/src/langbot/pkg/agent/runner/persistent_state_store.py
index 7477838b8..4665f127a 100644
--- a/src/langbot/pkg/agent/runner/persistent_state_store.py
+++ b/src/langbot/pkg/agent/runner/persistent_state_store.py
@@ -191,7 +191,12 @@ class PersistentStateStore:
         # Seed external.conversation_id from event.conversation_id if not set
         if self._check_scope_enabled('conversation', binding) and event.conversation_id:
             if 'external.conversation_id' not in snapshot['conversation']:
-                snapshot['conversation']['external.conversation_id'] = event.conversation_id
+                from .context_reset import get_reset_generation
+
+                generation = await get_reset_generation(
+                    self._db_engine, event.workspace_id, event.bot_id, event.conversation_id
+                )
+                snapshot['conversation']['external.conversation_id'] = generation or event.conversation_id
 
         return snapshot
 
diff --git a/src/langbot/pkg/agent/runner/platform_tools.py b/src/langbot/pkg/agent/runner/platform_tools.py
index eb7347a01..5b7dc6509 100644
--- a/src/langbot/pkg/agent/runner/platform_tools.py
+++ b/src/langbot/pkg/agent/runner/platform_tools.py
@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-from ...telemetry import diagnostics
-
 import copy
 import fnmatch
 import typing
@@ -467,7 +465,11 @@ def build_platform_tool_resources(
         reason = None
         if definition is None:
             reason = 'unknown_tool'
-        elif definition.api not in supported_apis:
+        elif definition.api not in supported_apis and not (
+            definition.name == 'event_reply'
+            and {'reply_message', 'get_message'} <= supported_apis
+            and (event.delivery.reply_target or {}).get('message_id')
+        ):
             reason = 'adapter_api_unsupported'
         elif definition.scope == 'event' and not _event_matches(event.event_type, definition.event_patterns):
             reason = 'event_incompatible'
@@ -669,7 +671,6 @@ def resolve_platform_api_call(session, bot_uuid, action, params, context_tool=No
     raise ValueError(f'Platform API {action} or its target is not authorized for this run')
 
 
-@diagnostics.observe('api', 'host.platform_tool', source='agent')
 async def execute_platform_tool(
     ap: typing.Any,
     execution_context: typing.Any,
@@ -687,33 +688,131 @@ async def execute_platform_tool(
     normalized = _normalize_platform_params(definition, parameters)
     if definition.scope == 'event':
         normalized = _event_params(definition, context, normalized)
-    # This flag is frozen by the Host from the synthetic debug envelope, not tool arguments.
-    if delivery.get('surface') == 'webui' and (delivery.get('platform_capabilities') or {}).get('debug_mock') is True:
-        diagnostics.annotate(source='webui_debug', attributes={'synthetic': True})
-        result = _execute_mock_platform_tool(definition, context, normalized)
-        if message_chain is not None:
-            result['parameters']['message'] = message_chain.model_dump(mode='json')
-        return result
-    bot_id = authorization.get('bot_id')
-    if not bot_id:
-        raise ValueError('This run is not associated with a platform bot')
-    bot = await ap.platform_mgr.get_bot_by_uuid(execution_context, bot_id)
-    if bot is None:
-        raise ValueError(f'Bot {bot_id} is not running')
-    if definition.api not in set(bot.adapter.get_supported_apis() or []):
-        raise ValueError(f'Platform API {definition.api} is no longer supported by bot {bot_id}')
-    api_func = getattr(bot.adapter, definition.api, None)
-    if not callable(api_func):
-        raise ValueError(f'Platform API {definition.api} is declared but not implemented')
-    if definition.api == 'send_message':
-        normalized = {
-            'target_type': _require_string(normalized, 'target_type'),
-            'target_id': _require_string(normalized, 'target_id'),
-            'message': message_chain
-            if message_chain is not None
-            else platform_message.MessageChain([platform_message.Plain(text=_require_string(normalized, 'text'))]),
-        }
-    return await api_func(**normalized)
+    # Plugin/RPC actions run outside the ingress context, so bind the owning
+    # execution id here: nested adapter observations then join its trace.
+    from ...telemetry.execution import record, reset_execution_id, set_execution_id
+
+    execution_id = str(session.get('run_id') or '').strip()
+    execution_token = set_execution_id(execution_id)
+    try:
+        # This flag is frozen by the Host from the synthetic debug envelope, not tool arguments.
+        mock = (
+            delivery.get('surface') == 'webui'
+            and (delivery.get('platform_capabilities') or {}).get('debug_mock') is True
+        )
+        if mock:
+            outcome = 'unknown'
+            error_detail = ''
+            try:
+                result = _execute_mock_platform_tool(definition, context, normalized)
+                if message_chain is not None:
+                    result['parameters']['message'] = message_chain.model_dump(mode='json')
+                outcome = 'success'
+            except Exception as exc:
+                outcome = 'failed'
+                error_detail = str(exc)
+                raise
+            finally:
+                record(
+                    ap,
+                    execution_context,
+                    family='platform_api',
+                    operation=definition.api,
+                    mode=authorization.get('processor_type', 'none'),
+                    synthetic=True,
+                    outcome=outcome,
+                    error=error_detail,
+                    execution_id=execution_id or None,
+                )
+            return result
+        bot_id = authorization.get('bot_id')
+        if not bot_id:
+            raise ValueError('This run is not associated with a platform bot')
+        bot = await ap.platform_mgr.get_bot_by_uuid(execution_context, bot_id)
+        if bot is None:
+            raise ValueError(f'Bot {bot_id} is not running')
+        supported_apis = set(bot.adapter.get_supported_apis() or [])
+        reply_via_source = (
+            definition.name == 'event_reply'
+            and 'send_message' not in supported_apis
+            and {'reply_message', 'get_message'} <= supported_apis
+            and (delivery.get('reply_target') or {}).get('message_id')
+        )
+        if definition.api not in supported_apis and not reply_via_source:
+            raise ValueError(f'Platform API {definition.api} is no longer supported by bot {bot_id}')
+        api_func = getattr(bot.adapter, definition.api, None)
+        if reply_via_source:
+
+            async def api_func(target_type, target_id, message):
+                source = await bot.adapter.get_message(
+                    chat_type=target_type,
+                    chat_id=target_id,
+                    message_id=delivery['reply_target']['message_id'],
+                )
+                return await bot.adapter.reply_message(source, message)
+
+        # Some platforms require a one-time event credential instead of an
+        # ordinary conversation send. Only pass the Host-frozen event context.
+        event_reply_handler = getattr(bot.adapter, 'reply_to_event', None)
+        if definition.name == 'event_reply' and callable(event_reply_handler):
+
+            async def api_func(target_type, target_id, message):
+                return await event_reply_handler(
+                    context=context, target_type=target_type, target_id=target_id, message=message
+                )
+
+        if not callable(api_func):
+            raise ValueError(f'Platform API {definition.api} is declared but not implemented')
+        if definition.api == 'send_message':
+            normalized = {
+                'target_type': _require_string(normalized, 'target_type'),
+                'target_id': _require_string(normalized, 'target_id'),
+                'message': message_chain
+                if message_chain is not None
+                else platform_message.MessageChain([platform_message.Plain(text=_require_string(normalized, 'text'))]),
+            }
+        from ...telemetry.platform import processing_mode
+
+        token = processing_mode.set(authorization.get('processor_type', 'none'))
+        delivery_error = None
+        try:
+            return await api_func(**normalized)
+        except Exception as exc:
+            delivery_error = str(exc)
+            raise
+        finally:
+            processing_mode.reset(token)
+            if definition.api == 'send_message' and getattr(ap, 'monitoring_service', None) is not None:
+                try:
+                    import json
+
+                    chain = normalized.get('message')
+                    content = chain.model_dump(mode='json') if hasattr(chain, 'model_dump') else str(chain or '')
+                    await ap.monitoring_service.record_message(
+                        execution_context,
+                        bot_id=bot_id,
+                        bot_name=str(getattr(getattr(bot, 'bot_entity', None), 'name', bot_id)),
+                        pipeline_id=str(authorization.get('processor_id') or ''),
+                        pipeline_name=str(authorization.get('processor_id') or authorization.get('runner_id') or ''),
+                        message_content=json.dumps(content, ensure_ascii=False),
+                        session_id=str(authorization.get('conversation_id') or ''),
+                        role='assistant',
+                        status='error' if delivery_error else 'success',
+                        level='error' if delivery_error else 'info',
+                        run_id=execution_id,
+                        parent_message_id=execution_id,
+                        variables=json.dumps(
+                            {
+                                'delivery_error': delivery_error,
+                                'target_type': normalized.get('target_type'),
+                                'target_id': normalized.get('target_id'),
+                            }
+                        ),
+                    )
+                except Exception as exc:
+                    ap.logger.warning(f'Failed to record platform delivery: {exc}')
+    finally:
+        reset_execution_id(execution_token)
 
 
 def _execute_mock_platform_tool(
diff --git a/src/langbot/pkg/agent/runner/query_bridge.py b/src/langbot/pkg/agent/runner/query_bridge.py
index 9932bf785..84923ce17 100644
--- a/src/langbot/pkg/agent/runner/query_bridge.py
+++ b/src/langbot/pkg/agent/runner/query_bridge.py
@@ -36,7 +36,9 @@ class QueryRunBridge:
         """Build an event-first run plan from a Pipeline Query."""
         runner_id = RunnerConfigResolver.resolve_runner_id(query.pipeline_config)
         if not runner_id:
-            raise RunnerNotFoundError('no runner configured')
+            raise RunnerNotFoundError(
+                'no runner configured; legacy (pre-4.11) pipelines must be migrated with the pipeline migration assistant'
+            )
 
         event = QueryEntryAdapter.query_to_event(query)
         agent_config = QueryEntryAdapter.config_to_agent_config(query, runner_id)
diff --git a/src/langbot/pkg/agent/runner/query_entry_adapter.py b/src/langbot/pkg/agent/runner/query_entry_adapter.py
index 7999f9efd..a3ecbfe16 100644
--- a/src/langbot/pkg/agent/runner/query_entry_adapter.py
+++ b/src/langbot/pkg/agent/runner/query_entry_adapter.py
@@ -335,6 +335,8 @@ class QueryEntryAdapter:
 
         # Handle launcher_id
         launcher_id = getattr(query, 'launcher_id', None)
+        if launcher_id is not None:
+            launcher_id = str(launcher_id)
 
         # Build session_id from launcher info if available
         session_id = None
diff --git a/src/langbot/pkg/agent/runner/registry.py b/src/langbot/pkg/agent/runner/registry.py
index 91877e5e7..3ef6319cf 100644
--- a/src/langbot/pkg/agent/runner/registry.py
+++ b/src/langbot/pkg/agent/runner/registry.py
@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-from ...telemetry import diagnostics
-
 import typing
 import asyncio
 
@@ -138,9 +136,6 @@ class RunnerRegistry:
             permissions=typed_manifest.permissions,
             raw_manifest=manifest,
         )
-        manager = getattr(self.ap, 'diagnostics', None)
-        if isinstance(manager, diagnostics.DiagnosticsManager) and manager.enabled:
-            diagnostics.declare_runner(descriptor)
         return descriptor
 
     async def refresh(self, context: TenantContext) -> None:
diff --git a/src/langbot/pkg/agent/runner/reply_stream.py b/src/langbot/pkg/agent/runner/reply_stream.py
index 789038ff3..aa2f3500f 100644
--- a/src/langbot/pkg/agent/runner/reply_stream.py
+++ b/src/langbot/pkg/agent/runner/reply_stream.py
@@ -7,8 +7,6 @@ from dataclasses import dataclass, field
 from typing import Literal
 from uuid import UUID, uuid4
 
-from ...telemetry import diagnostics
-
 import pydantic
 from langbot_plugin.api.entities.builtin.platform import events, message
 from langbot_plugin.api.entities.builtin.provider import message as provider_message
@@ -56,9 +54,6 @@ class ReplyStreamSession:
         self._closed = False
         self._active: set[asyncio.Task] = set()
 
-    @diagnostics.observe(
-        'delivery', 'reply_stream.apply', source='agent', fields=lambda b: {'attributes': {'stream': True}}
-    )
     async def apply(self, request: ReplyStreamRequest) -> dict:
         task = asyncio.create_task(self._apply(request))
         self._active.add(task)
diff --git a/src/langbot/pkg/agent/runner/resource_builder.py b/src/langbot/pkg/agent/runner/resource_builder.py
index 76d5dd1c4..4e113f8f0 100644
--- a/src/langbot/pkg/agent/runner/resource_builder.py
+++ b/src/langbot/pkg/agent/runner/resource_builder.py
@@ -106,6 +106,8 @@ class AgentResourceBuilder:
             descriptor,
             runner_config,
         )
+        if getattr(self.ap, 'skill_mgr', None) is not None:
+            await self.ap.skill_mgr.reload_skills(execution_context)
         skills = self._build_skills_from_binding(
             execution_context,
             resource_policy,
diff --git a/src/langbot/pkg/agent/runner/run_journal.py b/src/langbot/pkg/agent/runner/run_journal.py
index dd2bfbf27..3a0217f89 100644
--- a/src/langbot/pkg/agent/runner/run_journal.py
+++ b/src/langbot/pkg/agent/runner/run_journal.py
@@ -68,6 +68,26 @@ class AgentRunJournal:
     def _sanitize_attachments(cls, attachments: typing.Iterable[typing.Any]) -> list[dict[str, typing.Any]]:
         return [cls._sanitize_attachment_ref(attachment) for attachment in attachments]
 
+    @classmethod
+    def _sanitize_event_data(cls, value: typing.Any) -> typing.Any:
+        """Retain arbitrary event fields without copying inline binary payloads."""
+        if hasattr(value, 'model_dump'):
+            value = value.model_dump(mode='json')
+        if isinstance(value, list):
+            return [cls._sanitize_event_data(item) for item in value]
+        if not isinstance(value, dict):
+            return value
+        result = {}
+        for key, item in value.items():
+            if key in ('image_base64', 'file_base64', 'base64') and item:
+                result[key] = None
+                result['content_redacted'] = True
+            elif key == 'attachments' and isinstance(item, list):
+                result[key] = [cls._sanitize_event_data(cls._sanitize_attachment_ref(ref)) for ref in item]
+            else:
+                result[key] = cls._sanitize_event_data(item)
+        return result
+
     async def create_run(
         self,
         *,
@@ -78,6 +98,9 @@ class AgentRunJournal:
         authorization: dict[str, typing.Any],
     ) -> dict[str, typing.Any]:
         """Create the Host-owned run ledger record."""
+        from ...telemetry import trace
+
+        trace_state = trace.current()
         runtime = context.get('runtime') if isinstance(context, dict) else {}
         return await self._get_run_ledger_store().create_run(
             run_id=context['run_id'],
@@ -93,17 +116,21 @@ class AgentRunJournal:
             authorization=authorization,
             metadata={
                 'event_type': event.event_type,
+                'ingress_event_id': trace_state.event_id if trace_state else event.event_id,
                 'source': event.source,
                 'processor_id': binding.processor_id,
                 'processor_type': binding.processor_type,
                 **(
-                    {'input_event': event.data, 'delivery': event.delivery.model_dump(mode='json')}
+                    {
+                        'input_event': self._sanitize_event_data(event.data),
+                        'delivery': event.delivery.model_dump(mode='json'),
+                    }
                     if binding.processor_type == 'event_processor'
                     else {}
                 ),
                 **(
                     {
-                        **({'input_event': event.data} if event.event_type != 'message.received' else {}),
+                        'input_event': self._sanitize_event_data(event.data),
                         'input': {
                             'text': context.get('input', {}).get('text', ''),
                             'contents': self._sanitize_contents(context.get('input', {}).get('contents', [])),
@@ -260,7 +287,7 @@ class AgentRunJournal:
             event_time=(
                 datetime.datetime.fromtimestamp(event.event_time, datetime.timezone.utc) if event.event_time else None
             ),
-            metadata=metadata,
+            metadata={**(metadata or {}), 'input_event': self._sanitize_event_data(event.data)},
         )
 
     async def write_user_transcript(
diff --git a/src/langbot/pkg/agent/runner/run_ledger_store.py b/src/langbot/pkg/agent/runner/run_ledger_store.py
index 1cb3d476d..2a8b9c75b 100644
--- a/src/langbot/pkg/agent/runner/run_ledger_store.py
+++ b/src/langbot/pkg/agent/runner/run_ledger_store.py
@@ -12,6 +12,8 @@ from sqlalchemy.ext.asyncio import AsyncEngine, AsyncSession
 from sqlalchemy.orm import sessionmaker
 
 from ...persistence.datetime_utils import as_naive_utc
+from ...utils.inflight import inflight_hub
+from ...utils.execution_lease import host_execution_lease
 
 from ...entity.persistence.agent_run import AgentRun, AgentRunEvent, AgentRuntime
 
@@ -136,6 +138,7 @@ class RunLedgerStore:
                 thread_id=thread_id,
                 workspace_id=workspace_id,
                 bot_id=bot_id,
+                **(host_execution_lease() if status == 'running' and queue_name is None else {}),
                 status=status,
                 queue_name=queue_name,
                 priority=priority,
@@ -149,6 +152,7 @@ class RunLedgerStore:
             )
             session.add(run)
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run)
 
     async def claim_next_run(
@@ -206,6 +210,7 @@ class RunLedgerStore:
             run.last_claimed_at = as_naive_utc(now)
             run.updated_at = as_naive_utc(now)
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run, include_claim_token=True)
 
     async def renew_claim(
@@ -226,6 +231,7 @@ class RunLedgerStore:
             run.claim_lease_expires_at = as_naive_utc(now + datetime.timedelta(seconds=max(int(lease_seconds), 1)))
             run.updated_at = as_naive_utc(now)
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run)
 
     async def release_claim(
@@ -254,6 +260,7 @@ class RunLedgerStore:
             if status in TERMINAL_STATUSES:
                 run.finished_at = run.finished_at or as_naive_utc(now)
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run)
 
     async def release_expired_claims(
@@ -332,6 +339,7 @@ class RunLedgerStore:
             )
             session.add(row)
             await session.commit()
+            inflight_hub.notify_run(run_id, event_type)
             return self._event_to_dict(row)
 
     async def append_audit_event(
@@ -366,8 +374,36 @@ class RunLedgerStore:
             )
             session.add(row)
             await session.commit()
+            inflight_hub.notify_run(run_id, event_type)
             return self._event_to_dict(row)
 
+    async def pin_skill_revision(self, run_id: str, workspace_id: str, name: str, revision: str) -> str:
+        """Pin a Host-owned revision without overwriting concurrent run metadata."""
+        async with self._session_factory() as session:
+            run = await self._get_run_row(session, run_id)
+            if run is None or run.workspace_id != workspace_id or run.status in TERMINAL_STATUSES:
+                raise ValueError('Skill activation requires an active run in the current Workspace')
+            original = run.metadata_json
+            metadata = _json_loads(original, {})
+            bindings = metadata.setdefault('activated_skills', {})
+            if name in bindings:
+                return str(bindings[name])
+            bindings[name] = revision
+            result = await session.execute(
+                sqlalchemy.update(AgentRun)
+                .where(
+                    AgentRun.run_id == run_id,
+                    AgentRun.metadata_json == original,
+                    AgentRun.status.notin_(TERMINAL_STATUSES),
+                )
+                .values(metadata_json=_json_dumps(metadata))
+                .execution_options(synchronize_session=False)
+            )
+            if result.rowcount != 1:
+                raise ValueError('Concurrent run update; retry Skill activation')
+            await session.commit()
+            return revision
+
     async def finalize_run(
         self,
         *,
@@ -410,6 +446,7 @@ class RunLedgerStore:
                 else:
                     run.metadata_json = _json_dumps(metadata)
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run)
 
     async def validate_active_claim(
@@ -443,6 +480,7 @@ class RunLedgerStore:
             run.updated_at = as_naive_utc(now)
             run.status_reason = status_reason or run.status_reason
             await session.commit()
+            inflight_hub.notify(run.workspace_id)
             return self._run_to_dict(run)
 
     async def get_run(self, run_id: str) -> dict[str, typing.Any] | None:
diff --git a/src/langbot/pkg/agent/runner/session_registry.py b/src/langbot/pkg/agent/runner/session_registry.py
index d7e2910d8..e7c12716b 100644
--- a/src/langbot/pkg/agent/runner/session_registry.py
+++ b/src/langbot/pkg/agent/runner/session_registry.py
@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-from ...telemetry import diagnostics
-
 import asyncio
 import copy
 import typing
@@ -172,7 +170,6 @@ class AgentRunSessionRegistry:
             'query_id': query_id,
             'execution_query': execution_query,
             'reply_streams': reply_streams,
-            '_diagnostic_context': diagnostics.capture_context(),
             'plugin_identity': plugin_identity,
             'authorization': authorization,
             'status': {
diff --git a/src/langbot/pkg/agent/runner/transcript_store.py b/src/langbot/pkg/agent/runner/transcript_store.py
index aec02502b..426740746 100644
--- a/src/langbot/pkg/agent/runner/transcript_store.py
+++ b/src/langbot/pkg/agent/runner/transcript_store.py
@@ -335,6 +335,9 @@ class TranscriptStore:
         thread_id: str | None,
         strict_thread: bool,
     ) -> typing.Any:
+        from .context_reset import active_history_condition
+
+        query = query.where(active_history_condition(Transcript))
         if bot_id is not None:
             query = query.where(Transcript.bot_id == bot_id)
         if workspace_id is not None:
diff --git a/src/langbot/pkg/api/http/controller/group.py b/src/langbot/pkg/api/http/controller/group.py
index 181dbe806..6cafbdc8c 100644
--- a/src/langbot/pkg/api/http/controller/group.py
+++ b/src/langbot/pkg/api/http/controller/group.py
@@ -17,6 +17,8 @@ from ....workspace.errors import WorkspaceNotFoundError
 from ....cloud.entitlements import EntitlementUnavailableError
 from ....core.errors import TaskCapacityError
 from ....provider.modelmgr.codex_errors import CodexProviderError
+from ....i18n import DEFAULT_LANGUAGE, language, language_scope, t, error_text
+from ....i18n.workspace import get_workspace_language
 from ..authz import (
     AuthenticationDeniedError,
     AuthorizationError,
@@ -28,7 +30,6 @@ from ..authz import (
 )
 from ..context import PrincipalContext, PrincipalType, RequestContext, WorkspaceContext
 from ....cloud.support_admin import SupportAdminSessionError
-from ... import management_diagnostics as diagnostics
 
 if typing.TYPE_CHECKING:
     from ....core.app import Application
@@ -138,7 +139,7 @@ class RouterGroup(abc.ABC):
                         # receive RequestContext or enforce Workspace permissions.
                         self._inject_handler_context(f, kwargs, user_email, None, account)
                     except Exception as e:
-                        return self._auth_error_response(e)
+                        return await self._localized_auth_error_response(e)
 
                 elif auth_type == AuthType.USER_TOKEN:
                     # get token from Authorization header
@@ -161,7 +162,7 @@ class RouterGroup(abc.ABC):
                             require_permission(request_context, permission)
                         self._inject_handler_context(f, kwargs, user_email, request_context)
                     except Exception as e:
-                        return self._auth_error_response(e)
+                        return await self._localized_auth_error_response(e)
 
                 elif auth_type == AuthType.API_KEY:
                     # get API key from Authorization header or X-API-Key header
@@ -180,7 +181,7 @@ class RouterGroup(abc.ABC):
                             require_permission(request_context, permission)
                         self._inject_handler_context(f, kwargs, None, request_context)
                     except Exception as e:
-                        return self._auth_error_response(e)
+                        return await self._localized_auth_error_response(e)
 
                 elif auth_type == AuthType.USER_TOKEN_OR_API_KEY:
                     token = quart.request.headers.get('Authorization', '').replace('Bearer ', '')
@@ -194,7 +195,7 @@ class RouterGroup(abc.ABC):
                                 require_permission(request_context, permission)
                             self._inject_handler_context(f, kwargs, None, request_context)
                         except Exception as e:
-                            return self._auth_error_response(e)
+                            return await self._localized_auth_error_response(e)
                     # Try API key first (check X-API-Key header)
                     elif api_key := quart.request.headers.get('X-API-Key', ''):
                         # API key authentication
@@ -204,7 +205,7 @@ class RouterGroup(abc.ABC):
                                 require_permission(request_context, permission)
                             self._inject_handler_context(f, kwargs, None, request_context)
                         except Exception as e:
-                            return self._auth_error_response(e)
+                            return await self._localized_auth_error_response(e)
                     else:
                         # Try user token authentication (Authorization header)
                         if not token:
@@ -224,7 +225,7 @@ class RouterGroup(abc.ABC):
                             # Authentication succeeded and authorization was
                             # evaluated. Do not reinterpret a denied user token
                             # as an API key, which would mask the stable 403/404.
-                            return self._auth_error_response(e)
+                            return await self._localized_auth_error_response(e)
                         except Exception:
                             # If user token fails, maybe it's an API key in Authorization header
                             try:
@@ -233,12 +234,13 @@ class RouterGroup(abc.ABC):
                                     require_permission(request_context, permission)
                                 self._inject_handler_context(f, kwargs, None, request_context)
                             except Exception as e:
-                                return self._auth_error_response(e)
+                                return await self._localized_auth_error_response(e)
 
                 observed_at = time.monotonic()
                 try:
                     if request_context is not None:
-                        diagnostics.workspace(request_context)
+                        language.set(await get_workspace_language(self.ap, request_context.workspace_uuid))
+                    if request_context is not None:
                         with bounded_executor.blocking_work_scope(request_context.workspace_uuid):
                             persistence_mgr = getattr(
                                 self.ap,
@@ -294,6 +296,8 @@ class RouterGroup(abc.ABC):
                         )
                     if isinstance(e, CodexProviderError):
                         return self.http_status(e.status_code, e.error_code, str(e))
+                    if isinstance(e, EntitlementUnavailableError):
+                        return self.http_status(403, 'entitlement_unavailable', str(e))
                     if isinstance(e, AuthorizationError):
                         return self.http_status(e.status_code, e.error_code, str(e))
                     if isinstance(e, WorkspaceNotFoundError):
@@ -321,14 +325,11 @@ class RouterGroup(abc.ABC):
                     )
                     return self.internal_error_response(request_id)
 
-            # Observe outside authentication, using the registered Core handler
-            # identity rather than the URL (which can contain user identifiers).
-            new_f = diagnostics.observe(
-                diagnostics.operation_id('http', f, rule=rule, methods=options.get('methods')),
-                source='http',
-                ap=self.ap,
-                http=True,
-            )(handler_error)
+            async def new_f(*args, **kwargs):
+                # Reset even on errors; child tasks inherit this request's locale.
+                with language_scope(quart.request.headers.get('X-LangBot-Language', DEFAULT_LANGUAGE)):
+                    return await handler_error(*args, **kwargs)
+
             # Quart/Flask requires a unique endpoint name even when the same URL
             # intentionally has separate handlers for different HTTP methods.
             # Include the method set so CRUD routes can declare distinct
@@ -696,6 +697,14 @@ class RouterGroup(abc.ABC):
             'Invalid authentication credentials',
         )
 
+    async def _localized_auth_error_response(self, error: Exception):
+        # A resolved context is trusted even when its permissions deny this
+        # operation. Never look up a language using the unverified selector.
+        context = getattr(quart.g, 'request_context', None)
+        if context is not None:
+            language.set(await get_workspace_language(self.ap, context.workspace_uuid))
+        return self._auth_error_response(error)
+
     def request_id(self) -> str:
         """Return one stable request ID for authentication, logs, and errors."""
 
@@ -716,7 +725,7 @@ class RouterGroup(abc.ABC):
         response = quart.jsonify(
             {
                 'code': 'internal_error',
-                'msg': 'Internal server error',
+                'msg': t('Internal server error'),
                 'request_id': resolved_request_id,
             }
         )
@@ -735,12 +744,10 @@ class RouterGroup(abc.ABC):
 
     def fail(self, code: int | str, msg: str) -> quart.Response:
         """Return an error response"""
-
-        diagnostics.outcome('failed')
         return quart.jsonify(
             {
                 'code': code,
-                'msg': msg,
+                'msg': error_text(code, msg),
             }
         )
 
diff --git a/src/langbot/pkg/api/http/controller/groups/agent_debug_stream.py b/src/langbot/pkg/api/http/controller/groups/agent_debug_stream.py
index 272f7bd92..700eab509 100644
--- a/src/langbot/pkg/api/http/controller/groups/agent_debug_stream.py
+++ b/src/langbot/pkg/api/http/controller/groups/agent_debug_stream.py
@@ -2,12 +2,13 @@
 
 from __future__ import annotations
 
+from .....i18n import t
+
 import asyncio
 import contextlib
 import json
 
 import quart
-from .... import management_diagnostics as diagnostics
 
 from .....agent.runner.errors import (
     RunnerError,
@@ -26,27 +27,40 @@ def debug_stream_response(service, context, agent_uuid: str, payload: dict) -> q
             await queue.put({'kind': 'result', 'data': result})
 
         async def execute() -> None:
+            application = getattr(service, 'ap', None)
+            persistence_mgr = getattr(application, 'persistence_mgr', None)
+            tenant_scope = getattr(persistence_mgr, 'tenant_scope', None)
+            scope = tenant_scope(context.workspace_uuid) if callable(tenant_scope) else contextlib.nullcontext()
             try:
-                result = await service.debug_agent(context, agent_uuid, payload, on_result=on_result)
+                # The streamed body is emitted after the request handler returned,
+                # so the request's tenant scope has already closed: carry the
+                # trusted Workspace identity into the deferred execution.
+                async with scope:
+                    result = await service.debug_agent(
+                        context,
+                        agent_uuid,
+                        payload,
+                        on_result=on_result,
+                    )
                 await queue.put({'kind': 'completed', 'data': result})
             except Exception as exc:
-                # The stream still uses HTTP 200 when execution returns an
-                # error frame. Mark the inherited request, not its contents.
-                diagnostics.outcome('failed')
                 if isinstance(exc, RunnerExecutionError):
                     code, message = exc.error_code or 'runner_execution_failed', exc.message
                 elif isinstance(exc, RunnerNotFoundError):
-                    code, message = 'runner_not_found', 'The configured Agent runner is unavailable'
+                    code, message = 'runner_not_found', t('Agent runner temporarily unavailable. Please try again.')
                 elif isinstance(exc, RunnerNotAuthorizedError):
-                    code, message = 'runner_not_authorized', 'The configured Agent runner is not authorized'
+                    code, message = 'runner_not_authorized', t('Agent runner not authorized for this pipeline.')
                 elif isinstance(exc, RunnerProtocolError):
-                    code, message = 'runner_protocol_error', 'The Agent runner returned an invalid response'
+                    code, message = 'runner_protocol_error', t('The Agent runner returned an invalid response.')
                 elif isinstance(exc, ValueError):
                     code, message = 'invalid_request', str(exc)
                 elif isinstance(exc, RunnerError):
-                    code, message = 'runner_error', 'The Agent runner could not complete this test'
+                    code, message = 'runner_error', t('Agent runner execution failed.')
                 else:
-                    code, message = 'runner_error', 'The Agent debug execution failed'
+                    logger = getattr(application, 'logger', None)
+                    if logger is not None:
+                        logger.exception('Agent debug stream execution failed')
+                    code, message = 'runner_error', t('Agent runner execution failed.')
                 await queue.put({'kind': 'error', 'code': code, 'msg': message})
 
         task = asyncio.create_task(execute())
diff --git a/src/langbot/pkg/api/http/controller/groups/assistant.py b/src/langbot/pkg/api/http/controller/groups/assistant.py
index f902c6f44..7684db2e1 100644
--- a/src/langbot/pkg/api/http/controller/groups/assistant.py
+++ b/src/langbot/pkg/api/http/controller/groups/assistant.py
@@ -1,5 +1,8 @@
 """Web-session-only assistant endpoints; resource tools use the existing service layer."""
 
+import asyncio
+import json
+
 import quart
 from uuid import UUID
 from pydantic import BaseModel, ConfigDict, Field, ValidationError
@@ -10,6 +13,11 @@ from ...context import RequestContext
 from ...service.assistant import AssistantError, AssistantService
 
 
+class StopInput(BaseModel):
+    model_config = ConfigDict(extra='forbid')
+    revision: int = Field(ge=0, strict=True)
+
+
 class TurnInput(BaseModel):
     model_config = ConfigDict(extra='forbid')
     revision: int = Field(ge=0, strict=True)
@@ -23,6 +31,27 @@ class AssistantRouterGroup(group.RouterGroup):
     async def initialize(self):
         service = AssistantService(self.ap)
 
+        @self.route('/conversations', methods=['GET'], permission=Permission.RESOURCE_VIEW)
+        async def list_conversations(request_context: RequestContext):
+            return self.success(data={'conversations': await service.list_conversations(request_context)})
+
+        @self.route('/conversations/<conversation_id>/stop', methods=['POST'], permission=Permission.RUNTIME_OPERATE)
+        async def stop(conversation_id: str, request_context: RequestContext):
+            try:
+                body = StopInput.model_validate(await quart.request.get_json())
+                return self.success(data=await service.stop(request_context, conversation_id, body.revision))
+            except ValidationError:
+                return self.http_status(400, 'invalid_input', 'Invalid stop request')
+            except AssistantError as exc:
+                return self.http_status(exc.status, exc.code, exc.code)
+
+        @self.route('/recommended-model', methods=['GET'], permission=Permission.RUNTIME_OPERATE)
+        async def recommended_model(request_context: RequestContext):
+            try:
+                return self.success(data=await service.recommended_model(request_context))
+            except AssistantError as exc:
+                return self.http_status(exc.status, exc.code, exc.code)
+
         @self.route('/conversations', methods=['POST'], permission=Permission.RUNTIME_OPERATE)
         async def create(request_context: RequestContext):
             try:
@@ -33,7 +62,7 @@ class AssistantRouterGroup(group.RouterGroup):
         @self.route('/conversations/<conversation_id>', methods=['GET'], permission=Permission.RESOURCE_VIEW)
         async def get(conversation_id: str, request_context: RequestContext):
             try:
-                return self.success(data=service.public_view(await service.get(request_context, conversation_id)))
+                return self.success(data=await service.progress(request_context, conversation_id))
             except AssistantError as exc:
                 return self.http_status(exc.status, exc.code, exc.code)
 
@@ -56,3 +85,43 @@ class AssistantRouterGroup(group.RouterGroup):
                 return self.http_status(400, 'invalid_input', 'Invalid assistant request')
             except AssistantError as exc:
                 return self.http_status(exc.status, exc.code, exc.code)
+
+        @self.route(
+            '/conversations/<conversation_id>/turn/stream', methods=['POST'], permission=Permission.RUNTIME_OPERATE
+        )
+        async def stream_turn(conversation_id: str, request_context: RequestContext):
+            try:
+                body = TurnInput.model_validate(await quart.request.get_json())
+                if (body.text is None) == (body.approved is None) or (body.text is not None and not body.text.strip()):
+                    return self.http_status(400, 'invalid_input', 'Provide text or an approval decision')
+                await service.get(request_context, conversation_id)
+            except ValidationError:
+                return self.http_status(400, 'invalid_input', 'Invalid assistant request')
+            except AssistantError as exc:
+                return self.http_status(exc.status, exc.code, exc.code)
+            try:
+                return assistant_stream_response(service, request_context, conversation_id, body)
+            except AssistantError as exc:
+                return self.http_status(exc.status, exc.code, exc.code)
+
+
+def assistant_stream_response(service, context, conversation_id, body):
+    """Detach execution from the subscriber; only an explicit stop cancels a run."""
+    queue = service.start_run(context, conversation_id, body)
+
+    async def stream():
+        while True:
+            try:
+                frame = await asyncio.wait_for(queue.get(), timeout=15)
+            except TimeoutError:
+                yield '\n'
+                continue
+            yield json.dumps(frame, ensure_ascii=False) + '\n'
+            if frame['kind'] in {'completed', 'error'}:
+                break
+
+    response = quart.Response(stream(), content_type='application/x-ndjson; charset=utf-8')
+    response.timeout = None
+    response.headers['Cache-Control'] = 'no-store'
+    response.headers['X-Accel-Buffering'] = 'no'
+    return response
diff --git a/src/langbot/pkg/api/http/controller/groups/monitoring.py b/src/langbot/pkg/api/http/controller/groups/monitoring.py
index 01a901743..5cbdf56c6 100644
--- a/src/langbot/pkg/api/http/controller/groups/monitoring.py
+++ b/src/langbot/pkg/api/http/controller/groups/monitoring.py
@@ -24,6 +24,24 @@ def parse_iso_datetime(datetime_str: str | None) -> datetime.datetime | None:
     return dt
 
 
+def validate_monitoring_window(
+    start_time: datetime.datetime | None,
+    end_time: datetime.datetime | None,
+    *,
+    allow_point_window: bool = False,
+) -> None:
+    """Bound explicit dashboard windows to one year without silently truncating them."""
+    if start_time is None:
+        return
+    end = end_time or datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None)
+    if (
+        start_time > end
+        or (start_time == end and not allow_point_window)
+        or end - start_time > datetime.timedelta(days=365)
+    ):
+        quart.abort(400, description='Monitoring time range must be positive and no longer than 365 days')
+
+
 @group.group_class('monitoring', '/api/v1/monitoring')
 class MonitoringRouterGroup(group.RouterGroup):
     async def initialize(self) -> None:
@@ -39,6 +57,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             metrics = await self.ap.monitoring_service.get_overview_metrics(
                 request_context,
@@ -50,6 +69,80 @@ class MonitoringRouterGroup(group.RouterGroup):
 
             return self.success(data=metrics)
 
+        @self.route('/in-flight/stream', methods=['GET'], permission=Permission.RESOURCE_VIEW)
+        async def stream_inflight(request_context: RequestContext):
+            from .....utils.inflight import inflight_hub
+            from ...service.tenant import require_workspace_uuid
+            import json
+
+            workspace = require_workspace_uuid(request_context)
+            previous = []
+
+            async def snapshot():
+                nonlocal previous
+                result = await self.ap.monitoring_service.get_inflight_snapshot(request_context, previous)
+                previous = [row for row in result['items'] if row['status_group'] in {'running', 'queued'}]
+                return result
+
+            async def frames():
+                async for frame in inflight_hub.watch(workspace, snapshot):
+                    yield 'data: ' + json.dumps(frame, ensure_ascii=False) + '\n\n'
+
+            response = quart.Response(frames(), content_type='text/event-stream')
+            response.headers['Cache-Control'] = 'no-cache, no-store'
+            response.headers['X-Accel-Buffering'] = 'no'
+            response.timeout = None
+            return response
+
+        @self.route('/executions', methods=['GET'], permission=Permission.RESOURCE_VIEW)
+        async def get_executions(request_context: RequestContext) -> str:
+            """Unified execution list across agent runs and pipeline queries."""
+            bot_ids = quart.request.args.getlist('botId')
+            pipeline_ids = quart.request.args.getlist('pipelineId')
+            agent_ids = quart.request.args.getlist('agentId')
+            statuses = quart.request.args.getlist('status')
+            source = quart.request.args.get('source', 'all')
+            mode = quart.request.args.get('mode', 'all')
+            start_time = parse_iso_datetime(quart.request.args.get('startTime'))
+            end_time = parse_iso_datetime(quart.request.args.get('endTime'))
+            validate_monitoring_window(start_time, end_time)
+
+            result = await self.ap.monitoring_service.get_executions(
+                request_context,
+                bot_ids=bot_ids if bot_ids else None,
+                pipeline_ids=pipeline_ids if pipeline_ids else None,
+                agent_ids=agent_ids if agent_ids else None,
+                statuses=statuses if statuses else None,
+                source=source,
+                mode=mode,
+                start_time=start_time,
+                end_time=end_time,
+                limit=quart.request.args.get('limit', 50),
+                offset=quart.request.args.get('offset', 0),
+            )
+
+            return self.success(data=result)
+
+        @self.route(
+            '/executions/<source>/<execution_id>',
+            methods=['GET'],
+            permission=Permission.RESOURCE_VIEW,
+        )
+        async def get_execution_detail(source: str, execution_id: str, request_context: RequestContext) -> str:
+            """Full trace of a single execution (agent run or pipeline query)."""
+            try:
+                detail = await self.ap.monitoring_service.get_execution_detail(
+                    request_context,
+                    source,
+                    execution_id,
+                    section=quart.request.args.get('section'),
+                    offset=quart.request.args.get('offset', 0),
+                    limit=quart.request.args.get('limit', 100),
+                )
+            except ValueError as exc:
+                return self.fail(404, str(exc))
+            return self.success(data=detail)
+
         @self.route('/token-statistics', methods=['GET'], permission=Permission.RESOURCE_VIEW)
         async def get_token_statistics(request_context: RequestContext) -> str:
             """Get detailed token usage statistics (summary, per-model, timeseries)."""
@@ -63,9 +156,12 @@ class MonitoringRouterGroup(group.RouterGroup):
 
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             stats = await self.ap.monitoring_service.get_token_statistics(
                 request_context,
+                execution_statuses=quart.request.args.getlist('status') or None,
+                mode=quart.request.args.get('mode', 'all'),
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -96,6 +192,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             messages, total = await self.ap.monitoring_service.get_messages(
                 request_context,
@@ -137,6 +234,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             llm_calls, total = await self.ap.monitoring_service.get_llm_calls(
                 request_context,
@@ -176,6 +274,7 @@ class MonitoringRouterGroup(group.RouterGroup):
 
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             tool_calls, total = await self.ap.monitoring_service.get_tool_calls(
                 request_context,
@@ -216,6 +315,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             embedding_calls, total = await self.ap.monitoring_service.get_embedding_calls(
                 request_context,
@@ -257,6 +357,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             # Parse is_active
             is_active = None
@@ -304,6 +405,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             errors, total = await self.ap.monitoring_service.get_errors(
                 request_context,
@@ -327,6 +429,10 @@ class MonitoringRouterGroup(group.RouterGroup):
         @self.route('/data', methods=['GET'], permission=Permission.RESOURCE_VIEW)
         async def get_all_data(request_context: RequestContext) -> str:
             """Get all monitoring data in a single request"""
+            execution_filters = {
+                'execution_statuses': quart.request.args.getlist('status') or None,
+                'mode': quart.request.args.get('mode', 'all'),
+            }
             # Parse query parameters
             bot_ids = quart.request.args.getlist('botId')
             pipeline_ids = quart.request.args.getlist('pipelineId')
@@ -337,10 +443,12 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             # Get overview metrics
             overview = await self.ap.monitoring_service.get_overview_metrics(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -350,6 +458,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get messages
             messages, messages_total = await self.ap.monitoring_service.get_messages(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -361,6 +470,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get LLM calls
             llm_calls, llm_calls_total = await self.ap.monitoring_service.get_llm_calls(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -372,6 +482,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get tool calls
             tool_calls, tool_calls_total = await self.ap.monitoring_service.get_tool_calls(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -383,6 +494,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get sessions
             sessions, sessions_total = await self.ap.monitoring_service.get_sessions(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -395,6 +507,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get errors
             errors, errors_total = await self.ap.monitoring_service.get_errors(
                 request_context,
+                **execution_filters,
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -406,6 +519,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Get embedding calls
             embedding_calls, embedding_calls_total = await self.ap.monitoring_service.get_embedding_calls(
                 request_context,
+                **execution_filters,
                 start_time=start_time,
                 end_time=end_time,
                 limit=limit,
@@ -417,6 +531,7 @@ class MonitoringRouterGroup(group.RouterGroup):
                     'traffic': await get_traffic_series(
                         self.ap,
                         request_context,
+                        **execution_filters,
                         bot_ids=bot_ids or None,
                         pipeline_ids=pipeline_ids or None,
                         start_time=start_time,
@@ -440,6 +555,30 @@ class MonitoringRouterGroup(group.RouterGroup):
                 }
             )
 
+        @self.route(
+            '/sessions/reset-context',
+            methods=['POST'],
+            auth_type=group.AuthType.USER_TOKEN_OR_API_KEY,
+            permission=Permission.RESOURCE_MANAGE,
+        )
+        async def reset_session_context(request_context: RequestContext):
+            body = await quart.request.get_json()
+            if not isinstance(body, dict):
+                return self.http_status(400, 'invalid_request', 'Expected a JSON object')
+            try:
+                result = await self.ap.monitoring_service.reset_session_context(
+                    request_context,
+                    body.get('bot_id'),
+                    body.get('session_id'),
+                )
+            except ValueError as exc:
+                return self.http_status(400, 'invalid_request', str(exc))
+            except LookupError as exc:
+                return self.http_status(404, 'resource_not_found', str(exc))
+            except RuntimeError as exc:
+                return self.http_status(409, 'session_busy', str(exc))
+            return self.success(data=result)
+
         @self.route(
             '/sessions/<session_id>/analysis',
             methods=['GET'],
@@ -450,6 +589,8 @@ class MonitoringRouterGroup(group.RouterGroup):
             """Get detailed analysis for a specific session"""
             start_time = parse_iso_datetime(quart.request.args.get('startTime'))
             end_time = parse_iso_datetime(quart.request.args.get('endTime'))
+            # Message pages can contain one timestamp; tool-call bounds are inclusive.
+            validate_monitoring_window(start_time, end_time, allow_point_window=True)
             analysis = await self.ap.monitoring_service.get_session_analysis(
                 request_context,
                 session_id,
@@ -491,6 +632,7 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             # Get data based on export type
             if export_type == 'messages':
@@ -639,7 +781,7 @@ class MonitoringRouterGroup(group.RouterGroup):
                     'platform',
                 ]
             else:
-                return self.error(message=f'Invalid export type: {export_type}', code=400)
+                return self.fail(400, f'Invalid export type: {export_type}')
 
             # Generate CSV content with UTF-8 BOM for Excel compatibility
             import io
@@ -681,9 +823,12 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             stats = await self.ap.monitoring_service.get_feedback_stats(
                 request_context,
+                execution_statuses=quart.request.args.getlist('status') or None,
+                mode=quart.request.args.get('mode', 'all'),
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 start_time=start_time,
@@ -707,12 +852,15 @@ class MonitoringRouterGroup(group.RouterGroup):
             # Parse datetime
             start_time = parse_iso_datetime(start_time_str)
             end_time = parse_iso_datetime(end_time_str)
+            validate_monitoring_window(start_time, end_time)
 
             # Parse feedback type
             feedback_type = int(feedback_type_str) if feedback_type_str else None
 
             feedback_list, total = await self.ap.monitoring_service.get_feedback_list(
                 request_context,
+                execution_statuses=quart.request.args.getlist('status') or None,
+                mode=quart.request.args.get('mode', 'all'),
                 bot_ids=bot_ids if bot_ids else None,
                 pipeline_ids=pipeline_ids if pipeline_ids else None,
                 feedback_type=feedback_type,
diff --git a/src/langbot/pkg/api/http/controller/groups/pipelines/embed.py b/src/langbot/pkg/api/http/controller/groups/pipelines/embed.py
index 0017e194a..9f32a63d3 100644
--- a/src/langbot/pkg/api/http/controller/groups/pipelines/embed.py
+++ b/src/langbot/pkg/api/http/controller/groups/pipelines/embed.py
@@ -21,7 +21,8 @@ import httpx
 import quart
 
 from ... import group
-from ..... import management_diagnostics as diagnostics
+from ......i18n import language, t
+from ......i18n.workspace import get_workspace_language
 from ......utils import httpclient, paths
 from ......platform.sources.websocket_manager import WebSocketScope, is_valid_session_id, ws_connection_manager
 from .websocket_chat import create_scoped_duplex_tasks, wait_for_duplex_tasks
@@ -74,6 +75,7 @@ class EmbedRouterGroup(group.RouterGroup):
         bot = await self.ap.platform_mgr.resolve_public_bot(bot_uuid)
         pipeline_uuid = bot.get_pipeline_target_for_event_type('message.received') if bot is not None else None
         if bot is not None and bot.bot_entity.adapter == 'web_page_bot' and bot.bot_entity.enable and pipeline_uuid:
+            language.set(await get_workspace_language(self.ap, bot.execution_context.workspace_uuid))
             return bot, pipeline_uuid
         return None, None
 
@@ -285,7 +287,7 @@ class EmbedRouterGroup(group.RouterGroup):
                     return self.http_status(404, -1, 'WebSocket adapter not found')
 
                 websocket_adapter.reset_session(pipeline_uuid, session_type, session_id)
-                return self.success(data={'message': 'Session reset successfully'})
+                return self.success(data={'message': t('Session reset successfully')})
 
             except Exception as e:
                 logger.error(f'Failed to reset embed session: {e}', exc_info=True)
@@ -330,24 +332,22 @@ class EmbedRouterGroup(group.RouterGroup):
         # -- Embed WebSocket endpoint ----------------------------------------
 
         @self.quart_app.websocket(self.path + '/<bot_uuid>/ws/connect')
-        @diagnostics.observe('websocket.embed.session', source='websocket', ap=self.ap)
         async def embed_websocket_connect(bot_uuid: str):
             """WebSocket connection for embed widget, keyed by bot_uuid."""
             await quart.websocket.accept()
             if not _is_valid_uuid(bot_uuid):
                 await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Invalid bot_uuid format'}))
-                diagnostics.outcome('rejected')
                 return
 
             runtime_bot, pipeline_uuid = await self._resolve_bot(bot_uuid)
             if runtime_bot is None:
-                await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Bot not found or not available'}))
-                diagnostics.outcome('rejected')
+                await quart.websocket.send(
+                    json.dumps({'type': 'error', 'message': t('Bot not found or not available')})
+                )
                 return
 
             session_type = quart.websocket.args.get('session_type', 'person')
             if session_type not in ['person', 'group']:
-                diagnostics.outcome('rejected')
                 await quart.websocket.send(
                     json.dumps({'type': 'error', 'message': 'session_type must be person or group'})
                 )
@@ -356,24 +356,21 @@ class EmbedRouterGroup(group.RouterGroup):
             session_id = quart.websocket.args.get('session_id', '')
             if not is_valid_session_id(session_id):
                 await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Valid session_id is required'}))
-                diagnostics.outcome('rejected')
                 return
 
             try:
                 await self._authenticate_websocket(runtime_bot)
                 await self._assert_execution_active(runtime_bot)
             except Exception:
-                await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Unauthorized'}))
-                diagnostics.outcome('rejected')
+                await quart.websocket.send(json.dumps({'type': 'error', 'message': t('Unauthorized')}))
                 return
-
-            diagnostics.workspace(runtime_bot.execution_context)
             try:
                 proxy_bot = await self.ap.platform_mgr.get_websocket_proxy_bot(runtime_bot.execution_context)
                 websocket_adapter = proxy_bot.adapter
                 if not websocket_adapter:
-                    await quart.websocket.send(json.dumps({'type': 'error', 'message': 'WebSocket adapter not found'}))
-                    diagnostics.outcome('rejected')
+                    await quart.websocket.send(
+                        json.dumps({'type': 'error', 'message': t('WebSocket adapter not found')})
+                    )
                     return
 
                 connection = await ws_connection_manager.add_connection(
@@ -431,16 +428,14 @@ class EmbedRouterGroup(group.RouterGroup):
                 try:
                     await wait_for_duplex_tasks(receive_task, send_task)
                 except Exception as e:
-                    diagnostics.outcome('failed')
                     logger.error(f'Embed WebSocket task error: {e}')
                 finally:
                     await ws_connection_manager.remove_connection(connection.connection_id)
 
             except Exception as e:
-                diagnostics.outcome('failed')
                 logger.error(f'Embed WebSocket connection error: {e}', exc_info=True)
                 try:
-                    await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Internal server error'}))
+                    await quart.websocket.send(json.dumps({'type': 'error', 'message': t('Internal server error')}))
                 except Exception:
                     pass
 
@@ -451,36 +446,30 @@ class EmbedRouterGroup(group.RouterGroup):
             while connection.is_active:
                 message = await quart.websocket.receive()
                 await ws_connection_manager.update_activity(connection.connection_id)
+                try:
+                    data = await asyncio.to_thread(json.loads, message)
+                    message_type = data.get('type', 'message')
 
-                with diagnostics.scope(self.ap, 'websocket.embed.message', source='websocket'):
-                    diagnostics.workspace(getattr(owner_bot, 'execution_context', None))
-                    try:
-                        data = await asyncio.to_thread(json.loads, message)
-                        message_type = data.get('type', 'message')
-
-                        if message_type == 'ping':
-                            await connection.send_queue.put(
-                                {'type': 'pong', 'timestamp': datetime.datetime.now().isoformat()}
-                            )
-                        elif message_type == 'message':
-                            try:
-                                current_bot = await self._resolve_connected_bot(owner_bot, pipeline_uuid)
-                            except Exception:
-                                diagnostics.outcome('rejected')
-                                await connection.send_queue.put({'type': 'error', 'message': 'Bot is unavailable'})
-                                break
-                            await websocket_adapter.handle_websocket_message(connection, data, owner_bot=current_bot)
-                        elif message_type == 'disconnect':
+                    if message_type == 'ping':
+                        await connection.send_queue.put(
+                            {'type': 'pong', 'timestamp': datetime.datetime.now().isoformat()}
+                        )
+                    elif message_type == 'message':
+                        try:
+                            current_bot = await self._resolve_connected_bot(owner_bot, pipeline_uuid)
+                        except Exception:
+                            await connection.send_queue.put({'type': 'error', 'message': t('Bot is unavailable')})
                             break
-                        else:
-                            diagnostics.outcome('skipped')
+                        await websocket_adapter.handle_websocket_message(connection, data, owner_bot=current_bot)
+                    elif message_type == 'disconnect':
+                        break
+                    else:
+                        pass
 
-                    except json.JSONDecodeError:
-                        diagnostics.outcome('rejected')
-                        await connection.send_queue.put({'type': 'error', 'message': 'Invalid JSON format'})
+                except json.JSONDecodeError:
+                    await connection.send_queue.put({'type': 'error', 'message': t('Invalid JSON format')})
 
         except Exception as e:
-            diagnostics.outcome('failed')
             logger.error(f'Embed receive error: {e}', exc_info=True)
         finally:
             connection.is_active = False
@@ -496,13 +485,11 @@ class EmbedRouterGroup(group.RouterGroup):
                     message = await asyncio.wait_for(connection.send_queue.get(), timeout=1.0)
                     if message is None:
                         break
-                    with diagnostics.scope(self.ap, 'websocket.embed.send', source='websocket'):
-                        encoded = await asyncio.to_thread(json.dumps, message)
-                        await quart.websocket.send(encoded)
+                    encoded = await asyncio.to_thread(json.dumps, message)
+                    await quart.websocket.send(encoded)
                 except asyncio.TimeoutError:
                     continue
         except Exception as e:
-            diagnostics.outcome('failed')
             logger.error(f'Embed send error: {e}', exc_info=True)
         finally:
             connection.is_active = False
diff --git a/src/langbot/pkg/api/http/controller/groups/pipelines/websocket_chat.py b/src/langbot/pkg/api/http/controller/groups/pipelines/websocket_chat.py
index cc9582719..649293e10 100644
--- a/src/langbot/pkg/api/http/controller/groups/pipelines/websocket_chat.py
+++ b/src/langbot/pkg/api/http/controller/groups/pipelines/websocket_chat.py
@@ -2,6 +2,9 @@
 
 from __future__ import annotations
 
+from ......i18n import language, t
+from ......i18n.workspace import get_workspace_language
+
 import asyncio
 import datetime
 import json
@@ -14,7 +17,6 @@ import quart
 from ....authz import Permission, permissions_for_role, require_permission
 from ....context import PrincipalContext, PrincipalType, RequestContext, WorkspaceContext
 from ... import group
-from ..... import management_diagnostics as diagnostics
 from ......core.task_boundary import run_in_workspace_uow
 from ......platform.sources.websocket_manager import WebSocketScope, ws_connection_manager
 from ......utils import bounded_executor
@@ -211,7 +213,6 @@ class WebSocketChatRouterGroup(group.RouterGroup):
 
     async def initialize(self) -> None:
         @self.quart_app.websocket(self.path + '/connect')
-        @diagnostics.observe('websocket.dashboard.session', source='websocket', ap=self.ap)
         async def websocket_connect(pipeline_uuid: str):
             """Open one authenticated dashboard debug connection."""
 
@@ -219,14 +220,11 @@ class WebSocketChatRouterGroup(group.RouterGroup):
             try:
                 request_context, token = await self._authenticate_websocket()
             except Exception:
-                diagnostics.outcome('rejected')
-                await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Unauthorized'}))
+                await quart.websocket.send(json.dumps({'type': 'error', 'message': t('Unauthorized')}))
                 return
-
-            diagnostics.workspace(request_context)
+            language.set(await get_workspace_language(self.ap, request_context.workspace_uuid))
             session_type = quart.websocket.args.get('session_type', 'person')
             if session_type not in ['person', 'group']:
-                diagnostics.outcome('rejected')
                 await quart.websocket.send(
                     json.dumps({'type': 'error', 'message': 'session_type must be person or group'})
                 )
@@ -235,8 +233,7 @@ class WebSocketChatRouterGroup(group.RouterGroup):
             try:
                 websocket_adapter = await self._get_scoped_adapter(request_context, pipeline_uuid)
                 if websocket_adapter is None:
-                    diagnostics.outcome('rejected')
-                    await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Pipeline not found'}))
+                    await quart.websocket.send(json.dumps({'type': 'error', 'message': t('Pipeline not found')}))
                     return
 
                 connection = await ws_connection_manager.add_connection(
@@ -294,16 +291,14 @@ class WebSocketChatRouterGroup(group.RouterGroup):
                 try:
                     await wait_for_duplex_tasks(receive_task, send_task)
                 except Exception as exc:
-                    diagnostics.outcome('failed')
                     logger.error(f'WebSocket task execution error: {exc}')
                 finally:
                     await ws_connection_manager.remove_connection(connection.connection_id)
 
             except Exception:
-                diagnostics.outcome('failed')
                 logger.error('Dashboard WebSocket connection error', exc_info=True)
                 try:
-                    await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Internal server error'}))
+                    await quart.websocket.send(json.dumps({'type': 'error', 'message': t('Internal server error')}))
                 except Exception:
                     pass
 
@@ -417,35 +412,28 @@ class WebSocketChatRouterGroup(group.RouterGroup):
             while connection.is_active:
                 message = await quart.websocket.receive()
                 await ws_connection_manager.update_activity(connection.connection_id)
-
-                with diagnostics.scope(self.ap, 'websocket.dashboard.message', source='websocket'):
-                    diagnostics.workspace(request_context)
-                    try:
-                        data = await asyncio.to_thread(json.loads, message)
-                        message_type = data.get('type', 'message')
-                        if message_type == 'ping':
-                            await connection.send_queue.put(
-                                {'type': 'pong', 'timestamp': datetime.datetime.now().isoformat()}
-                            )
-                        elif message_type == 'message':
-                            try:
-                                request_context = await self._revalidate_websocket_authorization(request_context, token)
-                            except Exception:
-                                diagnostics.outcome('rejected')
-                                await connection.send_queue.put({'type': 'error', 'message': 'Unauthorized'})
-                                break
-                            await websocket_adapter.handle_websocket_message(connection, data)
-                        elif message_type == 'disconnect':
+                try:
+                    data = await asyncio.to_thread(json.loads, message)
+                    message_type = data.get('type', 'message')
+                    if message_type == 'ping':
+                        await connection.send_queue.put(
+                            {'type': 'pong', 'timestamp': datetime.datetime.now().isoformat()}
+                        )
+                    elif message_type == 'message':
+                        try:
+                            request_context = await self._revalidate_websocket_authorization(request_context, token)
+                        except Exception:
+                            await connection.send_queue.put({'type': 'error', 'message': t('Unauthorized')})
                             break
-                        else:
-                            diagnostics.outcome('skipped')
-                            logger.warning(f'Unknown WebSocket message type: {message_type}')
-                    except json.JSONDecodeError:
-                        diagnostics.outcome('rejected')
-                        await connection.send_queue.put({'type': 'error', 'message': 'Invalid JSON format'})
+                        await websocket_adapter.handle_websocket_message(connection, data)
+                    elif message_type == 'disconnect':
+                        break
+                    else:
+                        logger.warning(f'Unknown WebSocket message type: {message_type}')
+                except json.JSONDecodeError:
+                    await connection.send_queue.put({'type': 'error', 'message': t('Invalid JSON format')})
 
         except Exception:
-            diagnostics.outcome('failed')
             logger.error('Dashboard WebSocket receive error', exc_info=True)
         finally:
             connection.is_active = False
@@ -461,13 +449,11 @@ class WebSocketChatRouterGroup(group.RouterGroup):
                     message = await asyncio.wait_for(connection.send_queue.get(), timeout=1.0)
                     if message is None:
                         break
-                    with diagnostics.scope(self.ap, 'websocket.dashboard.send', source='websocket'):
-                        encoded = await asyncio.to_thread(json.dumps, message)
-                        await quart.websocket.send(encoded)
+                    encoded = await asyncio.to_thread(json.dumps, message)
+                    await quart.websocket.send(encoded)
                 except asyncio.TimeoutError:
                     continue
         except Exception:
-            diagnostics.outcome('failed')
             logger.error('Dashboard WebSocket send error', exc_info=True)
         finally:
             connection.is_active = False
diff --git a/src/langbot/pkg/api/http/controller/groups/platform/adapters.py b/src/langbot/pkg/api/http/controller/groups/platform/adapters.py
index 75a046e8c..d740221f1 100644
--- a/src/langbot/pkg/api/http/controller/groups/platform/adapters.py
+++ b/src/langbot/pkg/api/http/controller/groups/platform/adapters.py
@@ -1,3 +1,5 @@
+from langbot.pkg.i18n import t, DEFAULT_LANGUAGE
+from langbot.pkg.i18n.workspace import get_workspace_language
 import asyncio
 import dataclasses
 import mimetypes
@@ -6,6 +8,7 @@ import quart
 
 from langbot.pkg.api.http.authz import Permission
 from langbot.pkg.api.http.context import RequestContext
+from langbot.pkg.cloud.plan_policy import workspace_entitlement, platform_feature_names, require_workspace_platform
 from langbot.pkg.core.errors import TaskCapacityError
 from langbot.pkg.utils import httpclient, importutil
 
@@ -158,12 +161,98 @@ def _decrypt_qqofficial_secret(encrypted_b64: str, key: bytes) -> str:
 @group.group_class('adapters', '/api/v1/platform/adapters')
 class AdaptersRouterGroup(group.RouterGroup):
     async def initialize(self) -> None:
+        from ....service.slack_setup import get_slack_setup, SlackSetupError
+
+        slack_setup = get_slack_setup(self.ap)
+
+        @self.route(
+            '/slack/setup',
+            methods=['POST'],
+            permission=Permission.RESOURCE_MANAGE,
+            auth_type=group.AuthType.USER_TOKEN_OR_API_KEY,
+        )
+        async def slack_start(request_context: RequestContext):
+            data = await quart.request.get_json()
+            if not isinstance(data, dict):
+                return self.http_status(400, -1, 'invalid_input')
+            try:
+                result = await slack_setup.start(
+                    request_context,
+                    **{
+                        key: data[key]
+                        for key in (
+                            'bot_uuid',
+                            'access_token',
+                            'refresh_token',
+                            'webhook_url',
+                            'name',
+                            'socket_mode',
+                            'redirect_url',
+                        )
+                        if key in data
+                    },
+                )
+                return self.success(data=result)
+            except (SlackSetupError, TypeError) as exc:
+                return self.http_status(400, -1, str(exc) if isinstance(exc, SlackSetupError) else 'invalid_input')
+
+        @self.route(
+            '/slack/setup/<session_id>',
+            methods=['GET', 'DELETE'],
+            permission=Permission.RESOURCE_MANAGE,
+            auth_type=group.AuthType.USER_TOKEN_OR_API_KEY,
+        )
+        async def slack_status(session_id: str, request_context: RequestContext):
+            try:
+                slack_setup.owned(request_context, session_id)
+                if quart.request.method == 'DELETE':
+                    slack_setup.discard(session_id)
+                    return self.success(data={})
+                response = await quart.make_response(self.success(data=slack_setup.status(request_context, session_id)))
+                response.headers['Cache-Control'] = 'no-store'
+                return response
+            except SlackSetupError:
+                return self.http_status(404, -1, 'session_not_found')
+
+        @self.route('/slack/setup/callback', methods=['GET'], auth_type=group.AuthType.NONE)
+        async def slack_callback():
+            session = slack_setup.sessions.get(quart.request.args.get('state', ''))
+            locale = DEFAULT_LANGUAGE
+            if session is not None:
+                locale = await get_workspace_language(self.ap, session['context'].workspace_uuid)
+            try:
+                await slack_setup.finish(
+                    quart.request.args.get('state', ''),
+                    quart.request.args.get('code', ''),
+                    quart.request.args.get('error'),
+                )
+                message = t('Slack authorization completed. Return to LangBot to finish setup.', locale=locale)
+                status = 200
+            except SlackSetupError:
+                message = t('Slack authorization failed or expired. Return to LangBot for details.', locale=locale)
+                status = 400
+            return quart.Response(
+                message,
+                status=status,
+                content_type='text/plain; charset=utf-8',
+                headers={'Cache-Control': 'no-store', 'Referrer-Policy': 'no-referrer'},
+            )
+
         @self.route('', methods=['GET'])
-        async def _() -> str:
-            return self.success(data={'adapters': self.ap.platform_mgr.get_available_adapters_info()})
+        async def _(request_context: RequestContext) -> str:
+            snapshot = await workspace_entitlement(self.ap, request_context.workspace_uuid)
+            adapters = self.ap.platform_mgr.get_available_adapters_info()
+            if snapshot is not None:
+                adapters = [
+                    a
+                    for a in adapters
+                    if not any(snapshot.features.get(f) is False for f in platform_feature_names(a['name']))
+                ]
+            return self.success(data={'adapters': adapters})
 
         @self.route('/<adapter_name>', methods=['GET'])
-        async def _(adapter_name: str) -> str:
+        async def _(adapter_name: str, request_context: RequestContext) -> str:
+            await require_workspace_platform(self.ap, request_context.workspace_uuid, adapter_name)
             adapter_info = self.ap.platform_mgr.get_available_adapter_info_by_name(adapter_name)
 
             if adapter_info is None:
@@ -187,6 +276,15 @@ class AdaptersRouterGroup(group.RouterGroup):
                 importutil.read_resource_file_bytes(icon_path), mimetype=mimetypes.guess_type(icon_path)[0]
             )
 
+        @self.route('/dingtalk/stream-card-template', methods=['GET'], auth_type=group.AuthType.NONE)
+        async def _() -> quart.Response:
+            filename = 'dingtalk_stream_card.json'
+            response = quart.Response(
+                importutil.read_resource_file_bytes(f'templates/{filename}'), mimetype='application/json'
+            )
+            response.headers['Content-Disposition'] = f'attachment; filename={filename}'
+            return response
+
         @self.route('/dingtalk/human-input-card-template', methods=['GET'], auth_type=group.AuthType.NONE)
         async def _() -> quart.Response:
             filename = 'dingtalk_human_input_card.json'
diff --git a/src/langbot/pkg/api/http/controller/groups/plugins.py b/src/langbot/pkg/api/http/controller/groups/plugins.py
index 4bce79869..28024ca5d 100644
--- a/src/langbot/pkg/api/http/controller/groups/plugins.py
+++ b/src/langbot/pkg/api/http/controller/groups/plugins.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from .....i18n import t
+
 import asyncio
 import base64
 import collections.abc
@@ -165,6 +167,16 @@ except Exception:
     pass
 
 
+def _is_plugin_not_installed(exc: BaseException) -> bool:
+    """True when the connector reports a plugin absent from the resolved Workspace.
+
+    The connector raises a plain ``ValueError`` for this case, so resource routes
+    answer 404 instead of surfacing a 500 the WebUI would have to swallow.
+    """
+
+    return 'is not installed in this workspace' in str(exc).lower()
+
+
 def _normalize_plugin_asset_path(filepath: str) -> str | None:
     filepath = filepath.replace('\\', '/')
     if filepath.startswith('/'):
@@ -286,14 +298,20 @@ class PluginsRouterGroup(group.RouterGroup):
     async def _check_extensions_limit(self, request_context: RequestContext) -> str | None:
         """Check if extensions limit is reached. Returns error response if limit exceeded, None otherwise."""
         await self.ap.plugin_connector.require_workspace_context(request_context)
-        limitation = self.ap.instance_config.data.get('system', {}).get('limitation', {})
+        from ...service.workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, request_context.workspace_uuid)
         max_extensions = limitation.get('max_extensions', -1)
         if max_extensions >= 0:
             plugins = await self.ap.plugin_connector.list_plugins()
             mcp_servers = await self.ap.mcp_service.get_mcp_servers(request_context)
             total_extensions = len(plugins) + len(mcp_servers)
             if total_extensions >= max_extensions:
-                return self.http_status(400, -1, f'Maximum number of extensions ({max_extensions}) reached')
+                return self.http_status(
+                    400,
+                    -1,
+                    t('Maximum number of {resource} ({limit}) reached', resource=t('extensions'), limit=max_extensions),
+                )
         return None
 
     @staticmethod
@@ -596,7 +614,15 @@ class PluginsRouterGroup(group.RouterGroup):
             request_context: RequestContext,
         ) -> quart.Response:
             await self._require_authenticated_plugin_runtime_context(request_context)
-            icon_data = await self.ap.plugin_connector.get_plugin_icon(author, plugin_name)
+            try:
+                icon_data = await self.ap.plugin_connector.get_plugin_icon(author, plugin_name)
+            except ValueError as exc:
+                # A plugin that is not installed for this Workspace is a missing
+                # resource, not a server fault: the WebUI asks for icons by plugin
+                # identity and must be able to fall back silently.
+                if _is_plugin_not_installed(exc):
+                    return quart.Response('Icon not found', status=404)
+                raise
             icon_bytes = await asyncio.to_thread(base64.b64decode, icon_data['plugin_icon_base64'])
             return quart.Response(icon_bytes, mimetype=icon_data['mime_type'])
 
diff --git a/src/langbot/pkg/api/http/controller/groups/provider/models.py b/src/langbot/pkg/api/http/controller/groups/provider/models.py
index 32a15e947..678550b69 100644
--- a/src/langbot/pkg/api/http/controller/groups/provider/models.py
+++ b/src/langbot/pkg/api/http/controller/groups/provider/models.py
@@ -14,6 +14,26 @@ MODEL_TEST_ERROR_CODE = 'model_test_failed'
 @group.group_class('models/llm', '/api/v1/provider/models/llm')
 class LLMModelsRouterGroup(group.RouterGroup):
     async def initialize(self) -> None:
+        @self.route('/_/starred', methods=['GET'], permission=Permission.RESOURCE_VIEW)
+        async def get_starred(request_context: RequestContext) -> str:
+            return self.success(data={'uuid': await self.ap.llm_model_service.get_starred_model(request_context)})
+
+        @self.route('/_/starred', methods=['PUT'], permission=Permission.PROVIDER_SECRET_MANAGE)
+        async def set_starred(request_context: RequestContext) -> str:
+            body = await quart.request.json
+            model_uuid = body.get('uuid')
+            if model_uuid is not None and not isinstance(model_uuid, str):
+                return self.http_status(400, -1, 'uuid must be a string or null')
+            try:
+                await self.ap.llm_model_service.set_starred_model(request_context, model_uuid)
+            except ValueError as exc:
+                return self.http_status(400, -1, str(exc))
+            return self.success(data={})
+
+        @self.route('/_/default', methods=['GET'], permission=Permission.RESOURCE_MANAGE)
+        async def get_default(request_context: RequestContext) -> str:
+            return self.success(data={'uuid': await self.ap.llm_model_service.get_default_model(request_context)})
+
         @self.route(
             '',
             methods=['GET'],
diff --git a/src/langbot/pkg/api/http/controller/groups/system.py b/src/langbot/pkg/api/http/controller/groups/system.py
index 9efed6c77..56ca4f791 100644
--- a/src/langbot/pkg/api/http/controller/groups/system.py
+++ b/src/langbot/pkg/api/http/controller/groups/system.py
@@ -200,11 +200,25 @@ class SystemRouterGroup(group.RouterGroup):
             if invitation_delivery_service is None:
                 invitation_delivery_service = InvitationDeliveryService(self.ap)
 
+            # Release channel is independent of diagnostics or usage telemetry.
+            from packaging.version import Version, InvalidVersion
+
+            try:
+                release = Version(constants.semantic_version)
+                is_beta = bool(release.pre and release.pre[0] == 'b' and not release.local)
+            except InvalidVersion:
+                is_beta = False
+
+            # Detect Cloud deployment mode
+            is_cloud = getattr(getattr(self.ap, 'deployment', None), 'mode', 'oss') == 'cloud'
+
             return self.success(
                 data={
                     'version': constants.semantic_version,
                     'debug': constants.debug_mode,
                     'edition': constants.edition,
+                    'beta': is_beta,
+                    'deployment_mode': 'cloud' if is_cloud else 'oss',
                     'enable_marketplace': self.ap.instance_config.data.get('plugin', {}).get(
                         'enable_marketplace', True
                     ),
diff --git a/src/langbot/pkg/api/http/controller/groups/user.py b/src/langbot/pkg/api/http/controller/groups/user.py
index 240d7a213..ee2ee7dc9 100644
--- a/src/langbot/pkg/api/http/controller/groups/user.py
+++ b/src/langbot/pkg/api/http/controller/groups/user.py
@@ -267,8 +267,8 @@ class UserRouterGroup(group.RouterGroup):
 
             return self.success(data={'user': user_email})
 
-        @self.route('/change-password', methods=['POST'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str) -> str:
+        @self.route('/change-password', methods=['POST'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             # Check if password change is allowed
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
                 'allow_modify_login_info', True
@@ -282,13 +282,13 @@ class UserRouterGroup(group.RouterGroup):
             new_password = json_data['new_password']
 
             try:
-                await self.ap.user_service.change_password(user_email, current_password, new_password)
+                await self.ap.user_service.change_password(account.normalized_email, current_password, new_password)
             except argon2.exceptions.VerifyMismatchError:
                 return self.http_status(400, -1, 'Current password is incorrect')
             except ValueError as e:
                 return self.http_status(400, -1, str(e))
 
-            return self.success(data={'user': user_email})
+            return self.success(data={'user': account.user})
 
         # Space OAuth endpoints (redirect flow)
 
@@ -323,19 +323,19 @@ class UserRouterGroup(group.RouterGroup):
             except ValueError as e:
                 return self.fail(1, str(e))
 
-        @self.route('/space/bind-authorize-url', methods=['GET'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(request_context: RequestContext) -> str:
+        @self.route('/space/bind-authorize-url', methods=['GET'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             """Issue an account-bound, one-time Space OAuth redirect."""
             redirect_uri = quart.request.args.get('redirect_uri', '')
             if not redirect_uri:
                 return self.fail(1, 'Missing redirect_uri parameter')
-            if not request_context.account_uuid:
+            if not account.uuid:
                 return self.http_status(403, 'account_required', 'An Account is required')
             try:
                 redirect_uri = self._validate_space_redirect_uri(redirect_uri, bind=True)
                 state = await self.ap.user_service.issue_space_oauth_state(
                     'bind',
-                    account_uuid=request_context.account_uuid,
+                    account_uuid=account.uuid,
                 )
                 authorize_url = self.ap.space_service.get_oauth_authorize_url(redirect_uri, state)
                 return self.success(data={'authorize_url': authorize_url})
@@ -507,8 +507,8 @@ class UserRouterGroup(group.RouterGroup):
             capabilities['passkey_supported'] = True
             return self.success(data={'initialized': True, **capabilities})
 
-        @self.route('/set-password', methods=['POST'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str) -> str:
+        @self.route('/set-password', methods=['POST'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             """Set password for Space account (first time) or change password"""
             # Check if modifying login info is allowed
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
@@ -524,13 +524,9 @@ class UserRouterGroup(group.RouterGroup):
             if not new_password:
                 return self.http_status(400, -1, 'New password is required')
 
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
             try:
-                await self.ap.user_service.set_password(user_email, new_password, current_password)
-                return self.success(data={'user': user_email})
+                await self.ap.user_service.set_password(account.normalized_email, new_password, current_password)
+                return self.success(data={'user': account.user})
             except ValueError as e:
                 return self.http_status(400, -1, str(e))
             except argon2.exceptions.VerifyMismatchError:
@@ -595,8 +591,12 @@ class UserRouterGroup(group.RouterGroup):
             except Exception:
                 raise
 
-        @self.route('/passkey/register/options', methods=['POST'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str) -> str:
+        # Passkeys belong to the Account, not to a Workspace: these routes are
+        # ACCOUNT_TOKEN scoped for the same reason as the second-factor routes
+        # below. Resolving the caller through the login name would also miss every
+        # Account whose login name is not its email.
+        @self.route('/passkey/register/options', methods=['POST'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             """Generate WebAuthn registration options for current account."""
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
                 'allow_modify_login_info', True
@@ -604,16 +604,12 @@ class UserRouterGroup(group.RouterGroup):
             if not allow_modify_login_info:
                 return self.http_status(403, -1, 'Modifying login info is disabled')
 
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
             json_data = (await quart.request.json) or {}
             origin, rp_id = self._extract_origin_and_rp_id(json_data)
 
             try:
                 options, challenge_token = await self.ap.user_service.generate_passkey_registration_options(
-                    account_uuid=user_obj.uuid,
+                    account_uuid=account.uuid,
                     rp_id=rp_id,
                     origin=origin,
                     rp_name='LangBot',
@@ -622,8 +618,8 @@ class UserRouterGroup(group.RouterGroup):
             except Exception as e:
                 return self.fail(1, str(e))
 
-        @self.route('/passkey/register/verify', methods=['POST'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str) -> str:
+        @self.route('/passkey/register/verify', methods=['POST'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             """Verify WebAuthn registration response and save credential."""
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
                 'allow_modify_login_info', True
@@ -631,10 +627,6 @@ class UserRouterGroup(group.RouterGroup):
             if not allow_modify_login_info:
                 return self.http_status(403, -1, 'Modifying login info is disabled')
 
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
             json_data = await quart.request.json
             challenge_token = json_data.get('challenge_token')
             credential = json_data.get('credential') or json_data.get('response')
@@ -700,14 +692,10 @@ class UserRouterGroup(group.RouterGroup):
             except Exception as e:
                 return self.fail(1, str(e))
 
-        @self.route('/passkeys', methods=['GET'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str) -> str:
+        @self.route('/passkeys', methods=['GET'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account) -> str:
             """List registered passkeys for the current user."""
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
-            passkeys = await self.ap.user_service.get_user_passkeys(user_obj.uuid)
+            passkeys = await self.ap.user_service.get_user_passkeys(account.uuid)
             return self.success(
                 data=[
                     {
@@ -723,8 +711,8 @@ class UserRouterGroup(group.RouterGroup):
                 ]
             )
 
-        @self.route('/passkey/<passkey_uuid>', methods=['PATCH'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str, passkey_uuid: str) -> str:
+        @self.route('/passkey/<passkey_uuid>', methods=['PATCH'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account, passkey_uuid: str) -> str:
             """Rename a registered passkey."""
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
                 'allow_modify_login_info', True
@@ -732,17 +720,13 @@ class UserRouterGroup(group.RouterGroup):
             if not allow_modify_login_info:
                 return self.http_status(403, -1, 'Modifying login info is disabled')
 
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
             json_data = await quart.request.json
             name = (json_data.get('name') or '').strip()
             if not name:
                 return self.fail(1, 'Passkey name cannot be empty')
 
             updated = await self.ap.user_service.rename_user_passkey(
-                account_uuid=user_obj.uuid,
+                account_uuid=account.uuid,
                 passkey_uuid=passkey_uuid,
                 new_name=name,
             )
@@ -750,8 +734,8 @@ class UserRouterGroup(group.RouterGroup):
                 return self.http_status(404, -1, 'Passkey not found')
             return self.success(data={'uuid': updated.uuid, 'name': updated.name})
 
-        @self.route('/passkey/<passkey_uuid>', methods=['DELETE'], auth_type=group.AuthType.USER_TOKEN)
-        async def _(user_email: str, passkey_uuid: str) -> str:
+        @self.route('/passkey/<passkey_uuid>', methods=['DELETE'], auth_type=group.AuthType.ACCOUNT_TOKEN)
+        async def _(account, passkey_uuid: str) -> str:
             """Delete/revoke a registered passkey."""
             allow_modify_login_info = self.ap.instance_config.data.get('system', {}).get(
                 'allow_modify_login_info', True
@@ -759,12 +743,8 @@ class UserRouterGroup(group.RouterGroup):
             if not allow_modify_login_info:
                 return self.http_status(403, -1, 'Modifying login info is disabled')
 
-            user_obj = await self.ap.user_service.get_user_by_email(user_email)
-            if user_obj is None:
-                return self.http_status(404, -1, 'User not found')
-
             deleted = await self.ap.user_service.delete_user_passkey(
-                account_uuid=user_obj.uuid,
+                account_uuid=account.uuid,
                 passkey_uuid=passkey_uuid,
             )
             if not deleted:
@@ -914,22 +894,52 @@ class UserRouterGroup(group.RouterGroup):
             if access.membership.role not in ('owner', 'admin'):
                 raise PermissionError('Only Workspace owners and admins may manage other Accounts')
 
+        async def _require_target_in_workspace(
+            request_context: RequestContext,
+            target_account_uuid: str,
+        ) -> bool:
+            """Report whether the target Account belongs to the caller's Workspace.
+
+            A cross-Workspace Account is reported as absent, so the response never
+            confirms another tenant's Account existence.
+            """
+
+            try:
+                await self.ap.workspace_collaboration_service.resolve_account_workspace(
+                    target_account_uuid,
+                    request_context.workspace_uuid,
+                )
+            except Exception:
+                return False
+            return True
+
         @self.route('/totp/accounts', methods=['GET'], auth_type=group.AuthType.USER_TOKEN)
         async def _(request_context: RequestContext) -> str:
-            """List the second-factor state of every Account for owners/admins.
+            """List the second-factor state of the caller's Workspace Accounts.
 
-            Oversight is deliberately instance-wide: managing the second factor
-            of any Account is an owner/admin responsibility and is not scoped to
-            the caller's Workspace.
+            Oversight is a Workspace responsibility: owners and admins answer for
+            the Accounts that belong to their own Workspace and must never see
+            another tenant's Accounts.
             """
             try:
-                await _require_workspace_manager(request_context)
+                access = await self.ap.workspace_collaboration_service.resolve_account_workspace(
+                    request_context.account_uuid,
+                    request_context.workspace_uuid,
+                )
+                if access.membership.role not in ('owner', 'admin'):
+                    raise PermissionError('Only Workspace owners and admins may manage other Accounts')
+                members = await self.ap.workspace_collaboration_service.list_members(
+                    request_context.workspace_uuid,
+                    access.membership,
+                )
             except PermissionError as e:
                 return self.http_status(403, 'permission_denied', str(e))
             except Exception:
                 return self.http_status(403, 'permission_denied', 'Not permitted')
 
-            accounts = await self.ap.totp_service.list_account_states()
+            accounts = await self.ap.totp_service.list_account_states(
+                account_uuids=[member.membership.account_uuid for member in members],
+            )
             return self.success(data={'accounts': accounts})
 
         @self.route('/totp/accounts/<target_account_uuid>', methods=['DELETE'], auth_type=group.AuthType.USER_TOKEN)
@@ -948,6 +958,9 @@ class UserRouterGroup(group.RouterGroup):
             except Exception:
                 return self.http_status(403, 'permission_denied', 'Not permitted')
 
+            if not await _require_target_in_workspace(request_context, target_account_uuid):
+                return self.http_status(404, 'account_not_found', 'Account not found')
+
             revoked = await self.ap.totp_service.revoke_for_account(target_account_uuid)
             if not revoked:
                 return self.http_status(404, 'totp_not_enrolled', 'TOTP is not enabled for that Account')
@@ -978,6 +991,9 @@ class UserRouterGroup(group.RouterGroup):
             except Exception:
                 return self.http_status(403, 'permission_denied', 'Not permitted')
 
+            if not await _require_target_in_workspace(request_context, target_account_uuid):
+                return self.http_status(404, 'account_not_found', 'Account not found')
+
             target = await self.ap.totp_service.get_account(target_account_uuid)
             if target is None:
                 return self.http_status(404, 'account_not_found', 'Account not found')
@@ -1022,6 +1038,9 @@ class UserRouterGroup(group.RouterGroup):
             if not isinstance(code, str) or not code:
                 return self.fail(1, 'Missing code parameter')
 
+            if not await _require_target_in_workspace(request_context, target_account_uuid):
+                return self.http_status(404, 'account_not_found', 'Account not found')
+
             try:
                 result = await self.ap.totp_service.confirm_enrollment(target_account_uuid, code)
             except totp_module.TotpNotEnrolledError as e:
diff --git a/src/langbot/pkg/api/http/controller/groups/webhooks.py b/src/langbot/pkg/api/http/controller/groups/webhooks.py
index dbb9626cf..cbc886b15 100644
--- a/src/langbot/pkg/api/http/controller/groups/webhooks.py
+++ b/src/langbot/pkg/api/http/controller/groups/webhooks.py
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import quart
+import quart_cors
 import traceback
 
 from .. import group
@@ -10,6 +11,35 @@ from .....utils import bounded_executor
 @group.group_class('webhooks', '/bots')
 class WebhookRouterGroup(group.RouterGroup):
     async def initialize(self) -> None:
+        @self.quart_app.websocket(self.path + '/<bot_uuid>/onebot/v11/ws')
+        @quart_cors.cors_exempt
+        async def onebot_websocket(bot_uuid: str):
+            # Protocol clients omit Origin; bearer authentication protects this endpoint.
+            from .....platform.adapters.aiocqhttp.adapter import AiocqhttpAdapter
+
+            runtime = await self.ap.platform_mgr.resolve_public_bot(bot_uuid)
+            if runtime is None or not runtime.enable or not isinstance(runtime.adapter, AiocqhttpAdapter):
+                quart.abort(404)
+            adapter = runtime.adapter
+            socket = quart.websocket._get_current_object()
+            if adapter.config['connection_mode'] != 'reverse_ws':
+                quart.abort(409)
+            if not adapter.authenticate(socket.headers):
+                quart.abort(401)
+            if socket.headers.get('X-Client-Role', '').lower() != 'universal':
+                quart.abort(400)
+            self_id = socket.headers.get('X-Self-ID', '')
+            expected = str(adapter.config.get('self_id', '') or '')
+            if not self_id or (expected and self_id != expected):
+                quart.abort(400)
+            if adapter.bot.connection is not None:
+                quart.abort(409)
+            await socket.accept()
+            try:
+                await adapter.serve_connection(socket.send, socket.receive, socket.close, self_id)
+            except Exception:
+                await socket.close(1011)
+
         @self.route('/<bot_uuid>', methods=['GET', 'POST'], auth_type=group.AuthType.NONE)
         async def handle_webhook(bot_uuid: str):
             """处理 bot webhook 回调（无子路径）"""
@@ -31,6 +61,11 @@ class WebhookRouterGroup(group.RouterGroup):
             适配器返回的响应
         """
         try:
+            setup = getattr(self.ap, 'slack_setup_service', None)
+            if setup is not None and not path and quart.request.method == 'POST':
+                challenge = await setup.challenge(bot_uuid, quart.request)
+                if challenge is not None:
+                    return quart.jsonify(challenge)
             # Public ingress never accepts X-Workspace-Id.  The opaque bot UUID
             # is resolved against the already-bound runtime resource, which
             # carries the trusted Workspace and placement generation.
diff --git a/src/langbot/pkg/api/http/controller/groups/workspaces.py b/src/langbot/pkg/api/http/controller/groups/workspaces.py
index 7873e9f17..4f653a2e3 100644
--- a/src/langbot/pkg/api/http/controller/groups/workspaces.py
+++ b/src/langbot/pkg/api/http/controller/groups/workspaces.py
@@ -14,6 +14,8 @@ from .....workspace.errors import WorkspaceNotFoundError
 from .....workspace.invitation_delivery import InvitationDeliveryService
 from .....operation_trace import service as settings_service
 from .. import group
+from .....i18n import SUPPORTED_LANGUAGES
+from .....i18n.workspace import get_workspace_language, set_workspace_language
 
 
 def _workspace_payload(workspace: Workspace) -> dict[str, typing.Any]:
@@ -74,6 +76,34 @@ class WorkspacesRouterGroup(group.RouterGroup):
         return await operation()
 
     async def initialize(self) -> None:
+        @self.route(
+            '/current/language',
+            methods=['GET'],
+            auth_type=group.AuthType.USER_TOKEN_OR_API_KEY,
+            permission=Permission.WORKSPACE_VIEW,
+        )
+        async def get_language(request_context: RequestContext):
+            return self.success(
+                data={
+                    'language': await get_workspace_language(self.ap, request_context.workspace_uuid, refresh=True),
+                    'supported_languages': list(SUPPORTED_LANGUAGES),
+                }
+            )
+
+        @self.route(
+            '/current/language',
+            methods=['PUT'],
+            auth_type=group.AuthType.USER_TOKEN_OR_API_KEY,
+            permission=Permission.WORKSPACE_UPDATE,
+        )
+        async def update_language(request_context: RequestContext):
+            body = await quart.request.get_json()
+            locale = body.get('language') if isinstance(body, dict) else None
+            if locale not in SUPPORTED_LANGUAGES:
+                return self.http_status(400, 'invalid_language', 'Unsupported Workspace language')
+            locale = await set_workspace_language(self.ap, request_context.workspace_uuid, locale)
+            return self.success(data={'language': locale})
+
         @self.route('/bootstrap', methods=['GET'], auth_type=group.AuthType.ACCOUNT_TOKEN)
         async def _(account) -> typing.Any:
             """List the active Workspaces available to an authenticated Account.
@@ -129,6 +159,8 @@ class WorkspacesRouterGroup(group.RouterGroup):
             membership = quart.g.workspace_membership
             workspace = await self.ap.workspace_service.get_workspace(request_context.workspace_uuid)
             plan_name: str | None = None
+            limitation = None
+            current_plan = None
             resolver = getattr(self.ap, 'entitlement_resolver', None)
             if workspace.source == WorkspaceSource.CLOUD_PROJECTION.value and resolver is not None:
                 entitlement = await resolver.resolve(
@@ -136,6 +168,18 @@ class WorkspacesRouterGroup(group.RouterGroup):
                     minimum_revision=request_context.entitlement_revision,
                 )
                 plan_name = entitlement.plan_name
+                if entitlement.plan_revision is not None and entitlement.monthly_bonus_credits is not None:
+                    current_plan = {
+                        'name': entitlement.plan_name,
+                        'subscription_started_at': entitlement.subscription_started_at,
+                        'revision': entitlement.plan_revision,
+                        'monthly_bonus_credits': entitlement.monthly_bonus_credits,
+                        'limits': entitlement.limits,
+                        'entitlement_features': entitlement.features,
+                    }
+                from ...service.workspace_limits import LIMIT_KEYS
+
+                limitation = {key: entitlement.limit(value) for key, value in LIMIT_KEYS.items()}
             if request_context.principal.principal_type == PrincipalType.SUPPORT_ADMIN:
                 return self.success(
                     data={
@@ -154,6 +198,8 @@ class WorkspacesRouterGroup(group.RouterGroup):
                         'permissions': sorted(request_context.workspace.permissions),
                         'placement_generation': request_context.placement_generation,
                         'plan_name': plan_name,
+                        'current_plan': current_plan,
+                        'limitation': limitation,
                     }
                 )
             account = await self.ap.user_service.get_user_by_uuid(request_context.account_uuid)
@@ -170,6 +216,8 @@ class WorkspacesRouterGroup(group.RouterGroup):
                     'permissions': sorted(request_context.workspace.permissions),
                     'placement_generation': request_context.placement_generation,
                     'plan_name': plan_name,
+                    'current_plan': current_plan,
+                    'limitation': limitation,
                 }
             )
 
diff --git a/src/langbot/pkg/api/http/controller/main.py b/src/langbot/pkg/api/http/controller/main.py
index e19552c65..dfe8a685a 100644
--- a/src/langbot/pkg/api/http/controller/main.py
+++ b/src/langbot/pkg/api/http/controller/main.py
@@ -174,9 +174,21 @@ class HTTPController:
 
         frontend_path = paths.get_frontend_path()
 
+        async def frontend_index():
+            from .social_preview import cloud_preview_html
+
+            response = await quart.send_from_directory(
+                frontend_path, 'index.html', mimetype='text/html', conditional=False
+            )
+            if quart.request.host.split(':', 1)[0].lower() == 'cloud.langbot.app':
+                response.set_data(cloud_preview_html(await response.get_data(as_text=True), quart.request.host))
+                response.headers.pop('ETag', None)
+                response.headers.pop('Last-Modified', None)
+            return response
+
         @self.quart_app.route('/')
         async def index():
-            response = await quart.send_from_directory(frontend_path, 'index.html', mimetype='text/html')
+            response = await frontend_index()
             response.headers['Cache-Control'] = 'no-cache, no-store, must-revalidate'
             response.headers['Pragma'] = 'no-cache'
             response.headers['Expires'] = '0'
@@ -207,7 +219,7 @@ class HTTPController:
                                 return response
 
                     # Fallback to index.html for SPA client-side routing
-                    response = await quart.send_from_directory(frontend_path, 'index.html', mimetype='text/html')
+                    response = await frontend_index()
                     response.headers['Cache-Control'] = 'no-cache, no-store, must-revalidate'
                     response.headers['Pragma'] = 'no-cache'
                     response.headers['Expires'] = '0'
diff --git a/src/langbot/pkg/api/http/controller/social_preview.py b/src/langbot/pkg/api/http/controller/social_preview.py
new file mode 100644
index 000000000..03e530ee0
--- /dev/null
+++ b/src/langbot/pkg/api/http/controller/social_preview.py
@@ -0,0 +1,20 @@
+"""Crawler-visible preview for the hosted Cloud frontend only."""
+
+CLOUD_IMAGE = 'https://cloud.langbot.app/social/cloud-v3.png'
+
+
+def cloud_preview_html(html: str, hostname: str) -> str:
+    if hostname.split(':', 1)[0].lower() != 'cloud.langbot.app':
+        return html
+    tags = (
+        '<meta property="og:type" content="website">'
+        '<meta property="og:title" content="LangBot Cloud">'
+        '<meta property="og:description" content="Run LangBot in the cloud.">'
+        f'<meta property="og:image" content="{CLOUD_IMAGE}">'
+        '<meta property="og:image:width" content="1200">'
+        '<meta property="og:image:height" content="630">'
+        '<meta name="twitter:card" content="summary_large_image">'
+        '<meta name="twitter:title" content="LangBot Cloud">'
+        f'<meta name="twitter:image" content="{CLOUD_IMAGE}">'
+    )
+    return html.replace('</head>', tags + '</head>', 1)
diff --git a/src/langbot/pkg/api/http/service/agent.py b/src/langbot/pkg/api/http/service/agent.py
index 62a67c3ac..c0f0b34bf 100644
--- a/src/langbot/pkg/api/http/service/agent.py
+++ b/src/langbot/pkg/api/http/service/agent.py
@@ -1,12 +1,15 @@
 from __future__ import annotations
 
+from ....i18n import t
+
+from contextlib import aclosing
+
 import datetime
 import copy
 import fnmatch
 import time
 import uuid
 import typing
-from ... import management_diagnostics as diagnostics
 
 import sqlalchemy
 from langbot_plugin.api.entities.builtin.runner.delivery import DeliveryContext
@@ -46,7 +49,7 @@ from .tenant import TenantContext, require_workspace_uuid, scope_statement
 AGENT_KIND_AGENT = 'agent'
 AGENT_KIND_PIPELINE = 'pipeline'
 AGENT_KIND_EVENT_PROCESSOR = 'event_processor'
-PIPELINE_EVENT_PATTERNS = ['message.*']
+PIPELINE_EVENT_PATTERNS = ['message.received']
 AGENT_DEFAULT_EVENT_PATTERNS = ['*']
 
 
@@ -135,7 +138,6 @@ class AgentService:
 
         return None
 
-    @diagnostics.observe('http.agent.debug_agent', source='webui_debug')
     async def debug_agent(
         self,
         context: RequestContext,
@@ -150,7 +152,6 @@ class AgentService:
         delivers outputs to a real platform, and supports both message and
         non-message event envelopes.
         """
-        diagnostics.workspace(context)
         agent = await self.get_agent(context, agent_uuid)
         if agent is None or agent.get('kind') not in {AGENT_KIND_AGENT, AGENT_KIND_EVENT_PROCESSOR}:
             raise ValueError('Agent not found')
@@ -351,25 +352,45 @@ class AgentService:
                 execution_events.append(visible_result)
 
         final_text = ''
-        async for output in self.ap.agent_run_orchestrator.run(
-            event,
-            binding,
-            adapter_context={
-                '_query': execution_query,
-                '_execution_context': execution_context,
-                '_result_observer': observe_result,
-            },
+        # A debug run has no platform event: the synthetic event envelope is the
+        # execution's inbound boundary, so the chain is opened here with the real
+        # event id and marked as a debug WebUI/API execution. The Runner inside
+        # then joins this chain instead of becoming a trace origin of its own.
+        from ....telemetry.execution import ingress as execution_ingress
+
+        principal_type = str(getattr(getattr(context, 'principal', None), 'principal_type', '') or '')
+        origin = 'api' if principal_type == 'api_key' else 'webui'
+        with execution_ingress(
+            self.ap,
+            'event_done',
+            execution_context,
+            event_id,
+            debug=True,
+            origin=origin,
+            synthetic_event=event_type,
         ):
-            output_text = self._provider_output_to_text(output)
-            if output_text:
-                final_text = output_text
-            output_items.append(
-                {
-                    'kind': output.__class__.__name__,
-                    'role': str(getattr(output, 'role', '') or ''),
-                    'text': output_text,
-                }
-            )
+            async with aclosing(
+                self.ap.agent_run_orchestrator.run(
+                    event,
+                    binding,
+                    adapter_context={
+                        '_query': execution_query,
+                        '_execution_context': execution_context,
+                        '_result_observer': observe_result,
+                    },
+                )
+            ) as owned_stream:
+                async for output in owned_stream:
+                    output_text = self._provider_output_to_text(output)
+                    if output_text:
+                        final_text = output_text
+                    output_items.append(
+                        {
+                            'kind': output.__class__.__name__,
+                            'role': str(getattr(output, 'role', '') or ''),
+                            'text': output_text,
+                        }
+                    )
 
         return {
             'event_id': event_id,
@@ -413,7 +434,7 @@ class AgentService:
             pipeline_uuid = await self.ap.pipeline_service.create_pipeline(
                 context,
                 {
-                    'name': agent_data.get('name') or 'New Pipeline',
+                    'name': agent_data.get('name') or t('New Pipeline'),
                     'description': agent_data.get('description') or '',
                     'emoji': agent_data.get('emoji') or '⚙️',
                     'config': {},
@@ -430,12 +451,19 @@ class AgentService:
             config = agent_data['config'] if 'config' in agent_data else await self._get_default_agent_config(context)
             config, runner_id, _ = RunnerConfigResolver.resolve_agent_config(config)
             await self._validate_runner_for_agent(context, runner_id)
+            if runner_id:
+                descriptor = await self.ap.runner_registry.get(context, runner_id)
+                config['runner_config'][runner_id] = await self.ap.llm_model_service.apply_default_model(
+                    context,
+                    descriptor.config_schema,
+                    config['runner_config'][runner_id],
+                )
             patterns = agent_data.get('supported_event_patterns', AGENT_DEFAULT_EVENT_PATTERNS)
         new_uuid = str(uuid.uuid4())
         values = {
             'workspace_uuid': workspace_uuid,
             'uuid': new_uuid,
-            'name': agent_data.get('name') or 'New Agent',
+            'name': agent_data.get('name') or t('New Agent'),
             'description': agent_data.get('description') or '',
             'emoji': agent_data.get('emoji') or ('🧩' if kind == AGENT_KIND_EVENT_PROCESSOR else '🤖'),
             'kind': kind,
@@ -512,6 +540,7 @@ class AgentService:
 
     async def _prepare_event_processor(self, context, data, existing=None):
         """Resolve an installed component and keep its capability declaration authoritative."""
+        draft = existing is None and 'config' not in data and 'parameters' not in data
         config = copy.deepcopy(data.get('config', existing.config if existing is not None else {}))
         if not isinstance(config, dict):
             raise ValueError('Processor configuration must be an object')
@@ -540,13 +569,19 @@ class AgentService:
             parameters = runner_config.get(component_ref)
         if parameters is None:
             parameters = self.ap.pipeline_service._get_default_values_from_schema(descriptor.config_schema)
+        if draft:
+            parameters = await self.ap.llm_model_service.apply_default_model(
+                context, descriptor.config_schema, parameters
+            )
         if not isinstance(parameters, dict):
             raise ValueError('Processor parameters must be an object')
         for field in descriptor.config_schema:
-            if field.get('required') and parameters.get(field['name']) in (None, ''):
+            if not draft and field.get('required') and parameters.get(field['name']) in (None, ''):
                 raise ValueError(f'Required processor parameter: {field["name"]}')
         config['runner_config'] = {component_ref: parameters}
-        return config, component_ref, descriptor.supported_event_patterns
+        # Component-only creation keeps defaults for the detail form, but does
+        # not subscribe to events until configuration is saved and validated.
+        return config, component_ref, [] if draft else descriptor.supported_event_patterns
 
     async def get_processor_runs(self, context, processor_id, *, before_id=None):
         """Read only this Workspace's Agent or plugin processor runs."""
diff --git a/src/langbot/pkg/api/http/service/assistant.py b/src/langbot/pkg/api/http/service/assistant.py
index 74e7073ae..033b53750 100644
--- a/src/langbot/pkg/api/http/service/assistant.py
+++ b/src/langbot/pkg/api/http/service/assistant.py
@@ -1,6 +1,7 @@
 """In-process management assistant using the existing model and resource services."""
 
 import asyncio
+import contextlib
 import json
 import uuid
 
@@ -44,6 +45,187 @@ class AssistantService:
         self.ap = ap
         # ponytail: per-process admission; shared quotas if multiple workers need a global ceiling.
         self._slots = asyncio.Semaphore(4)
+        self._runs = {}
+
+    async def list_conversations(self, context):
+        self._scope(context, '')
+        result = await self.ap.persistence_mgr.execute_async(
+            sa.select(Conversation)
+            .where(
+                Conversation.workspace_uuid == context.workspace_uuid,
+                Conversation.account_uuid == context.account_uuid,
+            )
+            .order_by(Conversation.updated_at.desc(), Conversation.uuid)
+            .limit(100)
+        )
+        return [
+            dict(
+                uuid=row['uuid'],
+                status=row['status'],
+                revision=row['revision'],
+                title=next(
+                    (
+                        m.get('content', '')[:80]
+                        for m in row['messages']
+                        if m['role'] == 'user' and isinstance(m.get('content'), str)
+                    ),
+                    '',
+                ),
+            )
+            for row in result.mappings()
+        ]
+
+    async def progress(self, context, conversation_id):
+        conversation = await self.get(context, conversation_id)
+        run = self._runs.get(conversation_id)
+        if run:
+            return {**(run['snapshot'] or self.public_view(conversation)), 'progress': run['progress']}
+        return self.public_view(conversation)
+
+    def start_run(self, context, conversation_id, body):
+        """Own execution independently of any HTTP subscriber; never block on a slow browser."""
+        if conversation_id in self._runs:
+            raise AssistantError('stale_turn')
+        if len(self._runs) >= 4:
+            raise AssistantError('busy', 429)
+        queue = asyncio.Queue(maxsize=32)
+        run = {
+            'queue': queue,
+            'snapshot': None,
+            'progress': {},
+            'stopping': False,
+            'revision': body.revision + 1,
+            'started': asyncio.Event(),
+        }
+        self._runs[conversation_id] = run
+
+        async def emit(frame):
+            if frame['kind'] == 'snapshot':
+                run['snapshot'] = frame['data']
+                run['progress'] = {}
+            elif frame['kind'] == 'text':
+                run['progress']['text'] = frame['data']['text']
+            elif frame['kind'] == 'phase':
+                run['progress'].update(frame['data'])
+            if queue.full():
+                queue.get_nowait()
+            queue.put_nowait(frame)
+
+        async def execute():
+            run['started'].set()
+            tenant_scope = getattr(self.ap.persistence_mgr, 'tenant_scope', None)
+            scope = tenant_scope(context.workspace_uuid) if callable(tenant_scope) else contextlib.nullcontext()
+            try:
+                async with scope:
+                    try:
+                        conversation = await self.turn(
+                            context,
+                            conversation_id,
+                            body.revision,
+                            body.text,
+                            body.approved,
+                            str(body.model_uuid) if body.model_uuid else None,
+                            on_event=emit,
+                        )
+                    except asyncio.CancelledError:
+                        conversation = await self.get(context, conversation_id)
+                        # Preserve interrupted tool outcomes; stopping cannot undo an external write.
+                        if run['stopping'] and (
+                            conversation['error'] == 'result_unknown'
+                            or (conversation['status'] == 'running' and conversation['revision'] == run['revision'])
+                        ):
+                            await self._save(context, conversation, 'failed', 'stopped')
+                    await emit({'kind': 'completed', 'data': self.public_view(conversation)})
+            except AssistantError as exc:
+                await emit({'kind': 'error', 'code': exc.code})
+            except Exception:
+                self.ap.logger.exception('Assistant background run failed')
+                await emit({'kind': 'error', 'code': 'turn_failed'})
+            finally:
+                self._runs.pop(conversation_id, None)
+
+        run['task'] = asyncio.create_task(execute())
+        return queue
+
+    async def stop(self, context, conversation_id, revision):
+        require_permission(context, Permission.RUNTIME_OPERATE)
+        conversation = await self.get(context, conversation_id)
+        run = self._runs.get(conversation_id)
+        if run and run['revision'] != revision:
+            raise AssistantError('stale_turn')
+        if not run:
+            if conversation['revision'] != revision:
+                raise AssistantError('stale_turn')
+            return self.public_view(conversation)
+        await run['started'].wait()
+        if not run['stopping']:
+            run['stopping'] = True
+            run['task'].cancel()
+        await asyncio.shield(run['task'])
+        return self.public_view(await self.get(context, conversation_id))
+
+    async def recommended_model(self, context):
+        require_permission(context, Permission.RUNTIME_OPERATE)
+        try:
+            recommended = await self.ap.space_service.get_recommended_assistant_model(context)
+            model = await self.ap.model_mgr.get_model_by_uuid(
+                ExecutionContext.from_request(context), recommended['uuid']
+            )
+            if 'func_call' not in (model.model_entity.abilities or []):
+                raise ValueError('Recommended model does not support tools')
+            return {'uuid': recommended['uuid'], 'name': model.model_entity.name}
+        except Exception as exc:
+            raise AssistantError('model_unavailable', 503) from exc
+
+    async def _invoke(self, model, kwargs, on_event):
+        if on_event is None:
+            return await model.provider.invoke_llm(**kwargs)
+        message = {'role': 'assistant', 'content': ''}
+        calls = {}
+        finished = False
+        async for chunk in model.provider.invoke_llm_stream(**kwargs):
+            finished = finished or chunk.is_final
+            content = chunk.content or ''
+            if isinstance(content, list):
+                content = ''.join(item.text or '' for item in content if item.type == 'text')
+            if chunk.all_content is not None:
+                message['content'] = chunk.all_content
+            else:
+                message['content'] += content
+            for field in ('resp_message_id', 'provider_specific_fields'):
+                value = getattr(chunk, field, None)
+                if value is not None:
+                    if field == 'provider_specific_fields':
+                        message[field] = {**message.get(field, {}), **value}
+                    else:
+                        message[field] = value
+            for call in chunk.tool_calls or []:
+                value = call.model_dump(mode='json')
+                if call.id in calls:
+                    saved = calls[call.id]
+                    saved['function']['arguments'] += call.function.arguments or ''
+                    if call.function.name:
+                        saved['function']['name'] = call.function.name
+                    if value.get('provider_specific_fields'):
+                        saved['provider_specific_fields'] = {
+                            **(saved.get('provider_specific_fields') or {}),
+                            **value['provider_specific_fields'],
+                        }
+                else:
+                    calls[call.id] = value
+            if len(calls) > 8 or len(json.dumps(message, ensure_ascii=False)) + len(json.dumps(calls)) > 64000:
+                raise AssistantError('response_too_large')
+            if content or chunk.all_content is not None:
+                await on_event({'kind': 'text', 'data': {'text': message['content']}})
+        if not finished:
+            raise AssistantError('stream_interrupted')
+        if calls:
+            message['tool_calls'] = list(calls.values())
+        return Message.model_validate(message)
+
+    async def _snapshot(self, conversation, on_event):
+        if on_event is not None:
+            await on_event({'kind': 'snapshot', 'data': self.public_view(conversation)})
 
     @staticmethod
     def _scope(context, conversation_id):
@@ -150,7 +332,7 @@ class AssistantService:
             raise AssistantError('stale_turn')
         conversation.update(status=status, error=error)
 
-    async def turn(self, context, conversation_id, revision, text=None, approved=None, model_uuid=None):
+    async def turn(self, context, conversation_id, revision, text=None, approved=None, model_uuid=None, on_event=None):
         require_permission(context, Permission.RUNTIME_OPERATE)
         if model_uuid is not None and text is None:
             raise AssistantError('invalid_input', 400)
@@ -201,7 +383,7 @@ class AssistantService:
                     await self._save(context, conversation, 'running')
                     try:
                         if not conversation['model_uuid']:
-                            recommended = await self.ap.space_service.get_recommended_chat_model(context)
+                            recommended = await self.recommended_model(context)
                             conversation['model_uuid'] = recommended['uuid']
                         execution = ExecutionContext.from_request(context)
                         model = selected_model or await self.ap.model_mgr.get_model_by_uuid(
@@ -215,22 +397,30 @@ class AssistantService:
                         return conversation
                     if approved is not None:
                         calls = self._calls(conversation['messages'][-1])
-                        await self._execute(context, conversation, calls, approved)
-                    for _ in range(8):
-                        response = await model.provider.invoke_llm(
-                            query=None,
-                            model=model,
-                            messages=[Message(role='system', content=SYSTEM_PROMPT)]
-                            + [Message.model_validate(message) for message in conversation['messages']],
-                            funcs=tool_definitions(context),
-                            extra_args=model.model_entity.extra_args or {},
-                            remove_think=True,
-                            execution_context=execution,
+                        await self._execute(context, conversation, calls, approved, on_event)
+                    for round_index in range(8):
+                        await self._snapshot(conversation, on_event)
+                        if on_event is not None:
+                            await on_event({'kind': 'phase', 'data': {'phase': 'thinking', 'round': round_index + 1}})
+                        response = await self._invoke(
+                            model,
+                            dict(
+                                query=None,
+                                model=model,
+                                messages=[Message(role='system', content=SYSTEM_PROMPT)]
+                                + [Message.model_validate(message) for message in conversation['messages']],
+                                funcs=tool_definitions(context),
+                                extra_args=model.model_entity.extra_args or {},
+                                remove_think=True,
+                                execution_context=execution,
+                            ),
+                            on_event,
                         )
                         message = response.model_dump(mode='json')
                         if len(json.dumps(message, ensure_ascii=False)) > 64000:
                             raise AssistantError('response_too_large')
                         conversation['messages'].append(message)
+                        await self._snapshot(conversation, on_event)
                         calls = self._calls(message)
                         if not calls:
                             await self._save(context, conversation, 'ready')
@@ -249,9 +439,12 @@ class AssistantService:
                         if any(TOOLS[call['function']['name']][2] for call in calls):
                             await self._save(context, conversation, 'approval')
                             return conversation
-                        await self._execute(context, conversation, calls, True)
+                        await self._execute(context, conversation, calls, True, on_event)
                     await self._save(context, conversation, 'failed', 'round_limit')
             except asyncio.CancelledError:
+                partial = self._runs.get(conversation_id, {}).get('progress', {}).get('text')
+                if partial:
+                    conversation['messages'].append(Message(role='assistant', content=partial).model_dump(mode='json'))
                 await asyncio.shield(self._save(context, conversation, 'failed', 'result_unknown'))
                 raise
             except TimeoutError:
@@ -280,7 +473,7 @@ class AssistantService:
             ).model_dump(mode='json')
         )
 
-    async def _execute(self, context, conversation, calls, approved):
+    async def _execute(self, context, conversation, calls, approved, on_event=None):
         # Validate the complete saved batch before any write, including after approval.
         if approved:
             for call in calls:
@@ -293,6 +486,19 @@ class AssistantService:
             else:
                 # Persist before effects. A crash leaves a running/unknown operation, never a replayable approval.
                 await self._save(context, conversation, 'running')
+                if on_event is not None:
+                    await on_event(
+                        {
+                            'kind': 'phase',
+                            'data': {
+                                'phase': 'tool',
+                                'tool': {
+                                    'name': func['name'],
+                                    'arguments': redact_secrets(json.loads(func['arguments'] or '{}')),
+                                },
+                            },
+                        }
+                    )
                 try:
                     result = await execute_tool(self.ap, context, func['name'], json.loads(func['arguments'] or '{}'))
                 except Exception as exc:
@@ -308,3 +514,4 @@ class AssistantService:
                     raise AssistantError('tool_failed')
             self._append_result(conversation, call, result)
             await self._save(context, conversation, 'running')
+            await self._snapshot(conversation, on_event)
diff --git a/src/langbot/pkg/api/http/service/assistant_tools.py b/src/langbot/pkg/api/http/service/assistant_tools.py
index a850f428e..fd9599f64 100644
--- a/src/langbot/pkg/api/http/service/assistant_tools.py
+++ b/src/langbot/pkg/api/http/service/assistant_tools.py
@@ -312,12 +312,35 @@ async def execute_tool(ap, context, name: str, arguments: dict):
             if await ap.knowledge_service.get_knowledge_base(context, kb_id) is None:
                 raise ValueError('Knowledge base not found')
         config = copy.deepcopy(pipeline['config'])
-        config['ai']['runner']['runner'] = 'local-agent'
-        local = config['ai']['local-agent']
-        local['model']['primary'] = args['model_uuid']
+        runner_id = 'plugin:langbot-team/LocalAgent/default'
+        ai = config.get('ai') if isinstance(config.get('ai'), dict) else {}
+        runner = ai.get('runner') if isinstance(ai.get('runner'), dict) else {}
+        runner_configs = ai.get('runner_config') if isinstance(ai.get('runner_config'), dict) else {}
+        local = runner_configs.get(runner_id) if isinstance(runner_configs.get(runner_id), dict) else {}
+        model = local.get('model') if isinstance(local.get('model'), dict) else {}
+        local['model'] = {
+            'primary': args['model_uuid'],
+            'fallbacks': model.get('fallbacks') if isinstance(model.get('fallbacks'), list) else [],
+            'reasoning': model.get('reasoning') if isinstance(model.get('reasoning'), dict) else {},
+        }
         local['prompt'] = [{'role': 'system', 'content': args['system_prompt']}]
         local['knowledge-bases'] = args['knowledge_base_uuids']
-        await ap.pipeline_service.update_pipeline(context, args['pipeline_uuid'], {'config': config})
+        # 4.11 shape: ai.runner.id + ai.runner_config[id]; legacy keys are dropped.
+        ai['runner'] = {'id': runner_id, 'expire-time': runner.get('expire-time', 0)}
+        ai['runner_config'] = {**runner_configs, runner_id: local}
+        ai.pop('local-agent', None)
+        config['ai'] = ai
+        update = {'config': config}
+        prefs = pipeline.get('extensions_preferences')
+        if isinstance(prefs, dict) and prefs.get('enable_all_plugins') is False:
+            plugins = prefs.get('plugins') if isinstance(prefs.get('plugins'), list) else []
+            entry = {'author': 'langbot-team', 'name': 'LocalAgent'}
+            if not any(
+                isinstance(p, dict) and p.get('author') == entry['author'] and p.get('name') == entry['name']
+                for p in plugins
+            ):
+                update['extensions_preferences'] = {**prefs, 'plugins': [*plugins, entry]}
+        await ap.pipeline_service.update_pipeline(context, args['pipeline_uuid'], update)
         return {'uuid': args['pipeline_uuid'], 'url': f'/home/pipelines?id={args["pipeline_uuid"]}'}
     resource_id = await ap.knowledge_service.create_knowledge_base(context, args)
     return {'uuid': resource_id, 'url': f'/home/knowledge?id={resource_id}'}
diff --git a/src/langbot/pkg/api/http/service/bot.py b/src/langbot/pkg/api/http/service/bot.py
index ef0c680ed..766adbc87 100644
--- a/src/langbot/pkg/api/http/service/bot.py
+++ b/src/langbot/pkg/api/http/service/bot.py
@@ -16,6 +16,7 @@ from .tenant import TenantContext, require_workspace_uuid, scope_statement
 from ....utils import httpclient
 from ....platform.sources import http_bot_signing
 from ....platform.adapter_names import canonical_adapter_name
+from ....cloud.plan_policy import require_workspace_platform
 from ....agent.runner.errors import RunnerNotFoundError
 
 
@@ -52,19 +53,19 @@ class BotService:
         return None
 
     def _adapter_declares_webhook_url(self, adapter_name: str) -> bool:
-        """Whether the adapter manifest declares a generated webhook URL config item."""
+        """Whether the manifest needs a generated public bot ingress URL."""
         component = self._get_adapter_component(adapter_name)
         if component is None:
             return False
 
         for config_item in component.spec.get('config', []):
-            if config_item.get('type') == 'webhook-url':
+            if config_item.get('type') == 'webhook-url' or config_item.get('name') == '__system.onebot_ws_url':
                 return True
         return False
 
     @staticmethod
     def _is_message_event_pattern(event_pattern: str) -> bool:
-        return event_pattern == 'message.*' or event_pattern.startswith('message.')
+        return event_pattern == 'message.received'
 
     @staticmethod
     def _event_pattern_covers(supported_pattern: str, binding_pattern: str) -> bool:
@@ -507,7 +508,7 @@ class BotService:
 
             if target_type == 'pipeline':
                 if not self._is_message_event_pattern(event_pattern):
-                    raise ValueError('Pipeline can only be bound to message events')
+                    raise ValueError('Pipeline can only be bound to message.received')
                 result = await self.ap.persistence_mgr.execute_async(
                     scope_statement(
                         sqlalchemy.select(persistence_pipeline.LegacyPipeline.uuid).where(
@@ -685,7 +686,9 @@ class BotService:
         """Create bot"""
         workspace_uuid = require_workspace_uuid(context)
         # Check limitation
-        limitation = self.ap.instance_config.data.get('system', {}).get('limitation', {})
+        from .workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, workspace_uuid)
         max_bots = limitation.get('max_bots', -1)
         if max_bots >= 0:
             existing_bots = await self.get_bots(context)
@@ -694,6 +697,7 @@ class BotService:
 
         # TODO: 检查配置信息格式
         bot_data = await self._prepare_bot_data(context, bot_data, include_uuid=True)
+        await require_workspace_platform(self.ap, workspace_uuid, bot_data['adapter'])
         bot_data['uuid'] = str(uuid.uuid4())
         bot_data['workspace_uuid'] = workspace_uuid
         bot_data.setdefault('event_bindings', [])
@@ -704,7 +708,8 @@ class BotService:
         bot = await self.get_bot(context, bot_data['uuid'], include_secret=True)
 
         try:
-            await self.ap.platform_mgr.load_bot(context, bot)
+            if bot.get('enable'):
+                await self.ap.platform_mgr.load_bot(context, bot)
         except Exception as exc:
             raise BotApplyError(bot_error_message(exc, bot), bot['uuid']) from exc
 
@@ -715,6 +720,9 @@ class BotService:
         workspace_uuid = require_workspace_uuid(context)
         update_data = await self._prepare_bot_data(context, bot_data, include_uuid=False)
         update_data.pop('workspace_uuid', None)
+        if 'adapter' in update_data or update_data.get('enable'):
+            current = await self.get_bot(context, bot_uuid, include_secret=True)
+            await require_workspace_platform(self.ap, workspace_uuid, update_data.get('adapter', current['adapter']))
 
         result = await self.ap.persistence_mgr.execute_async(
             scope_statement(
@@ -740,8 +748,8 @@ class BotService:
         bot = await self.get_bot(context, bot_uuid, include_secret=True)
         try:
             await self.ap.platform_mgr.remove_bot(context, bot_uuid)
-            runtime_bot = await self.ap.platform_mgr.load_bot(context, bot)
-            if runtime_bot.enable:
+            if bot.get('enable'):
+                runtime_bot = await self.ap.platform_mgr.load_bot(context, bot)
                 await runtime_bot.run()
         except Exception as exc:
             raise BotApplyError(bot_error_message(exc, bot), bot['uuid']) from exc
@@ -775,7 +783,8 @@ class BotService:
             raise WorkspaceNotFoundError('Bot not found')
         runtime_bot = await self.ap.platform_mgr.get_bot_by_uuid(context, bot_uuid)
         if runtime_bot is None:
-            raise Exception('Bot not found')
+            # Saved but disabled/unconfigured bots have no runtime logs yet.
+            return [], 0
 
         logs, total_count = await runtime_bot.logger.get_logs(from_index, max_count)
 
diff --git a/src/langbot/pkg/api/http/service/knowledge.py b/src/langbot/pkg/api/http/service/knowledge.py
index 654bf2f17..e85d16fd8 100644
--- a/src/langbot/pkg/api/http/service/knowledge.py
+++ b/src/langbot/pkg/api/http/service/knowledge.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import t
+
 import sqlalchemy
 
 from ....api.http.authz import WorkspaceRequiredError
@@ -56,14 +58,20 @@ class KnowledgeService:
         require_workspace_uuid(context)
         # In new architecture, we delegate entirely to RAGManager which uses plugins.
         # Legacy internal KB creation is removed.
-        limitation = (
-            getattr(getattr(self.ap, 'instance_config', None), 'data', {}).get('system', {}).get('limitation', {})
-        )
+        from .workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, require_workspace_uuid(context))
         max_knowledge_bases = limitation.get('max_knowledge_bases', -1)
         if max_knowledge_bases >= 0:
             knowledge_bases = await self.ap.rag_mgr.get_all_knowledge_base_details(context)
             if len(knowledge_bases) >= max_knowledge_bases:
-                raise ValueError(f'Maximum number of knowledge bases ({max_knowledge_bases}) reached')
+                raise ValueError(
+                    t(
+                        'Maximum number of {resource} ({limit}) reached',
+                        resource=t('knowledge bases'),
+                        limit=max_knowledge_bases,
+                    )
+                )
 
         knowledge_engine_plugin_id = kb_data.get('knowledge_engine_plugin_id')
         if not knowledge_engine_plugin_id:
diff --git a/src/langbot/pkg/api/http/service/mcp.py b/src/langbot/pkg/api/http/service/mcp.py
index 67e8e14ae..3e5eb7339 100644
--- a/src/langbot/pkg/api/http/service/mcp.py
+++ b/src/langbot/pkg/api/http/service/mcp.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import t
+
 import copy
 import re
 import uuid
@@ -212,7 +214,9 @@ class MCPService:
         # stdio MCP even though Box Runtime itself remains available.
         require_stdio_mcp_enabled(self.ap, server_data)
 
-        limitation = self.ap.instance_config.data.get('system', {}).get('limitation', {})
+        from .workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, workspace_uuid)
         max_extensions = limitation.get('max_extensions', -1)
         if max_extensions >= 0:
             mcp_count_result = await self.ap.persistence_mgr.execute_async(
@@ -226,7 +230,9 @@ class MCPService:
                 .where(persistence_plugin.PluginSetting.workspace_uuid == workspace_uuid)
             )
             if (mcp_count_result.scalar() or 0) + (plugin_count_result.scalar() or 0) >= max_extensions:
-                raise ValueError(f'Maximum number of extensions ({max_extensions}) reached')
+                raise ValueError(
+                    t('Maximum number of {resource} ({limit}) reached', resource=t('extensions'), limit=max_extensions)
+                )
 
         payload = dict(server_data)
         payload.pop('workspace_uuid', None)
@@ -244,7 +250,7 @@ class MCPService:
             )
         )
         if existing_result.first() is not None:
-            raise ValueError(f'MCP server already exists: {server_name}')
+            raise ValueError(t('MCP server already exists: {name}', name=server_name))
 
         await self.ap.persistence_mgr.execute_async(sqlalchemy.insert(persistence_mcp.MCPServer).values(payload))
         created = await self._get_mcp_server_by_uuid_raw(execution_context, payload['uuid'])
@@ -330,7 +336,7 @@ class MCPService:
                 raise ValueError('MCP server name is required')
             duplicate = await self._get_mcp_server_by_name_raw(execution_context, payload['name'])
             if duplicate is not None and duplicate['uuid'] != server_uuid:
-                raise ValueError(f'MCP server already exists: {payload["name"]}')
+                raise ValueError(t('MCP server already exists: {name}', name=payload['name']))
 
         effective_server = {**old_server, **payload}
         # Existing disabled rows remain readable/deletable.  Switching away
diff --git a/src/langbot/pkg/api/http/service/model.py b/src/langbot/pkg/api/http/service/model.py
index bf8c6b196..bfd20bfaa 100644
--- a/src/langbot/pkg/api/http/service/model.py
+++ b/src/langbot/pkg/api/http/service/model.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import t
+
 import uuid
 
 import sqlalchemy
@@ -146,7 +148,13 @@ async def _validate_provider_supports(
     if not support_type:
         return
     if model_type not in support_type:
-        raise ValueError(f'Provider requester "{requester_name}" does not support {model_type} models')
+        raise ValueError(
+            t(
+                'Provider {provider} does not support {model_type} models',
+                provider=requester_name,
+                model_type=model_type,
+            )
+        )
 
 
 async def _require_workspace_provider(
@@ -196,6 +204,91 @@ class LLMModelsService:
     def __init__(self, ap: app.Application) -> None:
         self.ap = ap
 
+    async def get_starred_model(self, context: TenantContext) -> str | None:
+        from ....entity.persistence.metadata import WorkspaceMetadata
+
+        result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(WorkspaceMetadata.value).where(
+                WorkspaceMetadata.workspace_uuid == require_workspace_uuid(context),
+                WorkspaceMetadata.key == 'starred_llm_model',
+            )
+        )
+        row = result.first()
+        model_uuid = row[0] if row else None
+        if model_uuid and await self.get_llm_model(context, model_uuid) is not None:
+            return model_uuid
+        return None
+
+    async def set_starred_model(self, context: TenantContext, model_uuid: str | None) -> None:
+        from ....entity.persistence.metadata import WorkspaceMetadata
+
+        workspace_uuid = require_workspace_uuid(context)
+        if model_uuid and await self.get_llm_model(context, model_uuid) is None:
+            raise ValueError('Model not found in this Workspace')
+        # A single metadata row makes concurrent selections replace each other.
+        dialect = self.ap.persistence_mgr.get_db_engine().dialect.name
+        if dialect == 'postgresql':
+            from sqlalchemy.dialects.postgresql import insert
+        else:
+            from sqlalchemy.dialects.sqlite import insert
+        statement = insert(WorkspaceMetadata).values(
+            workspace_uuid=workspace_uuid,
+            key='starred_llm_model',
+            value=model_uuid or '',
+        )
+        await self.ap.persistence_mgr.execute_async(
+            statement.on_conflict_do_update(
+                index_elements=['workspace_uuid', 'key'],
+                set_={'value': model_uuid or ''},
+            )
+        )
+
+    async def get_default_model(self, context: TenantContext) -> str | None:
+        starred = await self.get_starred_model(context)
+        if starred:
+            return starred
+        if self.ap.instance_config.data.get('space', {}).get('disable_models_service', False):
+            return None
+        owner = await self.ap.user_service.get_workspace_owner(require_workspace_uuid(context))
+        cloud = getattr(getattr(self.ap, 'deployment', None), 'mode', 'oss') == 'cloud'
+        if not cloud and not (owner and owner.space_account_uuid):
+            return None
+        try:
+            return (await self.ap.space_service.get_recommended_chat_model(context))['uuid']
+        except Exception:
+            return None
+
+    async def apply_default_model(self, context: TenantContext, schema: list[dict], parameters: dict) -> dict:
+        import copy
+
+        fields = [
+            field
+            for field in schema
+            if field.get('type')
+            in (
+                'llm-model-selector',
+                'select-llm-model',
+                'model-fallback-selector',
+            )
+            and field.get('name')
+        ]
+        if not fields:
+            return parameters
+        model_uuid = await self.get_default_model(context)
+        if not model_uuid:
+            return parameters
+        result = copy.deepcopy(parameters)
+        for field in fields:
+            name = field['name']
+            if field['type'] == 'model-fallback-selector':
+                value = result.get(name)
+                value = value if isinstance(value, dict) else {}
+                if not value.get('primary'):
+                    result[name] = {**value, 'primary': model_uuid, 'fallbacks': value.get('fallbacks', [])}
+            elif not result.get(name):
+                result[name] = model_uuid
+        return result
+
     async def get_llm_models(self, context: TenantContext, include_secret: bool = False) -> list[dict]:
         """Get all LLM models with provider info"""
         result = await self.ap.persistence_mgr.execute_async(
diff --git a/src/langbot/pkg/api/http/service/monitoring.py b/src/langbot/pkg/api/http/service/monitoring.py
index 0bae983a6..335d334c6 100644
--- a/src/langbot/pkg/api/http/service/monitoring.py
+++ b/src/langbot/pkg/api/http/service/monitoring.py
@@ -10,10 +10,15 @@ from sqlalchemy.dialects import postgresql as postgresql_dialect
 from sqlalchemy.dialects import sqlite as sqlite_dialect
 
 from ....core import app
+from ....entity.persistence import agent as persistence_agent
 from ....entity.persistence import monitoring as persistence_monitoring
+from ....entity.persistence import agent_run as persistence_agent_run
+from ....entity.persistence import agent_interaction as persistence_agent_interaction
+from ....entity.persistence.event_log import EventLog
 from ..authz import WorkspaceRequiredError
 from ..context import ExecutionContext
 from .tenant import TenantContext, require_workspace_uuid
+from .monitoring_execution_details import ExecutionDetailsMixin
 
 
 _DEFAULT_MONITORING_PAGE_ROWS = 1000
@@ -29,6 +34,76 @@ _HARD_MAX_MONITORING_OFFSET = 10000000
 _DEFAULT_CLEANUP_BATCHES_PER_TABLE = 4
 _HARD_MAX_CLEANUP_BATCHES_PER_TABLE = 100
 
+# Canonical execution status groups shared by the agent ledger and pipeline query
+# records so both object types can be summarised with one vocabulary.
+_AGENT_STATUS_GROUP = {
+    'completed': 'completed',
+    'failed': 'failed',
+    'timeout': 'failed',
+    'cancelled': 'cancelled',
+    'created': 'queued',
+    'queued': 'queued',
+    'claimed': 'running',
+    'running': 'running',
+}
+_PIPELINE_STATUS_GROUP = {
+    'success': 'completed',
+    'error': 'failed',
+    'pending': 'running',
+    'discarded': 'ignored',
+    'not_matched': 'ignored',
+}
+
+
+def _status_group(status: str, source: str) -> str:
+    table = _AGENT_STATUS_GROUP if source == 'agent' else _PIPELINE_STATUS_GROUP
+    return table.get(status, status)
+
+
+def _epoch_ms(value: datetime.datetime | None) -> int | None:
+    """Epoch milliseconds for a DB timestamp, treating naive values as UTC."""
+    if value is None:
+        return None
+    if value.tzinfo is None:
+        value = value.replace(tzinfo=datetime.timezone.utc)
+    return int(round(value.timestamp() * 1000))
+
+
+def _percentile_ms(values: list[float], fraction: float) -> int | None:
+    """Nearest-rank percentile (ms) over a non-empty, prefix-sorted list."""
+    if not values:
+        return None
+    rank = max(1, min(len(values), int(round(fraction * len(values)))))
+    return int(round(values[rank - 1]))
+
+
+def _collect_text(node, out: list[str], depth: int = 0) -> None:
+    """Best-effort extraction of human text from a serialised message chain."""
+    if depth > 6:
+        return
+    if isinstance(node, dict):
+        for key, value in node.items():
+            if key == 'text' and isinstance(value, str):
+                out.append(value)
+            elif isinstance(value, (dict, list)):
+                _collect_text(value, out, depth + 1)
+    elif isinstance(node, list):
+        for item in node:
+            _collect_text(item, out, depth + 1)
+
+
+def _message_preview(raw: str | None, limit: int = 160) -> str:
+    if not raw:
+        return ''
+    try:
+        parsed = json.loads(raw)
+    except (TypeError, ValueError):
+        return raw[:limit]
+    parts: list[str] = []
+    _collect_text(parsed, parts)
+    text = ' '.join(part.strip() for part in parts if part.strip())
+    return (text or raw)[:limit]
+
 
 def _normalize_user_id(value: str | int | None) -> str | None:
     """Convert numeric platform IDs before binding a VARCHAR with asyncpg.
@@ -52,13 +127,19 @@ def _workspace_transaction(method):
         tenant_uow = getattr(self.ap.persistence_mgr, 'tenant_uow', None)
         if callable(tenant_uow):
             async with tenant_uow(workspace_uuid):
-                return await method(self, context, *args, **kwargs)
-        return await method(self, context, *args, **kwargs)
+                result = await method(self, context, *args, **kwargs)
+        else:
+            result = await method(self, context, *args, **kwargs)
+        if method.__name__ in {'record_message', 'update_message_status'}:
+            from ....utils.inflight import inflight_hub
+
+            inflight_hub.notify(workspace_uuid)
+        return result
 
     return wrapped
 
 
-class MonitoringService:
+class MonitoringService(ExecutionDetailsMixin):
     """Monitoring service"""
 
     ap: app.Application
@@ -179,6 +260,9 @@ class MonitoringService:
             A dict mapping table name to the number of deleted rows.
         """
         workspace_uuid = self._require_write_context(context)
+        from ....cloud.plan_policy import workspace_history_retention_days
+
+        retention_days = await workspace_history_retention_days(self.ap, workspace_uuid, retention_days)
         if retention_days < 1:
             raise ValueError('retention_days must be >= 1')
         if batch_size < 1:
@@ -462,12 +546,23 @@ class MonitoringService:
         runner_name: str | None = None,
         variables: str | None = None,
         role: str = 'user',
+        event_id: str | None = None,
+        run_id: str | None = None,
+        parent_message_id: str | None = None,
     ) -> str:
         """Record a message"""
         workspace_uuid = self._require_write_context(context)
+        from ....telemetry import trace
+
+        trace_state = trace.current()
+        event_id = event_id or (trace_state.event_id if trace_state else None)
+        run_id = run_id or trace.current_run() or None
         message_id = str(uuid.uuid4())
         message_content = self._sanitize_message_content(message_content)
+        from ....utils.execution_lease import host_execution_lease
+
         message_data = {
+            **(host_execution_lease() if status == 'pending' else {}),
             'id': message_id,
             'workspace_uuid': workspace_uuid,
             'timestamp': datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None),
@@ -485,6 +580,9 @@ class MonitoringService:
             'runner_name': runner_name,
             'variables': variables,
             'role': role,
+            'event_id': event_id,
+            'run_id': run_id,
+            'parent_message_id': parent_message_id,
         }
 
         await self.ap.persistence_mgr.execute_async(
@@ -510,6 +608,7 @@ class MonitoringService:
         cost: float | None = None,
         error_message: str | None = None,
         message_id: str | None = None,
+        details: dict | None = None,
     ) -> str:
         """Record an LLM call"""
         workspace_uuid = self._require_write_context(context)
@@ -521,7 +620,7 @@ class MonitoringService:
             'model_name': model_name,
             'input_tokens': input_tokens,
             'output_tokens': output_tokens,
-            'total_tokens': input_tokens + output_tokens,
+            'total_tokens': ((details or {}).get('usage') or {}).get('total_tokens') or (input_tokens + output_tokens),
             'duration': duration,
             'cost': cost,
             'status': status,
@@ -532,6 +631,7 @@ class MonitoringService:
             'session_id': session_id,
             'error_message': error_message,
             'message_id': message_id,
+            'details': details,
         }
 
         await self.ap.persistence_mgr.execute_async(
@@ -806,6 +906,8 @@ class MonitoringService:
         pipeline_ids: list[str] | None = None,
         start_time: datetime.datetime | None = None,
         end_time: datetime.datetime | None = None,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> dict:
         """Get overview metrics"""
         workspace_uuid = require_workspace_uuid(context)
@@ -815,15 +917,47 @@ class MonitoringService:
         embedding_conditions = [persistence_monitoring.MonitoringEmbeddingCall.workspace_uuid == workspace_uuid]
         session_conditions = [persistence_monitoring.MonitoringSession.workspace_uuid == workspace_uuid]
 
+        message_conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringMessage, workspace_uuid, execution_statuses, mode
+            )
+        )
+        llm_conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringLLMCall, workspace_uuid, execution_statuses, mode
+            )
+        )
+        session_conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringSession, workspace_uuid, execution_statuses, mode
+            )
+        )
+        embedding_conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringEmbeddingCall, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             message_conditions.append(persistence_monitoring.MonitoringMessage.bot_id.in_(bot_ids))
             llm_conditions.append(persistence_monitoring.MonitoringLLMCall.bot_id.in_(bot_ids))
             session_conditions.append(persistence_monitoring.MonitoringSession.bot_id.in_(bot_ids))
 
         if pipeline_ids:
-            message_conditions.append(persistence_monitoring.MonitoringMessage.pipeline_id.in_(pipeline_ids))
-            llm_conditions.append(persistence_monitoring.MonitoringLLMCall.pipeline_id.in_(pipeline_ids))
-            session_conditions.append(persistence_monitoring.MonitoringSession.pipeline_id.in_(pipeline_ids))
+            message_conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringMessage, workspace_uuid, pipeline_ids
+                )
+            )
+            llm_conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringLLMCall, workspace_uuid, pipeline_ids
+                )
+            )
+            session_conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringSession, workspace_uuid, pipeline_ids
+                )
+            )
 
         if start_time:
             message_conditions.append(persistence_monitoring.MonitoringMessage.timestamp >= start_time)
@@ -885,6 +1019,10 @@ class MonitoringService:
         active_sessions_result = await self.ap.persistence_mgr.execute_async(active_session_query)
         active_sessions = active_sessions_result.scalar() or 0
 
+        # Execution-first KPI values (runs + pipeline queries, latency, token
+        # coverage) are served by `get_executions`, which applies the execution
+        # view filters. This endpoint stays the message/traffic rollup so the
+        # dashboard does not compute the same rollup twice per page load.
         return {
             'total_messages': total_messages,
             'llm_calls': llm_calls,
@@ -894,6 +1032,742 @@ class MonitoringService:
             'active_sessions': active_sessions,
         }
 
+    # ========== Execution view ==========
+
+    @staticmethod
+    def _debug_run_condition(model):
+        return sqlalchemy.or_(
+            model.binding_id.like('debug:%'),
+            model.conversation_id.like('debug:%'),
+            model.bot_id == 'websocket-proxy-bot',
+        )
+
+    @staticmethod
+    def _debug_pipeline_condition(model):
+        # Dashboard/API pipeline debug uses this reserved proxy identity.
+        # Real Web Page bots have their own bot UUID, even on the same adapter.
+        return sqlalchemy.func.coalesce(model.bot_id == 'websocket-proxy-bot', False)
+
+    def _monitoring_execution_condition(self, model, workspace_uuid, statuses=None, mode='all'):
+        """Filter telemetry by its owning execution, not by the call's outcome."""
+        if not statuses and mode == 'all':
+            return sqlalchemy.true()
+        Run = persistence_agent_run.AgentRun
+        Message = persistence_monitoring.MonitoringMessage
+        roots = sqlalchemy.orm.aliased(Message)
+        run_conditions = [Run.workspace_id == workspace_uuid]
+        if statuses:
+            run_conditions.append(Run.status.in_(self._raw_agent_statuses(statuses)))
+        debug = self._debug_run_condition(Run)
+        if mode != 'all':
+            run_conditions.append(debug if mode == 'debug' else sqlalchemy.not_(sqlalchemy.func.coalesce(debug, False)))
+        runs = sqlalchemy.select(Run.run_id).where(*run_conditions)
+        root_conditions = [roots.workspace_uuid == workspace_uuid, roots.role != 'assistant']
+        if statuses:
+            root_conditions.append(roots.status.in_(self._raw_pipeline_statuses(statuses)))
+        if mode != 'all':
+            root_debug = self._debug_pipeline_condition(roots)
+            root_conditions.append(root_debug if mode == 'debug' else sqlalchemy.not_(root_debug))
+        root_ids = sqlalchemy.select(roots.id).where(*root_conditions)
+        linked = sqlalchemy.or_(
+            Message.id.in_(root_ids), Message.parent_message_id.in_(root_ids), Message.run_id.in_(runs)
+        )
+        message_ids = sqlalchemy.select(Message.id).where(Message.workspace_uuid == workspace_uuid, linked)
+        if model is Message:
+            return linked
+        if hasattr(model, 'message_id'):
+            return sqlalchemy.or_(model.message_id.in_(runs), model.message_id.in_(message_ids))
+        if model is persistence_monitoring.MonitoringSession:
+            return sqlalchemy.or_(
+                model.session_id.in_(sqlalchemy.select(roots.session_id).where(*root_conditions)),
+                model.session_id.in_(sqlalchemy.select(Run.conversation_id).where(*run_conditions)),
+            )
+        return sqlalchemy.false()
+
+    def _processor_runs(self, workspace_uuid, processor_ids):
+        Run = persistence_agent_run.AgentRun
+        return sqlalchemy.select(Run.run_id).where(
+            Run.workspace_id == workspace_uuid,
+            sqlalchemy.func.coalesce(self._execution_json_text(Run.metadata_json, 'processor_id'), Run.agent_id).in_(
+                processor_ids
+            ),
+        )
+
+    def _monitoring_processor_condition(self, model, workspace_uuid, processor_ids):
+        direct = model.pipeline_id.in_(processor_ids)
+        if hasattr(model, 'message_id'):
+            Message = persistence_monitoring.MonitoringMessage
+            return sqlalchemy.or_(
+                direct,
+                model.message_id.in_(self._processor_runs(workspace_uuid, processor_ids)),
+                model.message_id.in_(
+                    sqlalchemy.select(Message.id).where(
+                        Message.workspace_uuid == workspace_uuid, Message.pipeline_id.in_(processor_ids)
+                    )
+                ),
+            )
+        return direct
+
+    def _agent_run_conditions(
+        self,
+        workspace_uuid: str,
+        *,
+        bot_ids: list[str] | None = None,
+        agent_ids: list[str] | None = None,
+        processor_ids: list[str] | None = None,
+        statuses: list[str] | None = None,
+        start_time: datetime.datetime | None = None,
+        end_time: datetime.datetime | None = None,
+        debug: bool | None = None,
+    ) -> list:
+        AgentRun = persistence_agent_run.AgentRun
+        conditions = [
+            AgentRun.workspace_id == workspace_uuid,
+            ~sqlalchemy.exists(
+                sqlalchemy.select(persistence_monitoring.MonitoringMessage.id).where(
+                    persistence_monitoring.MonitoringMessage.workspace_uuid == workspace_uuid,
+                    persistence_monitoring.MonitoringMessage.run_id == AgentRun.run_id,
+                    persistence_monitoring.MonitoringMessage.role != 'assistant',
+                )
+            ),
+        ]
+        if bot_ids:
+            conditions.append(AgentRun.bot_id.in_(bot_ids))
+        if agent_ids:
+            conditions.append(AgentRun.agent_id.in_(agent_ids))
+        if processor_ids:
+            conditions.append(AgentRun.run_id.in_(self._processor_runs(workspace_uuid, processor_ids)))
+        if statuses is not None:
+            conditions.append(AgentRun.status.in_(statuses) if statuses else sqlalchemy.false())
+        if start_time:
+            conditions.append(AgentRun.created_at >= start_time)
+        if end_time:
+            conditions.append(AgentRun.created_at <= end_time)
+        if debug is not None:
+            debug_condition = self._debug_run_condition(AgentRun)
+            conditions.append(
+                debug_condition if debug else sqlalchemy.not_(sqlalchemy.func.coalesce(debug_condition, False))
+            )
+        return conditions
+
+    def _pipeline_execution_conditions(
+        self,
+        workspace_uuid: str,
+        *,
+        bot_ids: list[str] | None = None,
+        pipeline_ids: list[str] | None = None,
+        statuses: list[str] | None = None,
+        start_time: datetime.datetime | None = None,
+        end_time: datetime.datetime | None = None,
+        debug: bool | None = None,
+    ) -> list:
+        MonitoringMessage = persistence_monitoring.MonitoringMessage
+        conditions = [
+            MonitoringMessage.workspace_uuid == workspace_uuid,
+            # Assistant replies are outputs of a query, not an execution of their own.
+            sqlalchemy.or_(MonitoringMessage.role.is_(None), MonitoringMessage.role != 'assistant'),
+        ]
+        if bot_ids:
+            conditions.append(MonitoringMessage.bot_id.in_(bot_ids))
+        if pipeline_ids:
+            conditions.append(self._monitoring_processor_condition(MonitoringMessage, workspace_uuid, pipeline_ids))
+        if statuses is not None:
+            conditions.append(MonitoringMessage.status.in_(statuses) if statuses else sqlalchemy.false())
+        if start_time:
+            conditions.append(MonitoringMessage.timestamp >= start_time)
+        if end_time:
+            conditions.append(MonitoringMessage.timestamp <= end_time)
+        if debug is not None:
+            debug_condition = self._debug_pipeline_condition(MonitoringMessage)
+            conditions.append(debug_condition if debug else sqlalchemy.not_(debug_condition))
+        return conditions
+
+    @staticmethod
+    def _raw_agent_statuses(groups: list[str] | None) -> list[str] | None:
+        """Raw agent statuses for a group filter; empty means "no group applies"."""
+        if not groups:
+            return None
+        raw: list[str] = []
+        for group in groups:
+            raw.extend(status for status, mapped in _AGENT_STATUS_GROUP.items() if mapped == group)
+        return raw
+
+    @staticmethod
+    def _raw_pipeline_statuses(groups: list[str] | None) -> list[str] | None:
+        """Raw pipeline statuses for a group filter; empty means "no group applies"."""
+        if not groups:
+            return None
+        raw: list[str] = []
+        for group in groups:
+            raw.extend(status for status, mapped in _PIPELINE_STATUS_GROUP.items() if mapped == group)
+        return raw
+
+    @staticmethod
+    def _model_rows(result) -> list:
+        """Rows of one ``select(Model)`` executed through ``execute_async``.
+
+        The Core keeps the connection result contract, where ``select(Model)``
+        yields a *flat* row of the model's columns (``row.uuid``, ``row.status``)
+        rather than a one-item ORM row. ``.scalars()`` would therefore return the
+        first column - the id - instead of the row, so the rows are taken as-is.
+        """
+        return list(result.all())
+
+    @staticmethod
+    def _deserialize_json(raw: str | None) -> dict:
+        if not raw:
+            return {}
+        try:
+            value = json.loads(raw)
+        except (TypeError, ValueError):
+            return {}
+        return value if isinstance(value, dict) else {}
+
+    async def _agent_names(self, workspace_uuid: str, rows: list) -> dict[str, str]:
+        """The processor name of every Agent run in one page, in one query.
+
+        A run stores the Agent it executed as a uuid; the list has to name the
+        processor, so the names are resolved for the whole page at once.
+        """
+        agent_ids = {self._deserialize_json(row.metadata_json).get('processor_id') or row.agent_id for row in rows} - {
+            None,
+            '',
+        }
+        if not agent_ids:
+            return {}
+        Agent = persistence_agent.Agent
+        result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(Agent.uuid, Agent.name).where(
+                sqlalchemy.and_(
+                    Agent.workspace_uuid == workspace_uuid,
+                    Agent.uuid.in_(agent_ids),
+                )
+            )
+        )
+        return {row.uuid: row.name for row in self._model_rows(result)}
+
+    def _serialize_agent_execution(self, row, agent_names: dict[str, str] | None = None) -> dict:
+        metadata = self._deserialize_json(row.metadata_json)
+        duration_ms = None
+        if row.started_at and row.finished_at:
+            duration_ms = int(round((row.finished_at - row.started_at).total_seconds() * 1000))
+        created_at_ms = _epoch_ms(row.created_at)
+        started_at_ms = _epoch_ms(row.started_at)
+        finished_at_ms = _epoch_ms(row.finished_at)
+        is_debug = (
+            str(row.binding_id or '').startswith('debug:')
+            or str(row.conversation_id or '').startswith('debug:')
+            or row.bot_id == 'websocket-proxy-bot'
+        )
+        return {
+            'source': 'agent',
+            'id': row.run_id,
+            'event_id': row.event_id,
+            'status': row.status,
+            'status_group': _status_group(row.status, 'agent'),
+            'title': metadata.get('event_type') or row.queue_name or row.runner_id,
+            'target_kind': metadata.get('processor_type') or ('agent' if row.agent_id else 'processor'),
+            'target_id': metadata.get('processor_id') or row.agent_id or row.binding_id,
+            'target_name': metadata.get('agent_name')
+            or metadata.get('target_name')
+            or (agent_names or {}).get(metadata.get('processor_id') or row.agent_id or ''),
+            'bot_id': row.bot_id,
+            'bot_name': metadata.get('bot_name'),
+            'pipeline_id': None,
+            'pipeline_name': None,
+            'runner_id': row.runner_id,
+            'conversation_id': row.conversation_id,
+            'session_id': row.conversation_id,
+            'created_at_ms': created_at_ms,
+            'started_at_ms': started_at_ms,
+            'finished_at_ms': finished_at_ms,
+            'duration_ms': duration_ms,
+            'usage': self._deserialize_json(row.usage_json) or None,
+            'cost': self._deserialize_json(row.cost_json) or None,
+            'queue_name': row.queue_name,
+            'debug': is_debug,
+            'has_error': row.status in {'failed', 'timeout'},
+            'status_reason': row.status_reason,
+            'input_preview': _message_preview(json.dumps(metadata.get('input') or {}, ensure_ascii=False), 200),
+        }
+
+    def _serialize_pipeline_execution(self, row) -> dict:
+        created_at_ms = _epoch_ms(row.timestamp)
+        return {
+            'source': 'pipeline',
+            'id': row.id,
+            'event_id': row.event_id,
+            'status': row.status,
+            'status_group': _status_group(row.status, 'pipeline'),
+            'title': _message_preview(row.message_content),
+            'target_kind': 'pipeline',
+            'target_id': row.pipeline_id,
+            'target_name': row.pipeline_name,
+            'bot_id': row.bot_id,
+            'bot_name': row.bot_name,
+            'pipeline_id': row.pipeline_id,
+            'pipeline_name': row.pipeline_name,
+            'runner_id': row.runner_name,
+            'conversation_id': None,
+            'session_id': row.session_id,
+            'created_at_ms': created_at_ms,
+            'started_at_ms': None,
+            'finished_at_ms': None,
+            'duration_ms': None,
+            'usage': None,
+            'cost': None,
+            'queue_name': None,
+            'debug': row.bot_id == 'websocket-proxy-bot',
+            'has_error': row.status == 'error',
+            'input_preview': _message_preview(row.message_content),
+            'platform': row.platform,
+            'user_id': row.user_id,
+            'user_name': row.user_name,
+        }
+
+    async def _get_token_coverage(
+        self,
+        workspace_uuid: str,
+        *,
+        bot_ids: list[str] | None = None,
+        pipeline_ids: list[str] | None = None,
+        start_time: datetime.datetime | None = None,
+        end_time: datetime.datetime | None = None,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
+    ) -> dict:
+        LLMCall = persistence_monitoring.MonitoringLLMCall
+        conditions = [LLMCall.workspace_uuid == workspace_uuid]
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringLLMCall, workspace_uuid, execution_statuses, mode
+            )
+        )
+        if bot_ids:
+            conditions.append(LLMCall.bot_id.in_(bot_ids))
+        if pipeline_ids:
+            conditions.append(self._monitoring_processor_condition(LLMCall, workspace_uuid, pipeline_ids))
+        if start_time:
+            conditions.append(LLMCall.timestamp >= start_time)
+        if end_time:
+            conditions.append(LLMCall.timestamp <= end_time)
+        result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(
+                sqlalchemy.func.count(LLMCall.id),
+                sqlalchemy.func.coalesce(sqlalchemy.func.sum(LLMCall.total_tokens), 0),
+                sqlalchemy.func.sum(sqlalchemy.case((LLMCall.total_tokens > 0, 1), else_=0)),
+            ).where(sqlalchemy.and_(*conditions))
+        )
+        row = result.first()
+        calls, total, calls_with_usage = row if row else (0, 0, 0)
+        return {
+            'total_tokens': int(total or 0),
+            'calls': calls or 0,
+            'calls_with_usage': calls_with_usage or 0,
+        }
+
+    async def _get_execution_summary(
+        self,
+        workspace_uuid: str,
+        *,
+        bot_ids: list[str] | None = None,
+        pipeline_ids: list[str] | None = None,
+        start_time: datetime.datetime | None = None,
+        end_time: datetime.datetime | None = None,
+        statuses: list[str] | None = None,
+        mode: str = 'all',
+    ) -> dict:
+        AgentRun = persistence_agent_run.AgentRun
+        run_conditions = self._agent_run_conditions(
+            workspace_uuid,
+            bot_ids=bot_ids,
+            processor_ids=pipeline_ids,
+            start_time=start_time,
+            end_time=end_time,
+            statuses=self._raw_agent_statuses(statuses),
+            debug=True if mode == 'debug' else False if mode == 'real' else None,
+        )
+        run_status_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(AgentRun.status, sqlalchemy.func.count(AgentRun.id))
+            .where(sqlalchemy.and_(*run_conditions))
+            .group_by(AgentRun.status)
+        )
+        agent_counts = {row[0]: row[1] for row in run_status_result.all()}
+        agent_total = sum(agent_counts.values())
+
+        debug_count_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(sqlalchemy.func.count(AgentRun.id)).where(
+                sqlalchemy.and_(
+                    *run_conditions,
+                    self._debug_run_condition(AgentRun),
+                )
+            )
+        )
+        debug_count = debug_count_result.scalar() or 0
+
+        # Percentiles are computed in Python, so the sample must stay bounded:
+        # take the most recent completed runs of the window and label the count
+        # so the UI can state how many runs the numbers rest on.
+        timing_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(AgentRun.started_at, AgentRun.finished_at)
+            .where(
+                sqlalchemy.and_(
+                    *run_conditions,
+                    AgentRun.status == 'completed',
+                    AgentRun.started_at.is_not(None),
+                    AgentRun.finished_at.is_not(None),
+                )
+            )
+            .order_by(AgentRun.created_at.desc())
+            .limit(self._detail_limit())
+        )
+        durations = sorted((row[1] - row[0]).total_seconds() * 1000 for row in timing_result.all() if row[0] and row[1])
+
+        pipeline_conditions = self._pipeline_execution_conditions(
+            workspace_uuid,
+            bot_ids=bot_ids,
+            pipeline_ids=pipeline_ids,
+            statuses=self._raw_pipeline_statuses(statuses),
+            start_time=start_time,
+            end_time=end_time,
+            debug=True if mode == 'debug' else False if mode == 'real' else None,
+        )
+        Message = persistence_monitoring.MonitoringMessage
+        pipeline_timing = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(AgentRun.started_at, AgentRun.finished_at)
+            .select_from(Message)
+            .join(
+                AgentRun,
+                sqlalchemy.and_(
+                    AgentRun.workspace_id == workspace_uuid,
+                    AgentRun.run_id == Message.run_id,
+                ),
+            )
+            .where(
+                *pipeline_conditions,
+                Message.status == 'success',
+                AgentRun.started_at.is_not(None),
+                AgentRun.finished_at.is_not(None),
+            )
+            .order_by(Message.timestamp.desc())
+            .limit(self._detail_limit())
+        )
+        durations.extend((row[1] - row[0]).total_seconds() * 1000 for row in pipeline_timing.all() if row[0] and row[1])
+        durations.sort()
+        pipeline_status_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(
+                persistence_monitoring.MonitoringMessage.status,
+                sqlalchemy.func.count(persistence_monitoring.MonitoringMessage.id),
+            )
+            .where(sqlalchemy.and_(*pipeline_conditions))
+            .group_by(persistence_monitoring.MonitoringMessage.status)
+        )
+        pipeline_counts = {row[0]: row[1] for row in pipeline_status_result.all()}
+        pipeline_total = sum(pipeline_counts.values())
+        pipeline_debug_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(sqlalchemy.func.count(persistence_monitoring.MonitoringMessage.id)).where(
+                *pipeline_conditions, self._debug_pipeline_condition(persistence_monitoring.MonitoringMessage)
+            )
+        )
+        debug_count += pipeline_debug_result.scalar() or 0
+
+        # Pending interactions are gated by the same bot/window scope as the runs
+        # so "waiting input" cannot drift from the rest of the card.
+        waiting_conditions = [
+            persistence_agent_interaction.AgentInteraction.workspace_id == workspace_uuid,
+            persistence_agent_interaction.AgentInteraction.status == 'pending',
+        ]
+        if bot_ids:
+            waiting_conditions.append(persistence_agent_interaction.AgentInteraction.bot_id.in_(bot_ids))
+        if pipeline_ids:
+            waiting_conditions.append(
+                persistence_agent_interaction.AgentInteraction.run_id.in_(
+                    self._processor_runs(workspace_uuid, pipeline_ids)
+                )
+            )
+        if start_time:
+            waiting_conditions.append(persistence_agent_interaction.AgentInteraction.created_at >= start_time)
+        if end_time:
+            waiting_conditions.append(persistence_agent_interaction.AgentInteraction.created_at <= end_time)
+        waiting_conditions.append(
+            persistence_agent_interaction.AgentInteraction.run_id.in_(
+                sqlalchemy.select(AgentRun.run_id).where(*run_conditions)
+            )
+        )
+        waiting_result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(sqlalchemy.func.count(persistence_agent_interaction.AgentInteraction.id)).where(
+                sqlalchemy.and_(*waiting_conditions)
+            )
+        )
+        waiting = waiting_result.scalar() or 0
+
+        by_status: dict[str, int] = {}
+        for status, count in agent_counts.items():
+            group = _status_group(status, 'agent')
+            by_status[group] = by_status.get(group, 0) + count
+        for status, count in pipeline_counts.items():
+            group = _status_group(status, 'pipeline')
+            by_status[group] = by_status.get(group, 0) + count
+
+        completed = by_status.get('completed', 0)
+        failed = by_status.get('failed', 0)
+        denominator = completed + failed
+        total = agent_total + pipeline_total
+        return {
+            'total': total,
+            'by_source': {'agent': agent_total, 'pipeline': pipeline_total},
+            'by_status': by_status,
+            'completed': completed,
+            'failed': failed,
+            'running': by_status.get('running', 0),
+            'queued': by_status.get('queued', 0),
+            'cancelled': by_status.get('cancelled', 0),
+            'ignored': by_status.get('ignored', 0),
+            'waiting': waiting,
+            'success_rate': round(completed / denominator * 100, 2) if denominator > 0 else None,
+            'denominator': denominator,
+            'p50_duration_ms': _percentile_ms(durations, 0.5),
+            'p95_duration_ms': _percentile_ms(durations, 0.95),
+            'duration_sample': len(durations),
+            'debug': debug_count,
+            'real': total - debug_count,
+        }
+
+    async def get_executions(
+        self,
+        context: TenantContext,
+        *,
+        bot_ids: list[str] | None = None,
+        pipeline_ids: list[str] | None = None,
+        agent_ids: list[str] | None = None,
+        statuses: list[str] | None = None,
+        source: str = 'all',
+        mode: str = 'all',
+        start_time: datetime.datetime | None = None,
+        end_time: datetime.datetime | None = None,
+        limit: int = 50,
+        offset: int = 0,
+        include_summary: bool = True,
+    ) -> dict:
+        """List executions from both object types (agent runs and pipeline queries)."""
+        workspace_uuid = require_workspace_uuid(context)
+        limit, offset = self.normalize_page_window(limit, offset)
+        if source not in {'all', 'agent', 'pipeline', 'event'}:
+            source = 'all'
+        if mode not in {'all', 'real', 'debug'}:
+            mode = 'all'
+        debug = None if mode == 'all' else mode == 'debug'
+        branches = []
+        AgentRun = persistence_agent_run.AgentRun
+        Message = persistence_monitoring.MonitoringMessage
+        if source in {'all', 'agent'}:
+            conditions = self._agent_run_conditions(
+                workspace_uuid,
+                bot_ids=bot_ids,
+                agent_ids=agent_ids,
+                processor_ids=pipeline_ids,
+                statuses=self._raw_agent_statuses(statuses),
+                start_time=start_time,
+                end_time=end_time,
+                debug=debug,
+            )
+            branches.append(
+                sqlalchemy.select(
+                    sqlalchemy.literal('agent').label('source'),
+                    AgentRun.run_id.label('id'),
+                    AgentRun.created_at.label('time'),
+                ).where(*conditions)
+            )
+        if source in {'all', 'pipeline'} and (not agent_ids or pipeline_ids):
+            conditions = self._pipeline_execution_conditions(
+                workspace_uuid,
+                bot_ids=bot_ids,
+                pipeline_ids=pipeline_ids,
+                statuses=self._raw_pipeline_statuses(statuses),
+                start_time=start_time,
+                end_time=end_time,
+                debug=debug,
+            )
+            branches.append(
+                sqlalchemy.select(
+                    sqlalchemy.literal('pipeline').label('source'),
+                    Message.id.label('id'),
+                    Message.timestamp.label('time'),
+                ).where(*conditions)
+            )
+        if source in {'all', 'event'} and mode != 'debug' and not pipeline_ids and not agent_ids:
+            conditions = self._unhandled_event_conditions(workspace_uuid, bot_ids, start_time, end_time)
+            if statuses:
+                conditions.append(self._execution_json_text(EventLog.metadata_json, 'status').in_(statuses))
+            branches.append(
+                sqlalchemy.select(
+                    sqlalchemy.literal('event').label('source'),
+                    EventLog.event_id.label('id'),
+                    EventLog.created_at.label('time'),
+                ).where(*conditions)
+            )
+        total = 0
+        page = []
+        if branches:
+            combined = sqlalchemy.union_all(*branches).subquery()
+            count = await self.ap.persistence_mgr.execute_async(
+                sqlalchemy.select(sqlalchemy.func.count()).select_from(combined)
+            )
+            total = count.scalar() or 0
+            refs = (
+                await self.ap.persistence_mgr.execute_async(
+                    sqlalchemy.select(combined)
+                    .order_by(combined.c.time.desc(), combined.c.source, combined.c.id)
+                    .offset(offset)
+                    .limit(limit)
+                )
+            ).all()
+            by_key = {}
+            for kind, model, identity_column in [
+                ('agent', AgentRun, AgentRun.run_id),
+                ('pipeline', Message, Message.id),
+                ('event', EventLog, EventLog.event_id),
+            ]:
+                ids = [r.id for r in refs if r.source == kind]
+                if not ids:
+                    continue
+                workspace_column = model.workspace_uuid if kind == 'pipeline' else model.workspace_id
+                records = await self._execution_rows(
+                    workspace_uuid,
+                    sqlalchemy.select(model).where(workspace_column == workspace_uuid, identity_column.in_(ids)),
+                )
+                names = await self._agent_names(workspace_uuid, records) if kind == 'agent' else {}
+                for record in records:
+                    item = (
+                        self._serialize_agent_execution(record, names)
+                        if kind == 'agent'
+                        else self._serialize_pipeline_execution(record)
+                        if kind == 'pipeline'
+                        else self._serialize_event_execution(record)
+                    )
+                    by_key[(kind, item['id'])] = item
+            page = [by_key[(ref.source, ref.id)] for ref in refs if (ref.source, ref.id) in by_key]
+            await self.enrich_execution_rows(workspace_uuid, page)
+        if not include_summary:
+            return {'items': page, 'total': total, 'has_more': total > offset + len(page)}
+        return {
+            'items': page,
+            'total': total,
+            'limit': limit,
+            'offset': offset,
+            'has_more': total > offset + len(page),
+            'summary': {
+                'executions': await self._get_execution_summary(
+                    workspace_uuid,
+                    statuses=statuses,
+                    mode=mode,
+                    bot_ids=bot_ids,
+                    pipeline_ids=pipeline_ids,
+                    start_time=start_time,
+                    end_time=end_time,
+                ),
+                'tokens': await self._get_token_coverage(
+                    workspace_uuid,
+                    execution_statuses=statuses,
+                    mode=mode,
+                    bot_ids=bot_ids,
+                    pipeline_ids=pipeline_ids,
+                    start_time=start_time,
+                    end_time=end_time,
+                ),
+            },
+        }
+
+    async def get_inflight_snapshot(self, context: TenantContext, previous: list[dict] | None = None) -> dict:
+        """SSE producers outlive the HTTP handler's tenant scope."""
+        workspace = require_workspace_uuid(context)
+        tenant_scope = getattr(self.ap.persistence_mgr, 'tenant_scope', None)
+        if callable(tenant_scope):
+            async with tenant_scope(workspace):
+                return await self._get_inflight_snapshot(context, previous)
+        return await self._get_inflight_snapshot(context, previous)
+
+    async def _get_inflight_snapshot(self, context: TenantContext, previous: list[dict] | None = None) -> dict:
+        """Bounded Workspace snapshot without dashboard aggregates or token payloads."""
+        from ....utils.inflight import inflight_hub
+
+        workspace = require_workspace_uuid(context)
+        active = await self.get_executions(context, statuses=['running', 'queued'], limit=100, include_summary=False)
+        # Execution timestamps are stored as UTC without a timezone. asyncpg
+        # rejects aware datetimes for these columns, even though SQLite accepts them.
+        recent_start = datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None) - datetime.timedelta(
+            seconds=30
+        )
+        recent = await self.get_executions(
+            context,
+            start_time=recent_start,
+            limit=100,
+            include_summary=False,
+        )
+        items = {(r['source'], r['id']): r for r in recent['items'] + active['items'] if r['source'] != 'event'}
+        # Keep the actual terminal state of older runs that just left the active set.
+        for kind, model, column, scope in [
+            (
+                'agent',
+                persistence_agent_run.AgentRun,
+                persistence_agent_run.AgentRun.run_id,
+                persistence_agent_run.AgentRun.workspace_id,
+            ),
+            (
+                'pipeline',
+                persistence_monitoring.MonitoringMessage,
+                persistence_monitoring.MonitoringMessage.id,
+                persistence_monitoring.MonitoringMessage.workspace_uuid,
+            ),
+        ]:
+            missing = [r['id'] for r in (previous or [])[:100] if r['source'] == kind and (kind, r['id']) not in items]
+            if missing:
+                records = await self._execution_rows(
+                    workspace, sqlalchemy.select(model).where(scope == workspace, column.in_(missing))
+                )
+                names = await self._agent_names(workspace, records) if kind == 'agent' else {}
+                for record in records:
+                    row = (
+                        self._serialize_agent_execution(record, names)
+                        if kind == 'agent'
+                        else self._serialize_pipeline_execution(record)
+                    )
+                    items[(kind, row['id'])] = row
+        rows = list(items.values())
+        Run = persistence_agent_run.AgentRun
+        Message = persistence_monitoring.MonitoringMessage
+        agent_ids = [r['id'] for r in rows if r['source'] == 'agent']
+        pipeline_ids = [r['id'] for r in rows if r['source'] == 'pipeline']
+        linked = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(Run.run_id, Message.id.label('message_id'))
+            .outerjoin(Message, sqlalchemy.and_(Message.workspace_uuid == workspace, Message.run_id == Run.run_id))
+            .where(
+                Run.workspace_id == workspace, sqlalchemy.or_(Run.run_id.in_(agent_ids), Message.id.in_(pipeline_ids))
+            )
+        )
+        identities = {
+            r.run_id: ('pipeline', r.message_id) if r.message_id in pipeline_ids else ('agent', r.run_id)
+            for r in linked.all()
+        }
+        inflight_hub.track(workspace, identities)
+        Event = persistence_agent_run.AgentRunEvent
+        latest = (
+            sqlalchemy.select(Event.run_id, sqlalchemy.func.max(Event.sequence).label('seq'))
+            .where(Event.run_id.in_(identities))
+            .group_by(Event.run_id)
+            .subquery()
+        )
+        phases = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(Event.run_id, Event.type).join(
+                latest, sqlalchemy.and_(Event.run_id == latest.c.run_id, Event.sequence == latest.c.seq)
+            )
+        )
+        for run_id, event_type in phases.all():
+            row = items.get(identities[run_id])
+            if row is not None:
+                row['progress_event'] = event_type
+        return {'items': rows, 'active_total': active['total'], 'truncated': active['has_more'] or recent['has_more']}
+
     async def get_token_statistics(
         self,
         context: TenantContext,
@@ -902,6 +1776,8 @@ class MonitoringService:
         start_time: datetime.datetime | None = None,
         end_time: datetime.datetime | None = None,
         bucket: str = 'hour',
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> dict:
         """Get detailed token usage statistics for production observability.
 
@@ -920,10 +1796,15 @@ class MonitoringService:
             bucket = 'hour'
 
         conditions = [LLMCall.workspace_uuid == workspace_uuid]
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringLLMCall, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(LLMCall.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(LLMCall.pipeline_id.in_(pipeline_ids))
+            conditions.append(self._monitoring_processor_condition(LLMCall, workspace_uuid, pipeline_ids))
         if start_time:
             conditions.append(LLMCall.timestamp >= start_time)
         if end_time:
@@ -1104,16 +1985,40 @@ class MonitoringService:
         end_time: datetime.datetime | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get messages with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
+        if bot_ids and session_ids and not pipeline_ids and not execution_statuses and mode == 'all':
+            source = self._session_message_source(workspace_uuid, bot_ids)
+            query = sqlalchemy.select(source).where(source.c.session_id.in_(session_ids))
+            if start_time:
+                query = query.where(source.c.timestamp >= start_time)
+            if end_time:
+                query = query.where(source.c.timestamp <= end_time)
+            return await self._session_projection_page(
+                workspace_uuid,
+                query.order_by(source.c.timestamp.desc(), source.c.id.desc()),
+                limit,
+                offset,
+            )
         conditions = [persistence_monitoring.MonitoringMessage.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringMessage, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringMessage.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringMessage.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringMessage, workspace_uuid, pipeline_ids
+                )
+            )
         if session_ids:
             conditions.append(persistence_monitoring.MonitoringMessage.session_id.in_(session_ids))
         if start_time:
@@ -1159,16 +2064,27 @@ class MonitoringService:
         end_time: datetime.datetime | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get LLM calls with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringLLMCall.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringLLMCall, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringLLMCall.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringLLMCall.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringLLMCall, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringLLMCall.timestamp >= start_time)
         if end_time:
@@ -1214,16 +2130,27 @@ class MonitoringService:
         end_time: datetime.datetime | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get tool calls with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringToolCall.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringToolCall, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringToolCall.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringToolCall.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringToolCall, workspace_uuid, pipeline_ids
+                )
+            )
         if session_ids:
             conditions.append(persistence_monitoring.MonitoringToolCall.session_id.in_(session_ids))
         if start_time:
@@ -1267,12 +2194,19 @@ class MonitoringService:
         knowledge_base_id: str | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get embedding calls with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringEmbeddingCall.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringEmbeddingCall, workspace_uuid, execution_statuses, mode
+            )
+        )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringEmbeddingCall.timestamp >= start_time)
         if end_time:
@@ -1310,6 +2244,112 @@ class MonitoringService:
             total,
         )
 
+    @_workspace_transaction
+    async def reset_session_context(self, context: TenantContext, bot_id: str, session_id: str) -> dict:
+        """Reset model context without deleting monitoring history or other sessions."""
+        from ....agent.runner.context_reset import HOST_RESET_RUNNER, reset_scope
+        from ....entity.persistence.runner_state import RunnerState
+
+        if not isinstance(bot_id, str) or not bot_id or not isinstance(session_id, str) or not session_id:
+            raise ValueError('bot_id and session_id are required')
+        workspace = require_workspace_uuid(context)
+        execute = self.ap.persistence_mgr.execute_async
+        source = self._session_message_source(workspace, [bot_id])
+        exists = (
+            await execute(
+                sqlalchemy.select(source.c.session_id)
+                .where(
+                    source.c.session_id == session_id,
+                )
+                .limit(1)
+            )
+        ).first()
+        if exists is None:
+            session = persistence_monitoring.MonitoringSession
+            exists = (
+                await execute(
+                    sqlalchemy.select(session.session_id)
+                    .where(
+                        session.workspace_uuid == workspace,
+                        session.bot_id == bot_id,
+                        session.session_id == session_id,
+                    )
+                    .limit(1)
+                )
+            ).first()
+        if exists is None:
+            raise LookupError('Session not found')
+        # Legacy sessions use uppercase PERSON/GROUP; envelope IDs use lowercase.
+        conversation = re.sub(r'^(PERSON|GROUP)_', lambda match: match[1].lower() + '_', session_id)
+        run = persistence_agent_run.AgentRun
+        busy = (
+            await execute(
+                sqlalchemy.select(run.run_id)
+                .where(
+                    run.workspace_id == workspace,
+                    run.bot_id == bot_id,
+                    run.conversation_id == conversation,
+                    run.status.in_(['created', 'queued', 'claimed', 'running']),
+                )
+                .limit(1)
+            )
+        ).first()
+        message = persistence_monitoring.MonitoringMessage
+        pending = (
+            await execute(
+                sqlalchemy.select(message.id)
+                .where(
+                    message.workspace_uuid == workspace,
+                    message.bot_id == bot_id,
+                    message.session_id.in_([session_id, conversation]),
+                    message.status == 'pending',
+                )
+                .limit(1)
+            )
+        ).first()
+        interaction = persistence_agent_interaction.AgentInteraction
+        waiting = (
+            await execute(
+                sqlalchemy.select(interaction.id)
+                .where(
+                    interaction.workspace_id == workspace,
+                    interaction.bot_id == bot_id,
+                    interaction.conversation_id == conversation,
+                    interaction.status.in_(['pending', 'submitted']),
+                    sqlalchemy.or_(
+                        interaction.expires_at.is_(None), interaction.expires_at > datetime.datetime.utcnow()
+                    ),
+                )
+                .limit(1)
+            )
+        ).first()
+        if busy or pending or waiting:
+            raise RuntimeError('Session is processing a task; wait until it finishes')
+        await execute(
+            sqlalchemy.delete(RunnerState).where(
+                RunnerState.workspace_id == workspace,
+                RunnerState.bot_id == bot_id,
+                RunnerState.conversation_id == conversation,
+                RunnerState.scope == 'conversation',
+            )
+        )
+        generation = str(uuid.uuid4())
+        await execute(
+            sqlalchemy.insert(RunnerState).values(
+                runner_id=HOST_RESET_RUNNER,
+                binding_identity=HOST_RESET_RUNNER,
+                scope='conversation',
+                scope_key=reset_scope(workspace, bot_id, conversation),
+                state_key='generation',
+                value_json=json.dumps(generation),
+                workspace_id=workspace,
+                bot_id=bot_id,
+                conversation_id=conversation,
+                created_at=datetime.datetime.utcnow(),
+            )
+        )
+        return {'reset': True, 'session_id': session_id}
+
     async def get_sessions(
         self,
         context: TenantContext,
@@ -1321,16 +2361,38 @@ class MonitoringService:
         is_active: bool | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get sessions with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
+        if bot_ids and not pipeline_ids and not execution_statuses and mode == 'all':
+            return await self.get_bot_conversation_sessions(
+                workspace_uuid,
+                bot_ids,
+                start_time,
+                end_time,
+                user_query,
+                is_active,
+                limit,
+                offset,
+            )
         conditions = [persistence_monitoring.MonitoringSession.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringSession, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringSession.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringSession.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringSession, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringSession.last_activity >= start_time)
         if end_time:
@@ -1385,16 +2447,27 @@ class MonitoringService:
         end_time: datetime.datetime | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get errors with filters"""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringError.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringError, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringError.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringError.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringError, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringError.timestamp >= start_time)
         if end_time:
@@ -1860,7 +2933,11 @@ class MonitoringService:
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringMessage.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringMessage.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringMessage, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringMessage.timestamp >= start_time)
         if end_time:
@@ -1916,7 +2993,11 @@ class MonitoringService:
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringLLMCall.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringLLMCall.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringLLMCall, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringLLMCall.timestamp >= start_time)
         if end_time:
@@ -2023,7 +3104,11 @@ class MonitoringService:
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringError.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringError.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringError, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringError.timestamp >= start_time)
         if end_time:
@@ -2074,7 +3159,11 @@ class MonitoringService:
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringSession.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringSession.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringSession, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringSession.last_activity >= start_time)
         if end_time:
@@ -2227,6 +3316,8 @@ class MonitoringService:
         pipeline_ids: list[str] | None = None,
         start_time: datetime.datetime | None = None,
         end_time: datetime.datetime | None = None,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> dict:
         """Get feedback statistics.
 
@@ -2236,10 +3327,19 @@ class MonitoringService:
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringFeedback.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringFeedback, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringFeedback.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringFeedback.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringFeedback, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringFeedback.timestamp >= start_time)
         if end_time:
@@ -2320,16 +3420,27 @@ class MonitoringService:
         end_time: datetime.datetime | None = None,
         limit: int = 100,
         offset: int = 0,
+        execution_statuses: list[str] | None = None,
+        mode: str = 'all',
     ) -> tuple[list[dict], int]:
         """Get feedback list with filters."""
         limit, offset = self.normalize_page_window(limit, offset)
         workspace_uuid = require_workspace_uuid(context)
         conditions = [persistence_monitoring.MonitoringFeedback.workspace_uuid == workspace_uuid]
 
+        conditions.append(
+            self._monitoring_execution_condition(
+                persistence_monitoring.MonitoringFeedback, workspace_uuid, execution_statuses, mode
+            )
+        )
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringFeedback.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringFeedback.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringFeedback, workspace_uuid, pipeline_ids
+                )
+            )
         if feedback_type is not None:
             conditions.append(persistence_monitoring.MonitoringFeedback.feedback_type == feedback_type)
         if start_time:
@@ -2382,7 +3493,11 @@ class MonitoringService:
         if bot_ids:
             conditions.append(persistence_monitoring.MonitoringFeedback.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(persistence_monitoring.MonitoringFeedback.pipeline_id.in_(pipeline_ids))
+            conditions.append(
+                self._monitoring_processor_condition(
+                    persistence_monitoring.MonitoringFeedback, workspace_uuid, pipeline_ids
+                )
+            )
         if start_time:
             conditions.append(persistence_monitoring.MonitoringFeedback.timestamp >= start_time)
         if end_time:
diff --git a/src/langbot/pkg/api/http/service/monitoring_execution_details.py b/src/langbot/pkg/api/http/service/monitoring_execution_details.py
new file mode 100644
index 000000000..467a8026f
--- /dev/null
+++ b/src/langbot/pkg/api/http/service/monitoring_execution_details.py
@@ -0,0 +1,804 @@
+"""Tenant-scoped execution inputs, outputs, related runs and conversation history."""
+
+from __future__ import annotations
+
+import datetime
+import json
+from langbot_plugin.api.entities.builtin.platform.events import event_summary
+
+import sqlalchemy as sa
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from ....entity.persistence.agent_run import AgentRun, AgentRunEvent
+from ....entity.persistence.event_log import EventLog
+from ....entity.persistence.transcript import Transcript
+from ....entity.persistence import monitoring as models
+from .tenant import require_workspace_uuid
+
+
+def epoch(value):
+    if value is None:
+        return None
+    return int(value.replace(tzinfo=datetime.timezone.utc).timestamp() * 1000)
+
+
+def parsed(value, fallback=None):
+    try:
+        return json.loads(value) if value else fallback
+    except (TypeError, ValueError):
+        return fallback
+
+
+def event_content(content, event_data):
+    """Keep event-specific properties alongside the normalized message input."""
+    if not isinstance(event_data, dict) or not event_data:
+        return content
+    # Keep the original event intact; text is the existing normalized input field.
+    return {'text': event_summary(event_data), 'event': event_data, 'input': content}
+
+
+def event_preview(payload, fallback='', event_type=''):
+    """Never use serialized event metadata as a human-readable list preview."""
+    if event_type.startswith('wecomcs.'):
+        return ''
+    if isinstance(payload, dict):
+        if payload.get('type'):
+            if str(payload['type']).startswith('wecomcs.'):
+                return ''
+            return event_summary(payload)[:1000]
+        if isinstance(payload.get('text'), str):
+            return payload['text'][:1000]
+    if fallback and not isinstance(parsed(fallback), (dict, list)):
+        return fallback[:1000]
+    return ''
+
+
+def event_item(row):
+    metadata = parsed(row.metadata_json, {})
+    return {
+        'id': row.event_id,
+        'event_type': row.event_type,
+        'source': row.source,
+        'timestamp_ms': epoch(row.event_time or row.created_at),
+        'actor_id': row.actor_id,
+        'actor_name': row.actor_name,
+        'conversation_id': row.conversation_id,
+        'thread_id': row.thread_id,
+        'content': event_content(parsed(row.input_json, None) or row.input_summary, metadata.get('input_event')),
+        'metadata': metadata,
+    }
+
+
+def message_item(row):
+    return {
+        'id': row.id,
+        'role': row.role or 'user',
+        'timestamp_ms': epoch(row.timestamp),
+        'content': parsed(row.message_content, None) or row.message_content,
+        'actor_id': row.user_id,
+        'actor_name': row.user_name,
+        'status': row.status,
+        'event_id': row.event_id,
+        'run_id': row.run_id,
+        'origin': 'delivery' if row.role == 'assistant' else 'input',
+        'delivery': parsed(row.variables, {}) if row.role == 'assistant' else None,
+    }
+
+
+def transcript_item(row):
+    return {
+        'id': row.transcript_id,
+        'role': row.role,
+        'timestamp_ms': epoch(row.created_at),
+        'content': parsed(row.content_json, None) or row.content,
+        'attachments': parsed(row.attachment_refs_json, []),
+        'event_id': row.event_id,
+        'run_id': row.run_id,
+        'origin': 'generated' if row.role == 'assistant' else 'input',
+    }
+
+
+class ExecutionDetailsMixin:
+    def _session_message_source(self, workspace, bot_ids):
+        """Combine delivered messages with ingress missing from the legacy ledger."""
+        message = models.MonitoringMessage
+        conversation = sa.func.coalesce(
+            EventLog.conversation_id,
+            sa.select(AgentRun.conversation_id)
+            .where(
+                AgentRun.workspace_id == workspace,
+                AgentRun.bot_id == EventLog.bot_id,
+                AgentRun.event_id == EventLog.event_id,
+            )
+            .order_by(AgentRun.id)
+            .limit(1)
+            .scalar_subquery(),
+            sa.case(
+                (
+                    self._execution_json_text(EventLog.input_json, 'chat_type') == 'group',
+                    sa.literal('group_') + self._execution_json_text(EventLog.input_json, 'chat_id'),
+                ),
+                (
+                    self._execution_json_text(EventLog.input_json, 'chat_type') == 'private',
+                    sa.literal('person_') + self._execution_json_text(EventLog.input_json, ('sender', 'id')),
+                ),
+            ),
+        )
+        values = {
+            'id': EventLog.event_id,
+            'workspace_uuid': EventLog.workspace_id,
+            'timestamp': sa.func.coalesce(EventLog.event_time, EventLog.created_at),
+            'bot_id': EventLog.bot_id,
+            'bot_name': sa.func.coalesce(self._execution_json_text(EventLog.metadata_json, 'bot_name'), ''),
+            'pipeline_id': sa.func.coalesce(
+                self._execution_json_text(EventLog.metadata_json, ('routes', 0, 'target_uuid')), ''
+            ),
+            'pipeline_name': sa.func.coalesce(
+                self._execution_json_text(EventLog.metadata_json, ('routes', 0, 'target_type')), ''
+            ),
+            'message_content': sa.func.coalesce(EventLog.input_json, EventLog.input_summary, ''),
+            'session_id': conversation,
+            'status': sa.func.coalesce(self._execution_json_text(EventLog.metadata_json, 'status'), 'success'),
+            'level': sa.literal('info'),
+            'platform': self._execution_json_text(EventLog.metadata_json, 'platform'),
+            'user_id': EventLog.actor_id,
+            'user_name': EventLog.actor_name,
+            'role': sa.literal('user'),
+            'event_id': EventLog.event_id,
+            'run_id': EventLog.run_id,
+        }
+        columns = list(message.__table__.columns)
+        stored = sa.select(*columns).where(message.workspace_uuid == workspace, message.bot_id.in_(bot_ids))
+        incoming = sa.select(
+            *[values.get(column.name, sa.literal(None)).label(column.name) for column in columns]
+        ).where(
+            EventLog.workspace_id == workspace,
+            EventLog.bot_id.in_(bot_ids),
+            EventLog.source == 'platform',
+            EventLog.event_type == 'message.received',
+            conversation.is_not(None),
+            conversation != '',
+            ~sa.exists(
+                sa.select(message.id).where(
+                    message.workspace_uuid == workspace,
+                    message.bot_id == EventLog.bot_id,
+                    sa.func.coalesce(message.role, 'user') != 'assistant',
+                    sa.or_(message.event_id == EventLog.event_id, message.id == EventLog.event_id),
+                )
+            ),
+        )
+        return sa.union_all(stored, incoming).subquery()
+
+    async def _session_projection_page(self, workspace, statement, limit, offset):
+        count = (
+            await self._execution_query(workspace, sa.select(sa.func.count()).select_from(statement.subquery()))
+        ).scalar_one()
+        rows = (await self._execution_query(workspace, statement.limit(limit).offset(offset))).mappings().all()
+        return [
+            {key: value.isoformat() if isinstance(value, datetime.datetime) else value for key, value in row.items()}
+            for row in rows
+        ], count
+
+    async def get_bot_conversation_sessions(
+        self, workspace, bot_ids, start_time, end_time, user_query, is_active, limit, offset
+    ):
+        """Read session summaries across processor types without mutating history."""
+        messages = self._session_message_source(workspace, bot_ids)
+        key = [messages.c.workspace_uuid, messages.c.bot_id, messages.c.session_id]
+        ranked = (
+            sa.select(
+                *messages.c,
+                sa.func.row_number()
+                .over(
+                    partition_by=key,
+                    order_by=[
+                        sa.case((messages.c.role == 'user', 0), else_=1),
+                        messages.c.timestamp.desc(),
+                        messages.c.id.desc(),
+                    ],
+                )
+                .label('position'),
+                sa.func.min(messages.c.timestamp).over(partition_by=key).label('first_seen'),
+                sa.func.max(messages.c.timestamp).over(partition_by=key).label('last_seen'),
+                sa.func.count().over(partition_by=key).label('count'),
+            )
+            .where(messages.c.session_id != '')
+            .subquery()
+        )
+        session = models.MonitoringSession
+        fields = list(session.__table__.columns)
+        projected = {
+            'start_time': ranked.c.first_seen,
+            'last_activity': ranked.c.last_seen,
+            'message_count': ranked.c['count'],
+            'is_active': sa.literal(True),
+        }
+        derived = sa.select(
+            *[
+                (projected[field.name] if field.name in projected else ranked.c[field.name]).label(field.name)
+                for field in fields
+            ]
+        ).where(ranked.c.position == 1)
+        existing = sa.select(*fields).where(session.workspace_uuid == workspace, session.bot_id.in_(bot_ids))
+        combined = sa.union_all(existing, derived).subquery()
+        group = [combined.c.workspace_uuid, combined.c.bot_id, combined.c.session_id]
+        merged = sa.select(
+            *[
+                combined.c[field.name]
+                for field in fields
+                if field.name not in {'message_count', 'start_time', 'last_activity'}
+            ],
+            sa.func.max(combined.c.message_count).over(partition_by=group).label('message_count'),
+            sa.func.min(combined.c.start_time).over(partition_by=group).label('start_time'),
+            sa.func.max(combined.c.last_activity).over(partition_by=group).label('last_activity'),
+            sa.func.row_number().over(partition_by=group, order_by=combined.c.last_activity.desc()).label('position'),
+        ).subquery()
+        query = sa.select(*[merged.c[field.name] for field in fields]).where(merged.c.position == 1)
+        if start_time:
+            query = query.where(merged.c.last_activity >= start_time)
+        if end_time:
+            query = query.where(merged.c.last_activity <= end_time)
+        if user_query and user_query.strip():
+            pattern = f'%{user_query.strip()}%'
+            query = query.where(sa.or_(merged.c.user_id.ilike(pattern), merged.c.user_name.ilike(pattern)))
+        if is_active is not None:
+            query = query.where(merged.c.is_active == is_active)
+        return await self._session_projection_page(
+            workspace,
+            query.order_by(merged.c.last_activity.desc(), merged.c.session_id),
+            limit,
+            offset,
+        )
+
+    def _execution_json_text(self, column, key):
+        # SQLite JSON functions accept TEXT directly; PostgreSQL operators
+        # require an explicit JSON cast for the existing text-backed journals.
+        if self.ap.persistence_mgr.get_db_engine().dialect.name == 'sqlite':
+            expression = sa.type_coerce(column, sa.JSON)
+        else:
+            expression = sa.cast(column, sa.JSON)
+        return expression[key].as_string()
+
+    async def link_execution_message(self, context, message_id, run_id, event_id):
+        workspace = self._require_write_context(context)
+        statement = (
+            sa.update(models.MonitoringMessage)
+            .where(
+                models.MonitoringMessage.workspace_uuid == workspace,
+                models.MonitoringMessage.id == message_id,
+            )
+            .values(run_id=run_id, event_id=sa.func.coalesce(models.MonitoringMessage.event_id, event_id))
+        )
+        tenant_uow = getattr(self.ap.persistence_mgr, 'tenant_uow', None)
+        if callable(tenant_uow):
+            async with tenant_uow(workspace) as uow:
+                await uow.session.execute(statement)
+        else:
+            await self.ap.persistence_mgr.execute_async(statement)
+
+    async def record_ingress_event(
+        self,
+        context,
+        *,
+        event_id,
+        event_type,
+        bot_id,
+        bot_name,
+        platform,
+        payload,
+        actor_id=None,
+        actor_name=None,
+        conversation_id=None,
+        status='running',
+        routes=None,
+    ):
+        """Persist entry facts even when routing never creates a processor run."""
+        from ....agent.runner.event_log_store import EventLogStore
+
+        workspace = self._require_write_context(context)
+        store = EventLogStore(self.ap.persistence_mgr.get_db_engine())
+        # Reuse the message sanitizer (including inline-media handling) instead
+        # of storing an unbounded adapter object in the event journal.
+        content = parsed(self._sanitize_message_content(json.dumps(payload, ensure_ascii=False, default=str)), {})
+        summary = event_preview(content, event_type=event_type)
+        await store.append_event(
+            event_id=event_id,
+            event_type=event_type,
+            source='platform',
+            workspace_id=workspace,
+            bot_id=bot_id,
+            actor_id=actor_id,
+            actor_name=actor_name,
+            conversation_id=conversation_id,
+            actor_type='user' if actor_id else 'system',
+            input_summary=summary,
+            input_json=content,
+            metadata={'status': status, 'bot_name': bot_name, 'platform': platform},
+        )
+        if routes is not None:
+            safe_routes = []
+            for route in routes:
+                if isinstance(route, BaseException):
+                    safe_routes.append({'status': 'failed', 'reason': str(route) or type(route).__name__})
+                elif isinstance(route, dict):
+                    safe_routes.append(
+                        {
+                            key: route.get(key)
+                            for key in (
+                                'status',
+                                'reason',
+                                'target_type',
+                                'target_uuid',
+                                'run_id',
+                                'failure_code',
+                            )
+                        }
+                    )
+            metadata = {'status': status, 'bot_name': bot_name, 'platform': platform, 'routes': safe_routes}
+            async with AsyncSession(self.ap.persistence_mgr.get_db_engine()) as session:
+                await session.execute(
+                    sa.update(EventLog)
+                    .where(
+                        EventLog.workspace_id == workspace,
+                        EventLog.event_id == event_id,
+                    )
+                    .values(metadata_json=json.dumps(metadata, ensure_ascii=False))
+                )
+                await session.commit()
+
+    async def _execution_query(self, workspace, statement):
+        tenant_uow = getattr(self.ap.persistence_mgr, 'tenant_uow', None)
+        if callable(tenant_uow):
+            async with tenant_uow(workspace) as uow:
+                return await uow.session.execute(statement)
+        async with AsyncSession(self.ap.persistence_mgr.get_db_engine()) as session:
+            return await session.execute(statement)
+
+    async def _execution_rows(self, workspace, statement):
+        return list((await self._execution_query(workspace, statement)).scalars().all())
+
+    async def _execution_page(self, workspace, statement, convert, offset, limit):
+        rows = await self._execution_rows(workspace, statement.offset(offset).limit(limit + 1))
+        return {
+            'items': [convert(row) for row in rows[:limit]],
+            'has_more': len(rows) > limit,
+            'next_offset': offset + min(len(rows), limit),
+        }
+
+    def _unhandled_event_conditions(self, workspace, bot_ids=None, start_time=None, end_time=None):
+        # Absorbed steering events belong to an existing run, too.
+        conditions = [
+            EventLog.workspace_id == workspace,
+            # The primary route is first; plugin subscriptions follow it.
+            # Delivered pipeline ingress is not an unhandled execution, including
+            # older queued/aggregated messages whose trace identity was lost.
+            ~sa.and_(
+                sa.func.coalesce(self._execution_json_text(EventLog.metadata_json, ('routes', 0, 'target_type')), '')
+                == 'pipeline',
+                sa.func.coalesce(self._execution_json_text(EventLog.metadata_json, ('routes', 0, 'status')), '')
+                == 'delivered',
+            ),
+            ~sa.exists(
+                sa.select(AgentRun.id).where(
+                    AgentRun.workspace_id == workspace,
+                    sa.or_(
+                        AgentRun.event_id == EventLog.event_id,
+                        AgentRun.run_id == EventLog.run_id,
+                        self._execution_json_text(AgentRun.metadata_json, 'ingress_event_id') == EventLog.event_id,
+                    ),
+                )
+            ),
+            ~sa.exists(
+                sa.select(models.MonitoringMessage.id).where(
+                    models.MonitoringMessage.workspace_uuid == workspace,
+                    models.MonitoringMessage.event_id == EventLog.event_id,
+                )
+            ),
+        ]
+        if bot_ids:
+            conditions.append(EventLog.bot_id.in_(bot_ids))
+        if start_time:
+            conditions.append(EventLog.created_at >= start_time)
+        if end_time:
+            conditions.append(EventLog.created_at <= end_time)
+        return conditions
+
+    def _serialize_event_execution(self, row):
+        metadata = parsed(row.metadata_json, {})
+        status = metadata.get('status', 'ignored')
+        return {
+            'source': 'event',
+            'id': row.event_id,
+            'event_id': row.event_id,
+            'status': status,
+            'status_group': status,
+            'title': row.event_type,
+            'input_preview': event_preview(parsed(row.input_json), row.input_summary, row.event_type),
+            'target_kind': 'event',
+            'target_id': None,
+            'target_name': None,
+            'bot_id': row.bot_id,
+            'bot_name': metadata.get('bot_name'),
+            'platform': metadata.get('platform') or row.source,
+            'user_id': row.actor_id,
+            'user_name': row.actor_name,
+            'conversation_id': row.conversation_id,
+            'session_id': row.conversation_id,
+            'created_at_ms': epoch(row.created_at),
+            'started_at_ms': None,
+            'finished_at_ms': None,
+            'duration_ms': None,
+            'usage': None,
+            'cost': None,
+            'queue_name': None,
+            'pipeline_id': None,
+            'pipeline_name': None,
+            'runner_id': None,
+            'debug': row.source == 'debug',
+            'has_error': status == 'failed',
+            'status_reason': metadata.get('reason')
+            or '; '.join(
+                route.get('reason') or route.get('failure_code') or 'failure_reason_not_recorded'
+                for route in metadata.get('routes', [])
+                if isinstance(route, dict) and route.get('status') == 'failed'
+            )
+            or None,
+        }
+
+    async def enrich_execution_rows(self, workspace, items):
+        pipeline_items = {item['id']: item for item in items if item.get('source') == 'pipeline'}
+        if pipeline_items:
+            # Pipeline rows own the UI identity; their linked Runner ledger owns
+            # timing and usage. Read them in one batch for both list and detail.
+            linked = await self._execution_query(
+                workspace,
+                sa.select(models.MonitoringMessage.id, AgentRun)
+                .join(
+                    AgentRun,
+                    sa.and_(
+                        AgentRun.workspace_id == workspace,
+                        AgentRun.run_id == models.MonitoringMessage.run_id,
+                    ),
+                )
+                .where(
+                    models.MonitoringMessage.workspace_uuid == workspace,
+                    models.MonitoringMessage.id.in_(pipeline_items),
+                ),
+            )
+            for message_id, run in linked.all():
+                metrics = self._serialize_agent_execution(run)
+                pipeline_items[message_id].update(
+                    {key: metrics[key] for key in ('started_at_ms', 'finished_at_ms', 'duration_ms', 'usage', 'cost')}
+                )
+
+            calls = models.MonitoringLLMCall
+            usage_rows = await self._execution_query(
+                workspace,
+                sa.select(
+                    calls.message_id,
+                    sa.func.sum(calls.input_tokens),
+                    sa.func.sum(calls.output_tokens),
+                    sa.func.sum(calls.total_tokens),
+                )
+                .where(
+                    calls.workspace_uuid == workspace,
+                    calls.message_id.in_(pipeline_items),
+                )
+                .group_by(calls.message_id),
+            )
+            for message_id, input_tokens, output_tokens, total_tokens in usage_rows.all():
+                if not pipeline_items[message_id].get('usage'):
+                    pipeline_items[message_id]['usage'] = {
+                        'input_tokens': int(input_tokens or 0),
+                        'output_tokens': int(output_tokens or 0),
+                        'total_tokens': int(total_tokens or 0),
+                    }
+        ids = {item['event_id'] for item in items if item.get('event_id')}
+        if not ids:
+            return
+        events = await self._execution_rows(
+            workspace,
+            sa.select(EventLog).where(
+                EventLog.workspace_id == workspace,
+                EventLog.event_id.in_(ids),
+            ),
+        )
+        by_id = {row.event_id: row for row in events}
+        for item in items:
+            event = by_id.get(item.get('event_id'))
+            if event:
+                meta = parsed(event.metadata_json, {})
+                item.update(
+                    event_type=event.event_type,
+                    input_preview=event_preview(parsed(event.input_json), event.input_summary, event.event_type),
+                    user_id=event.actor_id,
+                    user_name=event.actor_name,
+                    platform=meta.get('platform') or event.source,
+                    bot_name=item.get('bot_name') or meta.get('bot_name'),
+                )
+
+    async def get_execution_detail(self, context, source, execution_id, *, section=None, offset=0, limit=100):
+        workspace = require_workspace_uuid(context)
+        limit, offset = self.normalize_page_window(limit, offset)
+        limit = min(limit, 100)
+        sections = {
+            'inputs',
+            'outputs',
+            'deliveries',
+            'conversation',
+            'related',
+            'events',
+            'llm_calls',
+            'tool_calls',
+            'errors',
+        }
+        if section is not None and section not in sections:
+            raise ValueError('Unknown execution section')
+
+        run = message = event = None
+        if source in {'agent', 'auto'}:
+            rows = await self._execution_rows(
+                workspace,
+                sa.select(AgentRun).where(
+                    AgentRun.workspace_id == workspace,
+                    AgentRun.run_id == execution_id,
+                ),
+            )
+            run = rows[0] if rows else None
+        if run is None and source in {'pipeline', 'auto'}:
+            rows = await self._execution_rows(
+                workspace,
+                sa.select(models.MonitoringMessage).where(
+                    models.MonitoringMessage.workspace_uuid == workspace,
+                    models.MonitoringMessage.id == execution_id,
+                ),
+            )
+            message = rows[0] if rows else None
+            if message and message.parent_message_id:
+                return await self.get_execution_detail(
+                    context, 'auto', message.parent_message_id, section=section, offset=offset, limit=limit
+                )
+        if run is None and message is None and source in {'event', 'auto'}:
+            rows = await self._execution_rows(
+                workspace,
+                sa.select(EventLog).where(
+                    EventLog.workspace_id == workspace,
+                    EventLog.event_id == execution_id,
+                ),
+            )
+            event = rows[0] if rows else None
+        if run is None and message is None and event is None:
+            raise ValueError('Execution not found')
+
+        if message and message.run_id:
+            linked = await self._execution_rows(
+                workspace,
+                sa.select(AgentRun).where(AgentRun.workspace_id == workspace, AgentRun.run_id == message.run_id),
+            )
+            if linked:
+                linked_run = linked[0]
+            else:
+                linked_run = None
+        else:
+            linked_run = None
+
+        if run:
+            names = await self._agent_names(workspace, [run])
+            row = self._serialize_agent_execution(run, names)
+            event_id, bot, conversation, thread = run.event_id, run.bot_id, run.conversation_id, run.thread_id
+            run_id = run.run_id
+        elif message:
+            row = self._serialize_pipeline_execution(message)
+            event_id, bot, conversation, thread = message.event_id, message.bot_id, message.session_id, None
+            run_id = message.run_id
+            if linked_run:
+                row['duration_ms'] = (
+                    int((linked_run.finished_at - linked_run.started_at).total_seconds() * 1000)
+                    if linked_run.started_at and linked_run.finished_at
+                    else None
+                )
+                row['usage'] = parsed(linked_run.usage_json)
+                row['status_reason'] = linked_run.status_reason
+        else:
+            row = self._serialize_event_execution(event)
+            event_id, bot, conversation, thread = event.event_id, event.bot_id, event.conversation_id, event.thread_id
+            run_id = event.run_id
+        await self.enrich_execution_rows(workspace, [row])
+
+        ledger = run or linked_run
+        ingress_id = parsed(ledger.metadata_json, {}).get('ingress_event_id') if ledger else None
+        event_conditions = [
+            EventLog.workspace_id == workspace,
+            sa.or_(
+                EventLog.event_id == ingress_id if ingress_id else sa.false(),
+                EventLog.event_id == ledger.event_id if ledger else sa.false(),
+                EventLog.event_id == event_id if event_id else sa.false(),
+                EventLog.run_id == run_id if run_id else sa.false(),
+            ),
+        ]
+        event_ids = sa.select(EventLog.event_id).where(*event_conditions)
+        # All joins use explicit identities. Session/time matching is reserved
+        # for the visibly separate conversation history section.
+        message_scope = [models.MonitoringMessage.workspace_uuid == workspace]
+        related_message = sa.or_(
+            models.MonitoringMessage.id == message.id if message else sa.false(),
+            models.MonitoringMessage.parent_message_id == (message.id if message else run_id)
+            if message or run_id
+            else sa.false(),
+            models.MonitoringMessage.run_id == run_id if run_id else sa.false(),
+        )
+        transcript_scope = [Transcript.workspace_id == workspace, Transcript.bot_id == bot]
+        transcripts = sa.select(Transcript).where(
+            *transcript_scope,
+            sa.or_(
+                Transcript.run_id == run_id if run_id else sa.false(),
+                sa.and_(Transcript.role == 'user', Transcript.event_id.in_(event_ids)),
+            ),
+        )
+        call_ids = sa.select(models.MonitoringMessage.id).where(*message_scope, related_message)
+        statements = {
+            'inputs': (sa.select(EventLog).where(*event_conditions).order_by(EventLog.id.asc()), event_item),
+            'outputs': (transcripts.where(Transcript.role != 'user').order_by(Transcript.seq.asc()), transcript_item),
+            'deliveries': (
+                sa.select(models.MonitoringMessage)
+                .where(*message_scope, related_message, models.MonitoringMessage.role == 'assistant')
+                .order_by(models.MonitoringMessage.timestamp.asc(), models.MonitoringMessage.id.asc()),
+                message_item,
+            ),
+            'events': (
+                sa.select(AgentRunEvent)
+                .where(AgentRunEvent.run_id == run_id if run_id else sa.false())
+                .order_by(AgentRunEvent.sequence.asc()),
+                lambda r: {
+                    'id': r.id,
+                    'sequence': r.sequence,
+                    'type': r.type,
+                    'data': parsed(r.data_json, {}),
+                    'timestamp_ms': epoch(r.created_at),
+                },
+            ),
+        }
+        for name, model in [
+            ('llm_calls', models.MonitoringLLMCall),
+            ('tool_calls', models.MonitoringToolCall),
+            ('errors', models.MonitoringError),
+        ]:
+            statements[name] = (
+                sa.select(model)
+                .where(
+                    model.workspace_uuid == workspace,
+                    sa.or_(
+                        model.message_id.in_(call_ids),
+                        model.message_id == run_id if run_id else sa.false(),
+                    ),
+                )
+                .order_by(model.timestamp.asc(), model.id.asc()),
+                lambda r, model=model: self.ap.persistence_mgr.serialize_model(model, r),
+            )
+        if message:
+            statements['conversation'] = (
+                sa.select(models.MonitoringMessage)
+                .where(
+                    *message_scope,
+                    models.MonitoringMessage.bot_id == bot,
+                    models.MonitoringMessage.session_id == conversation,
+                )
+                .order_by(models.MonitoringMessage.timestamp.desc(), models.MonitoringMessage.id.desc()),
+                message_item,
+            )
+        else:
+            statements['conversation'] = (
+                sa.select(Transcript)
+                .where(
+                    *transcript_scope,
+                    Transcript.conversation_id == conversation if conversation else sa.false(),
+                    Transcript.thread_id == thread,
+                )
+                .order_by(Transcript.seq.desc()),
+                transcript_item,
+            )
+
+        pages = {}
+        for name, (query, convert) in statements.items():
+            if section is None or section == name:
+                pages[name] = await self._execution_page(workspace, query, convert, offset, limit)
+        if run:
+            meta = parsed(run.metadata_json, {})
+            for item in pages.get('inputs', {}).get('items', []):
+                # Older event logs omitted data, but the root run may retain it.
+                # Do not apply the root snapshot to later steering events.
+                if item['id'] == run.event_id and not item.get('metadata', {}).get('input_event'):
+                    item['content'] = event_content(item['content'], meta.get('input_event'))
+        if message and (section is None or section == 'inputs'):
+            # The original rich message is authoritative even when the event
+            # log is absent or only retained a compact event summary.
+            if offset == 0:
+                pages['inputs']['items'].insert(0, message_item(message))
+        if (
+            run
+            and not pages.get('inputs', {}).get('items')
+            and offset == 0
+            and (section is None or section == 'inputs')
+        ):
+            meta = parsed(run.metadata_json, {})
+            content = event_content(meta.get('input'), meta.get('input_event'))
+            if content:
+                pages['inputs']['items'] = [
+                    {'id': run.event_id or run.run_id, 'content': content, 'timestamp_ms': epoch(run.created_at)}
+                ]
+
+        if section is None or section == 'related':
+            # Union of ledger executions and legacy pipeline roots. SQL paging
+            # avoids silently dropping siblings on highly fanned-out events.
+            related_runs = sa.select(
+                sa.literal('agent').label('source'), AgentRun.run_id.label('id'), AgentRun.created_at.label('time')
+            ).where(
+                AgentRun.workspace_id == workspace,
+                sa.or_(
+                    AgentRun.event_id.in_(event_ids),
+                    self._execution_json_text(AgentRun.metadata_json, 'ingress_event_id').in_(event_ids),
+                ),
+                AgentRun.run_id != (run_id or ''),
+                ~sa.exists().where(
+                    models.MonitoringMessage.workspace_uuid == workspace,
+                    models.MonitoringMessage.run_id == AgentRun.run_id,
+                    models.MonitoringMessage.role != 'assistant',
+                ),
+            )
+            related_messages = sa.select(
+                sa.literal('pipeline'), models.MonitoringMessage.id, models.MonitoringMessage.timestamp
+            ).where(
+                *message_scope,
+                models.MonitoringMessage.event_id.in_(event_ids),
+                models.MonitoringMessage.role != 'assistant',
+                models.MonitoringMessage.id != (message.id if message else ''),
+            )
+            union = sa.union_all(related_runs, related_messages).subquery()
+            refs = (
+                await self._execution_query(
+                    workspace,
+                    sa.select(union).order_by(union.c.time, union.c.source, union.c.id).offset(offset).limit(limit + 1),
+                )
+            ).all()
+            related = []
+            for ref in refs[:limit]:
+                if ref.source == 'agent':
+                    records = await self._execution_rows(
+                        workspace,
+                        sa.select(AgentRun).where(AgentRun.workspace_id == workspace, AgentRun.run_id == ref.id),
+                    )
+                    related.append(
+                        self._serialize_agent_execution(records[0], await self._agent_names(workspace, records))
+                    )
+                else:
+                    records = await self._execution_rows(
+                        workspace,
+                        sa.select(models.MonitoringMessage).where(
+                            *message_scope, models.MonitoringMessage.id == ref.id
+                        ),
+                    )
+                    related.append(self._serialize_pipeline_execution(records[0]))
+            pages['related'] = {
+                'items': related,
+                'has_more': len(refs) > limit,
+                'next_offset': offset + min(len(refs), limit),
+            }
+
+        result = {
+            'source': row['source'],
+            'row': row,
+            'pages': pages,
+            'legacy_context': bool(message and not message.event_id),
+        }
+        if run:
+            # Do not expose runtime authorization, leases or claim tokens.
+            result['run'] = {'run_id': run.run_id, 'status': run.status, 'status_reason': run.status_reason}
+        if message:
+            result['message'] = self.ap.persistence_mgr.serialize_model(models.MonitoringMessage, message)
+        # Backwards-compatible fields for existing API consumers.
+        for key in ('events', 'llm_calls', 'tool_calls', 'errors'):
+            if key in pages:
+                result[key] = pages[key]['items']
+        result['has_more'] = pages.get('events', {}).get('has_more', False)
+        result['next_cursor'] = pages.get('events', {}).get('next_offset')
+        return result
diff --git a/src/langbot/pkg/api/http/service/monitoring_traffic.py b/src/langbot/pkg/api/http/service/monitoring_traffic.py
index 702d34e01..c8469c488 100644
--- a/src/langbot/pkg/api/http/service/monitoring_traffic.py
+++ b/src/langbot/pkg/api/http/service/monitoring_traffic.py
@@ -24,9 +24,14 @@ async def get_traffic_series(
     pipeline_ids: list[str] | None = None,
     start_time: datetime.datetime | None = None,
     end_time: datetime.datetime | None = None,
+    execution_statuses: list[str] | None = None,
+    mode: str = 'all',
 ) -> dict:
     """Count all matching records in UTC buckets, returning at most 1000 points."""
     workspace_uuid = require_workspace_uuid(context)
+    from .monitoring import MonitoringService
+
+    service = MonitoringService(ap)
     bucket = 'hour' if start_time and end_time and end_time - start_time <= datetime.timedelta(days=7) else 'day'
     step = datetime.timedelta(hours=1) if bucket == 'hour' else datetime.timedelta(days=1)
     postgres = ap.persistence_mgr.get_db_engine().dialect.name == 'postgresql'
@@ -40,10 +45,11 @@ async def get_traffic_series(
             pattern = '%Y-%m-%dT%H:00:00' if bucket == 'hour' else '%Y-%m-%dT00:00:00'
             time_bucket = sqlalchemy.func.strftime(pattern, timestamp)
         conditions = [model.workspace_uuid == workspace_uuid]
+        conditions.append(service._monitoring_execution_condition(model, workspace_uuid, execution_statuses, mode))
         if bot_ids:
             conditions.append(model.bot_id.in_(bot_ids))
         if pipeline_ids:
-            conditions.append(model.pipeline_id.in_(pipeline_ids))
+            conditions.append(service._monitoring_processor_condition(model, workspace_uuid, pipeline_ids))
         if start_time is not None:
             conditions.append(timestamp >= start_time)
         if end_time is not None:
diff --git a/src/langbot/pkg/api/http/service/pipeline.py b/src/langbot/pkg/api/http/service/pipeline.py
index 54568433c..22a1c86c0 100644
--- a/src/langbot/pkg/api/http/service/pipeline.py
+++ b/src/langbot/pkg/api/http/service/pipeline.py
@@ -68,6 +68,11 @@ class PipelineService:
         runner_config['id'] = selected.id
         runner_config.setdefault('expire-time', 0)
         ai_config['runner_config'] = {selected.id: self._get_default_values_from_schema(selected.config_schema)}
+        ai_config['runner_config'][selected.id] = await self.ap.llm_model_service.apply_default_model(
+            context,
+            selected.config_schema,
+            ai_config['runner_config'][selected.id],
+        )
         return config
 
     async def get_pipeline_metadata(self, context: TenantContext) -> list[dict]:
@@ -168,7 +173,9 @@ class PipelineService:
         if 'config' in pipeline_data:
             RunnerConfigResolver.validate_pipeline_config(pipeline_data['config'])
         # Check limitation
-        limitation = self.ap.instance_config.data.get('system', {}).get('limitation', {})
+        from .workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, workspace_uuid)
         max_pipelines = limitation.get('max_pipelines', -1)
         if max_pipelines >= 0:
             existing_pipelines = await self.get_pipelines(context)
@@ -285,7 +292,9 @@ class PipelineService:
         """Copy a pipeline with all its configurations"""
         workspace_uuid = require_workspace_uuid(context)
         # Check limitation
-        limitation = self.ap.instance_config.data.get('system', {}).get('limitation', {})
+        from .workspace_limits import workspace_limits
+
+        limitation = await workspace_limits(self.ap, workspace_uuid)
         max_pipelines = limitation.get('max_pipelines', -1)
         if max_pipelines >= 0:
             existing_pipelines = await self.get_pipelines(context)
diff --git a/src/langbot/pkg/api/http/service/pipeline_migration.py b/src/langbot/pkg/api/http/service/pipeline_migration.py
index f51e5b609..5a3dc296a 100644
--- a/src/langbot/pkg/api/http/service/pipeline_migration.py
+++ b/src/langbot/pkg/api/http/service/pipeline_migration.py
@@ -152,6 +152,28 @@ def _json(value):
     return json.dumps(value, sort_keys=True, separators=(',', ':'), ensure_ascii=False, default=str)
 
 
+_SKIPPED_STATES = ('already_current', 'not_legacy', 'unsupported_runner')
+
+
+def _version_at_least(installed, minimum):
+    # Pins are reviewed minimums; never downgrade an installed newer plugin.
+    from packaging.version import InvalidVersion, Version
+
+    if installed == minimum:
+        return True
+    try:
+        return Version(str(installed)) >= Version(str(minimum))
+    except (InvalidVersion, TypeError):
+        return False
+
+
+def _candidate(row, plan):
+    candidate = {**row, 'config': plan.get('config')}
+    if plan.get('extensions_preferences') is not None:
+        candidate['extensions_preferences'] = plan['extensions_preferences']
+    return candidate
+
+
 def _fingerprint(row):
     # SQL timestamps are neither monotonic revisions nor portable CAS values.
     return hashlib.sha256(
@@ -312,7 +334,7 @@ class PipelineMigrationService:
                 pending = dict(snapshots[0]._mapping)
                 source = pending['source_snapshot']
                 plan = plan_legacy_pipeline(source['config'], source['extensions_preferences'])
-                candidate = {**row, 'config': plan.get('config')}
+                candidate = _candidate(row, plan)
                 if plan['state'] != 'ready' or _fingerprint(candidate) != _fingerprint(row):
                     raise MigrationError('planner_changed')
         facts = await self._plugin(ctx, plan)
@@ -439,7 +461,7 @@ class PipelineMigrationService:
                 sources = []
                 for row in await self._rows(ctx):
                     _, _, _, state, _ = await self._plan(ctx, row)
-                    if state not in ('already_current', 'not_legacy'):
+                    if state not in _SKIPPED_STATES:
                         sources.append(row)
             if not sources:
                 raise MigrationError('nothing_to_migrate')
@@ -482,8 +504,11 @@ class PipelineMigrationService:
                         hard_blockers = [b for b in blockers if b['code'] not in ('plugin_missing', 'plugin_disabled')]
                         if hard_blockers:
                             raise MigrationError(hard_blockers[0]['code'])
-                        if state in ('already_current', 'not_legacy'):
-                            result.update(state='already_current', code=None)
+                        if state in _SKIPPED_STATES:
+                            result.update(
+                                state='unsupported_runner' if state == 'unsupported_runner' else 'already_current',
+                                code=None,
+                            )
                             continue
                         target = plan.get('target_plugin')
                         if not target:
@@ -505,7 +530,7 @@ class PipelineMigrationService:
                                 facts
                                 and facts['enabled']
                                 and descriptor
-                                and descriptor.plugin_version == target['version']
+                                and _version_at_least(descriptor.plugin_version, target['version'])
                             )
                             if not ready:
                                 task_context.metadata['phase'] = 'installing'
@@ -516,15 +541,22 @@ class PipelineMigrationService:
                                 try:
                                     await self._authorize(ctx)
                                     await self.ap.plugin_connector.require_workspace_context(execution)
-                                    await self.ap.plugin_connector.install_plugin(
-                                        PluginInstallSource.MARKETPLACE,
-                                        {
-                                            'plugin_author': target['author'],
-                                            'plugin_name': target['name'],
-                                            'plugin_version': target['version'],
-                                        },
-                                        task_context=install_context,
-                                    )
+                                    spec = {'plugin_author': target['author'], 'plugin_name': target['name']}
+                                    try:
+                                        await self.ap.plugin_connector.install_plugin(
+                                            PluginInstallSource.MARKETPLACE,
+                                            {**spec, 'plugin_version': target['version']},
+                                            task_context=install_context,
+                                        )
+                                    except (MarketplacePluginVersionNotFoundError, httpx.HTTPStatusError) as exc:
+                                        if isinstance(exc, httpx.HTTPStatusError) and exc.response.status_code != 404:
+                                            raise
+                                        # The reviewed pin was withdrawn: fall back to the
+                                        # marketplace latest; _verify_runtime still checks
+                                        # the schema against the planned config.
+                                        await self.ap.plugin_connector.install_plugin(
+                                            PluginInstallSource.MARKETPLACE, spec, task_context=install_context
+                                        )
                                     install_code = None
                                 except PluginRuntimeNotConnectedError:
                                     install_code = 'plugin_runtime_unavailable'
@@ -595,7 +627,7 @@ class PipelineMigrationService:
         descriptor = next((r for r in runners if r.id == plan['target_runner_id']), None)
         if descriptor is None or 'agent' not in descriptor.usages:
             raise MigrationError('runner_unavailable')
-        if descriptor.plugin_version != plan['target_plugin']['version']:
+        if not _version_at_least(descriptor.plugin_version, plan['target_plugin']['version']):
             raise MigrationError('plugin_version_incompatible')
         config = plan['config']
         RunnerConfigResolver.validate_pipeline_config(config)
@@ -685,12 +717,23 @@ class PipelineMigrationService:
         )
 
     async def _verify_resources(self, execution, descriptor, runner_config):
-        """Fail closed on references the scoped Host cannot authorize.
+        """Fail closed on unauthorized grants; prune dangling references.
 
         This runs only on explicit execution, outside the mutation transaction.
         Resource use is still reauthorized by the Host on every eventual run.
         """
 
+        async def has_row(model, identifier, *, column=None):
+            column = model.uuid if column is None else column
+            return (
+                await self.pm.execute_async(
+                    sa.select(model.uuid).where(
+                        model.workspace_uuid == execution.workspace_uuid,
+                        column == identifier,
+                    )
+                )
+            ).first() is not None
+
         async def require_row(model, identifier, *, column=None):
             column = model.uuid if column is None else column
             found = (
@@ -715,20 +758,26 @@ class PipelineMigrationService:
             if value and value not in config_schema.NONE_SENTINELS:
                 await require_row(EmbeddingModel, value)
 
+        # Dangling knowledge base / tool / MCP references (deleted in 4.10)
+        # are pruned rather than blocking the migration.
         kb_ids = runner_config.get('knowledge-bases', [])
-        if not isinstance(kb_ids, list) or any(not isinstance(v, str) or not v for v in kb_ids):
-            raise MigrationError('runner_resource_unavailable')
+        if not isinstance(kb_ids, list):
+            kb_ids = []
+        kb_ids = [v for v in kb_ids if isinstance(v, str) and v and await has_row(KnowledgeBase, v)]
+        if 'knowledge-bases' in runner_config:
+            runner_config['knowledge-bases'] = kb_ids
         if kb_ids:
             if not config_schema.uses_host_knowledge_bases(descriptor) or 'retrieve' not in getattr(
                 permissions, 'knowledge_bases', []
             ):
                 raise MigrationError('runner_resource_unavailable')
-            for kb_id in kb_ids:
-                await require_row(KnowledgeBase, kb_id)
 
         tool_names = runner_config.get('tools', [])
-        if not isinstance(tool_names, list) or any(not isinstance(v, str) or not v for v in tool_names):
-            raise MigrationError('runner_resource_unavailable')
+        if not isinstance(tool_names, list):
+            tool_names = []
+        tool_names = [v for v in tool_names if isinstance(v, str) and v]
+        if 'tools' in runner_config:
+            runner_config['tools'] = tool_names
         tool_grant = (
             runner_config.get('enable-all-tools', False)
             or tool_names
@@ -745,20 +794,22 @@ class PipelineMigrationService:
                 descriptor,
                 runner_config,
             )
-            if not set(tool_names) <= {tool['tool_name'] for tool in tools}:
-                raise MigrationError('runner_resource_unavailable')
+            available = {tool['tool_name'] for tool in tools}
+            if 'tools' in runner_config:
+                runner_config['tools'] = [v for v in tool_names if v in available]
 
-        for attachment in runner_config.get('mcp-resources', []):
-            if not isinstance(attachment, dict):
-                raise MigrationError('runner_resource_unavailable')
-            identifier = attachment.get('server_uuid') or attachment.get('server_id')
-            name = attachment.get('server_name')
-            if identifier:
-                await require_row(MCPServer, identifier)
-            elif name:
-                await require_row(MCPServer, name, column=MCPServer.name)
-            else:
-                raise MigrationError('runner_resource_unavailable')
+        if 'mcp-resources' in runner_config:
+            kept = []
+            for attachment in runner_config.get('mcp-resources') or []:
+                if not isinstance(attachment, dict):
+                    continue
+                identifier = attachment.get('server_uuid') or attachment.get('server_id')
+                name = attachment.get('server_name')
+                if (identifier and await has_row(MCPServer, identifier)) or (
+                    not identifier and name and await has_row(MCPServer, name, column=MCPServer.name)
+                ):
+                    kept.append(attachment)
+            runner_config['mcp-resources'] = kept
 
     async def _lock(self, ctx, pipeline_uuid, session):
         # Every pending producer shares this row and revalidates its captured
@@ -792,11 +843,15 @@ class PipelineMigrationService:
             if pending:
                 return pending['uuid'], pending['target_fingerprint']
             target = {**row, 'config': candidate['config']}
+            values = {'config': candidate['config']}
+            if plan.get('extensions_preferences') is not None:
+                target['extensions_preferences'] = plan['extensions_preferences']
+                values['extensions_preferences'] = plan['extensions_preferences']
             t = LegacyPipeline.__table__
             cas = [t.c.workspace_uuid == ctx.workspace_uuid, t.c.uuid == row['uuid']]
             for key in ('config', 'stages', 'extensions_preferences'):
                 cas.append(sa.cast(t.c[key], sa.Text) == row['_raw_' + key])
-            result = await self.pm.execute_async(sa.update(t).where(*cas).values(config=candidate['config']))
+            result = await self.pm.execute_async(sa.update(t).where(*cas).values(**values))
             if result.rowcount != 1:
                 raise MigrationError('preview_stale')
             snapshot_uuid = str(uuid.uuid4())
@@ -861,6 +916,8 @@ class PipelineMigrationService:
                     RunnerConfigResolver.validate_pipeline_config(plan['config'])
                     candidate_entity = {k: copy.deepcopy(v) for k, v in row.items() if not k.startswith('_')}
                     candidate_entity['config'] = copy.deepcopy(plan['config'])
+                    if plan.get('extensions_preferences') is not None:
+                        candidate_entity['extensions_preferences'] = copy.deepcopy(plan['extensions_preferences'])
                     runtime = await self.ap.pipeline_mgr.prepare_pipeline(execution, copy.deepcopy(candidate_entity))
                     if not data_only and await self._verify_runtime(execution, plan) != runtime_schema:
                         raise MigrationError('runner_schema_changed')
diff --git a/src/langbot/pkg/api/http/service/slack_setup.py b/src/langbot/pkg/api/http/service/slack_setup.py
new file mode 100644
index 000000000..cc6938920
--- /dev/null
+++ b/src/langbot/pkg/api/http/service/slack_setup.py
@@ -0,0 +1,353 @@
+"""Short-lived Slack app provisioning and OAuth handoff sessions."""
+
+import asyncio
+import hashlib
+import hmac
+import ipaddress
+import json
+import re
+import secrets
+import time
+from urllib.parse import urlencode, urlsplit
+
+import aiohttp
+
+from langbot.pkg.utils import httpclient
+
+
+class SlackSetupError(ValueError):
+    pass
+
+
+def callback_url(webhook_url: str, bot_uuid: str) -> str:
+    parsed = urlsplit(webhook_url)
+    suffix = f'/bots/{bot_uuid}'
+    if (
+        parsed.scheme != 'https'
+        or not parsed.hostname
+        or parsed.username
+        or parsed.password
+        or parsed.query
+        or parsed.fragment
+        or not parsed.path.endswith(suffix)
+        or parsed.hostname.lower() == 'localhost'
+    ):
+        raise SlackSetupError('invalid_webhook_url')
+    try:
+        if not ipaddress.ip_address(parsed.hostname).is_global:
+            raise SlackSetupError('invalid_webhook_url')
+    except ValueError as exc:
+        if isinstance(exc, SlackSetupError):
+            raise
+    return webhook_url[: -len(suffix)] + '/api/v1/platform/adapters/slack/setup/callback'
+
+
+def manifest(name: str, webhook_url: str, redirect_url: str, *, events: bool, socket_mode=False) -> dict:
+    settings = {'socket_mode_enabled': socket_mode}
+    if events:
+        settings['event_subscriptions'] = {
+            **({} if socket_mode else {'request_url': webhook_url}),
+            'bot_events': [
+                'app_mention',
+                'message.im',
+                'message.channels',
+                'message.groups',
+                'message.mpim',
+                'reaction_added',
+                'reaction_removed',
+                'member_joined_channel',
+                'member_left_channel',
+                'channel_rename',
+                'group_rename',
+                'channel_archive',
+                'channel_unarchive',
+                'group_archive',
+                'group_unarchive',
+                'app_home_opened',
+                'app_uninstalled',
+                'tokens_revoked',
+            ],
+        }
+        settings['interactivity'] = {
+            'is_enabled': True,
+            **({} if socket_mode else {'request_url': webhook_url}),
+        }
+    return {
+        'display_information': {'name': name},
+        'features': {
+            'bot_user': {'display_name': name, 'always_online': False},
+            'app_home': {'messages_tab_enabled': True, 'messages_tab_read_only_enabled': False},
+        },
+        'oauth_config': {
+            'redirect_urls': [redirect_url],
+            'scopes': {
+                'bot': [
+                    'app_mentions:read',
+                    'reactions:read',
+                    'chat:write',
+                    'channels:history',
+                    'channels:read',
+                    'groups:history',
+                    'groups:read',
+                    'im:history',
+                    'im:read',
+                    'im:write',
+                    'mpim:history',
+                    'mpim:read',
+                    'users:read',
+                    'files:read',
+                    'files:write',
+                ]
+            },
+        },
+        'settings': settings,
+    }
+
+
+class SlackSetupService:
+    TTL = 900
+
+    def __init__(self, ap):
+        self.ap = ap
+        self.sessions = {}
+
+    @staticmethod
+    def scope(context):
+        return (context.instance_uuid, context.workspace_uuid, context.placement_generation, context.principal)
+
+    def owned(self, context, session_id):
+        session = self.sessions.get(session_id)
+        if not session or session['scope'] != self.scope(context):
+            raise SlackSetupError('session_not_found')
+        return session
+
+    def discard(self, session_id):
+        session = self.sessions.pop(session_id, None)
+        if session:
+            session['timer'].cancel()
+            task = session.get('task')
+            if task and not task.done():
+                task.cancel()
+            session.clear()
+
+    async def api(self, method, token=None, **params):
+        headers = {'Authorization': f'Bearer {token}'} if token else {}
+        try:
+            async with httpclient.get_session().post(
+                f'https://slack.com/api/{method}',
+                data=params,
+                headers=headers,
+                timeout=aiohttp.ClientTimeout(total=25),
+                allow_redirects=False,
+            ) as response:
+                if response.status != 200:
+                    raise SlackSetupError(f'slack_http_{response.status}')
+                data = await response.json()
+        except (aiohttp.ClientError, asyncio.TimeoutError) as exc:
+            raise SlackSetupError('slack_connection_failed') from exc
+        if not data.get('ok'):
+            # Never surface arbitrary upstream text or credentials in errors.
+            code = str(data.get('error', 'unknown_error'))
+            if not re.fullmatch(r'[a-z_]{1,80}', code):
+                code = 'unknown_error'
+            raise SlackSetupError(code)
+        return data
+
+    async def start(
+        self,
+        context,
+        *,
+        bot_uuid,
+        access_token,
+        refresh_token,
+        webhook_url='',
+        name='LangBot',
+        socket_mode=False,
+        redirect_url='',
+    ):
+        if not all(isinstance(value, str) for value in (bot_uuid, access_token, refresh_token, webhook_url, name)):
+            raise SlackSetupError('invalid_input')
+        access_token, refresh_token, webhook_url, name = (
+            access_token.strip(),
+            refresh_token.strip(),
+            webhook_url.strip(),
+            name.strip(),
+        )
+        if not access_token or not refresh_token or not 1 <= len(name) <= 35:
+            raise SlackSetupError('invalid_input')
+        if not isinstance(socket_mode, bool):
+            raise SlackSetupError('invalid_input')
+        if socket_mode:
+            parsed = urlsplit(redirect_url)
+            if (
+                not parsed.hostname
+                or parsed.username
+                or parsed.password
+                or parsed.query
+                or parsed.fragment
+                or not parsed.path.endswith('/api/v1/platform/adapters/slack/setup/callback')
+                or not (parsed.scheme == 'https' or (parsed.scheme == 'http' and parsed.hostname == 'localhost'))
+            ):
+                raise SlackSetupError('invalid_redirect_url')
+            redirect = redirect_url
+        else:
+            redirect = callback_url(webhook_url, bot_uuid)
+        bot = await self.ap.bot_service.get_bot(context, bot_uuid)
+        if not bot or bot['adapter'] not in ('slack', 'slack-omni'):
+            raise SlackSetupError('slack_bot_not_found')
+        if (
+            len(self.sessions) >= 100
+            or sum(s['context'].workspace_uuid == context.workspace_uuid for s in self.sessions.values()) >= 10
+        ):
+            raise SlackSetupError('too_many_sessions')
+        if any(s['bot_uuid'] == bot_uuid for s in self.sessions.values()):
+            raise SlackSetupError('setup_already_active')
+        sid = secrets.token_urlsafe(32)
+        session = {
+            'scope': self.scope(context),
+            'context': context,
+            'status': 'creating',
+            'bot_uuid': bot_uuid,
+            'redirect_url': redirect,
+            'webhook_url': webhook_url,
+            'socket_mode': socket_mode,
+            'timer': asyncio.get_running_loop().call_later(self.TTL, self.discard, sid),
+        }
+        self.sessions[sid] = session
+        coro = self._create(session, access_token, refresh_token, name, sid)
+        try:
+            wrapper = self.ap.task_mgr.create_user_task(
+                coro,
+                kind='platform-adapter-credential-exchange',
+                name=f'slack-setup-{sid}',
+                label='Slack app setup',
+                instance_uuid=context.instance_uuid,
+                workspace_uuid=context.workspace_uuid,
+                placement_generation=context.placement_generation,
+            )
+            session['task'] = wrapper.task
+        except Exception:
+            coro.close()
+            self.discard(sid)
+            raise SlackSetupError('too_many_sessions') from None
+        return {'session_id': sid}
+
+    async def _create(self, session, access, refresh, name, sid):
+        async def config_api(method, **params):
+            nonlocal access, refresh
+            try:
+                return await self.api(method, access, **params)
+            except SlackSetupError as exc:
+                if str(exc) not in ('token_expired', 'invalid_auth', 'token_revoked'):
+                    raise
+                rotated = await self.api('tooling.tokens.rotate', refresh_token=refresh)
+                access, refresh = rotated['token'], rotated['refresh_token']
+                return await self.api(method, access, **params)
+
+        try:
+            base = manifest(
+                name, session['webhook_url'], session['redirect_url'], events=False, socket_mode=session['socket_mode']
+            )
+            result = await config_api('apps.manifest.create', manifest=json.dumps(base))
+            session['app_id'] = result['app_id']
+            session['credentials'] = result['credentials']
+            # Register the signing secret before Slack validates the final event URL.
+            complete = manifest(
+                name, session['webhook_url'], session['redirect_url'], events=True, socket_mode=session['socket_mode']
+            )
+            await config_api('apps.manifest.update', app_id=result['app_id'], manifest=json.dumps(complete))
+            session['authorize_url'] = 'https://slack.com/oauth/v2/authorize?' + urlencode(
+                {
+                    'client_id': result['credentials']['client_id'],
+                    'scope': ','.join(complete['oauth_config']['scopes']['bot']),
+                    'redirect_uri': session['redirect_url'],
+                    'state': sid,
+                }
+            )
+            session['status'] = 'waiting'
+        except Exception as exc:
+            if session:
+                session['status'] = 'error'
+                session['error'] = str(exc) if isinstance(exc, SlackSetupError) else 'setup_failed'
+                session.pop('credentials', None)
+        finally:
+            access = refresh = ''
+
+    def status(self, context, sid):
+        session = self.owned(context, sid)
+        result = {key: session[key] for key in ('status', 'app_id', 'authorize_url', 'error') if key in session}
+        if session['status'] == 'success':
+            result['config'] = session['config']
+        return result
+
+    async def finish(self, sid, code, error=None):
+        session = self.sessions.get(sid)
+        if not session or session['status'] != 'waiting':
+            raise SlackSetupError('session_not_found')
+        session['status'] = 'authorizing'
+        try:
+            if error or not code:
+                raise SlackSetupError('authorization_denied')
+            context = session['context']
+            await self.ap.workspace_service.get_execution_binding(
+                context.workspace_uuid,
+                expected_generation=context.placement_generation,
+            )
+            creds = session['credentials']
+            result = await self.api(
+                'oauth.v2.access',
+                client_id=creds['client_id'],
+                client_secret=creds['client_secret'],
+                code=code,
+                redirect_uri=session['redirect_url'],
+            )
+            if result.get('app_id') != session['app_id'] or result.get('token_type') != 'bot':
+                raise SlackSetupError('unexpected_oauth_response')
+            session['config'] = {
+                'bot_token': result['access_token'],
+                'signing_secret': creds['signing_secret'],
+                'socket_mode': session['socket_mode'],
+            }
+            session['status'] = 'success'
+        except Exception as exc:
+            if session:
+                session['status'] = 'error'
+                session['error'] = str(exc) if isinstance(exc, SlackSetupError) else 'authorization_failed'
+            raise SlackSetupError('authorization_failed') from None
+        finally:
+            session.pop('credentials', None)
+            session.pop('authorize_url', None)
+
+    async def challenge(self, bot_uuid, request):
+        session = next((s for s in self.sessions.values() if s['bot_uuid'] == bot_uuid and s.get('credentials')), None)
+        if not session:
+            return None
+        body = await request.get_data()
+        timestamp = request.headers.get('X-Slack-Request-Timestamp', '')
+        if not timestamp.isdigit() or abs(time.time() - int(timestamp)) > 300:
+            return None
+        signature = (
+            'v0='
+            + hmac.new(
+                session['credentials']['signing_secret'].encode(),
+                b'v0:' + timestamp.encode() + b':' + body,
+                hashlib.sha256,
+            ).hexdigest()
+        )
+        if not hmac.compare_digest(signature, request.headers.get('X-Slack-Signature', '')):
+            return None
+        data = await request.get_json(silent=True)
+        if not isinstance(data, dict) or data.get('type') != 'url_verification':
+            return None
+        context = session['context']
+        await self.ap.workspace_service.get_execution_binding(
+            context.workspace_uuid,
+            expected_generation=context.placement_generation,
+        )
+        return {'challenge': data.get('challenge', '')}
+
+
+def get_slack_setup(ap):
+    if not hasattr(ap, 'slack_setup_service'):
+        ap.slack_setup_service = SlackSetupService(ap)
+    return ap.slack_setup_service
diff --git a/src/langbot/pkg/api/http/service/space.py b/src/langbot/pkg/api/http/service/space.py
index b07f4f1ed..c95713f56 100644
--- a/src/langbot/pkg/api/http/service/space.py
+++ b/src/langbot/pkg/api/http/service/space.py
@@ -290,7 +290,23 @@ class SpaceService:
         selection = await self.get_model_selection('chat')
         if not selection:
             raise ValueError('No recommended chat model is available')
-        recommended = selection[0]
+        return await self._resolve_recommended_model(context, selection[0])
+
+    async def get_recommended_assistant_model(self, context: typing.Any) -> dict:
+        """Resolve the dedicated, health-filtered assistant recommendation without fallback."""
+        space_url = self._get_space_config()['url']
+        session = httpclient.get_session()
+        async with session.get(f'{space_url}/api/v1/models/assistant/recommendation') as response:
+            if response.status != 200:
+                raise ValueError('No recommended assistant model is available')
+            payload = await httpclient.read_json_limited(response)
+            if payload.get('code') != 0:
+                raise ValueError('No recommended assistant model is available')
+            recommended = SpaceModel.model_validate(payload['data']['model'])
+        return await self._resolve_recommended_model(context, recommended)
+
+    async def _resolve_recommended_model(self, context: typing.Any, recommended: SpaceModel) -> dict:
+        """Map a public catalog model to this Workspace, refreshing the catalog once."""
 
         async def find_local_model():
             result = await self.ap.persistence_mgr.execute_async(
diff --git a/src/langbot/pkg/api/http/service/totp.py b/src/langbot/pkg/api/http/service/totp.py
index 5995e91ab..8fd4a1157 100644
--- a/src/langbot/pkg/api/http/service/totp.py
+++ b/src/langbot/pkg/api/http/service/totp.py
@@ -49,6 +49,17 @@ _TOTP_MAX_ATTEMPTS = 5
 _RECOVERY_ALPHABET = 'ABCDEFGHJKLMNPQRSTUVWXYZ23456789'
 
 
+def _utc_now_naive() -> datetime.datetime:
+    """UTC now without tzinfo, matching the credential ``timestamp`` columns.
+
+    The TOTP tables store ``timestamp without time zone``. asyncpg rejects an
+    offset-aware value for such a column (SQLite accepts it, which is why this
+    only surfaced on PostgreSQL), so every write converts at the boundary.
+    """
+
+    return datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None)
+
+
 class TotpError(ValueError):
     """Base class for TOTP second-factor failures."""
 
@@ -363,7 +374,7 @@ class TotpService:
                 await session.execute(
                     sqlalchemy.update(totp_entity.TotpCredential)
                     .where(totp_entity.TotpCredential.account_uuid == account_uuid)
-                    .values(disabled_at=datetime.datetime.now(datetime.timezone.utc))
+                    .values(disabled_at=_utc_now_naive())
                 )
                 await session.execute(
                     sqlalchemy.delete(totp_entity.TotpRecoveryCode).where(
@@ -409,7 +420,7 @@ class TotpService:
 
         counter = self._verify_counter(credential, code)
         codes = _generate_recovery_codes()
-        now = datetime.datetime.now(datetime.timezone.utc)
+        now = _utc_now_naive()
 
         async with self._session_factory()() as session:
             async with session.begin():
@@ -479,7 +490,7 @@ class TotpService:
         if not await self.verify_code(account_uuid, code, allow_recovery=True):
             raise TotpInvalidCodeError('Invalid TOTP code')
 
-        now = datetime.datetime.now(datetime.timezone.utc)
+        now = _utc_now_naive()
         async with self._session_factory()() as session:
             async with session.begin():
                 await session.execute(
@@ -497,11 +508,16 @@ class TotpService:
     # ------------------------------------------------------------------
     # owner/admin oversight
     # ------------------------------------------------------------------
-    async def list_account_states(self) -> list[dict[str, typing.Any]]:
-        """Second-factor state for every Account, for owner/admin oversight.
+    async def list_account_states(
+        self,
+        *,
+        account_uuids: typing.Sequence[str],
+    ) -> list[dict[str, typing.Any]]:
+        """Second-factor state for the Accounts a Workspace owner/admin may see.
 
-        Oversight is instance-wide by design: an owner/admin may manage the
-        second factor of any Account.
+        Oversight belongs to the Workspace, so the caller supplies the member
+        Account UUIDs of its own Workspace: an owner/admin must never observe or
+        manage another tenant's Accounts.
 
         Only state is returned: no secret and no recovery-code material.
         """
@@ -525,6 +541,7 @@ class TotpService:
                 credential.last_used_at,
                 sqlalchemy.func.coalesce(unused_codes.c.unused, 0),
             )
+            .where(user.User.uuid.in_(tuple(account_uuids)))
             .outerjoin(
                 credential,
                 sqlalchemy.and_(
@@ -558,7 +575,7 @@ class TotpService:
         outstanding recovery codes are destroyed.
         """
 
-        now = datetime.datetime.now(datetime.timezone.utc)
+        now = _utc_now_naive()
         async with self._session_factory()() as session:
             async with session.begin():
                 result = await session.execute(
@@ -630,7 +647,7 @@ class TotpService:
             raise
 
         if consume_counter:
-            now = datetime.datetime.now(datetime.timezone.utc)
+            now = _utc_now_naive()
             async with self._session_factory()() as session:
                 async with session.begin():
                     await session.execute(
@@ -680,7 +697,7 @@ class TotpService:
                     totp_entity.TotpRecoveryCode.id == matched_id,
                     totp_entity.TotpRecoveryCode.used_at.is_(None),
                 )
-                .values(used_at=datetime.datetime.now(datetime.timezone.utc))
+                .values(used_at=_utc_now_naive())
             )
             await session.commit()
         return bool(result.rowcount)
diff --git a/src/langbot/pkg/api/http/service/user.py b/src/langbot/pkg/api/http/service/user.py
index 5cd5766d0..60926178f 100644
--- a/src/langbot/pkg/api/http/service/user.py
+++ b/src/langbot/pkg/api/http/service/user.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import t
+
 import sqlalchemy
 import argon2
 import jwt
@@ -423,12 +425,12 @@ class UserService:
 
         user_obj = await self.get_user_by_email(user_email)
         if user_obj is None:
-            raise ValueError('用户不存在')
+            raise ValueError(t('Resource not found'))
         self._require_active_account(user_obj)
 
         # Check if this user has a local password set
         if not user_obj.password:
-            raise ValueError('请使用 LangBot 账号登录')
+            raise ValueError(t('Please sign in with your LangBot Account'))
 
         await self._verify_password(user_obj.password, password)
 
diff --git a/src/langbot/pkg/api/http/service/workspace_limits.py b/src/langbot/pkg/api/http/service/workspace_limits.py
new file mode 100644
index 000000000..dc41a2a2b
--- /dev/null
+++ b/src/langbot/pkg/api/http/service/workspace_limits.py
@@ -0,0 +1,20 @@
+"""Resolve resource limits at the Workspace service boundary."""
+
+from ....cloud.entitlements import EntitlementUnavailableError
+
+LIMIT_KEYS = {
+    'max_bots': 'bots.max',
+    'max_pipelines': 'pipelines.max',
+    'max_extensions': 'plugins.max',
+    'max_knowledge_bases': 'knowledge_bases.max',
+}
+
+
+async def workspace_limits(ap, workspace_uuid: str, *, minimum_revision: int = 0) -> dict[str, int]:
+    if getattr(getattr(ap, 'deployment', None), 'mode', None) != 'cloud':
+        return dict(ap.instance_config.data.get('system', {}).get('limitation', {}))
+    resolver = getattr(ap, 'entitlement_resolver', None)
+    if resolver is None:
+        raise EntitlementUnavailableError('Workspace entitlement resolver is unavailable')
+    snapshot = await resolver.resolve(workspace_uuid, minimum_revision=minimum_revision)
+    return {config_key: snapshot.limit(limit_key) for config_key, limit_key in LIMIT_KEYS.items()}
diff --git a/src/langbot/pkg/api/management_diagnostics.py b/src/langbot/pkg/api/management_diagnostics.py
deleted file mode 100644
index a69f37366..000000000
--- a/src/langbot/pkg/api/management_diagnostics.py
+++ /dev/null
@@ -1,225 +0,0 @@
-"""Content-free management boundaries. Diagnostics never participate in I/O.
-
-Only source-code identities and status categories enter these helpers. Never
-pass a URL, request, frame, token, tool argument, or serialized result to them.
-"""
-
-from __future__ import annotations
-
-import contextlib
-import contextvars
-import functools
-import re
-
-import quart
-from quart.wrappers.response import IterableBody, ResponseBody
-
-from ..telemetry import diagnostics as d
-from ..telemetry import diagnostic_privacy as privacy
-
-_ACTIVE_BOUNDARY = contextvars.ContextVar('management_diagnostic_boundary', default=None)
-
-
-def operation_id(source, fn, *, rule='', methods=()):
-    """Called at registration with a Core function, never a client tool name."""
-    module = fn.__module__.split('.groups.', 1)[-1]
-    if source != 'http':
-        module = ''
-    name = fn.__name__
-    if source == 'http' and name == '_':
-        # Many Core routes use the anonymous function name `_`. Disambiguate
-        # using ONLY the source-declared template and methods at registration.
-        # This is never quart.request.path, url_rule, endpoint or request.method.
-        template = re.sub(r'[^A-Za-z0-9_.:-]+', '.', rule).strip('.') or 'root'
-        name = '.'.join((*sorted(methods or ('GET',)), template))
-    return '.'.join(part for part in (source, module, name) if part)[:128]
-
-
-def outcome(value, reason_code='response_error'):
-    try:
-        d.set_outcome(value, reason_code=reason_code)
-    except Exception:
-        pass
-
-
-def workspace(context):
-    """Annotate only this boundary's owned span with a matching trusted context."""
-    try:
-        from .http.context import ExecutionContext, RequestContext
-
-        if not isinstance(context, (ExecutionContext, RequestContext)):
-            return
-        boundary = _ACTIVE_BOUNDARY.get()
-        span = d.current_span()
-        if boundary is None or span is None or span is not boundary.span:
-            return
-        if d._manager(boundary.ap) is not span.manager:
-            return
-        instance = getattr(
-            getattr(boundary.ap, 'workspace_service', None),
-            'instance_uuid',
-            getattr(span.manager, 'instance_id', None),
-        )
-        # Structural recorders need not expose instance metadata. Real managers
-        # do, and must never receive another instance's Workspace annotation.
-        if instance is not None and instance != context.instance_uuid:
-            return
-        d.annotate(workspace_uuid=context.workspace_uuid)
-    except Exception:
-        pass
-
-
-class Boundary:
-    def __init__(self, ap, operation, source):
-        self.ap = ap
-        self.span = None
-        try:
-            manager = d._manager(ap)
-            if manager is not None:
-                fields = {'source': source, 'stage': 'execute'}
-                if source == 'webui_debug':
-                    fields['attributes'] = {'synthetic': True}
-                self.span = d.Span(manager, 'api', operation, fields)
-        except Exception:
-            pass
-
-    @contextlib.contextmanager
-    def activate(self):
-        # None is an explicit inheritance barrier, not a no-op: disabled or
-        # broken B work must not borrow enabled A's span through ContextVars.
-        token = d._CURRENT.set(self.span)
-        boundary_token = _ACTIVE_BOUNDARY.set(self)
-        try:
-            yield
-        finally:
-            _ACTIVE_BOUNDARY.reset(boundary_token)
-            d._CURRENT.reset(token)
-
-    def finish(self, error=None):
-        if self.span is not None:
-            try:
-                self.span.finish(error)
-            except Exception:
-                pass
-
-
-class _ObservedBody(ResponseBody):
-    """Carry the HTTP parent through iteration, not generator suspension."""
-
-    def __init__(self, body, boundary):
-        self.body = body
-        self.boundary = boundary
-        self.iterator = None
-        self.exhausted = False
-
-    async def __aenter__(self):
-        try:
-            with self.boundary.activate():
-                entered = await self.body.__aenter__()
-                self.iterator = entered.__aiter__()
-            return self
-        except BaseException as exc:
-            self.boundary.finish(exc)
-            raise
-
-    async def __aexit__(self, exc_type, exc_value, tb):
-        try:
-            with self.boundary.activate():
-                result = await self.body.__aexit__(exc_type, exc_value, tb)
-        except BaseException as exc:
-            self.boundary.finish(exc)
-            raise
-        finally:
-            self.boundary.finish(exc_value if exc_value is not None else (None if self.exhausted else GeneratorExit()))
-        return result
-
-    async def __aiter__(self):
-        try:
-            if self.iterator is None:
-                with self.boundary.activate():
-                    self.iterator = self.body.__aiter__()
-            while True:
-                try:
-                    with self.boundary.activate():
-                        value = await anext(self.iterator)
-                except StopAsyncIteration:
-                    self.exhausted = True
-                    self.boundary.finish()
-                    return
-                yield value
-        except BaseException as exc:
-            self.boundary.finish(exc)
-            raise
-
-
-def _response_status(value):
-    """Inspect response metadata only; never read or deserialize a body."""
-    response = value[0] if isinstance(value, tuple) else value
-    status = response.status_code if isinstance(response, quart.Response) else 200
-    if isinstance(value, tuple) and len(value) > 1 and type(value[1]) is int:
-        status = value[1]
-    if status in (401, 403):
-        outcome('rejected')
-    elif status >= 400:
-        outcome('failed')
-    elif isinstance(response, dict) and 'code' in response and response['code'] != 0:
-        outcome('failed')
-    return response
-
-
-@contextlib.contextmanager
-def scope(ap, operation, *, source):
-    """A fixed operation around an existing non-generator statement block."""
-    try:
-        privacy.code_value('operation', operation)
-    except Exception:
-        pass
-    boundary = Boundary(ap, operation, source)
-    try:
-        with boundary.activate():
-            yield
-    except BaseException as exc:
-        boundary.finish(exc)
-        raise
-    finally:
-        boundary.finish()
-
-
-def observe(operation, *, source, ap=None, http=False):
-    # The operation is supplied by Core registration code, not request data.
-    try:
-        privacy.code_value('operation', operation)
-    except Exception:
-        pass
-
-    def decorator(fn):
-        @functools.wraps(fn)
-        async def wrapped(*args, **kwargs):
-            try:
-                owner = (ap() if callable(ap) else ap) if ap is not None else getattr(args[0], 'ap', None)
-            except Exception:
-                owner = None
-            boundary = Boundary(owner, operation, source)
-            streaming = False
-            try:
-                with boundary.activate():
-                    value = await fn(*args, **kwargs)
-                    if http:
-                        try:
-                            response = _response_status(value)
-                            if isinstance(response, quart.Response) and isinstance(response.response, IterableBody):
-                                response.response = _ObservedBody(response.response, boundary)
-                                streaming = True
-                        except Exception:
-                            pass
-                return value
-            except BaseException as exc:
-                boundary.finish(exc)
-                raise
-            finally:
-                if not streaming:
-                    boundary.finish()
-
-        return wrapped
-
-    return decorator
diff --git a/src/langbot/pkg/api/mcp/mount.py b/src/langbot/pkg/api/mcp/mount.py
index 6a07bd2e4..075237b57 100644
--- a/src/langbot/pkg/api/mcp/mount.py
+++ b/src/langbot/pkg/api/mcp/mount.py
@@ -24,7 +24,6 @@ import uuid
 from ..http.context import PrincipalContext, PrincipalType, RequestContext, WorkspaceContext
 from .context import bind_request_context, reset_request_context
 from .server import LangBotMCPServer
-from .. import management_diagnostics as diagnostics
 
 if typing.TYPE_CHECKING:
     from ...core import app as app_module
@@ -86,7 +85,6 @@ class MCPMount:
         authenticate_api_key = self.ap.apikey_service.authenticate_api_key
         is_mcp_path = self._is_mcp_path
 
-        @diagnostics.observe('mcp.request', source='mcp', ap=self.ap)
         async def dispatch_mcp(scope, receive, send):
             # Authenticate MCP HTTP requests with a LangBot API key.
             api_key = _extract_api_key(scope.get('headers', []))
@@ -96,7 +94,6 @@ class MCPMount:
                     identity = await authenticate_api_key(api_key)
 
             if identity is None:
-                diagnostics.outcome('rejected')
                 await send(
                     {
                         'type': 'http.response.start',
@@ -123,7 +120,6 @@ class MCPMount:
                     entitlement = await resolver.resolve(identity.workspace_uuid)
                     entitlement_revision = entitlement.entitlement_revision
             except Exception:
-                diagnostics.outcome('rejected')
                 await send(
                     {
                         'type': 'http.response.start',
@@ -151,7 +147,6 @@ class MCPMount:
                 ),
                 entitlement_revision=entitlement_revision,
             )
-            diagnostics.workspace(request_context)
             tenant_scope = getattr(self.ap.persistence_mgr, 'tenant_scope', None)
             if not callable(tenant_scope):
                 raise RuntimeError('MCP request persistence scope is unavailable')
diff --git a/src/langbot/pkg/api/mcp/server.py b/src/langbot/pkg/api/mcp/server.py
index d07edb7a9..d7ad628d6 100644
--- a/src/langbot/pkg/api/mcp/server.py
+++ b/src/langbot/pkg/api/mcp/server.py
@@ -19,12 +19,14 @@ from __future__ import annotations
 
 import json
 import typing
+import functools
 
 from mcp.server.fastmcp import FastMCP
 
 from ..http.authz import Permission, require_permission
 from .context import get_request_context
-from .. import management_diagnostics as diagnostics
+from ...i18n import language_scope
+from ...i18n.workspace import get_workspace_language, set_workspace_language
 
 if typing.TYPE_CHECKING:
     from ...core import app as app_module
@@ -53,7 +55,6 @@ def _dump(value: typing.Any) -> str:
 def _authorized(permission: Permission):
     context = get_request_context()
     require_permission(context, permission)
-    diagnostics.workspace(context)
     return context
 
 
@@ -80,6 +81,20 @@ class LangBotMCPServer:
         ap = self.ap
         mcp = self
 
+        @mcp.tool(description='Get the Workspace language used for backend messages and default bot replies.')
+        async def get_workspace_message_language() -> str:
+            context = _authorized(Permission.WORKSPACE_VIEW)
+            return _dump({'language': await get_workspace_language(ap, context.workspace_uuid, refresh=True)})
+
+        @mcp.tool(
+            description='Set the Workspace message language. Requires workspace.update; does not change personal UI language.'
+        )
+        async def set_workspace_message_language(
+            language: typing.Literal['en_US', 'zh_Hans', 'zh_Hant', 'ja_JP', 'es_ES', 'ru_RU', 'th_TH', 'vi_VN'],
+        ) -> str:
+            context = _authorized(Permission.WORKSPACE_UPDATE)
+            return _dump({'language': await set_workspace_language(ap, context.workspace_uuid, language)})
+
         # ----- System (read-only) -------------------------------------- #
         @mcp.tool(description='Get basic LangBot system/runtime information (version, edition).')
         async def get_system_info() -> str:
@@ -167,7 +182,7 @@ class LangBotMCPServer:
             return _dump(ap.box_service.get_recent_errors(context))
 
         @mcp.tool(
-            description='Read bounded Workspace runtime records. Filters follow the monitoring service: bot_ids, pipeline_ids, session_ids, start_time, end_time, knowledge_base_id, user_query, is_active as supported by the record kind.'
+            description='Read bounded Workspace runtime records. LLM calls include optional details (timing in ms, complete provider usage, structured error); older records have null details. Filters follow the monitoring service: bot_ids, pipeline_ids, session_ids, start_time, end_time, mode (all/real/debug), execution_statuses (owning execution status groups), knowledge_base_id, user_query, is_active as supported by the record kind.'
         )
         async def get_monitoring_records(
             kind: typing.Literal['messages', 'llm_calls', 'tool_calls', 'embedding_calls', 'sessions', 'errors'],
@@ -178,7 +193,7 @@ class LangBotMCPServer:
             import datetime
 
             context = _authorized(Permission.RESOURCE_VIEW)
-            allowed = {'start_time', 'end_time'}
+            allowed = {'start_time', 'end_time', 'mode', 'execution_statuses'}
             if kind != 'embedding_calls':
                 allowed |= {'bot_ids', 'pipeline_ids'}
             if kind in {'messages', 'tool_calls'}:
@@ -209,6 +224,116 @@ class LangBotMCPServer:
                 return _dump(await ap.monitoring_service.get_message_details(context, identifier))
             return _dump(await ap.monitoring_service.get_session_analysis(context, identifier))
 
+        @mcp.tool(
+            description='Read a bounded snapshot of active and recently started Workspace executions, including the latest progress event. No percentage is estimated.'
+        )
+        async def get_inflight_executions() -> str:
+            return _dump(await ap.monitoring_service.get_inflight_snapshot(_authorized(Permission.RESOURCE_VIEW)))
+
+        @mcp.tool(
+            description='Reset a bot session model context, preserving monitoring history. Refuses busy sessions.'
+        )
+        async def reset_session_context(bot_id: str, session_id: str) -> str:
+            return _dump(
+                await ap.monitoring_service.reset_session_context(
+                    _authorized(Permission.RESOURCE_MANAGE),
+                    bot_id,
+                    session_id,
+                )
+            )
+
+        @mcp.tool(
+            description='Create/configure a Slack app for an existing Workspace bot draft. Returns a setup session; poll it for the installation authorization URL. The user must approve Slack installation. Configuration tokens stay in memory only.'
+        )
+        async def start_slack_setup(
+            bot_uuid: str,
+            access_token: str,
+            refresh_token: str,
+            webhook_url: str = '',
+            name: str = 'LangBot',
+            socket_mode: bool = False,
+            redirect_url: str = '',
+        ) -> str:
+            from langbot.pkg.api.http.service.slack_setup import get_slack_setup
+
+            return _dump(
+                await get_slack_setup(ap).start(
+                    _authorized(Permission.RESOURCE_MANAGE),
+                    bot_uuid=bot_uuid,
+                    access_token=access_token,
+                    refresh_token=refresh_token,
+                    webhook_url=webhook_url,
+                    name=name,
+                    socket_mode=socket_mode,
+                    redirect_url=redirect_url,
+                )
+            )
+
+        @mcp.tool(
+            description='Read an owned Slack setup session. Success includes adapter credentials; save them with the existing bot update tool. Socket Mode additionally needs an App-Level Token with connections:write.'
+        )
+        async def get_slack_setup_status(session_id: str) -> str:
+            from langbot.pkg.api.http.service.slack_setup import get_slack_setup
+
+            return _dump(get_slack_setup(ap).status(_authorized(Permission.RESOURCE_MANAGE), session_id))
+
+        @mcp.tool(
+            description='Discard an owned Slack setup session and its temporary credentials. Does not delete the Slack app.'
+        )
+        async def cancel_slack_setup(session_id: str) -> str:
+            from langbot.pkg.api.http.service.slack_setup import get_slack_setup
+
+            service = get_slack_setup(ap)
+            service.owned(_authorized(Permission.RESOURCE_MANAGE), session_id)
+            service.discard(session_id)
+            return _dump({'cancelled': True})
+
+        @mcp.tool(
+            description='List Workspace executions, including events that did not start a processor. The legacy pipeline_ids filter accepts Agent, Pipeline and event processor IDs. Returns bounded rows and summary metrics.'
+        )
+        async def get_monitoring_executions(
+            limit: int = 50,
+            offset: int = 0,
+            bot_ids: list[str] | None = None,
+            pipeline_ids: list[str] | None = None,
+            statuses: list[str] | None = None,
+            mode: typing.Literal['all', 'real', 'debug'] = 'all',
+        ) -> str:
+            context = _authorized(Permission.RESOURCE_VIEW)
+            return _dump(
+                await ap.monitoring_service.get_executions(
+                    context,
+                    limit=limit,
+                    offset=offset,
+                    bot_ids=bot_ids,
+                    pipeline_ids=pipeline_ids,
+                    statuses=statuses,
+                    mode=mode,
+                )
+            )
+
+        @mcp.tool(
+            description='Read execution inputs, generated outputs, delivery records, calls, errors, related executions and conversation history. Each section is independently paginated; follow pages[section].next_offset when has_more is true.'
+        )
+        async def get_monitoring_execution_detail(
+            source: typing.Literal['auto', 'agent', 'pipeline', 'event'],
+            identifier: str,
+            section: str | None = None,
+            offset: int = 0,
+            limit: int = 100,
+        ) -> str:
+            context = _authorized(Permission.RESOURCE_VIEW)
+            return _dump(
+                await ap.monitoring_service.get_execution_detail(
+                    context,
+                    source,
+                    identifier,
+                    section=section,
+                    offset=offset,
+                    limit=limit,
+                )
+            )
+
         # ----- Bots ---------------------------------------------------- #
         @mcp.tool(description='List all messaging-platform bots. Secrets are redacted.')
         async def list_bots() -> str:
@@ -222,7 +347,7 @@ class LangBotMCPServer:
 
         @mcp.tool(
             description=(
-                'Create a bot. `bot_data` is a JSON object matching the LangBot '
+                'Create a bot. Cloud Workspace platform restrictions also apply to this tool. `bot_data` is a JSON object matching the LangBot '
                 'POST /api/v1/platform/bots body (e.g. name, adapter, config). '
                 'Use event_bindings for exclusive Agent/Pipeline routes; plugin_processors is an array '
                 'of {processor_uuid, enabled} subscriptions that automatically match Runner-declared events. '
@@ -380,12 +505,38 @@ class LangBotMCPServer:
             context = _authorized(Permission.RUNTIME_OPERATE)
             return _dump(await ap.agent_service.debug_agent(context, processor_uuid, payload))
 
+        @mcp.tool(description='Get the single starred LLM model UUID for the current Workspace.')
+        async def get_starred_model() -> str:
+            context = _authorized(Permission.RESOURCE_VIEW)
+            return _dump({'uuid': await ap.llm_model_service.get_starred_model(context)})
+
+        @mcp.tool(
+            description='Star one LLM model for this Workspace, replacing the previous choice. Pass null to clear.'
+        )
+        async def set_starred_model(model_uuid: str | None = None) -> str:
+            context = _authorized(Permission.PROVIDER_SECRET_MANAGE)
+            await ap.llm_model_service.set_starred_model(context, model_uuid)
+            return _dump({'uuid': model_uuid})
+
+        @mcp.tool(
+            description='Resolve the default LLM model: starred model, then the available Space wizard recommendation.'
+        )
+        async def get_default_model() -> str:
+            context = _authorized(Permission.RESOURCE_MANAGE)
+            return _dump({'uuid': await ap.llm_model_service.get_default_model(context)})
+
         @mcp.tool(description='List all configured LLM models. Secrets are redacted.')
         async def list_llm_models() -> str:
             context = _authorized(Permission.RESOURCE_VIEW)
             return _dump(await ap.llm_model_service.get_llm_models(context, include_secret=False))
 
-        @mcp.tool(description='Get a single LLM model by UUID.')
+        @mcp.tool(
+            description=(
+                'Get a single LLM model by UUID, including reasoning capabilities. '
+                'Optional budget_tokens min/max describes numeric thinking-budget support; '
+                'reasoning_config and extra_args retain the existing model API shape.'
+            )
+        )
         async def get_llm_model(model_uuid: str) -> str:
             context = _authorized(Permission.RESOURCE_VIEW)
             return _dump(await ap.llm_model_service.get_llm_model(context, model_uuid, include_secret=False))
@@ -443,7 +594,12 @@ class LangBotMCPServer:
         """Register a tool boundary before its authorization and service call."""
 
         def register(fn):
-            observed = diagnostics.observe(diagnostics.operation_id('mcp', fn), source='mcp', ap=self.ap)(fn)
+            @functools.wraps(fn)
+            async def observed(*args, **kwargs):
+                context = get_request_context()
+                with language_scope(await get_workspace_language(self.ap, context.workspace_uuid)):
+                    return await fn(*args, **kwargs)
+
             return self.mcp.tool(**options)(observed)
 
         return register
diff --git a/src/langbot/pkg/box/connector.py b/src/langbot/pkg/box/connector.py
index 0b736e308..c1771e2f6 100644
--- a/src/langbot/pkg/box/connector.py
+++ b/src/langbot/pkg/box/connector.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import contextlib
 import json
@@ -163,7 +161,6 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
     def _uses_websocket(self) -> bool:
         return self.uses_websocket()
 
-    @diagnostics.observe('lifecycle', 'box.initialize', source='runtime', stage='execute')
     async def initialize(self) -> None:
         async with self._lifecycle_lock:
             if self._closing:
diff --git a/src/langbot/pkg/box/runner.py b/src/langbot/pkg/box/runner.py
index 6be620f88..6e88cb842 100644
--- a/src/langbot/pkg/box/runner.py
+++ b/src/langbot/pkg/box/runner.py
@@ -24,6 +24,7 @@ class RunBoxBinding:
     exported: dict = field(default_factory=dict)
     submitted: set = field(default_factory=set)
     lock: asyncio.Lock = field(default_factory=asyncio.Lock)
+    read_only_mounts: list[dict] = field(default_factory=list)
 
 
 def prepare_input_files(query, agent_input):
@@ -152,7 +153,10 @@ class RunnerBoxService:
         if set(request.options) - {'image'}:
             raise BoxValidationError('Only the Box image can be specified; resource limits are Host-owned')
         spec = {'session_id': session_id, **request.options}
-        result = await self.box.create_session(context, spec)
+        from ..provider.tools.loaders.skill import build_execution_mounts
+
+        mounts = build_execution_mounts(self.box.ap, query) if query is not None else []
+        result = await self.box.create_session(context, spec, trusted_read_only_mounts=mounts)
         return self.public_session(result)
 
     async def bind(self, context, query, run_id, box_id):
@@ -172,7 +176,11 @@ class RunnerBoxService:
             return self.binding_info(current)
         fields = ('image', 'network', 'cpus', 'memory_mb', 'pids_limit', 'read_only_rootfs', 'persistent')
         spec = {key: item[key] for key in fields if key in item}
-        binding = RunBoxBinding(run_id, box_id, spec, run_id)
+        from ..provider.tools.loaders.skill import build_execution_mounts
+
+        binding = RunBoxBinding(
+            run_id, box_id, spec, run_id, read_only_mounts=build_execution_mounts(self.box.ap, query)
+        )
         object.__setattr__(query, '_box_binding', binding)
         return self.binding_info(binding)
 
diff --git a/src/langbot/pkg/box/service.py b/src/langbot/pkg/box/service.py
index 4802cc438..4527b954d 100644
--- a/src/langbot/pkg/box/service.py
+++ b/src/langbot/pkg/box/service.py
@@ -20,7 +20,6 @@ from langbot_plugin.box.security import BOX_SHARED_WORKSPACE_PROBE_PREFIX
 from .admission import SandboxAdmissionController, require_cloud_admission_policy
 from .connector import BoxRuntimeConnector, _get_box_config
 from . import secure_fs
-from ..telemetry import features as telemetry_features
 from ..utils import httpclient
 from ..api.http.context import ExecutionContext
 from ..api.http.service.tenant import TenantContext, require_workspace_uuid
@@ -591,6 +590,23 @@ class BoxService:
         skip_host_mount_validation: bool = False,
         trusted_read_only_mounts: list[dict] | None = None,
     ) -> dict:
+        result = await self._execute_spec_payload_result(
+            spec_payload,
+            query,
+            skip_host_mount_validation=skip_host_mount_validation,
+            trusted_read_only_mounts=trusted_read_only_mounts,
+        )
+        return self._serialize_result(result)
+
+    async def _execute_spec_payload_result(
+        self,
+        spec_payload: dict,
+        query: pipeline_query.Query,
+        *,
+        skip_host_mount_validation: bool = False,
+        trusted_read_only_mounts: list[dict] | None = None,
+    ) -> BoxExecutionResult:
+        """Execute within the bound Workspace and preserve the raw result for file transfers."""
         if not self._available:
             raise BoxError(
                 'Box runtime is not available. Configure an available Box backend before using Box features.'
@@ -598,6 +614,8 @@ class BoxService:
         from .runner import binding_for
 
         binding = binding_for(query)
+        if trusted_read_only_mounts is None:
+            trusted_read_only_mounts = binding.read_only_mounts
         if spec_payload.get('session_id') not in (None, '', binding.session_id):
             raise BoxValidationError('session_id must match the bound Box')
         spec_payload = {**binding.spec, **spec_payload, 'session_id': binding.session_id}
@@ -653,8 +671,7 @@ class BoxService:
             f'query_id={query.query_id} '
             f'summary={json.dumps(self._summarize_result(result), ensure_ascii=False)}'
         )
-        telemetry_features.increment(query, 'sandbox', 'execs')
-        return self._serialize_result(result)
+        return result
 
     def resolve_box_session_id(self, query: pipeline_query.Query) -> str:
         """Use the Box explicitly bound by the current Runner invocation."""
@@ -770,6 +787,10 @@ class BoxService:
         to the sandbox (and vice-versa). It is ``None`` / not a local dir for
         E2B and remote runtimes, where we must fall back to the exec channel.
         """
+        # secure_fs relies on POSIX openat/O_NOFOLLOW. On Windows use the
+        # sandbox execution channel rather than misclassifying EACCES as a link.
+        if os.open not in os.supports_dir_fd or not hasattr(os, 'O_NOFOLLOW'):
+            return None
         root = self._tenant_workspace(self._query_execution_context(query))
         if not root or not os.path.isdir(root) or os.path.islink(root):
             return None
@@ -1161,16 +1182,17 @@ class BoxService:
     async def _read_outbox_via_exec(self, query: pipeline_query.Query) -> list[dict]:
         """Fallback: read the outbox over the exec channel (E2B / remote).
 
-        Uses ``client.execute`` directly (bypassing ``_serialize_result``)
-        so stdout is NOT truncated by ``output_limit_chars`` - the raw
-        base64 payload can be far larger than the 4000-char display limit.
+        Use the same Workspace-scoped execution path as tools, retaining raw
+        stdout because base64 payloads can exceed the display truncation limit.
         """
         import json as _json
 
         target_dir = f'{self.OUTBOX_MOUNT_DIR}/{self._attachment_query_key(query)}'
-        max_file_bytes = self._EXEC_FALLBACK_MAX_BYTES
+        # Outbound bytes travel in raw stdout, not argv. Use the regular
+        # attachment limits rather than the smaller inbound command limit.
+        max_file_bytes = self._ATTACHMENT_MAX_BYTES
         max_files = self._ATTACHMENT_MAX_FILES
-        max_total_bytes = max_file_bytes * max_files
+        max_total_bytes = self._ATTACHMENT_MAX_TOTAL_BYTES
         max_scan_entries = 1000
         script = (
             'import base64, json, os\n'
@@ -1203,11 +1225,11 @@ class BoxService:
             '                    continue\n'
             '                size = entry.stat(follow_symlinks=False).st_size\n'
             '                if size > max_file_bytes or total_bytes + size > max_total_bytes:\n'
-            '                    continue\n'
+            "                    raise RuntimeError('Sandbox output attachment exceeds the export size limit')\n"
             "                with open(entry.path, 'rb') as f:\n"
             '                    data = f.read(max_file_bytes + 1)\n'
             '                if len(data) > max_file_bytes or total_bytes + len(data) > max_total_bytes:\n'
-            '                    continue\n'
+            "                    raise RuntimeError('Sandbox output attachment exceeds the export size limit')\n"
             '            except OSError:\n'
             '                continue\n'
             '            rel = os.path.relpath(entry.path, target)\n'
@@ -1222,17 +1244,26 @@ class BoxService:
             'timeout_sec': 120,
             'session_id': self.resolve_box_session_id(query),
         }
-        try:
-            spec = self.build_spec(spec_payload)
-            result = await self.client.execute(spec)
-        except Exception:
-            return []
+        result = await self._execute_spec_payload_result(spec_payload, query)
         if not result.ok:
-            return []
+            if 'Sandbox output attachment exceeds the export size limit' in (result.stderr or ''):
+                raise BoxValidationError(
+                    f'Sandbox output attachments exceed the export limit '
+                    f'({max_file_bytes} bytes per file, {max_total_bytes} bytes total); files were retained'
+                )
+            raise BoxValidationError(f'Failed to read sandbox outbox (exit code {result.exit_code})')
         try:
-            return _json.loads(str(result.stdout or '').strip().splitlines()[-1])
-        except Exception:
-            return []
+            entries = _json.loads(str(result.stdout or '').strip().splitlines()[-1])
+            if not isinstance(entries, list) or any(
+                not isinstance(entry, dict)
+                or not isinstance(entry.get('name'), str)
+                or not isinstance(entry.get('b64'), str)
+                for entry in entries
+            ):
+                raise ValueError('Invalid attachment list')
+            return entries
+        except (ValueError, IndexError) as exc:
+            raise BoxValidationError('Sandbox outbox returned invalid attachment data') from exc
 
     async def _clear_outbox(self, query: pipeline_query.Query, host_dir: str | None) -> None:
         """Empty the per-query outbox after collection.
@@ -1377,9 +1408,12 @@ class BoxService:
         spec_payload: dict,
         *,
         skip_host_mount_validation: bool = False,
+        trusted_read_only_mounts: list[dict] | None = None,
     ) -> dict:
         execution_context = await self._validated_execution_context(context)
-        spec_payload = self._managed_policy_payload(execution_context, spec_payload)
+        spec_payload = self._managed_policy_payload(
+            execution_context, spec_payload, trusted_read_only_mounts=trusted_read_only_mounts
+        )
         await self._require_validated_workspace_sandbox(execution_context)
         if spec_payload.get('host_path') in (None, ''):
             tenant_workspace = self._tenant_workspace(execution_context)
@@ -1838,7 +1872,6 @@ class BoxService:
     # ── Observability ─────────────────────────────────────────────────
 
     def _record_error(self, exc: Exception, query: pipeline_query.Query):
-        telemetry_features.increment(query, 'sandbox', 'errors')
         self._recent_errors.append(
             {
                 'timestamp': _dt.datetime.now(_UTC).isoformat(),
diff --git a/src/langbot/pkg/cloud/entitlements.py b/src/langbot/pkg/cloud/entitlements.py
index d6d34afd5..7787f7ae1 100644
--- a/src/langbot/pkg/cloud/entitlements.py
+++ b/src/langbot/pkg/cloud/entitlements.py
@@ -55,6 +55,10 @@ class EntitlementSnapshot(pydantic.BaseModel):
     # must continue to use generic ``features`` and ``limits`` exclusively.
     plan_name: str | None = pydantic.Field(default=None, min_length=1, max_length=128)
 
+    subscription_started_at: int | None = pydantic.Field(default=None, ge=0)
+    plan_revision: int | None = pydantic.Field(default=None, ge=1)
+    monthly_bonus_credits: int | None = pydantic.Field(default=None, ge=0)
+
     @pydantic.field_validator('features')
     @classmethod
     def _validate_feature_names(cls, value: dict[str, bool]) -> dict[str, bool]:
diff --git a/src/langbot/pkg/cloud/plan_policy.py b/src/langbot/pkg/cloud/plan_policy.py
new file mode 100644
index 000000000..dc64b32f5
--- /dev/null
+++ b/src/langbot/pkg/cloud/plan_policy.py
@@ -0,0 +1,42 @@
+"""Workspace plan policies evaluated from trusted entitlement projections."""
+
+from .entitlements import EntitlementFeatureUnavailableError, EntitlementUnavailableError
+from ..platform.adapter_names import canonical_adapter_name
+
+
+def platform_feature_names(adapter: str) -> set[str]:
+    canonical = canonical_adapter_name(adapter)
+    names = {adapter, canonical}
+    if adapter.lower() == 'line':
+        names.update({'line', 'LINE'})
+    if canonical.endswith('-omni'):
+        names.add(canonical[:-5])
+        names.add(canonical[:-5] + '-eba')
+    return {'platform.' + name for name in names}
+
+
+async def workspace_entitlement(ap, workspace_uuid: str):
+    if getattr(getattr(ap, 'deployment', None), 'mode', None) != 'cloud':
+        return None
+    resolver = getattr(ap, 'entitlement_resolver', None)
+    if resolver is None:
+        raise EntitlementUnavailableError('Workspace entitlement resolver is unavailable')
+    return await resolver.resolve(workspace_uuid)
+
+
+async def require_workspace_platform(ap, workspace_uuid: str, adapter: str) -> None:
+    snapshot = await workspace_entitlement(ap, workspace_uuid)
+    if snapshot is None:
+        return
+    for feature in platform_feature_names(adapter):
+        if snapshot.features.get(feature) is False:
+            raise EntitlementFeatureUnavailableError(feature, entitlement_revision=snapshot.entitlement_revision)
+
+
+async def workspace_history_retention_days(ap, workspace_uuid: str, fallback: int = 30) -> int:
+    snapshot = await workspace_entitlement(ap, workspace_uuid)
+    if snapshot is None:
+        return fallback
+    # Old signed snapshots inherit the existing 30-day policy. Never interpret
+    # missing entitlement data as an unlimited retention policy.
+    return max(1, snapshot.limits.get('chat_history.retention_days', 30))
diff --git a/src/langbot/pkg/core/app.py b/src/langbot/pkg/core/app.py
index 0e7f0e61f..97dce0da4 100644
--- a/src/langbot/pkg/core/app.py
+++ b/src/langbot/pkg/core/app.py
@@ -306,6 +306,19 @@ class Application:
         if callable(database_snapshot):
             database_stats = database_snapshot()
 
+        runtime_ops_stats = {}
+        runtime_ops_reporter = getattr(self.plugin_connector, 'runtime_ops_reporter', None)
+        if runtime_ops_reporter is not None:
+            runtime_ops_stats = runtime_ops_reporter.get_stats()
+
+        telemetry_stats = {}
+        telemetry_health = getattr(self.telemetry, 'health', None)
+        if callable(telemetry_health):
+            try:
+                telemetry_stats = telemetry_health()
+            except Exception:
+                telemetry_stats = {}
+
         return {
             'asyncio_tasks': asyncio_tasks,
             'event_loop': self.event_loop_monitor.snapshot(),
@@ -316,7 +329,8 @@ class Application:
             'query_pool': query_pool_stats,
             'models': model_stats,
             'runtimes': runtime_stats,
-            'telemetry_tasks': len(getattr(self.telemetry, 'send_tasks', ())),
+            'runtime_ops': runtime_ops_stats,
+            'telemetry': telemetry_stats,
         }
 
     async def initialize(self):
@@ -346,6 +360,13 @@ class Application:
 
     async def run(self):
         self.event_loop_monitor.start()
+        from ..utils.execution_lease import maintain_execution_leases
+
+        self.task_mgr.create_task(
+            maintain_execution_leases(self),
+            name='execution-lease-maintenance',
+            scopes=[core_entities.LifecycleControlScope.APPLICATION],
+        )
         try:
             if (
                 self.directory_projection_service is not None
@@ -408,7 +429,10 @@ class Application:
 
             monitoring_cfg = self.instance_config.data.get('monitoring', {})
             auto_cleanup_cfg = monitoring_cfg.get('auto_cleanup', {})
-            monitoring_enabled = auto_cleanup_cfg.get('enabled', True)
+            monitoring_enabled = (
+                auto_cleanup_cfg.get('enabled', True)
+                or getattr(getattr(self, 'deployment', None), 'mode', None) == 'cloud'
+            )
             retention_days = self._get_positive_int_config(
                 auto_cleanup_cfg.get('retention_days', 30),
                 default=30,
@@ -644,10 +668,6 @@ class Application:
             if self.telemetry is not None:
                 with contextlib.suppress(Exception):
                     await self.telemetry.shutdown()
-            diagnostics_manager = getattr(self, 'diagnostics', None)
-            if diagnostics_manager is not None:
-                with contextlib.suppress(Exception):
-                    await diagnostics_manager.shutdown()
             if self.vector_db_mgr is not None:
                 with contextlib.suppress(Exception):
                     await self.vector_db_mgr.shutdown()
diff --git a/src/langbot/pkg/core/boot.py b/src/langbot/pkg/core/boot.py
index 270610392..1b15e9922 100644
--- a/src/langbot/pkg/core/boot.py
+++ b/src/langbot/pkg/core/boot.py
@@ -7,7 +7,6 @@ import os
 
 from . import app
 from . import stage
-from ..telemetry import diagnostics
 from ..utils import constants, importutil
 
 # Import startup stage implementation to register
@@ -40,33 +39,7 @@ async def make_app(loop: asyncio.AbstractEventLoop) -> app.Application:
             stage_cls = stage.preregistered_stages[stage_name]
             stage_inst = stage_cls()
 
-            if stage_name == 'GenKeysStage':
-                # Optional diagnostics must not make startup depend on package
-                # metadata, session markers, or its background transport.
-                ap.diagnostics = None
-                space_config = ap.instance_config.data.get('space', {})
-                if not space_config.get('disable_telemetry', False) and not space_config.get(
-                    'disable_beta_diagnostics', False
-                ):
-                    manager = None
-                    try:
-                        manager = diagnostics.DiagnosticsManager(ap, marker_path='data/labels/beta_diagnostics_session')
-                        await manager.start_session()
-                        manager.start()
-                        ap.diagnostics = manager
-                    except BaseException as exc:
-                        if manager is not None:
-                            # Cleanup faults cannot replace the startup fault.
-                            try:
-                                await manager.shutdown(drain_timeout=0)
-                            except asyncio.CancelledError:
-                                if not isinstance(exc, asyncio.CancelledError):
-                                    raise
-                            except Exception:
-                                pass
-                        if not isinstance(exc, Exception):
-                            raise
-            await diagnostics.observe('lifecycle', 'startup.' + stage_name, source='startup', ap=ap)(stage_inst.run)(ap)
+            await stage_inst.run(ap)
 
         await ap.initialize()
     except BaseException:
diff --git a/src/langbot/pkg/core/stages/build_app.py b/src/langbot/pkg/core/stages/build_app.py
index f515500c9..dd2e060d5 100644
--- a/src/langbot/pkg/core/stages/build_app.py
+++ b/src/langbot/pkg/core/stages/build_app.py
@@ -164,12 +164,6 @@ class BuildAppStage(stage.BootingStage):
         ap.persistence_mgr = persistence_mgr_inst
         await persistence_mgr_inst.initialize()
 
-        # Persistence is ready now. Open the cheap global traceability gate if
-        # any Workspace already has tracing enabled, so a restart resumes
-        # recording without ever checking the database on the request path.
-        if ap.workspace_settings_service is not None:
-            await ap.workspace_settings_service.prime_global_flag()
-
         if deployment.multi_workspace_enabled:
             directory_projection_service = DirectoryProjectionService(
                 ap,
@@ -205,6 +199,13 @@ class BuildAppStage(stage.BootingStage):
             await cloud_model_catalog_service.initialize()
             ap.cloud_model_catalog_service = cloud_model_catalog_service
 
+        # Persistence and the Workspace directory are both ready now. Open the
+        # cheap global traceability gate if any Workspace already has tracing
+        # enabled, so a restart resumes recording without ever checking the
+        # database on the request path.
+        if ap.workspace_settings_service is not None:
+            await ap.workspace_settings_service.prime_global_flag()
+
         ap.workspace_collaboration_service = workspace_collaboration_module.WorkspaceCollaborationService(
             ap,
             workspace_service_inst,
diff --git a/src/langbot/pkg/core/stages/load_config.py b/src/langbot/pkg/core/stages/load_config.py
index bf76c3961..b41e2accd 100644
--- a/src/langbot/pkg/core/stages/load_config.py
+++ b/src/langbot/pkg/core/stages/load_config.py
@@ -59,6 +59,9 @@ _RUNTIME_POLICY_DEFAULTS = {
             'restart_failure_window_seconds': 30.0,
             'restart_circuit_open_seconds': 60.0,
             'require_hard_limits': False,
+            # Cloud-only optional overrides; null inherits the unified policy.
+            'dedicated': dict.fromkeys(('max_cpus', 'max_memory_mb', 'max_pids', 'max_open_files', 'max_file_size_mb')),
+            'shared': dict.fromkeys(('max_cpus', 'max_memory_mb', 'max_pids', 'max_open_files', 'max_file_size_mb')),
         },
     },
     'mcp': {'stdio': {'enabled': True}},
diff --git a/src/langbot/pkg/core/taskmgr.py b/src/langbot/pkg/core/taskmgr.py
index 6c5df7b81..b3ae3d1c7 100644
--- a/src/langbot/pkg/core/taskmgr.py
+++ b/src/langbot/pkg/core/taskmgr.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ..i18n import t
+
 import asyncio
 import json
 import typing
@@ -202,7 +204,7 @@ class TaskWrapper:
         """Return the stable task projection exposed to API-key callers."""
         if self.task.cancelled():
             status = 'cancelled'
-            error = {'type': 'task_cancelled', 'message': 'Task was cancelled'}
+            error = {'type': 'task_cancelled', 'message': t('Task was cancelled')}
             result = None
         elif not self.task.done():
             status = 'running'
@@ -212,7 +214,7 @@ class TaskWrapper:
             exception = self.assume_exception()
             if exception is not None:
                 status = 'failed'
-                error = {'type': 'task_failed', 'message': 'Task execution failed'}
+                error = {'type': 'task_failed', 'message': t('Task execution failed')}
                 result = None
             else:
                 status = 'succeeded'
diff --git a/src/langbot/pkg/entity/persistence/agent_run.py b/src/langbot/pkg/entity/persistence/agent_run.py
index 7d309110d..7d5dc0a8c 100644
--- a/src/langbot/pkg/entity/persistence/agent_run.py
+++ b/src/langbot/pkg/entity/persistence/agent_run.py
@@ -44,6 +44,10 @@ class AgentRun(Base):
     bot_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=True, index=True)
     """Bot UUID this run belongs to."""
 
+    execution_owner_id = sqlalchemy.Column(sqlalchemy.String(36), nullable=True, index=True)
+    execution_lease_expires_at = sqlalchemy.Column(sqlalchemy.DateTime, nullable=True, index=True)
+    """Host process lease, independent of execution duration and remote queue claims."""
+
     status = sqlalchemy.Column(sqlalchemy.String(50), nullable=False, index=True)
     """Run lifecycle status."""
 
diff --git a/src/langbot/pkg/entity/persistence/metadata.py b/src/langbot/pkg/entity/persistence/metadata.py
index ac22cd2e9..b60fd9dd4 100644
--- a/src/langbot/pkg/entity/persistence/metadata.py
+++ b/src/langbot/pkg/entity/persistence/metadata.py
@@ -9,7 +9,7 @@ class Metadata(Base):
     __tablename__ = 'metadata'
 
     key = sqlalchemy.Column(sqlalchemy.String(255), primary_key=True)
-    value = sqlalchemy.Column(sqlalchemy.String(255))
+    value = sqlalchemy.Column(sqlalchemy.Text())
 
 
 class WorkspaceMetadata(Base):
@@ -23,4 +23,4 @@ class WorkspaceMetadata(Base):
         primary_key=True,
     )
     key = sqlalchemy.Column(sqlalchemy.String(255), primary_key=True)
-    value = sqlalchemy.Column(sqlalchemy.String(255))
+    value = sqlalchemy.Column(sqlalchemy.Text())
diff --git a/src/langbot/pkg/entity/persistence/monitoring.py b/src/langbot/pkg/entity/persistence/monitoring.py
index 1d2e8ed83..ecf032986 100644
--- a/src/langbot/pkg/entity/persistence/monitoring.py
+++ b/src/langbot/pkg/entity/persistence/monitoring.py
@@ -21,6 +21,10 @@ class MonitoringMessage(Base):
     pipeline_name = sqlalchemy.Column(sqlalchemy.String(255), nullable=False)
     message_content = sqlalchemy.Column(sqlalchemy.Text, nullable=False)
     session_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=False, index=True)
+    execution_owner_id = sqlalchemy.Column(sqlalchemy.String(36), nullable=True, index=True)
+    execution_lease_expires_at = sqlalchemy.Column(sqlalchemy.DateTime, nullable=True, index=True)
+    """Host process lease, independent of execution duration and remote queue claims."""
+
     status = sqlalchemy.Column(sqlalchemy.String(50), nullable=False)  # success, error, pending
     level = sqlalchemy.Column(sqlalchemy.String(50), nullable=False)  # info, warning, error, debug
     platform = sqlalchemy.Column(sqlalchemy.String(255), nullable=True)
@@ -29,6 +33,9 @@ class MonitoringMessage(Base):
     runner_name = sqlalchemy.Column(sqlalchemy.String(255), nullable=True)  # Runner name for this query
     variables = sqlalchemy.Column(sqlalchemy.Text, nullable=True)  # Query variables as JSON string
     role = sqlalchemy.Column(sqlalchemy.String(50), nullable=True, default='user')  # user, assistant
+    event_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=True, index=True)
+    run_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=True, index=True)
+    parent_message_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=True, index=True)
 
     __table_args__ = (
         sqlalchemy.Index('ix_monitoring_messages_workspace_timestamp', 'workspace_uuid', 'timestamp'),
@@ -63,6 +70,8 @@ class MonitoringLLMCall(Base):
     error_message = sqlalchemy.Column(sqlalchemy.Text, nullable=True)
     message_id = sqlalchemy.Column(sqlalchemy.String(255), nullable=True, index=True)  # Associated message ID
 
+    details = sqlalchemy.Column(sqlalchemy.JSON, nullable=True)
+
     __table_args__ = (
         sqlalchemy.Index('ix_monitoring_llm_calls_workspace_timestamp', 'workspace_uuid', 'timestamp'),
         sqlalchemy.Index('ix_monitoring_llm_calls_workspace_session', 'workspace_uuid', 'session_id'),
diff --git a/src/langbot/pkg/i18n/__init__.py b/src/langbot/pkg/i18n/__init__.py
new file mode 100644
index 000000000..113b4f909
--- /dev/null
+++ b/src/langbot/pkg/i18n/__init__.py
@@ -0,0 +1,197 @@
+"""Small, dependency-free catalog for LangBot-owned runtime messages.
+
+English message templates are the stable catalog IDs. Diagnostic text from
+providers, plugins and validators is deliberately not translated.
+"""
+
+from __future__ import annotations
+
+import contextlib
+import contextvars
+import json
+from pathlib import Path
+
+DEFAULT_LANGUAGE = 'en_US'
+SUPPORTED_LANGUAGES = ('en_US', 'zh_Hans', 'zh_Hant', 'ja_JP', 'es_ES', 'ru_RU', 'th_TH', 'vi_VN')
+language = contextvars.ContextVar('langbot_message_language', default=DEFAULT_LANGUAGE)
+_catalogs = {
+    locale: json.loads((Path(__file__).parent / 'locales' / f'{locale}.json').read_text(encoding='utf-8'))
+    for locale in SUPPORTED_LANGUAGES
+}
+
+# Normalize existing API wording at the response boundary. Only exact owned
+# messages are matched; validators and third-party diagnostics pass through.
+_ALIASES = {
+    **{
+        name: 'Resource not found'
+        for name in (
+            'bot not found',
+            'model not found',
+            'provider not found',
+            'plugin not found',
+            'Server not found',
+            'Skill not found',
+            'Pipeline not found',
+            'pipeline not found',
+            'knowledge base not found',
+            'Knowledge base not found',
+            'User not found',
+            'Account not found',
+            'agent not found',
+            'API key not found',
+            'Message not found',
+            'Task not found',
+            'Passkey not found',
+            'Webhook not found',
+            'Bot not found or not available',
+            'Bot is unavailable',
+            'WebSocket adapter not found',
+            'adapter not found',
+            'requester not found',
+            'icon not found',
+            'Session not found',
+        )
+    },
+    'Unauthorized': 'Invalid authentication credentials',
+    'No valid user token provided': 'Invalid authentication credentials',
+    'No valid API key provided': 'Invalid authentication credentials',
+    'Invalid username or password': 'Invalid authentication credentials',
+    'Invalid TOTP code': 'Invalid TOTP or recovery code',
+    'Too many attempts, try again later': 'Too many requests. Please try again later.',
+    'The configured Agent runner is unavailable': 'Agent runner temporarily unavailable. Please try again.',
+    'The configured Agent runner is not authorized': 'Agent runner not authorized for this pipeline.',
+    'The Agent runner could not complete this test': 'Agent runner execution failed.',
+    'Forbidden': 'Permission denied',
+    'Member removal permission is required': 'Permission denied',
+    'The Agent runner returned an invalid response': 'The Agent runner returned an invalid response.',
+    'Invalid or expired state': 'Authorization failed. Please try signing in again.',
+    'Space OAuth Workspace binding mismatch': 'Authorization failed. Please try signing in again.',
+    'Space OAuth response is missing the Cloud Workspace binding': 'Authorization failed. Please try signing in again.',
+    'Space launch requires Cloud mode': 'This feature is unavailable in this Workspace.',
+    'An Account is required': 'Please sign in with your LangBot Account',
+    'ChatGPT authorization requires an authorized workspace user': 'Permission denied',
+    'ChatGPT authorization is busy. Please retry shortly.': 'Server is busy. Please try again later.',
+    'ChatGPT authorization network error. Please retry.': 'Connection failed',
+    'ChatGPT credential storage failed. Please retry.': 'Failed to save ChatGPT credentials. Please retry.',
+    'ChatGPT Codex response was not completed': 'ChatGPT Codex response failed or was incomplete. Please retry.',
+    'ChatGPT Codex stream ended before completion. Please retry.': 'ChatGPT Codex response failed or was incomplete. Please retry.',
+    'Not permitted': 'Permission denied',
+    'Unauthorized or session expired': 'Invalid authentication credentials',
+    'TOTP is not enabled for that Account': 'TOTP is not enabled for this Account',
+    'No pending TOTP enrolment for this Account': 'TOTP is not enabled for this Account',
+    'Too many active credential exchanges': 'Server is busy. Please try again later.',
+    'Timeout waiting for QR code': 'Request timed out',
+    'Model availability is unavailable': 'This feature is unavailable in this Workspace.',
+    'Survey not available': 'This feature is unavailable in this Workspace.',
+    **{
+        message: 'Submission failed. Please try again.'
+        for message in (
+            'Failed to submit response',
+            'Failed to submit feedback',
+            'Failed to dismiss',
+        )
+    },
+    **{
+        message: 'File size exceeds the limit.'
+        for message in (
+            'Image size exceeds 10MB limit.',
+            'file size exceeds 10MB limit',
+            'attachment too large',
+        )
+    },
+    **{
+        message: 'Authorization failed. Please try signing in again.'
+        for message in (
+            'Space launch failed',
+            'Space OAuth failed',
+            'LangBot Account binding failed',
+            'Failed to get access token from Space',
+            'ChatGPT returned an incomplete authorization. Please sign in again.',
+            'ChatGPT authorization was cancelled or replaced. Please retry.',
+            'ChatGPT returned an invalid authorization response',
+            'ChatGPT returned an invalid device code',
+            'ChatGPT returned an invalid device code expiry',
+            'ChatGPT device authorization failed. Please start again.',
+            'ChatGPT returned an incomplete device authorization',
+            'ChatGPT token exchange failed. Please start sign-in again.',
+        )
+    },
+}
+
+
+def normalize_language(value: str | None) -> str:
+    value = (value or '').replace('-', '_')
+    aliases = {
+        'zh_CN': 'zh_Hans',
+        'zh_TW': 'zh_Hant',
+        'en': 'en_US',
+        'ja': 'ja_JP',
+        'es': 'es_ES',
+        'ru': 'ru_RU',
+        'th': 'th_TH',
+        'vi': 'vi_VN',
+    }
+    value = aliases.get(value, value)
+    return value if value in SUPPORTED_LANGUAGES else DEFAULT_LANGUAGE
+
+
+@contextlib.contextmanager
+def language_scope(locale: str):
+    token = language.set(normalize_language(locale))
+    try:
+        yield
+    finally:
+        language.reset(token)
+
+
+def t(message: str, *, locale: str | None = None, **params) -> str:
+    selected = normalize_language(locale or language.get())
+    # Preserve established English diagnostic wording and resource names.
+    # Aliases only share translations; they do not rewrite the source language.
+    key = _ALIASES.get(message, message)
+    text = message if selected == DEFAULT_LANGUAGE else _catalogs[selected].get(key, message)
+    return text.format(**params) if params else text
+
+
+def error_text(code: int | str, message: str) -> str:
+    """Stable business codes take precedence over variable diagnostic wording."""
+    if language.get() == DEFAULT_LANGUAGE:
+        return message
+    messages = {
+        'permission_denied': 'Permission denied',
+        'resource_not_found': 'Resource not found',
+        'membership_not_found': 'Resource not found',
+        'invalid_authentication': 'Invalid authentication credentials',
+        'task_capacity_exceeded': 'Server is busy. Please try again later.',
+        'blocking_work_capacity_exceeded': 'Server is busy. Please try again later.',
+        'edition_limit': 'This feature is unavailable in this Workspace.',
+        'entitlement_unavailable': 'This feature is unavailable in this Workspace.',
+        'last_owner_required': 'The Workspace must retain an owner.',
+        'already_a_member': 'This user is already a member.',
+        'invitation_email_mismatch': 'The invitation email does not match your Account',
+        'totp_invalid_code': 'Invalid TOTP or recovery code',
+        'totp_challenge_invalid': 'Invalid TOTP or recovery code',
+        'totp_already_enrolled': 'TOTP is already enabled for this Account',
+        'totp_not_enrolled': 'TOTP is not enabled for this Account',
+    }
+    if code in ('invitation_invalid', 'invitation_expired', 'invitation_revoked', 'invitation_used'):
+        return t('Invalid or expired invitation')
+    return t(messages.get(code, message))
+
+
+def pipeline_notice(query, name: str, default: str, **params) -> str:
+    """Empty overrides inherit the Workspace language; user text stays literal.
+
+    Only documented placeholders are substituted, so arbitrary braces in custom
+    messages cannot cause a formatting exception.
+    """
+    config = query.pipeline_config.get('output', {}).get('misc', {})
+    custom = config.get(name)
+    # This was the shipped default before language inheritance existed.
+    if name == 'failure-hint' and custom == 'Request failed.':
+        custom = None
+    if not isinstance(custom, str) or not custom.strip():
+        return t(default, **params)
+    for key, value in params.items():
+        custom = custom.replace('{' + key + '}', str(value))
+    return custom
diff --git a/src/langbot/pkg/i18n/adapters.py b/src/langbot/pkg/i18n/adapters.py
new file mode 100644
index 000000000..899b57a19
--- /dev/null
+++ b/src/langbot/pkg/i18n/adapters.py
@@ -0,0 +1,31 @@
+"""Language boundaries for adapter-owned replies outside the Pipeline."""
+
+import functools
+
+from . import DEFAULT_LANGUAGE, language_scope, t
+from .workspace import _cache, get_workspace_language
+
+
+def _owner(adapter):
+    logger = getattr(adapter, 'logger', None)
+    context = getattr(logger, 'execution_context', None)
+    workspace_uuid = getattr(context, 'workspace_uuid', None)
+    return getattr(logger, 'ap', None), workspace_uuid if isinstance(workspace_uuid, str) else None
+
+
+def adapter_t(adapter, message: str, **params) -> str:
+    """Synchronous SDK callbacks use the bot's own cached language, never a peer's."""
+    ap, workspace_uuid = _owner(adapter)
+    cached = _cache(ap).get(workspace_uuid) if ap is not None and workspace_uuid else None
+    return t(message, locale=cached[0] if cached else DEFAULT_LANGUAGE, **params)
+
+
+def localized_adapter(method):
+    @functools.wraps(method)
+    async def wrapped(self, *args, **kwargs):
+        ap, workspace_uuid = _owner(self)
+        locale = await get_workspace_language(ap, workspace_uuid) if ap is not None else DEFAULT_LANGUAGE
+        with language_scope(locale):
+            return await method(self, *args, **kwargs)
+
+    return wrapped
diff --git a/src/langbot/pkg/i18n/locales/en_US.json b/src/langbot/pkg/i18n/locales/en_US.json
new file mode 100644
index 000000000..dd07ff255
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/en_US.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "Request failed.",
+  "Too many requests. Please try again later.": "Too many requests. Please try again later.",
+  "Your message contains inappropriate content. Please revise it.": "Your message contains inappropriate content. Please revise it.",
+  "Content checking failed. Please contact the administrator.": "Content checking failed. Please contact the administrator.",
+  "Calling {tools}...": "Calling {tools}...",
+  "Agent runner not found: {runner}": "Agent runner not found: {runner}",
+  "Agent runner not authorized for this pipeline.": "Agent runner not authorized for this pipeline.",
+  "Agent runner temporarily unavailable. Please try again.": "Agent runner temporarily unavailable. Please try again.",
+  "Agent runner execution failed.": "Agent runner execution failed.",
+  "The Agent runner returned an invalid response.": "The Agent runner returned an invalid response.",
+  "Model request failed": "Model request failed",
+  "Context length exceeded": "Context length exceeded",
+  "Invalid request parameters": "Invalid request parameters",
+  "Invalid API key": "Invalid API key",
+  "Invalid model or endpoint": "Invalid model or endpoint",
+  "Rate limit exceeded or insufficient balance": "Rate limit exceeded or insufficient balance",
+  "Request timed out": "Request timed out",
+  "Connection failed": "Connection failed",
+  "API error": "API error",
+  "Unknown error": "Unknown error",
+  "Internal server error": "Internal server error",
+  "Resource not found": "Resource not found",
+  "Permission denied": "Permission denied",
+  "Invalid authentication credentials": "Invalid authentication credentials",
+  "MCP runtime failed": "MCP runtime failed",
+  "Task execution failed": "Task execution failed",
+  "Task was cancelled": "Task was cancelled",
+  "Maximum number of {resource} ({limit}) reached": "Maximum number of {resource} ({limit}) reached",
+  "extensions": "extensions",
+  "knowledge bases": "knowledge bases",
+  "MCP server already exists: {name}": "MCP server already exists: {name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models is managed by Cloud and cannot be modified",
+  "Please sign in with your LangBot Account": "Please sign in with your LangBot Account",
+  "Current password is incorrect": "Current password is incorrect",
+  "Invalid TOTP or recovery code": "Invalid TOTP or recovery code",
+  "No local password set, please set a password first": "No local password set, please set a password first",
+  "This LangBot Account is already bound to another user": "This LangBot Account is already bound to another user",
+  "Invalid recovery key": "Invalid recovery key",
+  "Done": "Done",
+  "Thank you for your feedback": "Thank you for your feedback",
+  "Failed to process feedback": "Failed to process feedback",
+  "Helpful": "Helpful",
+  "Not helpful": "Not helpful",
+  "Select an option": "Select an option",
+  "Enter a value": "Enter a value",
+  "Submit": "Submit",
+  "Submitted": "Submitted",
+  "Thinking": "Thinking",
+  "AI-generated": "AI-generated",
+  "AI-generated content. Refer to the cited sources for details.": "AI-generated content. Refer to the cited sources for details.",
+  "No bridges configured.": "No bridges configured.",
+  "Signing in to all bridges again...": "Signing in to all bridges again...",
+  "Bridge status:": "Bridge status:",
+  "Signed in ✓": "Signed in ✓",
+  "Not signed in ✗": "Not signed in ✗",
+  "None": "None",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "Bridge {bridge}: skipped (login command or DM room missing).",
+  "Bridge {bridge}: sending {command}...": "Bridge {bridge}: sending {command}...",
+  "Bridge {bridge}: sign-out failed.": "Bridge {bridge}: sign-out failed.",
+  "Bridge {bridge}: sign-in started.": "Bridge {bridge}: sign-in started.",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Slack authorization completed. Return to LangBot to finish setup.",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Slack authorization failed or expired. Return to LangBot for details.",
+  "TOTP is already enabled for this Account": "TOTP is already enabled for this Account",
+  "TOTP is not enabled for this Account": "TOTP is not enabled for this Account",
+  "Invalid or expired passkey challenge": "Invalid or expired passkey challenge",
+  "Passkey credential not recognized": "Passkey credential not recognized",
+  "Account is disabled": "Account is disabled",
+  "Modifying login info is disabled": "Modifying login info is disabled",
+  "Invalid GitHub repository URL": "Invalid GitHub repository URL",
+  "Server is busy. Please try again later.": "Server is busy. Please try again later.",
+  "This feature is unavailable in this Workspace.": "This feature is unavailable in this Workspace.",
+  "The Workspace must retain an owner.": "The Workspace must retain an owner.",
+  "Invalid or expired invitation": "Invalid or expired invitation",
+  "This user is already a member.": "This user is already a member.",
+  "Voice connection failed": "Voice connection failed",
+  "User is not in the voice channel": "User is not in the voice channel",
+  "Voice service is unavailable": "Voice service is unavailable",
+  "Submission failed. Please try again.": "Submission failed. Please try again.",
+  "File size exceeds the limit.": "File size exceeds the limit.",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "File size exceeds 10MB limit. Please split large files into smaller parts.",
+  "Authorization failed. Please try signing in again.": "Authorization failed. Please try signing in again.",
+  "Bind the LangBot Account with the same email as this local Account": "Bind the LangBot Account with the same email as this local Account",
+  "Only local accounts can bind to Space": "Only local accounts can bind to Space",
+  "System already initialized": "System already initialized",
+  "System not initialized": "System not initialized",
+  "Password login is disabled on LangBot Cloud": "Password login is disabled on LangBot Cloud",
+  "This edition supports one Workspace per instance": "This edition supports one Workspace per instance",
+  "Cloud Workspaces are created by the SaaS control plane": "Cloud Workspaces are created by the SaaS control plane",
+  "Login with your LangBot Account to accept this invitation": "Login with your LangBot Account to accept this invitation",
+  "RAG migration is not needed": "RAG migration is not needed",
+  "New Agent": "New Agent",
+  "New Pipeline": "New Pipeline",
+  "Provider {provider} does not support {model_type} models": "Provider {provider} does not support {model_type} models",
+  "Invalid JSON format": "Invalid JSON format",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "ChatGPT sign-in required. Open this provider and sign in again.",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT denied this request. Check subscription and workspace permissions.",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT rejected the model or request. Check the selected model and request settings.",
+  "ChatGPT Codex upstream request failed. Please retry later.": "ChatGPT Codex upstream request failed. Please retry later.",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "ChatGPT subscription usage limit reached. Please retry later or check your plan.",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "ChatGPT Codex response failed or was incomplete. Please retry.",
+  "ChatGPT token refresh temporarily failed. Please retry.": "ChatGPT token refresh temporarily failed. Please retry.",
+  "A second factor is required": "A second factor is required",
+  "Sign in or provide registration details to accept this invitation": "Sign in or provide registration details to accept this invitation",
+  "Sign out before creating the invited local Account": "Sign out before creating the invited local Account",
+  "Turnstile verification failed": "Turnstile verification failed",
+  "The invitation email does not match your Account": "The invitation email does not match your Account",
+  "Failed to save ChatGPT credentials. Please retry.": "Failed to save ChatGPT credentials. Please retry.",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.",
+  "Option {number}": "Option {number}",
+  "Confirmation needed": "Confirmation needed",
+  "Session reset successfully": "Session reset successfully"
+}
diff --git a/src/langbot/pkg/i18n/locales/es_ES.json b/src/langbot/pkg/i18n/locales/es_ES.json
new file mode 100644
index 000000000..116aa09a8
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/es_ES.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "La solicitud ha fallado.",
+  "Too many requests. Please try again later.": "Demasiadas solicitudes. Inténtalo más tarde.",
+  "Your message contains inappropriate content. Please revise it.": "Tu mensaje contiene contenido inapropiado. Modifícalo.",
+  "Content checking failed. Please contact the administrator.": "No se pudo verificar el contenido. Contacta con el administrador.",
+  "Calling {tools}...": "Llamando a {tools}…",
+  "Agent runner not found: {runner}": "No se encontró el ejecutor del agente: {runner}",
+  "Agent runner not authorized for this pipeline.": "Este flujo no tiene autorización para usar el ejecutor del agente.",
+  "Agent runner temporarily unavailable. Please try again.": "El ejecutor del agente no está disponible temporalmente. Reinténtalo.",
+  "Agent runner execution failed.": "La ejecución del agente ha fallado.",
+  "The Agent runner returned an invalid response.": "El ejecutor del agente devolvió una respuesta no válida.",
+  "Model request failed": "Error en la solicitud al modelo",
+  "Context length exceeded": "Se superó el límite de contexto",
+  "Invalid request parameters": "Parámetros de solicitud no válidos",
+  "Invalid API key": "Clave API no válida",
+  "Invalid model or endpoint": "Modelo o endpoint no válido",
+  "Rate limit exceeded or insufficient balance": "Límite de solicitudes excedido o saldo insuficiente",
+  "Request timed out": "La solicitud agotó el tiempo de espera",
+  "Connection failed": "Error de conexión",
+  "API error": "Error de API",
+  "Unknown error": "Error desconocido",
+  "Internal server error": "Error interno del servidor",
+  "Resource not found": "Recurso no encontrado",
+  "Permission denied": "No tienes permiso para realizar esta acción",
+  "Invalid authentication credentials": "Error de autenticación. Vuelve a iniciar sesión",
+  "MCP runtime failed": "Error de ejecución de MCP",
+  "Task execution failed": "La tarea ha fallado",
+  "Task was cancelled": "La tarea se ha cancelado",
+  "Maximum number of {resource} ({limit}) reached": "Se alcanzó el máximo de {resource} ({limit})",
+  "extensions": "extensiones",
+  "knowledge bases": "bases de conocimiento",
+  "MCP server already exists: {name}": "El servidor MCP ya existe: {name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models está gestionado por Cloud y no se puede modificar",
+  "Please sign in with your LangBot Account": "Inicia sesión con tu cuenta de LangBot",
+  "Current password is incorrect": "La contraseña actual es incorrecta",
+  "Invalid TOTP or recovery code": "Código TOTP o de recuperación no válido",
+  "No local password set, please set a password first": "No hay contraseña local. Configura una primero",
+  "This LangBot Account is already bound to another user": "Esta cuenta de LangBot ya está vinculada a otro usuario",
+  "Invalid recovery key": "Clave de recuperación no válida",
+  "Done": "Hecho",
+  "Thank you for your feedback": "Gracias por tus comentarios",
+  "Failed to process feedback": "No se pudieron procesar los comentarios",
+  "Helpful": "Útil",
+  "Not helpful": "No es útil",
+  "Select an option": "Selecciona una opción",
+  "Enter a value": "Introduce un valor",
+  "Submit": "Enviar",
+  "Submitted": "Enviado",
+  "Thinking": "Razonamiento",
+  "AI-generated": "Generado por IA",
+  "AI-generated content. Refer to the cited sources for details.": "Contenido generado por IA. Consulta las fuentes citadas para más detalles.",
+  "No bridges configured.": "No hay puentes configurados.",
+  "Signing in to all bridges again...": "Iniciando sesión de nuevo en todos los puentes…",
+  "Bridge status:": "Estado de los puentes:",
+  "Signed in ✓": "Sesión iniciada ✓",
+  "Not signed in ✗": "Sin sesión ✗",
+  "None": "Ninguno",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "Puente {bridge}: omitido (falta comando de inicio de sesión o sala DM).",
+  "Bridge {bridge}: sending {command}...": "Puente {bridge}: enviando {command}…",
+  "Bridge {bridge}: sign-out failed.": "Puente {bridge}: no se pudo cerrar sesión.",
+  "Bridge {bridge}: sign-in started.": "Puente {bridge}: inicio de sesión en curso.",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Autorización de Slack completada. Vuelve a LangBot para terminar la configuración.",
+  "Slack authorization failed or expired. Return to LangBot for details.": "La autorización de Slack falló o caducó. Consulta los detalles en LangBot.",
+  "TOTP is already enabled for this Account": "Esta cuenta ya tiene TOTP activado",
+  "TOTP is not enabled for this Account": "Esta cuenta no tiene TOTP activado",
+  "Invalid or expired passkey challenge": "El desafío de clave de acceso no es válido o ha caducado",
+  "Passkey credential not recognized": "No se reconoce la clave de acceso",
+  "Account is disabled": "La cuenta está desactivada",
+  "Modifying login info is disabled": "No se permite modificar los datos de inicio de sesión",
+  "Invalid GitHub repository URL": "URL del repositorio GitHub no válida",
+  "Server is busy. Please try again later.": "El servidor está ocupado. Inténtalo más tarde.",
+  "This feature is unavailable in this Workspace.": "Esta función no está disponible en este espacio de trabajo.",
+  "The Workspace must retain an owner.": "El espacio de trabajo debe conservar un propietario.",
+  "Invalid or expired invitation": "Invitación no válida o caducada",
+  "This user is already a member.": "Este usuario ya es miembro.",
+  "Voice connection failed": "Error de conexión de voz",
+  "User is not in the voice channel": "El usuario no está en el canal de voz",
+  "Voice service is unavailable": "El servicio de voz no está disponible",
+  "Submission failed. Please try again.": "El envío ha fallado. Reinténtalo.",
+  "File size exceeds the limit.": "El archivo supera el tamaño permitido.",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "El archivo supera los 10 MB. Divídelo en partes más pequeñas.",
+  "Authorization failed. Please try signing in again.": "La autorización ha fallado. Vuelve a iniciar sesión.",
+  "Bind the LangBot Account with the same email as this local Account": "Vincula una cuenta de LangBot con el mismo correo que esta cuenta local",
+  "Only local accounts can bind to Space": "Solo las cuentas locales pueden vincularse a Space",
+  "System already initialized": "El sistema ya está inicializado",
+  "System not initialized": "El sistema no está inicializado",
+  "Password login is disabled on LangBot Cloud": "El acceso con contraseña está desactivado en LangBot Cloud",
+  "This edition supports one Workspace per instance": "Esta edición admite un espacio de trabajo por instancia",
+  "Cloud Workspaces are created by the SaaS control plane": "Los espacios de trabajo Cloud se crean desde el panel SaaS",
+  "Login with your LangBot Account to accept this invitation": "Inicia sesión con tu cuenta de LangBot para aceptar la invitación",
+  "RAG migration is not needed": "No es necesaria la migración de RAG",
+  "New Agent": "Nuevo agente",
+  "New Pipeline": "Nuevo flujo",
+  "Provider {provider} does not support {model_type} models": "El proveedor {provider} no admite modelos {model_type}",
+  "Invalid JSON format": "Formato JSON no válido",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "Debes iniciar sesión en ChatGPT. Abre este proveedor y vuelve a iniciar sesión.",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "ChatGPT limitó la solicitud por frecuencia o uso. Inténtalo más tarde o revisa tu plan.",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT rechazó la solicitud. Revisa la suscripción y los permisos del espacio de trabajo.",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT rechazó el modelo o la solicitud. Revisa el modelo seleccionado y la configuración.",
+  "ChatGPT Codex upstream request failed. Please retry later.": "Falló la solicitud a ChatGPT Codex. Inténtalo más tarde.",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "Se alcanzó el límite de uso de tu suscripción de ChatGPT. Inténtalo más tarde o revisa tu plan.",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "La respuesta de ChatGPT Codex falló o está incompleta. Inténtalo de nuevo.",
+  "ChatGPT token refresh temporarily failed. Please retry.": "No se pudo renovar el token de ChatGPT temporalmente. Inténtalo de nuevo.",
+  "A second factor is required": "Se requiere un segundo factor de autenticación",
+  "Sign in or provide registration details to accept this invitation": "Inicia sesión o proporciona los datos de registro para aceptar esta invitación",
+  "Sign out before creating the invited local Account": "Cierra sesión antes de crear la cuenta local invitada",
+  "Turnstile verification failed": "La verificación de Turnstile falló. Inténtalo de nuevo",
+  "The invitation email does not match your Account": "El correo de la invitación no coincide con tu cuenta",
+  "Failed to save ChatGPT credentials. Please retry.": "No se pudieron guardar las credenciales de ChatGPT. Inténtalo de nuevo.",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "No se pudo iniciar el acceso por dispositivo. Activa el inicio de sesión con código de dispositivo en la configuración de ChatGPT.",
+  "Option {number}": "Opción {number}",
+  "Confirmation needed": "Se requiere confirmación",
+  "Session reset successfully": "Sesión restablecida"
+}
diff --git a/src/langbot/pkg/i18n/locales/ja_JP.json b/src/langbot/pkg/i18n/locales/ja_JP.json
new file mode 100644
index 000000000..292bb1b56
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/ja_JP.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "リクエストに失敗しました。",
+  "Too many requests. Please try again later.": "リクエストが多すぎます。しばらくしてから再試行してください。",
+  "Your message contains inappropriate content. Please revise it.": "不適切な内容が含まれています。メッセージを修正してください。",
+  "Content checking failed. Please contact the administrator.": "コンテンツの確認に失敗しました。管理者にお問い合わせください。",
+  "Calling {tools}...": "{tools} を呼び出しています…",
+  "Agent runner not found: {runner}": "Agent ランナーが見つかりません：{runner}",
+  "Agent runner not authorized for this pipeline.": "このパイプラインには Agent ランナーの使用権限がありません。",
+  "Agent runner temporarily unavailable. Please try again.": "Agent ランナーは一時的に利用できません。再試行してください。",
+  "Agent runner execution failed.": "Agent の実行に失敗しました。",
+  "The Agent runner returned an invalid response.": "Agent ランナーが無効な応答を返しました。",
+  "Model request failed": "モデルへのリクエストに失敗しました",
+  "Context length exceeded": "コンテキスト長の上限を超えました",
+  "Invalid request parameters": "リクエストのパラメータが無効です",
+  "Invalid API key": "API キーが無効です",
+  "Invalid model or endpoint": "モデルまたはエンドポイントが無効です",
+  "Rate limit exceeded or insufficient balance": "リクエスト制限を超えたか、残高が不足しています",
+  "Request timed out": "リクエストがタイムアウトしました",
+  "Connection failed": "接続に失敗しました",
+  "API error": "API エラー",
+  "Unknown error": "不明なエラー",
+  "Internal server error": "サーバー内部エラー",
+  "Resource not found": "リソースが見つかりません",
+  "Permission denied": "この操作の権限がありません",
+  "Invalid authentication credentials": "認証に失敗しました。再度ログインしてください",
+  "MCP runtime failed": "MCP の実行に失敗しました",
+  "Task execution failed": "タスクの実行に失敗しました",
+  "Task was cancelled": "タスクがキャンセルされました",
+  "Maximum number of {resource} ({limit}) reached": "{resource}の上限数（{limit}）に達しました",
+  "extensions": "拡張機能",
+  "knowledge bases": "ナレッジベース",
+  "MCP server already exists: {name}": "MCP サーバーは既に存在します：{name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models はクラウドで管理されているため変更できません",
+  "Please sign in with your LangBot Account": "LangBot アカウントでログインしてください",
+  "Current password is incorrect": "現在のパスワードが正しくありません",
+  "Invalid TOTP or recovery code": "認証コードまたはリカバリーコードが無効です",
+  "No local password set, please set a password first": "ローカルパスワードが未設定です。先に設定してください",
+  "This LangBot Account is already bound to another user": "この LangBot アカウントは別のユーザーに連携されています",
+  "Invalid recovery key": "リカバリーキーが無効です",
+  "Done": "完了しました",
+  "Thank you for your feedback": "フィードバックありがとうございます",
+  "Failed to process feedback": "フィードバックの処理に失敗しました",
+  "Helpful": "役に立った",
+  "Not helpful": "役に立たなかった",
+  "Select an option": "選択してください",
+  "Enter a value": "入力してください",
+  "Submit": "送信",
+  "Submitted": "送信済み",
+  "Thinking": "思考過程",
+  "AI-generated": "AI による生成コンテンツ",
+  "AI-generated content. Refer to the cited sources for details.": "AI が生成した内容です。詳細は引用元を参照してください。",
+  "No bridges configured.": "ブリッジが設定されていません。",
+  "Signing in to all bridges again...": "すべてのブリッジに再ログインしています…",
+  "Bridge status:": "ブリッジの状態：",
+  "Signed in ✓": "ログイン済み ✓",
+  "Not signed in ✗": "未ログイン ✗",
+  "None": "なし",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "ブリッジ {bridge}：スキップしました（ログインコマンドまたは DM ルームがありません）。",
+  "Bridge {bridge}: sending {command}...": "ブリッジ {bridge}：{command} を送信中…",
+  "Bridge {bridge}: sign-out failed.": "ブリッジ {bridge}：ログアウトに失敗しました。",
+  "Bridge {bridge}: sign-in started.": "ブリッジ {bridge}：ログインを開始しました。",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Slack の認証が完了しました。LangBot に戻って設定を完了してください。",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Slack の認証に失敗したか期限切れです。詳細は LangBot で確認してください。",
+  "TOTP is already enabled for this Account": "このアカウントでは二段階認証が有効です",
+  "TOTP is not enabled for this Account": "このアカウントでは二段階認証が無効です",
+  "Invalid or expired passkey challenge": "パスキーの認証要求が無効または期限切れです",
+  "Passkey credential not recognized": "このパスキーは認識できません",
+  "Account is disabled": "アカウントは無効です",
+  "Modifying login info is disabled": "ログイン情報の変更は無効です",
+  "Invalid GitHub repository URL": "GitHub リポジトリの URL が無効です",
+  "Server is busy. Please try again later.": "サーバーが混雑しています。後で再試行してください。",
+  "This feature is unavailable in this Workspace.": "このワークスペースではこの機能を利用できません。",
+  "The Workspace must retain an owner.": "ワークスペースにはオーナーが必要です。",
+  "Invalid or expired invitation": "招待が無効または期限切れです",
+  "This user is already a member.": "このユーザーは既にメンバーです。",
+  "Voice connection failed": "音声接続に失敗しました",
+  "User is not in the voice channel": "ユーザーは音声チャンネルに参加していません",
+  "Voice service is unavailable": "音声サービスを利用できません",
+  "Submission failed. Please try again.": "送信に失敗しました。再試行してください。",
+  "File size exceeds the limit.": "ファイルサイズが上限を超えています。",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "ファイルが 10MB を超えています。分割してアップロードしてください。",
+  "Authorization failed. Please try signing in again.": "認証に失敗しました。再度ログインしてください。",
+  "Bind the LangBot Account with the same email as this local Account": "このローカルアカウントと同じメールアドレスの LangBot アカウントを連携してください",
+  "Only local accounts can bind to Space": "Space と連携できるのはローカルアカウントのみです",
+  "System already initialized": "システムは初期化済みです",
+  "System not initialized": "システムが初期化されていません",
+  "Password login is disabled on LangBot Cloud": "LangBot Cloud ではパスワードログインは無効です",
+  "This edition supports one Workspace per instance": "このエディションではインスタンスごとにワークスペースは1つです",
+  "Cloud Workspaces are created by the SaaS control plane": "クラウドのワークスペースは SaaS 管理画面で作成します",
+  "Login with your LangBot Account to accept this invitation": "招待を承諾するには LangBot アカウントでログインしてください",
+  "RAG migration is not needed": "RAG の移行は不要です",
+  "New Agent": "新しい Agent",
+  "New Pipeline": "新しいパイプライン",
+  "Provider {provider} does not support {model_type} models": "プロバイダー {provider} は {model_type} モデルに対応していません",
+  "Invalid JSON format": "JSON 形式が無効です",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "ChatGPT へのログインが必要です。このプロバイダーを開いて再ログインしてください。",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "ChatGPT のリクエストが制限されました（頻度または使用量）。後で再試行するか、プランを確認してください。",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT がリクエストを拒否しました。サブスクリプションとワークスペースの権限を確認してください。",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT がモデルまたはリクエストを拒否しました。選択したモデルとリクエスト設定を確認してください。",
+  "ChatGPT Codex upstream request failed. Please retry later.": "ChatGPT Codex の上流リクエストに失敗しました。後で再試行してください。",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "ChatGPT のサブスクリプション使用量が上限に達しました。後で再試行するか、プランを確認してください。",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "ChatGPT Codex の応答に失敗したか、応答が不完全です。再試行してください。",
+  "ChatGPT token refresh temporarily failed. Please retry.": "ChatGPT トークンの更新に一時的に失敗しました。再試行してください。",
+  "A second factor is required": "二要素認証が必要です",
+  "Sign in or provide registration details to accept this invitation": "招待を承諾するにはログインするか、登録情報を入力してください",
+  "Sign out before creating the invited local Account": "招待されたローカルアカウントを作成する前にログアウトしてください",
+  "Turnstile verification failed": "人間確認に失敗しました。再試行してください",
+  "The invitation email does not match your Account": "招待のメールアドレスがアカウントと一致しません",
+  "Failed to save ChatGPT credentials. Please retry.": "ChatGPT 認証情報の保存に失敗しました。再試行してください。",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "ChatGPT のデバイスログインを開始できません。ChatGPT の設定でデバイスコードログインを有効にしてください。",
+  "Option {number}": "選択肢 {number}",
+  "Confirmation needed": "確認が必要です",
+  "Session reset successfully": "セッションをリセットしました"
+}
diff --git a/src/langbot/pkg/i18n/locales/ru_RU.json b/src/langbot/pkg/i18n/locales/ru_RU.json
new file mode 100644
index 000000000..ed9529018
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/ru_RU.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "Запрос не выполнен.",
+  "Too many requests. Please try again later.": "Слишком много запросов. Повторите позже.",
+  "Your message contains inappropriate content. Please revise it.": "Сообщение содержит недопустимый контент. Измените его.",
+  "Content checking failed. Please contact the administrator.": "Ошибка проверки содержимого. Обратитесь к администратору.",
+  "Calling {tools}...": "Вызов {tools}…",
+  "Agent runner not found: {runner}": "Исполнитель агента не найден: {runner}",
+  "Agent runner not authorized for this pipeline.": "У конвейера нет разрешения на использование исполнителя агента.",
+  "Agent runner temporarily unavailable. Please try again.": "Исполнитель агента временно недоступен. Повторите попытку.",
+  "Agent runner execution failed.": "Ошибка выполнения агента.",
+  "The Agent runner returned an invalid response.": "Исполнитель агента вернул некорректный ответ.",
+  "Model request failed": "Ошибка запроса к модели",
+  "Context length exceeded": "Превышена длина контекста",
+  "Invalid request parameters": "Некорректные параметры запроса",
+  "Invalid API key": "Недействительный API-ключ",
+  "Invalid model or endpoint": "Некорректная модель или адрес API",
+  "Rate limit exceeded or insufficient balance": "Превышен лимит запросов или недостаточно средств",
+  "Request timed out": "Время ожидания запроса истекло",
+  "Connection failed": "Ошибка подключения",
+  "API error": "Ошибка API",
+  "Unknown error": "Неизвестная ошибка",
+  "Internal server error": "Внутренняя ошибка сервера",
+  "Resource not found": "Ресурс не найден",
+  "Permission denied": "Нет разрешения на это действие",
+  "Invalid authentication credentials": "Ошибка аутентификации. Войдите снова",
+  "MCP runtime failed": "Ошибка выполнения MCP",
+  "Task execution failed": "Ошибка выполнения задачи",
+  "Task was cancelled": "Задача отменена",
+  "Maximum number of {resource} ({limit}) reached": "Достигнуто максимальное число {resource} ({limit})",
+  "extensions": "расширений",
+  "knowledge bases": "баз знаний",
+  "MCP server already exists: {name}": "Сервер MCP уже существует: {name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models управляется облаком и не может быть изменён",
+  "Please sign in with your LangBot Account": "Войдите с учётной записью LangBot",
+  "Current password is incorrect": "Текущий пароль неверен",
+  "Invalid TOTP or recovery code": "Неверный код TOTP или восстановления",
+  "No local password set, please set a password first": "Локальный пароль не задан. Сначала задайте пароль",
+  "This LangBot Account is already bound to another user": "Эта учётная запись LangBot уже связана с другим пользователем",
+  "Invalid recovery key": "Недействительный ключ восстановления",
+  "Done": "Готово",
+  "Thank you for your feedback": "Спасибо за отзыв",
+  "Failed to process feedback": "Не удалось обработать отзыв",
+  "Helpful": "Полезно",
+  "Not helpful": "Не полезно",
+  "Select an option": "Выберите вариант",
+  "Enter a value": "Введите значение",
+  "Submit": "Отправить",
+  "Submitted": "Отправлено",
+  "Thinking": "Ход рассуждений",
+  "AI-generated": "Создано ИИ",
+  "AI-generated content. Refer to the cited sources for details.": "Контент создан ИИ. Подробности смотрите в указанных источниках.",
+  "No bridges configured.": "Мосты не настроены.",
+  "Signing in to all bridges again...": "Повторный вход во все мосты…",
+  "Bridge status:": "Состояние мостов:",
+  "Signed in ✓": "Вход выполнен ✓",
+  "Not signed in ✗": "Вход не выполнен ✗",
+  "None": "Нет",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "Мост {bridge}: пропущен (нет команды входа или комнаты DM).",
+  "Bridge {bridge}: sending {command}...": "Мост {bridge}: отправка {command}…",
+  "Bridge {bridge}: sign-out failed.": "Мост {bridge}: ошибка выхода.",
+  "Bridge {bridge}: sign-in started.": "Мост {bridge}: вход начат.",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Авторизация Slack завершена. Вернитесь в LangBot для завершения настройки.",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Авторизация Slack не удалась или истекла. Подробности в LangBot.",
+  "TOTP is already enabled for this Account": "Для этой учётной записи TOTP уже включён",
+  "TOTP is not enabled for this Account": "Для этой учётной записи TOTP не включён",
+  "Invalid or expired passkey challenge": "Запрос проверки ключа доступа недействителен или истёк",
+  "Passkey credential not recognized": "Ключ доступа не распознан",
+  "Account is disabled": "Учётная запись отключена",
+  "Modifying login info is disabled": "Изменение данных входа отключено",
+  "Invalid GitHub repository URL": "Некорректный URL репозитория GitHub",
+  "Server is busy. Please try again later.": "Сервер занят. Повторите позже.",
+  "This feature is unavailable in this Workspace.": "Эта функция недоступна в данном рабочем пространстве.",
+  "The Workspace must retain an owner.": "У рабочего пространства должен остаться владелец.",
+  "Invalid or expired invitation": "Приглашение недействительно или истекло",
+  "This user is already a member.": "Этот пользователь уже является участником.",
+  "Voice connection failed": "Ошибка голосового соединения",
+  "User is not in the voice channel": "Пользователь не в голосовом канале",
+  "Voice service is unavailable": "Голосовая служба недоступна",
+  "Submission failed. Please try again.": "Не удалось отправить. Повторите попытку.",
+  "File size exceeds the limit.": "Размер файла превышает ограничение.",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "Размер файла превышает 10 МБ. Разделите его на части.",
+  "Authorization failed. Please try signing in again.": "Авторизация не удалась. Войдите снова.",
+  "Bind the LangBot Account with the same email as this local Account": "Свяжите учётную запись LangBot с тем же адресом почты, что и локальная",
+  "Only local accounts can bind to Space": "К Space можно привязать только локальные учётные записи",
+  "System already initialized": "Система уже инициализирована",
+  "System not initialized": "Система не инициализирована",
+  "Password login is disabled on LangBot Cloud": "В LangBot Cloud вход по паролю отключён",
+  "This edition supports one Workspace per instance": "Эта редакция поддерживает одно рабочее пространство на экземпляр",
+  "Cloud Workspaces are created by the SaaS control plane": "Облачные рабочие пространства создаются в панели SaaS",
+  "Login with your LangBot Account to accept this invitation": "Войдите в LangBot, чтобы принять приглашение",
+  "RAG migration is not needed": "Миграция RAG не требуется",
+  "New Agent": "Новый агент",
+  "New Pipeline": "Новый конвейер",
+  "Provider {provider} does not support {model_type} models": "Провайдер {provider} не поддерживает модели {model_type}",
+  "Invalid JSON format": "Некорректный формат JSON",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "Требуется вход в ChatGPT. Откройте этого провайдера и войдите снова.",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "Запрос ChatGPT ограничен по частоте или использованию. Повторите позже или проверьте тариф.",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT отклонил запрос. Проверьте подписку и права рабочего пространства.",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT отклонил модель или запрос. Проверьте выбранную модель и настройки запроса.",
+  "ChatGPT Codex upstream request failed. Please retry later.": "Ошибка запроса к ChatGPT Codex. Повторите позже.",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "Достигнут лимит использования подписки ChatGPT. Повторите позже или проверьте тариф.",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "Ошибка или неполный ответ ChatGPT Codex. Повторите запрос.",
+  "ChatGPT token refresh temporarily failed. Please retry.": "Временная ошибка обновления токена ChatGPT. Повторите попытку.",
+  "A second factor is required": "Требуется второй фактор аутентификации",
+  "Sign in or provide registration details to accept this invitation": "Войдите или укажите данные регистрации, чтобы принять приглашение",
+  "Sign out before creating the invited local Account": "Выйдите перед созданием приглашённой локальной учётной записи",
+  "Turnstile verification failed": "Проверка Turnstile не пройдена. Повторите попытку",
+  "The invitation email does not match your Account": "Адрес приглашения не совпадает с вашей учётной записью",
+  "Failed to save ChatGPT credentials. Please retry.": "Не удалось сохранить учётные данные ChatGPT. Повторите попытку.",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "Не удалось начать вход с устройства. Включите вход по коду устройства в настройках ChatGPT.",
+  "Option {number}": "Вариант {number}",
+  "Confirmation needed": "Требуется подтверждение",
+  "Session reset successfully": "Сессия сброшена"
+}
diff --git a/src/langbot/pkg/i18n/locales/th_TH.json b/src/langbot/pkg/i18n/locales/th_TH.json
new file mode 100644
index 000000000..ff40282cd
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/th_TH.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "คำขอล้มเหลว",
+  "Too many requests. Please try again later.": "มีคำขอมากเกินไป โปรดลองอีกครั้งภายหลัง",
+  "Your message contains inappropriate content. Please revise it.": "ข้อความมีเนื้อหาไม่เหมาะสม โปรดแก้ไข",
+  "Content checking failed. Please contact the administrator.": "ตรวจสอบเนื้อหาไม่สำเร็จ โปรดติดต่อผู้ดูแลระบบ",
+  "Calling {tools}...": "กำลังเรียกใช้ {tools}…",
+  "Agent runner not found: {runner}": "ไม่พบตัวรัน Agent: {runner}",
+  "Agent runner not authorized for this pipeline.": "ไปป์ไลน์นี้ไม่ได้รับอนุญาตให้ใช้ตัวรัน Agent",
+  "Agent runner temporarily unavailable. Please try again.": "ตัวรัน Agent ไม่พร้อมใช้งานชั่วคราว โปรดลองอีกครั้ง",
+  "Agent runner execution failed.": "การทำงานของ Agent ล้มเหลว",
+  "The Agent runner returned an invalid response.": "ตัวรัน Agent ส่งการตอบกลับที่ไม่ถูกต้อง",
+  "Model request failed": "คำขอไปยังโมเดลล้มเหลว",
+  "Context length exceeded": "ความยาวบริบทเกินขีดจำกัด",
+  "Invalid request parameters": "พารามิเตอร์คำขอไม่ถูกต้อง",
+  "Invalid API key": "คีย์ API ไม่ถูกต้อง",
+  "Invalid model or endpoint": "โมเดลหรือจุดเชื่อมต่อไม่ถูกต้อง",
+  "Rate limit exceeded or insufficient balance": "เกินขีดจำกัดคำขอหรือยอดเงินไม่เพียงพอ",
+  "Request timed out": "คำขอหมดเวลา",
+  "Connection failed": "การเชื่อมต่อล้มเหลว",
+  "API error": "ข้อผิดพลาด API",
+  "Unknown error": "ข้อผิดพลาดที่ไม่ทราบสาเหตุ",
+  "Internal server error": "ข้อผิดพลาดภายในเซิร์ฟเวอร์",
+  "Resource not found": "ไม่พบทรัพยากร",
+  "Permission denied": "ไม่มีสิทธิ์ดำเนินการนี้",
+  "Invalid authentication credentials": "ยืนยันตัวตนไม่สำเร็จ โปรดเข้าสู่ระบบอีกครั้ง",
+  "MCP runtime failed": "การทำงานของ MCP ล้มเหลว",
+  "Task execution failed": "การทำงานของงานล้มเหลว",
+  "Task was cancelled": "งานถูกยกเลิกแล้ว",
+  "Maximum number of {resource} ({limit}) reached": "ถึงจำนวนสูงสุดของ{resource} ({limit}) แล้ว",
+  "extensions": "ส่วนขยาย",
+  "knowledge bases": "ฐานความรู้",
+  "MCP server already exists: {name}": "มีเซิร์ฟเวอร์ MCP อยู่แล้ว: {name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models จัดการโดย Cloud และไม่สามารถแก้ไขได้",
+  "Please sign in with your LangBot Account": "โปรดเข้าสู่ระบบด้วยบัญชี LangBot",
+  "Current password is incorrect": "รหัสผ่านปัจจุบันไม่ถูกต้อง",
+  "Invalid TOTP or recovery code": "รหัส TOTP หรือรหัสกู้คืนไม่ถูกต้อง",
+  "No local password set, please set a password first": "ยังไม่ได้ตั้งรหัสผ่านในระบบ โปรดตั้งรหัสผ่านก่อน",
+  "This LangBot Account is already bound to another user": "บัญชี LangBot นี้เชื่อมโยงกับผู้ใช้อื่นแล้ว",
+  "Invalid recovery key": "คีย์กู้คืนไม่ถูกต้อง",
+  "Done": "เสร็จแล้ว",
+  "Thank you for your feedback": "ขอบคุณสำหรับความคิดเห็น",
+  "Failed to process feedback": "ประมวลผลความคิดเห็นไม่สำเร็จ",
+  "Helpful": "มีประโยชน์",
+  "Not helpful": "ไม่มีประโยชน์",
+  "Select an option": "โปรดเลือก",
+  "Enter a value": "โปรดป้อนค่า",
+  "Submit": "ส่ง",
+  "Submitted": "ส่งแล้ว",
+  "Thinking": "กระบวนการคิด",
+  "AI-generated": "สร้างโดย AI",
+  "AI-generated content. Refer to the cited sources for details.": "เนื้อหาสร้างโดย AI ดูรายละเอียดจากแหล่งอ้างอิง",
+  "No bridges configured.": "ยังไม่ได้ตั้งค่าบริดจ์",
+  "Signing in to all bridges again...": "กำลังเข้าสู่ระบบบริดจ์ทั้งหมดอีกครั้ง…",
+  "Bridge status:": "สถานะบริดจ์:",
+  "Signed in ✓": "เข้าสู่ระบบแล้ว ✓",
+  "Not signed in ✗": "ยังไม่ได้เข้าสู่ระบบ ✗",
+  "None": "ไม่มี",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "บริดจ์ {bridge}: ข้ามแล้ว (ไม่มีคำสั่งเข้าสู่ระบบหรือห้อง DM)",
+  "Bridge {bridge}: sending {command}...": "บริดจ์ {bridge}: กำลังส่ง {command}…",
+  "Bridge {bridge}: sign-out failed.": "บริดจ์ {bridge}: ออกจากระบบไม่สำเร็จ",
+  "Bridge {bridge}: sign-in started.": "บริดจ์ {bridge}: เริ่มเข้าสู่ระบบแล้ว",
+  "Slack authorization completed. Return to LangBot to finish setup.": "อนุญาต Slack สำเร็จแล้ว กลับไปที่ LangBot เพื่อดำเนินการตั้งค่าให้เสร็จ",
+  "Slack authorization failed or expired. Return to LangBot for details.": "อนุญาต Slack ไม่สำเร็จหรือหมดอายุ ดูรายละเอียดใน LangBot",
+  "TOTP is already enabled for this Account": "บัญชีนี้เปิดใช้ TOTP แล้ว",
+  "TOTP is not enabled for this Account": "บัญชีนี้ยังไม่ได้เปิดใช้ TOTP",
+  "Invalid or expired passkey challenge": "คำขอยืนยัน Passkey ไม่ถูกต้องหรือหมดอายุ",
+  "Passkey credential not recognized": "ไม่รู้จัก Passkey นี้",
+  "Account is disabled": "บัญชีถูกปิดใช้งาน",
+  "Modifying login info is disabled": "ไม่อนุญาตให้แก้ไขข้อมูลเข้าสู่ระบบ",
+  "Invalid GitHub repository URL": "URL ที่เก็บ GitHub ไม่ถูกต้อง",
+  "Server is busy. Please try again later.": "เซิร์ฟเวอร์ไม่ว่าง โปรดลองอีกครั้งภายหลัง",
+  "This feature is unavailable in this Workspace.": "พื้นที่ทำงานนี้ไม่สามารถใช้ฟีเจอร์นี้ได้",
+  "The Workspace must retain an owner.": "พื้นที่ทำงานต้องมีเจ้าของอย่างน้อยหนึ่งคน",
+  "Invalid or expired invitation": "คำเชิญไม่ถูกต้องหรือหมดอายุ",
+  "This user is already a member.": "ผู้ใช้นี้เป็นสมาชิกอยู่แล้ว",
+  "Voice connection failed": "การเชื่อมต่อเสียงล้มเหลว",
+  "User is not in the voice channel": "ผู้ใช้ไม่ได้อยู่ในช่องเสียง",
+  "Voice service is unavailable": "บริการเสียงไม่พร้อมใช้งาน",
+  "Submission failed. Please try again.": "ส่งไม่สำเร็จ โปรดลองอีกครั้ง",
+  "File size exceeds the limit.": "ขนาดไฟล์เกินขีดจำกัด",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "ไฟล์ใหญ่กว่า 10MB โปรดแบ่งเป็นไฟล์ย่อย",
+  "Authorization failed. Please try signing in again.": "การอนุญาตล้มเหลว โปรดเข้าสู่ระบบอีกครั้ง",
+  "Bind the LangBot Account with the same email as this local Account": "เชื่อมโยงบัญชี LangBot ที่ใช้อีเมลเดียวกับบัญชีในระบบนี้",
+  "Only local accounts can bind to Space": "เฉพาะบัญชีในระบบเท่านั้นที่เชื่อมโยงกับ Space ได้",
+  "System already initialized": "ระบบเริ่มต้นแล้ว",
+  "System not initialized": "ระบบยังไม่ได้เริ่มต้น",
+  "Password login is disabled on LangBot Cloud": "LangBot Cloud ปิดการเข้าสู่ระบบด้วยรหัสผ่าน",
+  "This edition supports one Workspace per instance": "รุ่นนี้รองรับพื้นที่ทำงานหนึ่งแห่งต่ออินสแตนซ์",
+  "Cloud Workspaces are created by the SaaS control plane": "พื้นที่ทำงาน Cloud สร้างจากแผงควบคุม SaaS",
+  "Login with your LangBot Account to accept this invitation": "เข้าสู่ระบบด้วยบัญชี LangBot เพื่อรับคำเชิญ",
+  "RAG migration is not needed": "ไม่จำเป็นต้องย้าย RAG",
+  "New Agent": "Agent ใหม่",
+  "New Pipeline": "ไปป์ไลน์ใหม่",
+  "Provider {provider} does not support {model_type} models": "ผู้ให้บริการ {provider} ไม่รองรับโมเดล {model_type}",
+  "Invalid JSON format": "รูปแบบ JSON ไม่ถูกต้อง",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "ต้องเข้าสู่ระบบ ChatGPT โปรดเปิดผู้ให้บริการนี้แล้วเข้าสู่ระบบอีกครั้ง",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "คำขอ ChatGPT ถูกจำกัดตามความถี่หรือปริมาณการใช้งาน โปรดลองภายหลังหรือตรวจสอบแผนของคุณ",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT ปฏิเสธคำขอ โปรดตรวจสอบการสมัครสมาชิกและสิทธิ์ของพื้นที่ทำงาน",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT ปฏิเสธโมเดลหรือคำขอ โปรดตรวจสอบโมเดลและการตั้งค่าคำขอ",
+  "ChatGPT Codex upstream request failed. Please retry later.": "คำขอไปยัง ChatGPT Codex ล้มเหลว โปรดลองอีกครั้งภายหลัง",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "การใช้งานสมาชิก ChatGPT ถึงขีดจำกัดแล้ว โปรดลองภายหลังหรือตรวจสอบแผนของคุณ",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "การตอบกลับของ ChatGPT Codex ล้มเหลวหรือไม่ครบถ้วน โปรดลองอีกครั้ง",
+  "ChatGPT token refresh temporarily failed. Please retry.": "ไม่สามารถรีเฟรชโทเค็น ChatGPT ได้ชั่วคราว โปรดลองอีกครั้ง",
+  "A second factor is required": "ต้องยืนยันตัวตนสองขั้นตอน",
+  "Sign in or provide registration details to accept this invitation": "เข้าสู่ระบบหรือกรอกข้อมูลลงทะเบียนเพื่อยอมรับคำเชิญนี้",
+  "Sign out before creating the invited local Account": "ออกจากระบบก่อนสร้างบัญชีภายในที่ได้รับเชิญ",
+  "Turnstile verification failed": "การยืนยัน Turnstile ล้มเหลว โปรดลองอีกครั้ง",
+  "The invitation email does not match your Account": "อีเมลคำเชิญไม่ตรงกับบัญชีของคุณ",
+  "Failed to save ChatGPT credentials. Please retry.": "บันทึกข้อมูลรับรอง ChatGPT ไม่สำเร็จ โปรดลองอีกครั้ง",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "ไม่สามารถเริ่มเข้าสู่ระบบด้วยอุปกรณ์ได้ โปรดเปิดการเข้าสู่ระบบด้วยรหัสอุปกรณ์ในการตั้งค่า ChatGPT",
+  "Option {number}": "ตัวเลือก {number}",
+  "Confirmation needed": "ต้องยืนยัน",
+  "Session reset successfully": "รีเซ็ตเซสชันแล้ว"
+}
diff --git a/src/langbot/pkg/i18n/locales/vi_VN.json b/src/langbot/pkg/i18n/locales/vi_VN.json
new file mode 100644
index 000000000..39b3110a8
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/vi_VN.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "Yêu cầu thất bại.",
+  "Too many requests. Please try again later.": "Quá nhiều yêu cầu. Vui lòng thử lại sau.",
+  "Your message contains inappropriate content. Please revise it.": "Tin nhắn chứa nội dung không phù hợp. Vui lòng chỉnh sửa.",
+  "Content checking failed. Please contact the administrator.": "Kiểm tra nội dung thất bại. Vui lòng liên hệ quản trị viên.",
+  "Calling {tools}...": "Đang gọi {tools}…",
+  "Agent runner not found: {runner}": "Không tìm thấy trình chạy Agent: {runner}",
+  "Agent runner not authorized for this pipeline.": "Pipeline này không được phép sử dụng trình chạy Agent.",
+  "Agent runner temporarily unavailable. Please try again.": "Trình chạy Agent tạm thời không khả dụng. Vui lòng thử lại.",
+  "Agent runner execution failed.": "Thực thi Agent thất bại.",
+  "The Agent runner returned an invalid response.": "Trình chạy Agent trả về phản hồi không hợp lệ.",
+  "Model request failed": "Yêu cầu đến mô hình thất bại",
+  "Context length exceeded": "Đã vượt quá độ dài ngữ cảnh",
+  "Invalid request parameters": "Tham số yêu cầu không hợp lệ",
+  "Invalid API key": "Khóa API không hợp lệ",
+  "Invalid model or endpoint": "Mô hình hoặc địa chỉ API không hợp lệ",
+  "Rate limit exceeded or insufficient balance": "Vượt giới hạn yêu cầu hoặc số dư không đủ",
+  "Request timed out": "Yêu cầu đã hết thời gian chờ",
+  "Connection failed": "Kết nối thất bại",
+  "API error": "Lỗi API",
+  "Unknown error": "Lỗi không xác định",
+  "Internal server error": "Lỗi máy chủ nội bộ",
+  "Resource not found": "Không tìm thấy tài nguyên",
+  "Permission denied": "Bạn không có quyền thực hiện thao tác này",
+  "Invalid authentication credentials": "Xác thực thất bại. Vui lòng đăng nhập lại",
+  "MCP runtime failed": "Thực thi MCP thất bại",
+  "Task execution failed": "Thực thi tác vụ thất bại",
+  "Task was cancelled": "Tác vụ đã bị hủy",
+  "Maximum number of {resource} ({limit}) reached": "Đã đạt số lượng tối đa của {resource} ({limit})",
+  "extensions": "tiện ích mở rộng",
+  "knowledge bases": "cơ sở tri thức",
+  "MCP server already exists: {name}": "Máy chủ MCP đã tồn tại: {name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models được quản lý bởi Cloud và không thể chỉnh sửa",
+  "Please sign in with your LangBot Account": "Vui lòng đăng nhập bằng tài khoản LangBot",
+  "Current password is incorrect": "Mật khẩu hiện tại không đúng",
+  "Invalid TOTP or recovery code": "Mã TOTP hoặc mã khôi phục không hợp lệ",
+  "No local password set, please set a password first": "Chưa có mật khẩu cục bộ. Vui lòng đặt mật khẩu trước",
+  "This LangBot Account is already bound to another user": "Tài khoản LangBot này đã được liên kết với người dùng khác",
+  "Invalid recovery key": "Khóa khôi phục không hợp lệ",
+  "Done": "Hoàn tất",
+  "Thank you for your feedback": "Cảm ơn bạn đã phản hồi",
+  "Failed to process feedback": "Không thể xử lý phản hồi",
+  "Helpful": "Hữu ích",
+  "Not helpful": "Không hữu ích",
+  "Select an option": "Chọn một mục",
+  "Enter a value": "Nhập giá trị",
+  "Submit": "Gửi",
+  "Submitted": "Đã gửi",
+  "Thinking": "Quá trình suy nghĩ",
+  "AI-generated": "Được tạo bởi AI",
+  "AI-generated content. Refer to the cited sources for details.": "Nội dung do AI tạo. Xem nguồn trích dẫn để biết chi tiết.",
+  "No bridges configured.": "Chưa cấu hình cầu nối nào.",
+  "Signing in to all bridges again...": "Đang đăng nhập lại vào tất cả cầu nối…",
+  "Bridge status:": "Trạng thái cầu nối:",
+  "Signed in ✓": "Đã đăng nhập ✓",
+  "Not signed in ✗": "Chưa đăng nhập ✗",
+  "None": "Không có",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "Cầu nối {bridge}: đã bỏ qua (thiếu lệnh đăng nhập hoặc phòng DM).",
+  "Bridge {bridge}: sending {command}...": "Cầu nối {bridge}: đang gửi {command}…",
+  "Bridge {bridge}: sign-out failed.": "Cầu nối {bridge}: đăng xuất thất bại.",
+  "Bridge {bridge}: sign-in started.": "Cầu nối {bridge}: đã bắt đầu đăng nhập.",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Đã cấp quyền Slack. Quay lại LangBot để hoàn tất thiết lập.",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Cấp quyền Slack thất bại hoặc đã hết hạn. Xem chi tiết trong LangBot.",
+  "TOTP is already enabled for this Account": "Tài khoản này đã bật TOTP",
+  "TOTP is not enabled for this Account": "Tài khoản này chưa bật TOTP",
+  "Invalid or expired passkey challenge": "Yêu cầu xác minh passkey không hợp lệ hoặc đã hết hạn",
+  "Passkey credential not recognized": "Không nhận diện được passkey",
+  "Account is disabled": "Tài khoản đã bị vô hiệu hóa",
+  "Modifying login info is disabled": "Không được phép sửa thông tin đăng nhập",
+  "Invalid GitHub repository URL": "URL kho GitHub không hợp lệ",
+  "Server is busy. Please try again later.": "Máy chủ đang bận. Vui lòng thử lại sau.",
+  "This feature is unavailable in this Workspace.": "Tính năng này không khả dụng trong không gian làm việc này.",
+  "The Workspace must retain an owner.": "Không gian làm việc phải có ít nhất một chủ sở hữu.",
+  "Invalid or expired invitation": "Lời mời không hợp lệ hoặc đã hết hạn",
+  "This user is already a member.": "Người dùng này đã là thành viên.",
+  "Voice connection failed": "Kết nối thoại thất bại",
+  "User is not in the voice channel": "Người dùng không ở trong kênh thoại",
+  "Voice service is unavailable": "Dịch vụ thoại không khả dụng",
+  "Submission failed. Please try again.": "Gửi thất bại. Vui lòng thử lại.",
+  "File size exceeds the limit.": "Kích thước tệp vượt quá giới hạn.",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "Tệp vượt quá 10MB. Vui lòng chia thành các phần nhỏ hơn.",
+  "Authorization failed. Please try signing in again.": "Cấp quyền thất bại. Vui lòng đăng nhập lại.",
+  "Bind the LangBot Account with the same email as this local Account": "Liên kết tài khoản LangBot có cùng email với tài khoản cục bộ này",
+  "Only local accounts can bind to Space": "Chỉ tài khoản cục bộ mới có thể liên kết với Space",
+  "System already initialized": "Hệ thống đã được khởi tạo",
+  "System not initialized": "Hệ thống chưa được khởi tạo",
+  "Password login is disabled on LangBot Cloud": "LangBot Cloud không cho phép đăng nhập bằng mật khẩu",
+  "This edition supports one Workspace per instance": "Phiên bản này hỗ trợ một không gian làm việc cho mỗi phiên bản triển khai",
+  "Cloud Workspaces are created by the SaaS control plane": "Không gian làm việc Cloud được tạo từ bảng điều khiển SaaS",
+  "Login with your LangBot Account to accept this invitation": "Đăng nhập bằng tài khoản LangBot để chấp nhận lời mời",
+  "RAG migration is not needed": "Không cần di chuyển RAG",
+  "New Agent": "Agent mới",
+  "New Pipeline": "Pipeline mới",
+  "Provider {provider} does not support {model_type} models": "Nhà cung cấp {provider} không hỗ trợ mô hình {model_type}",
+  "Invalid JSON format": "Định dạng JSON không hợp lệ",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "Cần đăng nhập ChatGPT. Mở nhà cung cấp này và đăng nhập lại.",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "Yêu cầu ChatGPT bị giới hạn tần suất hoặc mức sử dụng. Hãy thử lại sau hoặc kiểm tra gói của bạn.",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT đã từ chối yêu cầu. Kiểm tra gói đăng ký và quyền của không gian làm việc.",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT đã từ chối mô hình hoặc yêu cầu. Kiểm tra mô hình đã chọn và cài đặt yêu cầu.",
+  "ChatGPT Codex upstream request failed. Please retry later.": "Yêu cầu tới ChatGPT Codex thất bại. Vui lòng thử lại sau.",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "Đã đạt giới hạn sử dụng gói ChatGPT. Hãy thử lại sau hoặc kiểm tra gói của bạn.",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "Phản hồi ChatGPT Codex thất bại hoặc chưa hoàn tất. Vui lòng thử lại.",
+  "ChatGPT token refresh temporarily failed. Please retry.": "Tạm thời không thể làm mới token ChatGPT. Vui lòng thử lại.",
+  "A second factor is required": "Cần xác thực hai yếu tố",
+  "Sign in or provide registration details to accept this invitation": "Đăng nhập hoặc cung cấp thông tin đăng ký để chấp nhận lời mời",
+  "Sign out before creating the invited local Account": "Đăng xuất trước khi tạo tài khoản cục bộ được mời",
+  "Turnstile verification failed": "Xác minh Turnstile thất bại. Vui lòng thử lại.",
+  "The invitation email does not match your Account": "Email trong lời mời không khớp với tài khoản của bạn",
+  "Failed to save ChatGPT credentials. Please retry.": "Không thể lưu thông tin xác thực ChatGPT. Vui lòng thử lại.",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "Không thể bắt đầu đăng nhập thiết bị ChatGPT. Hãy bật đăng nhập bằng mã thiết bị trong cài đặt ChatGPT.",
+  "Option {number}": "Lựa chọn {number}",
+  "Confirmation needed": "Cần xác nhận",
+  "Session reset successfully": "Đã đặt lại phiên"
+}
diff --git a/src/langbot/pkg/i18n/locales/zh_Hans.json b/src/langbot/pkg/i18n/locales/zh_Hans.json
new file mode 100644
index 000000000..29a12e840
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/zh_Hans.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "请求失败。",
+  "Too many requests. Please try again later.": "请求过于频繁，请稍后重试。",
+  "Your message contains inappropriate content. Please revise it.": "消息中存在不合适的内容，请修改。",
+  "Content checking failed. Please contact the administrator.": "内容检查失败，请联系管理员。",
+  "Calling {tools}...": "正在调用 {tools}…",
+  "Agent runner not found: {runner}": "找不到 Agent 运行器：{runner}",
+  "Agent runner not authorized for this pipeline.": "此流水线未获授权使用该 Agent 运行器。",
+  "Agent runner temporarily unavailable. Please try again.": "Agent 运行器暂时不可用，请重试。",
+  "Agent runner execution failed.": "Agent 执行失败。",
+  "The Agent runner returned an invalid response.": "Agent 运行器返回了无效响应。",
+  "Model request failed": "模型请求失败",
+  "Context length exceeded": "上下文长度超限",
+  "Invalid request parameters": "请求参数错误",
+  "Invalid API key": "API 密钥无效",
+  "Invalid model or endpoint": "模型或接口地址无效",
+  "Rate limit exceeded or insufficient balance": "请求过于频繁或余额不足",
+  "Request timed out": "请求超时",
+  "Connection failed": "连接失败",
+  "API error": "API 错误",
+  "Unknown error": "未知错误",
+  "Internal server error": "服务器内部错误",
+  "Resource not found": "资源不存在",
+  "Permission denied": "没有执行此操作的权限",
+  "Invalid authentication credentials": "身份验证失败，请重新登录",
+  "MCP runtime failed": "MCP 运行失败",
+  "Task execution failed": "任务执行失败",
+  "Task was cancelled": "任务已取消",
+  "Maximum number of {resource} ({limit}) reached": "{resource}数量已达到上限（{limit}）",
+  "extensions": "扩展",
+  "knowledge bases": "知识库",
+  "MCP server already exists: {name}": "MCP 服务器已存在：{name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models 由云端管理，无法修改",
+  "Please sign in with your LangBot Account": "请使用 LangBot 账号登录",
+  "Current password is incorrect": "当前密码不正确",
+  "Invalid TOTP or recovery code": "验证码或恢复码无效",
+  "No local password set, please set a password first": "尚未设置本地密码，请先设置密码",
+  "This LangBot Account is already bound to another user": "此 LangBot 账号已绑定其他用户",
+  "Invalid recovery key": "恢复密钥无效",
+  "Done": "操作成功",
+  "Thank you for your feedback": "感谢您的反馈",
+  "Failed to process feedback": "反馈处理失败",
+  "Helpful": "有帮助",
+  "Not helpful": "无帮助",
+  "Select an option": "请选择",
+  "Enter a value": "请输入",
+  "Submit": "提交",
+  "Submitted": "已提交",
+  "Thinking": "思考过程",
+  "AI-generated": "内容由 AI 生成",
+  "AI-generated content. Refer to the cited sources for details.": "内容由 AI 生成，详情请参考引用来源。",
+  "No bridges configured.": "没有配置任何桥。",
+  "Signing in to all bridges again...": "正在重新登录所有桥…",
+  "Bridge status:": "桥状态：",
+  "Signed in ✓": "已登录 ✓",
+  "Not signed in ✗": "未登录 ✗",
+  "None": "无",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "桥 {bridge}：已跳过（缺少登录命令或私聊房间）。",
+  "Bridge {bridge}: sending {command}...": "桥 {bridge}：正在发送 {command}…",
+  "Bridge {bridge}: sign-out failed.": "桥 {bridge}：退出登录失败。",
+  "Bridge {bridge}: sign-in started.": "桥 {bridge}：已开始登录。",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Slack 授权完成，请返回 LangBot 完成配置。",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Slack 授权失败或已过期，请返回 LangBot 查看详情。",
+  "TOTP is already enabled for this Account": "此账号已启用两步验证",
+  "TOTP is not enabled for this Account": "此账号尚未启用两步验证",
+  "Invalid or expired passkey challenge": "通行密钥验证请求无效或已过期",
+  "Passkey credential not recognized": "无法识别此通行密钥",
+  "Account is disabled": "账号已停用",
+  "Modifying login info is disabled": "不允许修改登录信息",
+  "Invalid GitHub repository URL": "GitHub 仓库地址无效",
+  "Server is busy. Please try again later.": "服务器繁忙，请稍后重试。",
+  "This feature is unavailable in this Workspace.": "此工作区无法使用该功能。",
+  "The Workspace must retain an owner.": "工作区必须保留一位所有者。",
+  "Invalid or expired invitation": "邀请无效或已过期",
+  "This user is already a member.": "此用户已是工作区成员。",
+  "Voice connection failed": "语音连接失败",
+  "User is not in the voice channel": "用户不在语音频道中",
+  "Voice service is unavailable": "语音服务不可用",
+  "Submission failed. Please try again.": "提交失败，请重试。",
+  "File size exceeds the limit.": "文件大小超过限制。",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "文件大小超过 10MB，请拆分后上传。",
+  "Authorization failed. Please try signing in again.": "授权失败，请重新登录。",
+  "Bind the LangBot Account with the same email as this local Account": "请绑定与此本地账号邮箱相同的 LangBot 账号",
+  "Only local accounts can bind to Space": "只有本地账号可以绑定 Space",
+  "System already initialized": "系统已初始化",
+  "System not initialized": "系统尚未初始化",
+  "Password login is disabled on LangBot Cloud": "LangBot Cloud 不支持密码登录",
+  "This edition supports one Workspace per instance": "此版本每个实例仅支持一个工作区",
+  "Cloud Workspaces are created by the SaaS control plane": "云工作区由 SaaS 控制台创建",
+  "Login with your LangBot Account to accept this invitation": "请登录 LangBot 账号以接受邀请",
+  "RAG migration is not needed": "无需迁移知识库",
+  "New Agent": "新 Agent",
+  "New Pipeline": "新流水线",
+  "Provider {provider} does not support {model_type} models": "提供商 {provider} 不支持 {model_type} 模型",
+  "Invalid JSON format": "JSON 格式无效",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "需要登录 ChatGPT。请打开此提供商并重新登录。",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "ChatGPT 请求受限（请求频率或用量限制）。请稍后重试或检查套餐。",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT 拒绝了此请求。请检查订阅和工作区权限。",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT 拒绝了模型或请求。请检查所选模型和请求设置。",
+  "ChatGPT Codex upstream request failed. Please retry later.": "ChatGPT Codex 上游请求失败，请稍后重试。",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "ChatGPT 订阅用量已达上限。请稍后重试或检查套餐。",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "ChatGPT Codex 响应失败或不完整，请重试。",
+  "ChatGPT token refresh temporarily failed. Please retry.": "ChatGPT 令牌暂时刷新失败，请重试。",
+  "A second factor is required": "需要二次验证",
+  "Sign in or provide registration details to accept this invitation": "请登录或填写注册信息以接受此邀请",
+  "Sign out before creating the invited local Account": "请先退出登录，再创建受邀本地账号",
+  "Turnstile verification failed": "人机验证失败，请重试",
+  "The invitation email does not match your Account": "邀请邮箱与您的账号不匹配",
+  "Failed to save ChatGPT credentials. Please retry.": "保存 ChatGPT 凭据失败，请重试。",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "无法启动 ChatGPT 设备登录。请在 ChatGPT 设置中启用设备代码登录。",
+  "Option {number}": "选项 {number}",
+  "Confirmation needed": "需要确认",
+  "Session reset successfully": "会话已重置"
+}
diff --git a/src/langbot/pkg/i18n/locales/zh_Hant.json b/src/langbot/pkg/i18n/locales/zh_Hant.json
new file mode 100644
index 000000000..15f5e90c3
--- /dev/null
+++ b/src/langbot/pkg/i18n/locales/zh_Hant.json
@@ -0,0 +1,114 @@
+{
+  "Request failed.": "請求失敗。",
+  "Too many requests. Please try again later.": "請求過於頻繁，請稍後重試。",
+  "Your message contains inappropriate content. Please revise it.": "訊息中存在不適當的內容，請修改。",
+  "Content checking failed. Please contact the administrator.": "內容檢查失敗，請聯絡管理員。",
+  "Calling {tools}...": "正在呼叫 {tools}…",
+  "Agent runner not found: {runner}": "找不到 Agent 執行器：{runner}",
+  "Agent runner not authorized for this pipeline.": "此流水線未獲授權使用該 Agent 執行器。",
+  "Agent runner temporarily unavailable. Please try again.": "Agent 執行器暫時無法使用，請重試。",
+  "Agent runner execution failed.": "Agent 執行失敗。",
+  "The Agent runner returned an invalid response.": "Agent 執行器傳回了無效回應。",
+  "Model request failed": "模型請求失敗",
+  "Context length exceeded": "上下文長度超出限制",
+  "Invalid request parameters": "請求參數錯誤",
+  "Invalid API key": "API 金鑰無效",
+  "Invalid model or endpoint": "模型或介面位址無效",
+  "Rate limit exceeded or insufficient balance": "請求過於頻繁或餘額不足",
+  "Request timed out": "請求逾時",
+  "Connection failed": "連線失敗",
+  "API error": "API 錯誤",
+  "Unknown error": "未知錯誤",
+  "Internal server error": "伺服器內部錯誤",
+  "Resource not found": "資源不存在",
+  "Permission denied": "沒有執行此操作的權限",
+  "Invalid authentication credentials": "身分驗證失敗，請重新登入",
+  "MCP runtime failed": "MCP 執行失敗",
+  "Task execution failed": "任務執行失敗",
+  "Task was cancelled": "任務已取消",
+  "Maximum number of {resource} ({limit}) reached": "{resource}數量已達上限（{limit}）",
+  "extensions": "擴充功能",
+  "knowledge bases": "知識庫",
+  "MCP server already exists: {name}": "MCP 伺服器已存在：{name}",
+  "LangBot Models is managed by Cloud and cannot be modified": "LangBot Models 由雲端管理，無法修改",
+  "Please sign in with your LangBot Account": "請使用 LangBot 帳號登入",
+  "Current password is incorrect": "目前密碼不正確",
+  "Invalid TOTP or recovery code": "驗證碼或復原碼無效",
+  "No local password set, please set a password first": "尚未設定本機密碼，請先設定密碼",
+  "This LangBot Account is already bound to another user": "此 LangBot 帳號已綁定其他使用者",
+  "Invalid recovery key": "復原金鑰無效",
+  "Done": "操作成功",
+  "Thank you for your feedback": "感謝您的回饋",
+  "Failed to process feedback": "回饋處理失敗",
+  "Helpful": "有幫助",
+  "Not helpful": "沒有幫助",
+  "Select an option": "請選擇",
+  "Enter a value": "請輸入",
+  "Submit": "提交",
+  "Submitted": "已提交",
+  "Thinking": "思考過程",
+  "AI-generated": "內容由 AI 產生",
+  "AI-generated content. Refer to the cited sources for details.": "內容由 AI 產生，詳情請參考引用來源。",
+  "No bridges configured.": "尚未設定任何橋接。",
+  "Signing in to all bridges again...": "正在重新登入所有橋接…",
+  "Bridge status:": "橋接狀態：",
+  "Signed in ✓": "已登入 ✓",
+  "Not signed in ✗": "未登入 ✗",
+  "None": "無",
+  "Bridge {bridge}: skipped (login command or DM room missing).": "橋接 {bridge}：已略過（缺少登入命令或私訊房間）。",
+  "Bridge {bridge}: sending {command}...": "橋接 {bridge}：正在傳送 {command}…",
+  "Bridge {bridge}: sign-out failed.": "橋接 {bridge}：登出失敗。",
+  "Bridge {bridge}: sign-in started.": "橋接 {bridge}：已開始登入。",
+  "Slack authorization completed. Return to LangBot to finish setup.": "Slack 授權完成，請返回 LangBot 完成設定。",
+  "Slack authorization failed or expired. Return to LangBot for details.": "Slack 授權失敗或已過期，請返回 LangBot 查看詳情。",
+  "TOTP is already enabled for this Account": "此帳號已啟用兩步驗證",
+  "TOTP is not enabled for this Account": "此帳號尚未啟用兩步驗證",
+  "Invalid or expired passkey challenge": "通行金鑰驗證請求無效或已過期",
+  "Passkey credential not recognized": "無法識別此通行金鑰",
+  "Account is disabled": "帳號已停用",
+  "Modifying login info is disabled": "不允許修改登入資訊",
+  "Invalid GitHub repository URL": "GitHub 儲存庫網址無效",
+  "Server is busy. Please try again later.": "伺服器忙碌中，請稍後重試。",
+  "This feature is unavailable in this Workspace.": "此工作區無法使用該功能。",
+  "The Workspace must retain an owner.": "工作區必須保留一位擁有者。",
+  "Invalid or expired invitation": "邀請無效或已過期",
+  "This user is already a member.": "此使用者已是工作區成員。",
+  "Voice connection failed": "語音連線失敗",
+  "User is not in the voice channel": "使用者不在語音頻道中",
+  "Voice service is unavailable": "語音服務無法使用",
+  "Submission failed. Please try again.": "提交失敗，請重試。",
+  "File size exceeds the limit.": "檔案大小超過限制。",
+  "File size exceeds 10MB limit. Please split large files into smaller parts.": "檔案大小超過 10MB，請分割後上傳。",
+  "Authorization failed. Please try signing in again.": "授權失敗，請重新登入。",
+  "Bind the LangBot Account with the same email as this local Account": "請綁定與此本機帳號電子郵件相同的 LangBot 帳號",
+  "Only local accounts can bind to Space": "只有本機帳號可以綁定 Space",
+  "System already initialized": "系統已初始化",
+  "System not initialized": "系統尚未初始化",
+  "Password login is disabled on LangBot Cloud": "LangBot Cloud 不支援密碼登入",
+  "This edition supports one Workspace per instance": "此版本每個執行個體僅支援一個工作區",
+  "Cloud Workspaces are created by the SaaS control plane": "雲端工作區由 SaaS 控制台建立",
+  "Login with your LangBot Account to accept this invitation": "請登入 LangBot 帳號以接受邀請",
+  "RAG migration is not needed": "無需遷移知識庫",
+  "New Agent": "新 Agent",
+  "New Pipeline": "新流水線",
+  "Provider {provider} does not support {model_type} models": "提供商 {provider} 不支援 {model_type} 模型",
+  "Invalid JSON format": "JSON 格式無效",
+  "ChatGPT sign-in required. Open this provider and sign in again.": "需要登入 ChatGPT。請開啟此供應商並重新登入。",
+  "ChatGPT request was limited (rate limit or usage restriction). Please retry later or check your plan.": "ChatGPT 請求受限（請求頻率或用量限制）。請稍後重試或檢查方案。",
+  "ChatGPT denied this request. Check subscription and workspace permissions.": "ChatGPT 拒絕了此請求。請檢查訂閱和工作區權限。",
+  "ChatGPT rejected the model or request. Check the selected model and request settings.": "ChatGPT 拒絕了模型或請求。請檢查所選模型和請求設定。",
+  "ChatGPT Codex upstream request failed. Please retry later.": "ChatGPT Codex 上游請求失敗，請稍後重試。",
+  "ChatGPT subscription usage limit reached. Please retry later or check your plan.": "ChatGPT 訂閱用量已達上限。請稍後重試或檢查方案。",
+  "ChatGPT Codex response failed or was incomplete. Please retry.": "ChatGPT Codex 回應失敗或不完整，請重試。",
+  "ChatGPT token refresh temporarily failed. Please retry.": "ChatGPT 權杖暫時更新失敗，請重試。",
+  "A second factor is required": "需要雙重驗證",
+  "Sign in or provide registration details to accept this invitation": "請登入或填寫註冊資料以接受此邀請",
+  "Sign out before creating the invited local Account": "請先登出，再建立受邀本機帳號",
+  "Turnstile verification failed": "人機驗證失敗，請重試",
+  "The invitation email does not match your Account": "邀請信箱與您的帳號不符",
+  "Failed to save ChatGPT credentials. Please retry.": "儲存 ChatGPT 憑證失敗，請重試。",
+  "Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.": "無法啟動 ChatGPT 裝置登入。請在 ChatGPT 設定中啟用裝置代碼登入。",
+  "Option {number}": "選項 {number}",
+  "Confirmation needed": "需要確認",
+  "Session reset successfully": "對話已重設"
+}
diff --git a/src/langbot/pkg/i18n/workspace.py b/src/langbot/pkg/i18n/workspace.py
new file mode 100644
index 000000000..0527b818b
--- /dev/null
+++ b/src/langbot/pkg/i18n/workspace.py
@@ -0,0 +1,71 @@
+"""Workspace-owned message language, stored in the existing tenant metadata."""
+
+from __future__ import annotations
+
+import time
+from collections import OrderedDict
+
+import sqlalchemy as sa
+from sqlalchemy.dialects.postgresql import insert as pg_insert
+from sqlalchemy.dialects.sqlite import insert as sqlite_insert
+
+from ..entity.persistence.metadata import WorkspaceMetadata
+from . import DEFAULT_LANGUAGE, SUPPORTED_LANGUAGES, normalize_language
+
+LANGUAGE_KEY = 'message_language'
+_CACHE_TTL = 30
+_CACHE_SIZE = 256
+
+
+def _cache(ap) -> OrderedDict:
+    cache = getattr(ap, '_message_language_cache', None)
+    if cache is None:
+        cache = OrderedDict()
+        ap._message_language_cache = cache
+    return cache
+
+
+async def get_workspace_language(ap, workspace_uuid: str, *, refresh: bool = False) -> str:
+    if not workspace_uuid:
+        return DEFAULT_LANGUAGE
+    cache = _cache(ap)
+    cached = cache.get(workspace_uuid)
+    if not refresh and cached and cached[1] > time.monotonic():
+        return cached[0]
+    manager = getattr(ap, 'persistence_mgr', None)
+    if manager is None or not callable(getattr(manager, 'tenant_uow', None)):
+        return DEFAULT_LANGUAGE
+    async with manager.tenant_uow(workspace_uuid) as uow:
+        result = await uow.execute(
+            sa.select(WorkspaceMetadata.value).where(
+                WorkspaceMetadata.workspace_uuid == workspace_uuid,
+                WorkspaceMetadata.key == LANGUAGE_KEY,
+            )
+        )
+        locale = normalize_language(result.scalar_one_or_none())
+    cache[workspace_uuid] = (locale, time.monotonic() + _CACHE_TTL)
+    cache.move_to_end(workspace_uuid)
+    while len(cache) > _CACHE_SIZE:
+        cache.popitem(last=False)
+    return locale
+
+
+async def set_workspace_language(ap, workspace_uuid: str, locale: str) -> str:
+    if locale not in SUPPORTED_LANGUAGES:
+        raise ValueError('Unsupported Workspace language')
+    async with ap.persistence_mgr.tenant_uow(workspace_uuid) as uow:
+        dialect = ap.persistence_mgr.get_db_engine().dialect.name
+        insert = pg_insert if dialect == 'postgresql' else sqlite_insert
+        statement = insert(WorkspaceMetadata).values(workspace_uuid=workspace_uuid, key=LANGUAGE_KEY, value=locale)
+        await uow.execute(
+            statement.on_conflict_do_update(
+                index_elements=['workspace_uuid', 'key'],
+                set_={'value': locale},
+            )
+        )
+    # Synchronous platform callbacks may run in SDK threads. Publish the new
+    # value only after commit, so they see changes without restarting a bot.
+    _cache(ap)[workspace_uuid] = (locale, time.monotonic() + _CACHE_TTL)
+    while len(_cache(ap)) > _CACHE_SIZE:
+        _cache(ap).popitem(last=False)
+    return locale
diff --git a/src/langbot/pkg/operation_trace/service.py b/src/langbot/pkg/operation_trace/service.py
index 82be57fcc..1d771645e 100644
--- a/src/langbot/pkg/operation_trace/service.py
+++ b/src/langbot/pkg/operation_trace/service.py
@@ -23,6 +23,8 @@ Human-readable labels are never returned from this module. The persisted
 from __future__ import annotations
 
 import asyncio
+import contextlib
+import contextvars
 import dataclasses
 import datetime
 import enum
@@ -1198,6 +1200,9 @@ class WorkspaceSettingsService:
         # refresh and page turn, which is what made reading the log feel slow.
         self._integrity_cache: dict[str, dict[str, typing.Any]] = {}
         self._dropped_count = 0
+        # Failed persistence attempts, so a silently broken writer (a missing
+        # tenant scope, an RLS rejection) is reported instead of only counted.
+        self._write_failures = 0
         # Set while the queue is empty; lets tests/shutdown await a drain.
         self._idle = asyncio.Event()
         self._idle.set()
@@ -1232,59 +1237,116 @@ class WorkspaceSettingsService:
         enable_globally()
         self.ap.operation_trace_active = True
 
+    def _workspace_scope(self, workspace_uuid: str) -> typing.AsyncContextManager[typing.Any]:
+        """Carry a Workspace scope into persistence work that runs off-request.
+
+        Cloud runtime rejects every database call made without an explicit scope,
+        and the audit table is RLS-protected on ``langbot.workspace_uuid``, so the
+        background writer and the retention loop have to establish their own
+        scope. On the request path the caller already bound the Workspace: that
+        scope is the correct one, so the boundary is skipped rather than nested.
+        """
+
+        manager = getattr(self.ap, 'persistence_mgr', None)
+        tenant_scope = getattr(manager, 'tenant_scope', None)
+        current_scope = getattr(manager, 'current_scope', None)
+        if not callable(tenant_scope) or not workspace_uuid:
+            return contextlib.nullcontext()
+        if callable(current_scope):
+            try:
+                active = current_scope()
+            except Exception:
+                # A child task that inherited its parent's scope cannot use it;
+                # an explicit boundary of its own is exactly what it needs.
+                active = None
+            if active is not None:
+                return contextlib.nullcontext()
+        return tenant_scope(workspace_uuid)
+
     async def prime_global_flag(self) -> None:
         """Open the cheap global gate at startup if a Workspace already opted in.
 
-        Runs one bounded query during application build, never per request, so a
+        Runs once during application build, never per request, so a
         previously-enabled Workspace keeps recording after a restart while a
-        disabled instance still pays nothing on the hot path.
+        disabled instance still pays nothing on the hot path. Workspaces are
+        enumerated through the execution bindings this instance owns, because the
+        metadata table is RLS-protected per Workspace and a single cross-Workspace
+        scan can never see them in cloud runtime.
         """
 
         try:
-            result = await self.ap.persistence_mgr.execute_async(
-                sqlalchemy.select(persistence_metadata.WorkspaceMetadata.workspace_uuid)
-                .where(
-                    persistence_metadata.WorkspaceMetadata.key == OPERATION_LEVEL_KEY,
-                    persistence_metadata.WorkspaceMetadata.value != str(OPERATION_LEVEL_NONE),
-                )
-                .limit(1)
-            )
-            if result.first() is not None:
-                self._activate()
+            for workspace_uuid in await self._traced_workspace_candidates():
+                async with self._workspace_scope(workspace_uuid):
+                    level = clamp_level(await self._read_metadata(workspace_uuid, OPERATION_LEVEL_KEY))
+                if level > OPERATION_LEVEL_NONE:
+                    self._activate()
+                    return
         except Exception as exc:  # pragma: no cover - defensive
             self.ap.logger.debug(f'Operation trace global gate prime skipped: {exc}')
 
+    async def _traced_workspace_candidates(self) -> list[str]:
+        """Return the Workspaces that may have tracing enabled.
+
+        A multi-Workspace runtime enumerates the execution bindings discovered for
+        this instance. A single-database deployment can scan the metadata table
+        directly, which is also the path a lightweight test double takes.
+        """
+
+        list_bindings = getattr(
+            getattr(self.ap, 'workspace_service', None),
+            'list_active_execution_bindings',
+            None,
+        )
+        if callable(list_bindings):
+            try:
+                bindings = await list_bindings()
+            except Exception as exc:  # pragma: no cover - discovery is optional
+                self.ap.logger.debug(f'Operation trace Workspace discovery skipped: {exc}')
+            else:
+                return [str(binding.workspace_uuid) for binding in bindings]
+        result = await self.ap.persistence_mgr.execute_async(
+            sqlalchemy.select(persistence_metadata.WorkspaceMetadata.workspace_uuid).where(
+                persistence_metadata.WorkspaceMetadata.key == OPERATION_LEVEL_KEY,
+                persistence_metadata.WorkspaceMetadata.value != str(OPERATION_LEVEL_NONE),
+            )
+        )
+        return [str(workspace_uuid) for workspace_uuid in result.scalars().all()]
+
     async def _read_metadata(self, workspace_uuid: str, key: str) -> str | None:
         try:
-            result = await self.ap.persistence_mgr.execute_async(
-                sqlalchemy.select(persistence_metadata.WorkspaceMetadata.value).where(
-                    persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
-                    persistence_metadata.WorkspaceMetadata.key == key,
+            # Resolved outside the request scope when the route wrapper records a
+            # request, so the read has to carry the Workspace itself.
+            async with self._workspace_scope(workspace_uuid):
+                result = await self.ap.persistence_mgr.execute_async(
+                    sqlalchemy.select(persistence_metadata.WorkspaceMetadata.value).where(
+                        persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
+                        persistence_metadata.WorkspaceMetadata.key == key,
+                    )
                 )
-            )
-            return result.scalar_one_or_none()
+                return result.scalar_one_or_none()
         except Exception as exc:  # pragma: no cover - defensive
             self.ap.logger.debug(f'Operation log metadata read failed for {key}: {exc}')
             return None
 
     async def _write_metadata(self, workspace_uuid: str, values: dict[str, str]) -> None:
-        for key, value in values.items():
-            try:
-                await self.ap.persistence_mgr.execute_async(
-                    sqlalchemy.delete(persistence_metadata.WorkspaceMetadata).where(
-                        persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
-                        persistence_metadata.WorkspaceMetadata.key == key,
+        async with self._workspace_scope(workspace_uuid):
+            for key, value in values.items():
+                try:
+                    await self.ap.persistence_mgr.execute_async(
+                        sqlalchemy.delete(persistence_metadata.WorkspaceMetadata).where(
+                            persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
+                            persistence_metadata.WorkspaceMetadata.key == key,
+                        )
                     )
-                )
-                await self.ap.persistence_mgr.execute_async(
-                    sqlalchemy.insert(persistence_metadata.WorkspaceMetadata).values(
-                        workspace_uuid=workspace_uuid,
-                        key=key,
-                        value=value,
+                    await self.ap.persistence_mgr.execute_async(
+                        sqlalchemy.insert(persistence_metadata.WorkspaceMetadata).values(
+                            workspace_uuid=workspace_uuid,
+                            key=key,
+                            value=value,
+                        )
                     )
-                )
-            except Exception as exc:  # pragma: no cover - defensive
-                self.ap.logger.warning(f'Operation log metadata write failed for {key}: {exc}')
+                except Exception as exc:  # pragma: no cover - defensive
+                    self.ap.logger.warning(f'Operation log metadata write failed for {key}: {exc}')
 
     async def _read_metadata_values(self, workspace_uuid: str, keys: typing.Iterable[str]) -> dict[str, str]:
         """Read several metadata rows for one Workspace in a single query."""
@@ -1293,16 +1355,17 @@ class WorkspaceSettingsService:
         if not wanted:
             return {}
         try:
-            result = await self.ap.persistence_mgr.execute_async(
-                sqlalchemy.select(
-                    persistence_metadata.WorkspaceMetadata.key,
-                    persistence_metadata.WorkspaceMetadata.value,
-                ).where(
-                    persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
-                    persistence_metadata.WorkspaceMetadata.key.in_(wanted),
+            async with self._workspace_scope(workspace_uuid):
+                result = await self.ap.persistence_mgr.execute_async(
+                    sqlalchemy.select(
+                        persistence_metadata.WorkspaceMetadata.key,
+                        persistence_metadata.WorkspaceMetadata.value,
+                    ).where(
+                        persistence_metadata.WorkspaceMetadata.workspace_uuid == workspace_uuid,
+                        persistence_metadata.WorkspaceMetadata.key.in_(wanted),
+                    )
                 )
-            )
-            return {row[0]: row[1] for row in result.all()}
+                return {row[0]: row[1] for row in result.all()}
         except Exception as exc:  # pragma: no cover - defensive
             self.ap.logger.debug(f'Operation log metadata batch read failed: {exc}')
             return {}
@@ -1583,7 +1646,13 @@ class WorkspaceSettingsService:
 
         if self._writer_task is not None and not self._writer_task.done():
             return
-        self._writer_task = asyncio.get_running_loop().create_task(self._writer_loop())
+        # A fresh context: the writer outlives the request that first enqueued a
+        # trace. Inheriting that request's persistence scope would make every
+        # later write fail, because a child task may not use its parent's scope.
+        self._writer_task = asyncio.get_running_loop().create_task(
+            self._writer_loop(),
+            context=contextvars.Context(),
+        )
 
     async def _writer_loop(self) -> None:
         """Drain the queue and persist traces sequentially.
@@ -1627,23 +1696,42 @@ class WorkspaceSettingsService:
 
         try:
             workspace_uuid = record_fields['workspace_uuid']
-            # Off the request path: hash the dedupe material, resolve the actor
-            # display name, then link and insert this row.
-            record_fields['dedupe_key'] = hash_dedupe_material(record_fields.pop('_dedupe_material'))
-            if record_fields.get('actor_name') is None and record_fields.get('actor_account_uuid'):
-                await self._fill_actor_name(record_fields)
-            if await self._is_duplicate_observation(workspace_uuid, record_fields['dedupe_key']):
-                return False
-            record_fields['prev_hash'] = await self._latest_record_hash(workspace_uuid)
-            record_fields['record_hash'] = compute_record_hash(record_fields)
-            await self.ap.persistence_mgr.execute_async(
-                sqlalchemy.insert(persistence_operation_log.WorkspaceOperationLog).values(**record_fields)
-            )
-            return True
+            # The writer drains the queue outside any request, so it has to carry
+            # its own Workspace scope: cloud runtime refuses unscoped access and
+            # the audit table only accepts rows that match the bound Workspace.
+            async with self._workspace_scope(workspace_uuid):
+                # Off the request path: hash the dedupe material, resolve the actor
+                # display name, then link and insert this row.
+                record_fields['dedupe_key'] = hash_dedupe_material(record_fields.pop('_dedupe_material'))
+                if record_fields.get('actor_name') is None and record_fields.get('actor_account_uuid'):
+                    await self._fill_actor_name(record_fields)
+                if await self._is_duplicate_observation(workspace_uuid, record_fields['dedupe_key']):
+                    return False
+                record_fields['prev_hash'] = await self._latest_record_hash(workspace_uuid)
+                record_fields['record_hash'] = compute_record_hash(record_fields)
+                await self.ap.persistence_mgr.execute_async(
+                    sqlalchemy.insert(persistence_operation_log.WorkspaceOperationLog).values(**record_fields)
+                )
+                return True
         except Exception as exc:  # pragma: no cover - auditing is best effort
-            self.ap.logger.debug(f'Operation log write skipped: {exc}')
+            self._note_write_failure(exc)
             return False
 
+    def _note_write_failure(self, exc: BaseException) -> None:
+        """Report a dropped trace without letting it break the caller.
+
+        Recording is best effort, but an audit trail that silently stops growing
+        is indistinguishable from an idle Workspace. The first failure and then
+        every hundredth are surfaced; the rest stay at debug so a persistent
+        failure cannot flood the log.
+        """
+
+        self._write_failures += 1
+        if self._write_failures == 1 or self._write_failures % 100 == 0:
+            self.ap.logger.warning(f'Operation log write failed ({self._write_failures} so far): {exc}')
+        else:
+            self.ap.logger.debug(f'Operation log write skipped: {exc}')
+
     async def _fill_actor_name(self, record_fields: dict[str, typing.Any]) -> None:
         """Resolve the actor display name off the request path."""
 
@@ -2344,7 +2432,9 @@ class WorkspaceSettingsService:
             self._insert_counters[workspace_uuid] = pending
             return
         self._insert_counters[workspace_uuid] = 0
-        await self._enforce_row_budget(workspace_uuid)
+        # Runs from the writer task, which holds no scope of its own.
+        async with self._workspace_scope(workspace_uuid):
+            await self._enforce_row_budget(workspace_uuid)
 
     async def _enforce_row_budget(self, workspace_uuid: str) -> None:
         """Drop the oldest rows when the Workspace row budget is exceeded."""
@@ -2396,67 +2486,39 @@ class WorkspaceSettingsService:
         by a privileged click.
         """
 
-        days = (
-            await self.get_retention_days(workspace_uuid) if retention_days is None else clamp_retention(retention_days)
-        )
-        budget = await self.get_max_rows(workspace_uuid) if max_rows is None else clamp_max_rows(max_rows)
-        cutoff = _utcnow() - datetime.timedelta(days=days)
-
-        expired = 0
-        try:
-            model = persistence_operation_log.WorkspaceOperationLog
-            result = await self.ap.persistence_mgr.execute_async(
-                sqlalchemy.delete(model).where(
-                    model.workspace_uuid == workspace_uuid,
-                    model.created_at < cutoff,
-                )
+        # The maintenance loop runs per Workspace without holding a scope, so
+        # retention has to bind one itself; on the settings route the request
+        # scope is already active and this is a no-op.
+        async with self._workspace_scope(workspace_uuid):
+            days = (
+                await self.get_retention_days(workspace_uuid)
+                if retention_days is None
+                else clamp_retention(retention_days)
             )
-            expired = int(result.rowcount or 0)
-        except Exception as exc:  # pragma: no cover - defensive
-            self.ap.logger.debug(f'Operation log retention prune skipped: {exc}')
+            budget = await self.get_max_rows(workspace_uuid) if max_rows is None else clamp_max_rows(max_rows)
+            cutoff = _utcnow() - datetime.timedelta(days=days)
 
-        if expired > 0:
-            # Age-based retention removed the oldest rows, so any cached
-            # verification -- including its boundary baseline link -- no longer
-            # describes the surviving chain.
-            self._invalidate_integrity_cache(workspace_uuid)
-
-        total = await self.count_logs(workspace_uuid)
-        trimmed = await self._delete_oldest(workspace_uuid, max(total - budget, 0))
-        # The budget was just enforced in full, so the deferred counter restarts.
-        self._insert_counters[workspace_uuid] = 0
-        return {'expired': expired, 'trimmed': trimmed, 'remaining': await self.count_logs(workspace_uuid)}
-
-    async def prune_all_workspaces(self) -> dict[str, int]:
-        """Run retention pruning for every Workspace that enabled tracing.
-
-        Returns a coarse report; individual Workspace failures are contained.
-        """
-
-        processed = 0
-        removed = 0
-        try:
-            workspaces = (
-                (
-                    await self.ap.persistence_mgr.execute_async(
-                        sqlalchemy.select(persistence_metadata.WorkspaceMetadata.workspace_uuid).where(
-                            persistence_metadata.WorkspaceMetadata.key == OPERATION_LEVEL_KEY
-                        )
+            expired = 0
+            try:
+                model = persistence_operation_log.WorkspaceOperationLog
+                result = await self.ap.persistence_mgr.execute_async(
+                    sqlalchemy.delete(model).where(
+                        model.workspace_uuid == workspace_uuid,
+                        model.created_at < cutoff,
                     )
                 )
-                .scalars()
-                .all()
-            )
-        except Exception as exc:  # pragma: no cover - defensive
-            self.ap.logger.debug(f'Operation log workspace enumeration skipped: {exc}')
-            return {'workspaces': 0, 'removed': 0}
-
-        for workspace_uuid in {item for item in workspaces if item}:
-            try:
-                report = await self.prune(workspace_uuid)
+                expired = int(result.rowcount or 0)
             except Exception as exc:  # pragma: no cover - defensive
-                self.ap.logger.debug(f'Operation log prune failed for {workspace_uuid}: {exc}')
-                continue
-            processed += 1
-            removed += report['expired'] + report['trimmed']
-        return {'workspaces': processed, 'removed': removed}
+                self.ap.logger.debug(f'Operation log retention prune skipped: {exc}')
+
+            if expired > 0:
+                # Age-based retention removed the oldest rows, so any cached
+                # verification -- including its boundary baseline link -- no longer
+                # describes the surviving chain.
+                self._invalidate_integrity_cache(workspace_uuid)
+
+            total = await self.count_logs(workspace_uuid)
+            trimmed = await self._delete_oldest(workspace_uuid, max(total - budget, 0))
+            # The budget was just enforced in full, so the deferred counter restarts.
+            self._insert_counters[workspace_uuid] = 0
+            return {'expired': expired, 'trimmed': trimmed, 'remaining': await self.count_logs(workspace_uuid)}
diff --git a/src/langbot/pkg/persistence/alembic/versions/0034_monitoring_execution_links.py b/src/langbot/pkg/persistence/alembic/versions/0034_monitoring_execution_links.py
new file mode 100644
index 000000000..4cc720d07
--- /dev/null
+++ b/src/langbot/pkg/persistence/alembic/versions/0034_monitoring_execution_links.py
@@ -0,0 +1,36 @@
+"""Link monitoring inputs and deliveries to events and runs."""
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = '0034_monitoring_execution_links'
+down_revision = '0033_operation_traceability'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    inspector = sa.inspect(op.get_bind())
+    if not inspector.has_table('monitoring_messages'):
+        return
+    columns = {c['name'] for c in inspector.get_columns('monitoring_messages')}
+    indexes = {i['name'] for i in inspector.get_indexes('monitoring_messages')}
+    for name in ('event_id', 'run_id', 'parent_message_id'):
+        if name not in columns:
+            op.add_column('monitoring_messages', sa.Column(name, sa.String(255), nullable=True))
+        index = f'ix_monitoring_messages_{name}'
+        if index not in indexes:
+            op.create_index(index, 'monitoring_messages', [name])
+
+
+def downgrade():
+    inspector = sa.inspect(op.get_bind())
+    if not inspector.has_table('monitoring_messages'):
+        return
+    columns = {c['name'] for c in inspector.get_columns('monitoring_messages')}
+    indexes = {i['name'] for i in inspector.get_indexes('monitoring_messages')}
+    for name in ('parent_message_id', 'run_id', 'event_id'):
+        if f'ix_monitoring_messages_{name}' in indexes:
+            op.drop_index(f'ix_monitoring_messages_{name}', table_name='monitoring_messages')
+        if name in columns:
+            op.drop_column('monitoring_messages', name)
diff --git a/src/langbot/pkg/persistence/alembic/versions/0035_pipeline_received_only.py b/src/langbot/pkg/persistence/alembic/versions/0035_pipeline_received_only.py
new file mode 100644
index 000000000..146d0c754
--- /dev/null
+++ b/src/langbot/pkg/persistence/alembic/versions/0035_pipeline_received_only.py
@@ -0,0 +1,39 @@
+"""Narrow legacy Pipeline wildcard routes to incoming messages."""
+
+import json
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = '0035_pipeline_received_only'
+down_revision = '0034_monitoring_execution_links'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    connection = op.get_bind()
+    if not sa.inspect(connection).has_table('bots'):
+        return
+    bots = sa.table('bots', sa.column('uuid', sa.String()), sa.column('event_bindings', sa.JSON()))
+    for bot_uuid, bindings in connection.execute(sa.select(bots.c.uuid, bots.c.event_bindings)).all():
+        if isinstance(bindings, str):
+            bindings = json.loads(bindings)
+        if not isinstance(bindings, list):
+            continue
+        changed = False
+        for binding in bindings:
+            if (
+                isinstance(binding, dict)
+                and binding.get('target_type') == 'pipeline'
+                and binding.get('event_pattern') == 'message.*'
+            ):
+                binding['event_pattern'] = 'message.received'
+                changed = True
+        if changed:
+            connection.execute(bots.update().where(bots.c.uuid == bot_uuid).values(event_bindings=bindings))
+
+
+def downgrade():
+    # An exact incoming-message route remains valid with older versions.
+    pass
diff --git a/src/langbot/pkg/persistence/alembic/versions/0036_execution_host_leases.py b/src/langbot/pkg/persistence/alembic/versions/0036_execution_host_leases.py
new file mode 100644
index 000000000..120d77275
--- /dev/null
+++ b/src/langbot/pkg/persistence/alembic/versions/0036_execution_host_leases.py
@@ -0,0 +1,41 @@
+"""Track the liveness of hosts owning direct executions."""
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = '0036_execution_host_leases'
+down_revision = '0035_pipeline_received_only'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    inspector = sa.inspect(op.get_bind())
+    for table in ('agent_run', 'monitoring_messages'):
+        if not inspector.has_table(table):
+            continue
+        columns = {c['name'] for c in inspector.get_columns(table)}
+        indexes = {i['name'] for i in inspector.get_indexes(table)}
+        for column, column_type in (
+            ('execution_owner_id', sa.String(36)),
+            ('execution_lease_expires_at', sa.DateTime()),
+        ):
+            if column not in columns:
+                op.add_column(table, sa.Column(column, column_type, nullable=True))
+            index = f'ix_{table}_{column}'
+            if index not in indexes:
+                op.create_index(index, table, [column])
+
+
+def downgrade():
+    inspector = sa.inspect(op.get_bind())
+    for table in ('agent_run', 'monitoring_messages'):
+        if not inspector.has_table(table):
+            continue
+        columns = {c['name'] for c in inspector.get_columns(table)}
+        indexes = {i['name'] for i in inspector.get_indexes(table)}
+        for column in ('execution_owner_id', 'execution_lease_expires_at'):
+            if f'ix_{table}_{column}' in indexes:
+                op.drop_index(f'ix_{table}_{column}', table_name=table)
+            if column in columns:
+                op.drop_column(table, column)
diff --git a/src/langbot/pkg/persistence/alembic/versions/0037_metadata_text.py b/src/langbot/pkg/persistence/alembic/versions/0037_metadata_text.py
new file mode 100644
index 000000000..dd6922cb4
--- /dev/null
+++ b/src/langbot/pkg/persistence/alembic/versions/0037_metadata_text.py
@@ -0,0 +1,22 @@
+"""Allow structured wizard progress in metadata values."""
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = '0037_metadata_text'
+down_revision = '0036_execution_host_leases'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    inspector = sa.inspect(op.get_bind())
+    for table in ('metadata', 'workspace_metadata'):
+        if inspector.has_table(table):
+            with op.batch_alter_table(table) as batch:
+                batch.alter_column('value', existing_type=sa.String(255), type_=sa.Text(), existing_nullable=True)
+
+
+def downgrade():
+    # Keep the widened type: narrowing would destroy valid saved progress.
+    pass
diff --git a/src/langbot/pkg/persistence/alembic/versions/0038_llm_call_details.py b/src/langbot/pkg/persistence/alembic/versions/0038_llm_call_details.py
new file mode 100644
index 000000000..a91469f23
--- /dev/null
+++ b/src/langbot/pkg/persistence/alembic/versions/0038_llm_call_details.py
@@ -0,0 +1,25 @@
+"""Persist additive LLM timing, usage and structured failure diagnostics."""
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = '0038_llm_call_details'
+down_revision = '0037_metadata_text'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    inspector = sa.inspect(op.get_bind())
+    if inspector.has_table('monitoring_llm_calls'):
+        columns = {column['name'] for column in inspector.get_columns('monitoring_llm_calls')}
+        if 'details' not in columns:
+            op.add_column('monitoring_llm_calls', sa.Column('details', sa.JSON(), nullable=True))
+
+
+def downgrade():
+    inspector = sa.inspect(op.get_bind())
+    if inspector.has_table('monitoring_llm_calls'):
+        columns = {column['name'] for column in inspector.get_columns('monitoring_llm_calls')}
+        if 'details' in columns:
+            op.drop_column('monitoring_llm_calls', 'details')
diff --git a/src/langbot/pkg/persistence/databases/sqlite.py b/src/langbot/pkg/persistence/databases/sqlite.py
index ed096040f..8080d20d8 100644
--- a/src/langbot/pkg/persistence/databases/sqlite.py
+++ b/src/langbot/pkg/persistence/databases/sqlite.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import sqlalchemy
 import sqlalchemy.ext.asyncio as sqlalchemy_asyncio
 
 from .. import database
@@ -13,3 +14,8 @@ class SQLiteDatabaseManager(database.BaseDatabaseManager):
         db_file_path = self.ap.instance_config.data.get('database', {}).get('sqlite', {}).get('path', 'data/langbot.db')
         engine_url = f'sqlite+aiosqlite:///{db_file_path}'
         self.engine = sqlalchemy_asyncio.create_async_engine(engine_url)
+        # Reader transactions must not block independent plugin callback commits.
+        # Configure the persistent journal before serving requests, not per write.
+        async with self.engine.connect() as connection:
+            await connection.execute(sqlalchemy.text('PRAGMA journal_mode=WAL'))
+            await connection.commit()
diff --git a/src/langbot/pkg/persistence/mgr.py b/src/langbot/pkg/persistence/mgr.py
index d303d88d1..89a59eeba 100644
--- a/src/langbot/pkg/persistence/mgr.py
+++ b/src/langbot/pkg/persistence/mgr.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import datetime
 import enum
 import sqlite3
@@ -162,7 +160,6 @@ class PersistenceManager:
             default=None,
         )
 
-    @diagnostics.observe('lifecycle', 'persistence.initialize', source='startup', stage='execute')
     async def initialize(self):
         database_type = self.ap.instance_config.data.get('database', {}).get('use', 'sqlite')
         self.ap.logger.info(f'Initializing database type: {database_type}...')
@@ -1658,7 +1655,6 @@ class PersistenceManager:
 
     # =================================
 
-    @diagnostics.observe('lifecycle', 'persistence.run_alembic_migrations', source='startup', stage='execute')
     async def _run_alembic_migrations(self, target_revision: str = 'head'):
         """Run the supported Alembic-based 4.x migrations."""
         from . import alembic_runner
diff --git a/src/langbot/pkg/persistence/tenant_uow.py b/src/langbot/pkg/persistence/tenant_uow.py
index 0eacc85c5..396b20103 100644
--- a/src/langbot/pkg/persistence/tenant_uow.py
+++ b/src/langbot/pkg/persistence/tenant_uow.py
@@ -218,7 +218,7 @@ _ALLOWED_SCOPED_BUILTIN_FUNCTION_TYPES = {
     'now': sqlalchemy.sql.functions.now,
     'sum': sqlalchemy.sql.functions.sum,
 }
-_ALLOWED_SCOPED_GENERIC_FUNCTIONS = frozenset({'date_trunc', 'length', 'nullif', 'strftime'})
+_ALLOWED_SCOPED_GENERIC_FUNCTIONS = frozenset({'date_trunc', 'length', 'nullif', 'row_number', 'strftime'})
 _ALLOWED_SCOPED_CUSTOM_OPERATORS = frozenset({'<=>'})
 _ALLOWED_SCOPED_STATEMENT_TYPES = (
     sqlalchemy.sql.dml.UpdateBase,
@@ -483,9 +483,18 @@ def _validate_scoped_statement_call(args: tuple[typing.Any, ...], kwargs: dict[s
                 and getattr(getattr(source, 'table', None), 'name', None) == 'legacy_pipelines'
                 and source.name in {'config', 'stages', 'extensions_preferences'}
             )
-            if not pipeline_json_cas:
+            monitoring_metadata = (
+                type(element.type) is sqlalchemy.JSON
+                and isinstance(source, sqlalchemy.Column)
+                and type(source.type) is sqlalchemy.Text
+                and getattr(getattr(source, 'table', None), 'name', None) in {'event_log', 'agent_run'}
+                and (
+                    source.name == 'metadata_json' or (source.table.name == 'event_log' and source.name == 'input_json')
+                )
+            )
+            if not (pipeline_json_cas or monitoring_metadata):
                 raise ScopedSessionTransactionError(
-                    'TenantUnitOfWork only allows trusted pgvector and Pipeline JSON CAS casts'
+                    'TenantUnitOfWork only allows trusted pgvector, Pipeline JSON CAS and monitoring metadata casts'
                 )
 
         if isinstance(element, sqlalchemy.sql.functions.FunctionElement):
diff --git a/src/langbot/pkg/pipeline/aggregator.py b/src/langbot/pkg/pipeline/aggregator.py
index acf0a29f7..32ec748b6 100644
--- a/src/langbot/pkg/pipeline/aggregator.py
+++ b/src/langbot/pkg/pipeline/aggregator.py
@@ -162,6 +162,13 @@ class MessageAggregator:
         if variables:
             enabled = False
 
+        # Preserve ingress identity across the queue and detached debounce task.
+        from ..telemetry import trace
+
+        ingress_trace = trace.current()
+        if ingress_trace is not None:
+            variables = {**(variables or {}), '_ingress_event_id': ingress_trace.event_id}
+
         if not enabled:
             await self.ap.query_pool.add_query(
                 bot_uuid=bot_uuid,
diff --git a/src/langbot/pkg/pipeline/cntfilter/filters/baiduexamine.py b/src/langbot/pkg/pipeline/cntfilter/filters/baiduexamine.py
index 8c74150b0..444fd4798 100644
--- a/src/langbot/pkg/pipeline/cntfilter/filters/baiduexamine.py
+++ b/src/langbot/pkg/pipeline/cntfilter/filters/baiduexamine.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import pipeline_notice
+
 from .. import entities
 from .. import filter as filter_model
 import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
@@ -58,6 +60,10 @@ class BaiduCloudExamine(filter_model.ContentFilter):
                     return entities.FilterResult(
                         level=entities.ResultLevel.BLOCK,
                         replacement=message,
-                        user_notice='消息中存在不合适的内容, 请修改',
+                        user_notice=pipeline_notice(
+                            query,
+                            'content-filter-hint',
+                            'Your message contains inappropriate content. Please revise it.',
+                        ),
                         console_notice=f'百度云判定结果：{conclusion}',
                     )
diff --git a/src/langbot/pkg/pipeline/cntfilter/filters/banwords.py b/src/langbot/pkg/pipeline/cntfilter/filters/banwords.py
index bfd8ca772..32a3a5f30 100644
--- a/src/langbot/pkg/pipeline/cntfilter/filters/banwords.py
+++ b/src/langbot/pkg/pipeline/cntfilter/filters/banwords.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import pipeline_notice
+
 from .. import filter as filter_model
 from .. import entities
 import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
@@ -35,13 +37,19 @@ class BanWordFilter(filter_model.ContentFilter):
             return entities.FilterResult(
                 level=entities.ResultLevel.BLOCK,
                 replacement='',
-                user_notice='内容检查规则执行失败，请联系管理员',
+                user_notice=pipeline_notice(
+                    query, 'content-filter-error-hint', 'Content checking failed. Please contact the administrator.'
+                ),
                 console_notice=f'Sensitive-word regex rejected: {exc}',
             )
 
         return entities.FilterResult(
             level=entities.ResultLevel.MASKED if found else entities.ResultLevel.PASS,
             replacement=current,
-            user_notice='消息中存在不合适的内容, 请修改' if found else '',
+            user_notice=pipeline_notice(
+                query, 'content-filter-hint', 'Your message contains inappropriate content. Please revise it.'
+            )
+            if found
+            else '',
             console_notice='',
         )
diff --git a/src/langbot/pkg/pipeline/legacy_config_migration.py b/src/langbot/pkg/pipeline/legacy_config_migration.py
index a3df33af0..7c44d05c0 100644
--- a/src/langbot/pkg/pipeline/legacy_config_migration.py
+++ b/src/langbot/pkg/pipeline/legacy_config_migration.py
@@ -19,18 +19,21 @@ import re
 from urllib.parse import urlsplit
 from string import Formatter
 
-PLANNER_VERSION = '4'
+PLANNER_VERSION = '5'
 
+# Minimum reviewed plugin versions. The marketplace delists old releases, so the
+# installer falls back to the latest release when a pin is gone; the Host then
+# validates the emitted config against the installed runner schema.
 _TARGETS = {
-    'local-agent': ('LocalAgent', '0.1.10'),
-    'dify-service-api': ('DifyAgent', '0.1.10'),
-    'coze-api': ('CozeAgent', '0.1.10'),
-    'dashscope-app-api': ('DashScopeAgent', '0.1.10'),
-    'n8n-service-api': ('N8nAgent', '0.1.10'),
-    'langflow-api': ('LangflowAgent', '0.1.10'),
-    'deerflow-api': ('DeerFlowAgent', '0.1.10'),
-    'tbox-app-api': ('TboxAgent', '0.1.8'),
-    'weknora-api': ('WeKnoraAgent', '0.1.10'),
+    'local-agent': ('LocalAgent', '0.2.1'),
+    'dify-service-api': ('DifyAgent', '0.2.3'),
+    'coze-api': ('CozeAgent', '0.2.0'),
+    'dashscope-app-api': ('DashScopeAgent', '0.2.2'),
+    'n8n-service-api': ('N8nAgent', '0.2.0'),
+    'langflow-api': ('LangflowAgent', '0.2.2'),
+    'deerflow-api': ('DeerFlowAgent', '0.2.1'),
+    'tbox-app-api': ('TboxAgent', '0.2.1'),
+    'weknora-api': ('WeKnoraAgent', '0.2.1'),
 }
 _DEERFLOW_FIELDS = {
     'api-base',
@@ -192,6 +195,51 @@ _DEFAULTS = {
         'advanced-settings': False,
     },
 }
+# Missing values for these fields cannot be invented: they are credentials or
+# endpoints. Everything else is repaired from reviewed 4.10 defaults.
+_CRITICAL = {
+    'local-agent': ('model',),
+    'dify-service-api': ('base-url', 'api-key'),
+    'coze-api': ('api-key', 'bot-id'),
+    'dashscope-app-api': ('api-key', 'app-id'),
+    'n8n-service-api': ('webhook-url',),
+    'langflow-api': ('base-url', 'api-key', 'flow-id'),
+    'deerflow-api': ('api-base',),
+    'tbox-app-api': ('api-key', 'app-id'),
+    'weknora-api': ('base-url', 'api-key'),
+}
+_LOCAL_PROMPT = [{'role': 'system', 'content': 'You are a helpful assistant.'}]
+# 4.10 UI defaults for required, non-critical fields.
+_FILL = {
+    'local-agent': {'prompt': _LOCAL_PROMPT, 'box-session-id-template': '{launcher_type}_{launcher_id}'},
+    'dify-service-api': {
+        'app-type': 'chat',
+        'base-prompt': (
+            'When the file content is readable, please read the content of this file. '
+            'When the file is an image, describe the content of this image.'
+        ),
+    },
+    'coze-api': {'api-base': 'https://api.coze.cn', 'timeout': 120},
+    'dashscope-app-api': {'app-type': 'agent'},
+    'weknora-api': {'app-type': 'chat'},
+}
+_INT_TIMEOUT = {'deerflow-api', 'n8n-service-api', 'dify-service-api'}
+_TIMEOUT_DEFAULT = {
+    'dify-service-api': 30,
+    'coze-api': 120,
+    'dashscope-app-api': 120,
+    'n8n-service-api': 120,
+    'deerflow-api': 300,
+    'tbox-app-api': 120,
+    'weknora-api': 120,
+    'local-agent': 300,
+}
+_ALIASES = {
+    # legacy runner: (UI/hyphen key, alternate key, default when both missing)
+    'coze-api': [('auto-save-history', 'auto_save_history')],
+    'dashscope-app-api': [('references-quote', 'references_quote')],
+    'langflow-api': [('input-type', 'input_type'), ('output-type', 'output_type')],
+}
 _ASSET_RUNNERS = {'dify-service-api', 'coze-api', 'dashscope-app-api', 'n8n-service-api', 'langflow-api'}
 _REMOVE_THINK_RUNNERS = {'local-agent', 'dify-service-api', 'coze-api', 'dashscope-app-api', 'tbox-app-api'}
 _REASONING_LEVELS = {'provider_default', 'disabled', 'enabled', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'}
@@ -240,38 +288,6 @@ def _attachments(value):
     )
 
 
-def _validate_preferences(result, preferences, name):
-    if preferences is None:
-        return
-    if type(preferences) is not dict or not _strict_json(preferences):
-        _block(result, 'invalid_extension_preferences', 'extensions_preferences')
-        return
-    for field in (
-        'enable_all_plugins',
-        'enable_all_mcp_servers',
-        'enable_all_skills',
-        'mcp_resource_agent_read_enabled',
-    ):
-        if field in preferences and type(preferences[field]) is not bool:
-            _block(result, 'invalid_extension_preferences', f'extensions_preferences.{field}')
-    for field in ('mcp_servers', 'skills'):
-        if field in preferences and not _names(preferences[field]):
-            _block(result, 'invalid_extension_preferences', f'extensions_preferences.{field}')
-    if 'mcp_resources' in preferences and not _attachments(preferences['mcp_resources']):
-        _block(result, 'invalid_extension_preferences', 'extensions_preferences.mcp_resources')
-    plugins = preferences.get('plugins', [])
-    if type(plugins) is not list or not all(
-        type(item) is dict and all(type(item.get(k)) is str and item[k] for k in ('author', 'name')) for item in plugins
-    ):
-        _block(result, 'invalid_extension_preferences', 'extensions_preferences.plugins')
-    elif preferences.get('enable_all_plugins', True) is False and not any(
-        item['author'] == 'langbot-team' and item['name'] == name for item in plugins
-    ):
-        # The immutable result contract contains no extension-preference patch.
-        # Require explicit binding instead of silently flipping allow-all.
-        _block(result, 'extensions.runner_excluded', 'extensions_preferences.plugins')
-
-
 def _warn(result, code, field):
     diagnostic = {'code': code, 'field': field}
     if diagnostic not in result['warnings']:
@@ -321,42 +337,138 @@ def _valid_prompt(prompt):
     return type(prompt) is list and all(_message_shape(item, fields, ('role',)) for item in prompt)
 
 
-def _validate_local(result, section):
+_TRUE = {'true', '1', 'yes', 'on'}
+_FALSE = {'false', '0', 'no', 'off'}
+_CUSTOM_RUNNER_NAME = re.compile(r'[A-Za-z0-9_.\-]{1,64}')
+_DEFAULT_BOX_TEMPLATE = '{launcher_type}_{launcher_id}'
+_HEADER_NAME = re.compile(r"[!#$%&'*+.^_`|~0-9A-Za-z-]+")
+
+
+def _coerce(kind, value):
+    """Best-effort, lossless coercion. Return (ok, value)."""
+    if kind == 'bool':
+        if type(value) is bool:
+            return True, value
+        if type(value) is int and value in (0, 1):
+            return True, bool(value)
+        if type(value) is str and value.strip().lower() in _TRUE:
+            return True, True
+        if type(value) is str and value.strip().lower() in _FALSE:
+            return True, False
+        return False, None
+    if kind in ('int', 'number'):
+        if type(value) is int:
+            return True, value
+        number = value
+        if type(value) is str and value.strip():
+            text = value.strip()
+            try:
+                return True, int(text)
+            except ValueError:
+                pass
+            try:
+                number = float(text)
+            except ValueError:
+                return False, None
+        if type(number) is not float or not math.isfinite(number):
+            return False, None
+        if number.is_integer():
+            return True, int(number)
+        return (kind == 'number'), (number if kind == 'number' else None)
+    if kind == 'names':
+        if type(value) is str:
+            return True, ([value.strip()] if value.strip() else [])
+        if type(value) is list:
+            return True, [v.strip() for v in value if type(v) is str and v.strip()]
+        return False, None
+    if kind == 'attachments':
+        if type(value) is list:
+            return True, [v for v in value if type(v) is dict and ('enabled' not in v or type(v['enabled']) is bool)]
+        return False, None
+    if type(value) is str:
+        return True, value
+    if type(value) in (int, float):
+        return True, str(value)
+    return False, None
+
+
+def _kind(legacy, field):
+    if field in _BOOLEANS:
+        return 'bool'
+    if field in _INTEGERS:
+        return 'int'
+    if field == 'timeout':
+        return 'int' if legacy in _INT_TIMEOUT else 'number'
+    if field in _NAME_LISTS:
+        return 'names'
+    if field == 'mcp-resources':
+        return 'attachments'
+    return 'str'
+
+
+def _normalize_local(result, section):
+    """Repair the LocalAgent model/prompt/template; never block on the model."""
     prefix = 'ai.local-agent'
     model = section.get('model')
     if type(model) is dict:
         if set(model) - {'primary', 'fallbacks', 'reasoning'}:
-            _block(result, 'unknown_field', f'{prefix}.model')
-        if type(model.get('primary')) is not str or not model['primary'] or not _names(model.get('fallbacks', [])):
-            _block(result, 'invalid_type', f'{prefix}.model')
+            _warn(result, 'migration.unknown_field_dropped', f'{prefix}.model')
+        primary = model.get('primary')
+        if type(primary) is not str:
+            _warn(result, 'migration.missing_default_filled', f'{prefix}.model')
+            primary = ''
+        raw_fallbacks = model.get('fallbacks', [])
+        ok, fallbacks = _coerce('names', raw_fallbacks)
+        if not ok or fallbacks != raw_fallbacks:
+            _warn(result, 'migration.value_coerced', f'{prefix}.model.fallbacks')
         reasoning = model.get('reasoning', {})
-        if type(reasoning) is not dict or not all(type(v) is str for v in reasoning.values()):
-            _block(result, 'invalid_type', f'{prefix}.model.reasoning')
-        elif any(value not in _REASONING_LEVELS for value in reasoning.values()):
-            _block(result, 'local.reasoning_value', f'{prefix}.model')
-        elif reasoning:
+        if type(reasoning) is not dict:
+            _warn(result, 'local.reasoning_dropped', f'{prefix}.model.reasoning')
+            reasoning = {}
+        valid = {k: v for k, v in reasoning.items() if type(v) is str and v in _REASONING_LEVELS}
+        if len(valid) != len(reasoning):
+            _warn(result, 'local.reasoning_dropped', f'{prefix}.model.reasoning')
+        if valid:
             _warn(result, 'local.model_reasoning', f'{prefix}.model.reasoning')
-    elif type(model) is not str or not model:
-        _block(result, 'invalid_type', f'{prefix}.model')
-    if not _valid_prompt(section.get('prompt')):
-        _block(result, 'local.prompt_shape', f'{prefix}.prompt')
+        section['model'] = {'primary': primary, 'fallbacks': fallbacks or [], 'reasoning': valid}
+    elif type(model) is str and model.strip():
+        section['model'] = {'primary': model, 'fallbacks': [], 'reasoning': {}}
+    else:
+        # No usable model: migrate anyway; the user picks one after migration.
+        _warn(result, 'migration.missing_default_filled', f'{prefix}.model')
+        section['model'] = {'primary': '', 'fallbacks': [], 'reasoning': {}}
+    prompt = section.get('prompt')
+    if prompt is None:
+        section['prompt'] = copy.deepcopy(_LOCAL_PROMPT)
+        _warn(result, 'migration.missing_default_filled', f'{prefix}.prompt')
+    elif type(prompt) is str:
+        section['prompt'] = [{'role': 'system', 'content': prompt}]
+        _warn(result, 'local.prompt_normalized', f'{prefix}.prompt')
+    elif not _valid_prompt(prompt):
+        kept = []
+        if type(prompt) is list:
+            for item in prompt:
+                if type(item) is dict and type(item.get('role')) is str and type(item.get('content')) is str:
+                    kept.append({'role': item['role'], 'content': item['content']})
+        section['prompt'] = kept if kept and _valid_prompt(kept) else copy.deepcopy(_LOCAL_PROMPT)
+        _warn(result, 'local.prompt_normalized', f'{prefix}.prompt')
     _warn(result, 'local.context_defaults', f'{prefix}.max-round')
     _warn(result, 'local.serial_tools_preserved', f'{prefix}.tools')
     _warn(result, 'local.retrieval_defaults', f'{prefix}.knowledge-bases')
-    template = section.get('box-session-id-template', '{launcher_type}_{launcher_id}')
-    if type(template) is not str or not template.strip():
-        _block(result, 'local.box_template_invalid', f'{prefix}.box-session-id-template')
-    else:
+    template = section.get('box-session-id-template', _DEFAULT_BOX_TEMPLATE)
+    valid_template = type(template) is str and bool(template.strip())
+    if valid_template:
         try:
-            fields = list(Formatter().parse(template))
-            if any(
+            valid_template = not any(
                 name is not None
                 and (not name or name.isdecimal() or re.search(r'[.\[\]{}]', name) or spec or conversion)
-                for _, name, spec, conversion in fields
-            ):
-                raise ValueError('Only named interpolation variables are supported')
+                for _, name, spec, conversion in Formatter().parse(template)
+            )
         except ValueError:
-            _block(result, 'local.box_template_invalid', f'{prefix}.box-session-id-template')
+            valid_template = False
+    if not valid_template:
+        section['box-session-id-template'] = _DEFAULT_BOX_TEMPLATE
+        _warn(result, 'local.box_template_defaulted', f'{prefix}.box-session-id-template')
     _warn(result, 'local.box_state_reset', f'{prefix}.box-session-id-template')
 
 
@@ -380,36 +492,28 @@ def _parse_tweaks(raw):
     return decoded
 
 
-def _validate_external(result, legacy, section):
+def _default_for(legacy, field):
+    for source in (_FILL.get(legacy, {}), _DEFAULTS[legacy]):
+        if field in source:
+            return True, copy.deepcopy(source[field])
+    return False, None
+
+
+def _normalize_external(result, legacy, section):
     prefix = f'ai.{legacy}'
     app_types = {
         'dify-service-api': ('chat', 'agent', 'workflow', 'chatflow'),
         'dashscope-app-api': ('agent', 'workflow'),
         'weknora-api': ('chat', 'agent'),
     }
-    required_strings = {
-        'dify-service-api': ('api-key', 'base-url'),
-        'coze-api': ('api-key', 'bot-id'),
-        'dashscope-app-api': ('api-key', 'app-id'),
-        'n8n-service-api': ('webhook-url',),
-        'langflow-api': ('base-url', 'api-key', 'flow-id'),
-        'tbox-app-api': ('api-key', 'app-id'),
-        'weknora-api': ('api-key', 'base-url'),
-    }
-    for field in required_strings.get(legacy, ()):
+    for field in _CRITICAL[legacy]:
         value = section.get(field)
-        if type(value) is str and (not value or (legacy == 'weknora-api' and not value.strip())):
+        if type(value) is str and not value.strip():
             _block(result, 'invalid_value', f'{prefix}.{field}')
-    if legacy in app_types and 'app-type' in section and section['app-type'] not in app_types[legacy]:
-        _block(result, 'invalid_value', f'{prefix}.app-type')
+    if legacy in app_types and section.get('app-type') not in app_types[legacy]:
+        section['app-type'] = _FILL[legacy]['app-type']
+        _warn(result, 'migration.invalid_value_defaulted', f'{prefix}.app-type')
     if legacy == 'coze-api':
-        actual, alias = section.get('auto_save_history'), section.get('auto-save-history')
-        if type(actual) is bool and type(alias) is bool and actual != alias:
-            _block(result, 'coze.history_alias', f'{prefix}.auto_save_history')
-        if 'auto_save_history' in section or 'auto-save-history' in section:
-            _warn(result, 'migration.alias_repaired', f'{prefix}.auto_save_history')
-        if actual is None and alias is None:
-            _warn(result, 'migration.null_default', f'{prefix}.auto_save_history')
         base = section.get('api-base')
         try:
             parsed = urlsplit(base) if type(base) is str else None
@@ -428,48 +532,35 @@ def _validate_external(result, legacy, section):
         except ValueError:
             _block(result, 'coze.custom_endpoint', f'{prefix}.api-base')
         _warn(result, 'coze.persistent_history', prefix)
-    elif legacy == 'dashscope-app-api':
-        if 'references-quote' in section:
-            if 'references_quote' in section and section['references_quote'] != section['references-quote']:
-                _block(result, 'dashscope.references_alias', f'{prefix}.references_quote')
-            else:
-                _warn(result, 'migration.alias_repaired', f'{prefix}.references-quote')
     elif legacy == 'langflow-api':
-        for axis in ('input', 'output'):
-            actual, ui = f'{axis}_type', f'{axis}-type'
-            if ui in section and section[ui] != section.get(actual, 'chat'):
-                _block(result, 'langflow.io_alias', f'{prefix}.{actual}')
-            if actual in section:
-                _warn(result, 'migration.alias_repaired', f'{prefix}.{actual}')
-        try:
-            _parse_tweaks(section.get('tweaks'))
-        except (ValueError, RecursionError):
-            _block(result, 'langflow.invalid_tweaks', f'{prefix}.tweaks')
-        if section.get('tweaks') is None or (
-            type(section.get('tweaks')) is str
-            and (not section['tweaks'].strip() or section['tweaks'].strip() == 'null')
-        ):
+        raw = section.get('tweaks')
+        if raw is None or (type(raw) is str and (not raw.strip() or raw.strip() == 'null')):
             _warn(result, 'langflow.tweaks_default', f'{prefix}.tweaks')
+        try:
+            section['tweaks'] = copy.deepcopy(_parse_tweaks(raw))
+        except (ValueError, RecursionError):
+            section['tweaks'] = {}
+            _warn(result, 'langflow.tweaks_invalid_defaulted', f'{prefix}.tweaks')
         _warn(result, 'langflow.persistent_history', prefix)
     elif legacy == 'dify-service-api':
         _warn(result, 'dify.timeout_default', f'{prefix}.timeout')
     elif legacy == 'n8n-service-api':
-        mode = section.get('auth-type', 'none')
         auth_fields = {
             'none': (),
             'basic': ('basic-username', 'basic-password'),
             'jwt': ('jwt-secret',),
             'header': ('header-name', 'header-value'),
         }
-        if type(mode) is not str or mode not in auth_fields:
-            _block(result, 'invalid_value', f'{prefix}.auth-type')
-        else:
-            for field in auth_fields[mode]:
-                if field not in section:
-                    _block(result, 'missing_field', f'{prefix}.{field}')
-        response = section.get('response-handling', 'reply')
-        if response not in ('reply', 'ignore'):
-            _block(result, 'invalid_value', f'{prefix}.response-handling')
+        mode = section.get('auth-type', 'none')
+        if mode not in auth_fields:
+            mode = section['auth-type'] = 'none'
+            _warn(result, 'migration.invalid_value_defaulted', f'{prefix}.auth-type')
+        for field in auth_fields[mode]:
+            if field not in section:
+                _block(result, 'missing_field', f'{prefix}.{field}')
+        if section.get('response-handling', 'reply') not in ('reply', 'ignore'):
+            section['response-handling'] = 'reply'
+            _warn(result, 'migration.invalid_value_defaulted', f'{prefix}.response-handling')
         if mode == 'basic':
             for field in ('basic-username', 'basic-password'):
                 value = section.get(field)
@@ -481,75 +572,133 @@ def _validate_external(result, legacy, section):
                     except (UnicodeEncodeError, ValueError):
                         _block(result, 'n8n.basic_encoding', f'{prefix}.{field}')
         if mode == 'header' and type(section.get('header-name')) is str:
-            if not re.fullmatch(r"[!#$%&'*+.^_`|~0-9A-Za-z-]+", section['header-name']):
+            if not _HEADER_NAME.fullmatch(section['header-name']):
                 _block(result, 'n8n.header_name', f'{prefix}.header-name')
         _warn(result, 'n8n.session_ids_reset', prefix)
-    elif legacy == 'weknora-api':
-        knowledge = section.get('knowledge-base-ids')
-        if type(knowledge) is list and any(type(v) is str and not v.strip() for v in knowledge):
-            _block(result, 'invalid_value', f'{prefix}.knowledge-base-ids')
-    if legacy in ('coze-api', 'n8n-service-api', 'weknora-api'):
-        timeout = section.get('timeout', 120)
-        if type(timeout) in (int, float) and timeout <= 0:
-            _block(result, 'invalid_value', f'{prefix}.timeout')
-
-
-def _validate_section(result, legacy, section):
-    prefix = f'ai.{legacy}'
-    if set(section) - _FIELDS[legacy]:
-        _block(result, 'unknown_field', prefix)
-    for field in _REQUIRED[legacy]:
-        if field not in section:
-            _block(result, 'missing_field', f'{prefix}.{field}')
-    # Iterate only known keys, never put an arbitrary saved key in diagnostics.
-    for field in sorted(_FIELDS[legacy] & section.keys()):
-        value = section[field]
-        if field in ('model', 'prompt', 'tweaks') or (legacy == 'dify-service-api' and field == 'timeout'):
-            continue
-        if value is None and (
-            (legacy == 'coze-api' and field in ('auto_save_history', 'auto-save-history'))
-            or (legacy == 'weknora-api' and field in ('agent-id', 'knowledge-base-ids'))
-        ):
-            continue
-        if field in _BOOLEANS:
-            valid = type(value) is bool
-        elif field in _INTEGERS or (field == 'timeout' and legacy == 'deerflow-api'):
-            valid = type(value) is int
-        elif field == 'timeout':
-            valid = type(value) in (int, float)
-        elif field in _NAME_LISTS:
-            valid = _names(value)
-        elif field == 'mcp-resources':
-            valid = _attachments(value)
-        else:
-            valid = type(value) is str
-        if not valid:
-            _block(result, 'invalid_type', f'{prefix}.{field}')
-    if legacy == 'local-agent':
-        _validate_local(result, section)
-    else:
-        _validate_external(result, legacy, section)
-    if legacy == 'deerflow-api':
+    elif legacy == 'deerflow-api':
         base = section.get('api-base')
         if type(base) is str and not base.strip().startswith(('http://', 'https://')):
             _block(result, 'invalid_value', f'{prefix}.api-base')
 
 
-def _validate_output(result, config, legacy):
+def _normalize_section(result, legacy, section):
+    """Return a repaired copy of a legacy runner section (value-free diagnostics)."""
+    prefix = f'ai.{legacy}'
+    critical = _CRITICAL[legacy]
+    if set(section) - _FIELDS[legacy]:
+        _warn(result, 'migration.unknown_field_dropped', prefix)
+    # Iterate only known keys, never put an arbitrary saved key in diagnostics.
+    known = {k: copy.deepcopy(section[k]) for k in sorted(_FIELDS[legacy] & section.keys())}
+    for preferred, other in _ALIASES.get(legacy, []):
+        # The UI (hyphen) key wins; dashscope's plugin schema keeps the underscore.
+        output = 'references_quote' if legacy == 'dashscope-app-api' else preferred
+        drop = other if output == preferred else preferred
+        if drop not in known:
+            continue
+        a, b = known.get(preferred), known.get(other)
+        if a is not None and b is not None and a != b:
+            _warn(result, 'migration.alias_conflict_resolved', f'{prefix}.{drop}')
+        else:
+            _warn(result, 'migration.alias_repaired', f'{prefix}.{drop}')
+        known.pop(drop)
+        known[output] = a if a is not None else b
+    if legacy == 'coze-api' and known.get('auto-save-history') is None:
+        known['auto-save-history'] = True
+        _warn(result, 'migration.null_default', f'{prefix}.auto_save_history')
+    for field in _REQUIRED[legacy]:
+        if field not in known and field in critical:
+            _block(result, 'missing_field', f'{prefix}.{field}')
+    for field in sorted(known):
+        value = known[field]
+        if field in ('model', 'prompt', 'tweaks'):
+            continue
+        if value is None and legacy == 'weknora-api' and field in ('agent-id', 'knowledge-base-ids'):
+            continue
+        ok, coerced = (False, None) if value is None else _coerce(_kind(legacy, field), value)
+        if ok and field == 'timeout' and coerced <= 0:
+            ok = False
+        if ok:
+            if coerced != value or type(coerced) is not type(value):
+                _warn(result, 'migration.value_coerced', f'{prefix}.{field}')
+            known[field] = coerced
+        elif field in critical:
+            _block(result, 'invalid_type', f'{prefix}.{field}')
+        else:
+            known.pop(field)
+            _warn(result, 'migration.invalid_value_defaulted', f'{prefix}.{field}')
+    for field in _REQUIRED[legacy]:
+        if field not in known and field not in critical:
+            has_default, default = _default_for(legacy, field)
+            if has_default:
+                known[field] = default
+                if field != 'prompt':
+                    _warn(result, 'migration.missing_default_filled', f'{prefix}.{field}')
+    if legacy == 'local-agent':
+        _normalize_local(result, known)
+    else:
+        _normalize_external(result, legacy, known)
+    return known
+
+
+def _normalize_preferences(result, preferences, name):
+    """Ignore malformed preference fields; return a patch only to bind the runner."""
+    if preferences is None:
+        return None
+    if type(preferences) is not dict or not _strict_json(preferences):
+        _warn(result, 'invalid_extension_preferences', 'extensions_preferences')
+        return None
+    for field in (
+        'enable_all_plugins',
+        'enable_all_mcp_servers',
+        'enable_all_skills',
+        'mcp_resource_agent_read_enabled',
+    ):
+        if field in preferences and type(preferences[field]) is not bool:
+            _warn(result, 'invalid_extension_preferences', f'extensions_preferences.{field}')
+    for field in ('mcp_servers', 'skills'):
+        if field in preferences and not _names(preferences[field]):
+            _warn(result, 'invalid_extension_preferences', f'extensions_preferences.{field}')
+    if 'mcp_resources' in preferences and not _attachments(preferences['mcp_resources']):
+        _warn(result, 'invalid_extension_preferences', 'extensions_preferences.mcp_resources')
+    plugins = preferences.get('plugins', [])
+
+    def valid(item):
+        return type(item) is dict and all(type(item.get(k)) is str and item[k] for k in ('author', 'name'))
+
+    kept = [item for item in plugins if valid(item)] if type(plugins) is list else []
+    if type(plugins) is not list or len(kept) != len(plugins):
+        _warn(result, 'invalid_extension_preferences', 'extensions_preferences.plugins')
+    if preferences.get('enable_all_plugins', True) is False and not any(
+        item['author'] == 'langbot-team' and item['name'] == name for item in kept
+    ):
+        # Bind exactly the migrated runner plugin; never flip allow-all.
+        patched = copy.deepcopy(preferences)
+        patched['plugins'] = [*copy.deepcopy(kept), {'author': 'langbot-team', 'name': name}]
+        _warn(result, 'extensions.runner_added', 'extensions_preferences.plugins')
+        return patched
+    return None
+
+
+def _normalize_output(result, config):
     output = config.get('output', {})
     if type(output) is not dict:
-        _block(result, 'invalid_type', 'output')
-        return
+        _warn(result, 'migration.invalid_value_defaulted', 'output')
+        return False
     misc = output.get('misc', {})
     if type(misc) is not dict:
-        _block(result, 'invalid_type', 'output.misc')
-        return
+        _warn(result, 'migration.invalid_value_defaulted', 'output.misc')
+        return False
     value = misc.get('remove-think', False)
+    ok, coerced = _coerce('bool', value)
+    if not ok:
+        _warn(result, 'migration.invalid_value_defaulted', 'output.misc.remove-think')
+        return False
     if type(value) is not bool:
-        _block(result, 'invalid_type', 'output.misc.remove-think')
+        _warn(result, 'migration.value_coerced', 'output.misc.remove-think')
+    return coerced
 
 
-def _assemble(result, legacy, section, config, preferences):
+def _assemble(result, legacy, section, remove_think, preferences):
     selected = {**copy.deepcopy(_DEFAULTS[legacy]), **copy.deepcopy(section)}
     prefix = f'ai.{legacy}'
     if legacy in _IDENTITY_SOURCES:
@@ -557,18 +706,18 @@ def _assemble(result, legacy, section, config, preferences):
     if legacy == 'local-agent':
         selected.pop('max-round', None)
         singular = selected.pop('knowledge-base', '')
-        if not selected['knowledge-bases'] and singular and singular != '__none__':
+        if not selected['knowledge-bases'] and type(singular) is str and singular and singular != '__none__':
             selected['knowledge-bases'] = [singular]
             _warn(result, 'migration.alias_repaired', f'{prefix}.knowledge-base')
-        model = selected['model']
-        selected['model'] = (
-            {'primary': model, 'fallbacks': [], 'reasoning': {}}
-            if type(model) is str
-            else {'fallbacks': [], 'reasoning': {}, **model}
+        preferences = preferences if type(preferences) is dict else {}
+        resources = preferences.get('mcp_resources', [])
+        read = preferences.get('mcp_resource_agent_read_enabled', True)
+        selected['mcp-resources'] = copy.deepcopy(
+            section.get('mcp-resources', resources if _attachments(resources) else [])
+        )
+        selected['mcp-resource-agent-read-enabled'] = section.get(
+            'mcp-resource-agent-read-enabled', read if type(read) is bool else True
         )
-        preferences = preferences or {}
-        for field, default in (('mcp-resources', []), ('mcp-resource-agent-read-enabled', True)):
-            selected[field] = copy.deepcopy(section.get(field, preferences.get(field.replace('-', '_'), default)))
     else:
         selected.update(
             {
@@ -582,26 +731,15 @@ def _assemble(result, legacy, section, config, preferences):
     if legacy in _ASSET_RUNNERS:
         selected['langbot-assets-enabled'] = False
     if legacy in _REMOVE_THINK_RUNNERS:
-        selected['remove-think'] = config.get('output', {}).get('misc', {}).get('remove-think', False)
+        selected['remove-think'] = remove_think
         _warn(result, 'migration.output_policy_copied', 'output.misc.remove-think')
     if legacy == 'dify-service-api':
         selected['timeout'] = 30
-    elif legacy == 'coze-api':
-        actual = selected.pop('auto_save_history', None)
-        selected['auto-save-history'] = actual if type(actual) is bool else section.get('auto-save-history')
-        if selected['auto-save-history'] is None:
-            selected['auto-save-history'] = True
-    elif legacy == 'dashscope-app-api':
-        if 'references-quote' in selected:
-            selected['references_quote'] = selected.pop('references-quote')
-    elif legacy == 'langflow-api':
-        for axis in ('input', 'output'):
-            selected[f'{axis}-type'] = selected.pop(f'{axis}_type', selected[f'{axis}-type'])
-        selected['tweaks'] = copy.deepcopy(_parse_tweaks(section.get('tweaks')))
     elif legacy == 'weknora-api':
-        selected.setdefault(
-            'agent-id', 'builtin-quick-answer' if selected['app-type'] == 'chat' else 'builtin-smart-reasoning'
-        )
+        if selected.get('agent-id') is None:
+            selected['agent-id'] = (
+                'builtin-quick-answer' if selected['app-type'] == 'chat' else 'builtin-smart-reasoning'
+            )
         if selected['knowledge-base-ids'] is None:
             selected['knowledge-base-ids'] = []
         _warn(result, 'weknora.session_title_changed', prefix)
@@ -616,6 +754,7 @@ def plan_legacy_pipeline(config, extensions_preferences=None) -> dict:
         'target_runner_id': None,
         'target_plugin': None,
         'config': None,
+        'extensions_preferences': None,
         'changed_paths': [],
         'blockers': [],
         'warnings': [],
@@ -630,26 +769,43 @@ def plan_legacy_pipeline(config, extensions_preferences=None) -> dict:
     selection = ai.get('runner', {})
     if type(selection) is not dict:
         return _block(result, 'invalid_type', 'ai.runner')
-    if 'id' in selection and 'runner' in selection:
-        return _block(result, 'mixed_runner_selection', 'ai.runner')
+    legacy = selection.get('runner')
+    known_legacy = type(legacy) is str and legacy in _TARGETS
     if 'id' in selection:
         current = selection['id']
-        # A blank runner id paired with no legacy runner section means the saved
-        # pipeline simply has no runner selected. There is nothing to migrate and
-        # nothing to convert, so report not_legacy instead of failing the whole
-        # batch with a malformed-id blocker.
-        if type(current) is str and not current.strip() and not any(legacy in ai for legacy in _TARGETS):
+        valid_id = type(current) is str and bool(re.fullmatch(r'plugin:[^/\s]+/[^/\s]+/[^/\s]+', current))
+        if 'runner' in selection:
+            # Prefer a valid current id; otherwise migrate the legacy selection.
+            if valid_id:
+                _warn(result, 'migration.mixed_runner_selection', 'ai.runner')
+                result['state'] = 'already_current'
+                return result
+            if not known_legacy:
+                return _block(result, 'mixed_runner_selection', 'ai.runner')
+            _warn(result, 'migration.mixed_runner_selection', 'ai.runner')
+        else:
+            # A blank runner id paired with no legacy runner section means the
+            # saved pipeline simply has no runner selected: nothing to migrate.
+            if type(current) is str and not current.strip() and not any(name in ai for name in _TARGETS):
+                return result
+            if not valid_id:
+                return _block(result, 'invalid_runner_id', 'ai.runner.id')
+            result['state'] = 'already_current'
             return result
-        if type(current) is not str or not re.fullmatch(r'plugin:[^/\s]+/[^/\s]+/[^/\s]+', current):
-            return _block(result, 'invalid_runner_id', 'ai.runner.id')
-        result['state'] = 'already_current'
-        return result
     if 'runner' not in selection:
         return result
-    legacy = selection['runner']
     if type(legacy) is not str:
         return _block(result, 'invalid_type', 'ai.runner.runner')
-    if legacy not in _TARGETS:
+    if not known_legacy:
+        if not legacy.strip():
+            return result
+        # 4.10 allowed in-tree custom runners (@runner_class). They have no
+        # plugin equivalent; report them without failing other pipelines.
+        result.update(
+            state='unsupported_runner',
+            legacy_runner=legacy if _CUSTOM_RUNNER_NAME.fullmatch(legacy) else None,
+        )
+        _warn(result, 'legacy.custom_runner_unsupported', 'ai.runner.runner')
         return result
     name, version = _TARGETS[legacy]
     target = f'plugin:langbot-team/{name}/default'
@@ -658,27 +814,27 @@ def plan_legacy_pipeline(config, extensions_preferences=None) -> dict:
         target_runner_id=target,
         target_plugin={'author': 'langbot-team', 'name': name, 'version': version},
     )
-    if 'runner_config' in ai:
-        if type(ai['runner_config']) is not dict:
-            return _block(result, 'invalid_type', 'ai.runner_config')
-        if ai['runner_config']:
-            return _block(result, 'mixed_runner_config', 'ai.runner_config')
-    if 'expire-time' in selection:
-        expiry = selection['expire-time']
-        if type(expiry) is not int or expiry < 0:
-            _block(result, 'invalid_expiry', 'ai.runner.expire-time')
+    if 'runner_config' in ai and (type(ai['runner_config']) is not dict or ai['runner_config']):
+        _warn(result, 'migration.runner_config_replaced', 'ai.runner_config')
+    expiry = selection.get('expire-time', 0)
+    ok, expiry_value = _coerce('int', expiry)
+    if not ok or expiry_value < 0:
+        expiry_value = 0
+        _warn(result, 'migration.invalid_value_defaulted', 'ai.runner.expire-time')
+    elif type(expiry) is not int:
+        _warn(result, 'migration.value_coerced', 'ai.runner.expire-time')
     if set(ai) - set(_TARGETS) - {'runner', 'runner_config'}:
-        _block(result, 'unknown_field', 'ai')
-    if set(selection) - {'runner', 'expire-time'}:
-        _block(result, 'unknown_field', 'ai.runner')
+        _warn(result, 'migration.unknown_section_archived', 'ai')
+    if set(selection) - {'runner', 'expire-time', 'id'}:
+        _warn(result, 'migration.unknown_field_dropped', 'ai.runner')
     if legacy not in ai:
         return _block(result, 'missing_field', f'ai.{legacy}')
     section = ai[legacy]
     if type(section) is not dict:
         return _block(result, 'invalid_type', f'ai.{legacy}')
-    _validate_section(result, legacy, section)
-    _validate_preferences(result, extensions_preferences, name)
-    _validate_output(result, config, legacy)
+    section = _normalize_section(result, legacy, section)
+    patched_preferences = _normalize_preferences(result, extensions_preferences, name)
+    remove_think = _normalize_output(result, config)
     _warn(result, 'external.state_validation_required', 'ai.runner')
     _warn(result, 'migration.history_reset', 'ai.runner')
     _warn(result, 'migration.new_defaults', f'ai.{legacy}')
@@ -689,22 +845,22 @@ def plan_legacy_pipeline(config, extensions_preferences=None) -> dict:
         _warn(result, 'migration.identity_preserved', f'ai.{legacy}')
     if result['blockers']:
         return result
+    selected = _assemble(result, legacy, section, remove_think, extensions_preferences)
     candidate = copy.deepcopy(config)
-    candidate['ai']['runner'].pop('runner')
-    candidate['ai']['runner']['id'] = target
-    selected = _assemble(result, legacy, section, config, extensions_preferences)
-    candidate['ai']['runner_config'] = {target: selected}
-    # The durable transaction stores the full source, including inactive legacy
-    # credentials. The active configuration must be canonical so the guarded
-    # editor never needs to interpret mixed legacy/current containers.
-    archived_paths = []
-    for old_runner in _TARGETS:
-        if old_runner in candidate['ai']:
-            candidate['ai'].pop(old_runner)
-            archived_paths.append(f'ai.{old_runner}')
+    # The durable snapshot stores the full source, including inactive legacy
+    # credentials and unknown leftovers. The active configuration is canonical.
+    candidate['ai'] = {
+        'runner': {'id': target, 'expire-time': expiry_value},
+        'runner_config': {target: selected},
+    }
+    archived_paths = [f'ai.{old}' for old in _TARGETS if old in ai]
+    changed = ['ai.runner.runner', 'ai.runner.id', 'ai.runner_config', *archived_paths]
+    if patched_preferences is not None:
+        changed.append('extensions_preferences.plugins')
     result.update(
         state='ready',
         config=candidate,
-        changed_paths=['ai.runner.runner', 'ai.runner.id', 'ai.runner_config', *archived_paths],
+        extensions_preferences=patched_preferences,
+        changed_paths=changed,
     )
     return result
diff --git a/src/langbot/pkg/pipeline/monitoring_helper.py b/src/langbot/pkg/pipeline/monitoring_helper.py
index 15ae38f47..4efb21498 100644
--- a/src/langbot/pkg/pipeline/monitoring_helper.py
+++ b/src/langbot/pkg/pipeline/monitoring_helper.py
@@ -18,6 +18,103 @@ if typing.TYPE_CHECKING:
 from .pool import get_query_execution_context
 
 
+async def resolve_processor_name(
+    ap: app.Application,
+    workspace_uuid: str,
+    processor_type: str,
+    processor_id: str | None,
+    runner_id: str = '',
+) -> str:
+    """Display name of the processor kind that handled a run.
+
+    Monitoring attributed every call to a Pipeline, so an Agent or a plugin event
+    processor recorded "Unknown" for the entity that actually ran. Each kind is
+    resolved from the table that owns it, falling back to the identity the
+    binding carried.
+    """
+    import sqlalchemy
+
+    from ..entity.persistence import agent as persistence_agent
+    from ..entity.persistence import pipeline as persistence_pipeline
+
+    identifier = str(processor_id or '').strip()
+    if identifier:
+        try:
+            model = persistence_pipeline.LegacyPipeline if processor_type == 'pipeline' else persistence_agent.Agent
+            result = await ap.persistence_mgr.execute_async(
+                sqlalchemy.select(model.name).where(
+                    sqlalchemy.and_(
+                        model.uuid == identifier,
+                        model.workspace_uuid == workspace_uuid,
+                    )
+                )
+            )
+            name = str(result.scalar() or '').strip()
+            if name:
+                return name
+        except Exception as e:
+            ap.logger.warning(f'Failed to resolve monitoring processor name: {e}')
+    # A deleted Agent or an unregistered plugin processor still names the
+    # component that ran, never a bare "Unknown".
+    return identifier or runner_id or processor_type
+
+
+async def prepare_monitoring_identity(
+    ap: app.Application,
+    query: pipeline_query.Query,
+    *,
+    processor_type: str,
+    processor_id: str | None,
+    runner_id: str = '',
+    execution_record_id: str = '',
+) -> None:
+    """Pin the monitoring identity of a runner-owned execution.
+
+    A Pipeline pins these variables when it starts a query; a Runner started by
+    an Agent or a plugin event processor has to pin them itself, otherwise every
+    LLM/tool call it makes is attributed to "Unknown". An enclosing Pipeline
+    already named the run, so its label is never overwritten.
+
+    ``_monitoring_message_id`` is the execution record the calls belong to: the
+    Pipeline's message, or the Agent run's id. Both are the id the executions
+    list shows, which is what lets a call jump back to its record.
+    """
+    try:
+        variables = getattr(query, 'variables', None)
+        if not isinstance(variables, dict):
+            return
+        if variables.get('_monitoring_pipeline_name'):
+            return
+
+        bot_name = 'WebChat'
+        bot_uuid = str(getattr(query, 'bot_uuid', '') or '')
+        workspace_uuid = str(getattr(query, 'workspace_uuid', '') or '')
+        if bot_uuid and workspace_uuid:
+            try:
+                bot = await ap.bot_service.get_bot(
+                    workspace_uuid,
+                    bot_uuid,
+                    include_secret=False,
+                )
+                if bot:
+                    bot_name = bot.get('name', bot_name)
+            except Exception as e:
+                ap.logger.warning(f'Failed to resolve monitoring bot name: {e}')
+
+        variables['_monitoring_bot_name'] = bot_name
+        variables['_monitoring_pipeline_name'] = await resolve_processor_name(
+            ap,
+            workspace_uuid,
+            processor_type,
+            processor_id,
+            runner_id,
+        )
+        if execution_record_id:
+            variables['_monitoring_message_id'] = execution_record_id
+    except Exception as e:
+        ap.logger.error(f'Failed to prepare monitoring identity: {e}')
+
+
 class MonitoringHelper:
     """Helper class for monitoring operations"""
 
@@ -199,6 +296,7 @@ class MonitoringHelper:
                 user_name=sender_name,
                 runner_name=runner_name,
                 role='assistant',
+                parent_message_id=(query.variables or {}).get('_monitoring_message_id'),
             )
         except Exception as e:
             ap.logger.error(f'Failed to record query response: {e}')
@@ -235,6 +333,7 @@ class MonitoringHelper:
                 pipeline_id=pipeline_id,
                 pipeline_name=pipeline_name,
                 message_content=f'Error: {str(error)}',
+                role='assistant',
                 session_id=session_id,
                 status='error',
                 level='error',
@@ -281,6 +380,7 @@ class MonitoringHelper:
         cost: float | None = None,
         error_message: str | None = None,
         message_id: str | None = None,
+        details: dict | None = None,
     ):
         """Record LLM call"""
         try:
@@ -301,6 +401,7 @@ class MonitoringHelper:
                 cost=cost,
                 error_message=error_message,
                 message_id=message_id,
+                details=details,
             )
         except Exception as e:
             ap.logger.error(f'Failed to record LLM call: {e}')
diff --git a/src/langbot/pkg/pipeline/pipelinemgr.py b/src/langbot/pkg/pipeline/pipelinemgr.py
index e838d72fc..fc7aaaf31 100644
--- a/src/langbot/pkg/pipeline/pipelinemgr.py
+++ b/src/langbot/pkg/pipeline/pipelinemgr.py
@@ -1,10 +1,9 @@
 from __future__ import annotations
 
-from ..telemetry import diagnostics
-
 import dataclasses
 import typing
 import traceback
+import asyncio
 
 import sqlalchemy
 
@@ -21,6 +20,8 @@ from ..api.http.context import ExecutionContext, PrincipalContext, PrincipalType
 from ..workspace.errors import WorkspaceError, WorkspaceInvariantError
 from .config_coercion import coerce_pipeline_config
 from .pool import get_query_execution_context
+from ..i18n import language_scope
+from ..i18n.workspace import get_workspace_language
 from ..agent.runner.config_resolver import RunnerConfigResolver
 from .extension_preferences import normalize_extension_preferences
 
@@ -180,7 +181,6 @@ class RuntimePipeline:
             raise WorkspaceInvariantError('RuntimePipeline instance does not match the active Workspace binding')
         return execution_context
 
-    @diagnostics.observe('run', 'pipeline.run', source='pipeline', fields=lambda b: {'processor_type': 'pipeline'})
     async def run(self, query: pipeline_query.Query):
         if (
             query.instance_uuid != self.execution_context.instance_uuid
@@ -196,6 +196,19 @@ class RuntimePipeline:
         query.variables['_pipeline_bound_mcp_servers'] = self.bound_mcp_servers
         query.variables['_pipeline_mcp_resource_attachments'] = self.mcp_resource_attachments
         query.variables['_pipeline_mcp_resource_agent_read_enabled'] = self.mcp_resource_agent_read_enabled
+        # Telemetry: the bindings this lane runs with. None means "all enabled",
+        # which the *_all flags make explicit in the chain record.
+        try:
+            from ..telemetry import resources as telemetry_resources
+
+            telemetry_resources.note_pipeline(
+                plugins=self.bound_plugins,
+                plugins_all=self.enable_all_plugins,
+                mcp_servers=self.bound_mcp_servers,
+                mcp_all=self.enable_all_mcp_servers,
+            )
+        except Exception:
+            pass
 
         # Record query start for monitoring
         try:
@@ -257,10 +270,12 @@ class RuntimePipeline:
         if result.console_notice:
             self.ap.logger.info(result.console_notice)
         if result.error_notice:
-            diagnostics.set_outcome('failed', reason_code='response_error')
             self.ap.logger.error(result.error_notice)
             # Mark query as having error
             query.variables['_monitoring_has_error'] = True
+            # The lane reports failures as a value instead of raising, so record the
+            # reason here: without it the uploaded trace would not explain the break.
+            self._record_lane_failure(query, str(result.error_notice))
             # Record error to monitoring system
             try:
                 await self._assert_execution_active(query)
@@ -370,7 +385,118 @@ class RuntimePipeline:
 
             i += 1
 
+    @staticmethod
+    def _synthetic_origin(query: pipeline_query.Query) -> str:
+        """Origin of a lane that starts without an inbound platform event."""
+        try:
+            variables = getattr(query, 'variables', None)
+            declared = str(variables.get('_telemetry_origin') or '').strip() if isinstance(variables, dict) else ''
+        except Exception:
+            declared = ''
+        if declared in ('webui', 'api'):
+            return declared
+        adapter = getattr(query, 'adapter', None)
+        if adapter is not None and adapter.__class__.__name__ == 'WebSocketAdapter':
+            # WebChat and the CLI diagnostic surface both speak through the
+            # WebSocket proxy adapter; the CLI marks itself in its variables.
+            try:
+                variables = getattr(query, 'variables', None)
+                if isinstance(variables, dict) and '_cli_run_status' in variables:
+                    return 'api'
+            except Exception:
+                pass
+            return 'webui'
+        return 'api'
+
+    @staticmethod
+    def _synthetic_event_type(query: pipeline_query.Query) -> str:
+        """Event type of the virtual inbound event a synthetic lane stands for."""
+        event = getattr(query, 'message_event', None)
+        event_type = str(getattr(event, 'type', None) or '').strip()
+        return event_type[:160] or 'message.received'
+
     async def process_query(self, query: pipeline_query.Query):
+        workspace_uuid = getattr(query, 'workspace_uuid', None) or getattr(self, 'workspace_uuid', '')
+        with language_scope(await get_workspace_language(self.ap, workspace_uuid)):
+            return await self._process_query_with_language(query)
+
+    async def _process_query_with_language(self, query: pipeline_query.Query):
+        from ..telemetry import trace as trace_mod
+        from ..telemetry.execution import ingress
+        from ..telemetry.execution import record as record_execution
+        from ..telemetry.platform import processing_mode
+        from ..telemetry.trace import stage_scope
+
+        token = processing_mode.set('pipeline')
+        # The Workspace-scoped opaque query uuid is the execution identity Space
+        # shows for this lane; it is a stable UUID for every pooled query.
+        ingress_event_id = (query.variables or {}).get('_ingress_event_id')
+        execution_id = ingress_event_id or str(getattr(query, 'query_uuid', '') or '').strip() or str(query.query_id)
+        # A lane with no inbound platform event in flight (WebChat, WebUI debug,
+        # HTTP pipeline run) synthesizes its virtual inbound event here, so every
+        # execution starts at an event boundary and the Runner is never the trace
+        # origin. A lane running under a platform route nests inside that ingress.
+        synthesizing = trace_mod.current() is None and not ingress_event_id
+        try:
+            with ingress(
+                self.ap,
+                'pipeline_done',
+                getattr(self, 'execution_context', None),
+                execution_id=execution_id,
+                debug=synthesizing,
+                origin=self._synthetic_origin(query) if synthesizing else 'platform',
+                synthetic_event=self._synthetic_event_type(query) if synthesizing else '',
+            ):
+                # The lane is its own workflow step: acknowledgements and replies it
+                # sends directly, and the runner it drives, hang under this node.
+                # When the lane runs under a platform route the node nests there.
+                with stage_scope() as node:
+                    lane_outcome = 'success'
+                    lane_error = ''
+                    try:
+                        from ..platform.processing_indicator import processing_indicator
+
+                        async with processing_indicator(
+                            query.adapter, query.launcher_type.value.lower(), query.launcher_id
+                        ):
+                            return await self._process_query(query)
+                    except asyncio.CancelledError:
+                        lane_outcome = 'cancelled'
+                        lane_error = 'cancelled'
+                        raise
+                    except BaseException as exc:
+                        lane_outcome = 'failed'
+                        lane_error = str(exc) or type(exc).__name__
+                        raise
+                    finally:
+                        try:
+                            variables = getattr(query, 'variables', None) or {}
+                            lane_has_error = bool(variables.get('_monitoring_has_error'))
+                        except Exception:
+                            lane_has_error = False
+                        if lane_outcome == 'success' and lane_has_error:
+                            # The lane reported the failure as a value, not an exception.
+                            lane_outcome = 'failed'
+                        config = getattr(query, 'pipeline_config', None)
+                        try:
+                            runner_id = (RunnerConfigResolver.resolve_runner_id(config) or '') if config else ''
+                        except Exception:
+                            runner_id = ''
+                        record_execution(
+                            self.ap,
+                            getattr(self, 'execution_context', None),
+                            family='pipeline',
+                            operation='run',
+                            mode='pipeline',
+                            runner=runner_id,
+                            outcome=lane_outcome,
+                            error=lane_error,
+                            node=node,
+                        )
+        finally:
+            processing_mode.reset(token)
+
+    async def _process_query(self, query: pipeline_query.Query):
         """处理请求"""
         await self._assert_execution_active(query)
         # Get monitoring metadata
@@ -430,7 +556,6 @@ class RuntimePipeline:
             await self._assert_execution_active(query)
 
             if event_ctx.is_prevented_default():
-                diagnostics.set_outcome('skipped', reason_code='discarded')
                 self.ap.logger.debug(
                     f'MessageReceived event prevented default for query {query.query_id}, pipeline={pipeline_name}'
                 )
@@ -477,14 +602,14 @@ class RuntimePipeline:
                     self.ap.logger.error(f'Failed to record query response: {e}')
 
         except WorkspaceError as e:
-            diagnostics.set_outcome('rejected', reason_code='processor_incompatible')
             self.ap.logger.info(f'Dropped query {query.query_id} because its Workspace execution binding is stale: {e}')
         except Exception as e:
-            diagnostics.set_outcome('failed', reason_code='response_error')
-            diagnostics.annotate(error=e)
             inst_name = query.current_stage_name if query.current_stage_name else 'unknown'
             self.ap.logger.error(f'Error processing query {query.query_id} stage={inst_name} : {e}')
             self.ap.logger.error(f'Traceback: {traceback.format_exc()}')
+            # The lane itself broke: land the reason on the execution trace so the
+            # chain is uploaded even though the error never became a StageProcessResult.
+            self._record_lane_failure(query, str(e) or type(e).__name__)
 
             # Record query error
             try:
@@ -508,6 +633,40 @@ class RuntimePipeline:
             self.ap.logger.debug(f'Query {query.query_id} processed')
             await self.ap.query_pool.remove_query(query)
 
+    def _record_lane_failure(self, query: pipeline_query.Query, reason: str) -> None:
+        """Land one failed ``runner/execute`` stage on the owning execution trace."""
+        try:
+            from ..telemetry.execution import record as record_execution
+            from ..telemetry.trace import current as current_trace
+            from ..telemetry.trace import stage_scope
+
+            detail = reason or 'pipeline_lane_failed'
+            state = current_trace()
+            # The runner orchestrator already lands its own failed ``runner/execute``
+            # node when the break happens inside a runner; in that case only carry the
+            # reason over instead of repeating the node on the chain.
+            already_recorded = bool(state is not None and state.runner_failed)
+            if not already_recorded:
+                # The lane failure is its own workflow step, hanging under whatever
+                # step routed into it (or a root when nothing did).
+                with stage_scope() as node:
+                    record_execution(
+                        self.ap,
+                        getattr(self, 'execution_context', None),
+                        family='runner',
+                        operation='execute',
+                        mode='pipeline',
+                        runner=(RunnerConfigResolver.resolve_runner_id(query.pipeline_config) or ''),
+                        outcome='failed',
+                        error=detail,
+                        node=node,
+                    )
+                state = current_trace()
+            if state is not None:
+                state.mark_failure(detail)
+        except Exception:
+            pass
+
 
 class PipelineManager:
     """流水线管理器"""
diff --git a/src/langbot/pkg/pipeline/pool.py b/src/langbot/pkg/pipeline/pool.py
index c2c05169b..b39b9a3a4 100644
--- a/src/langbot/pkg/pipeline/pool.py
+++ b/src/langbot/pkg/pipeline/pool.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from ..telemetry import diagnostics
-
 import asyncio
 import dataclasses
 import inspect
@@ -310,7 +308,6 @@ class QueryPool:
             )
             object.__setattr__(query, 'query_uuid', query_uuid)
             object.__setattr__(query, '_execution_context', execution_context)
-            object.__setattr__(query, '_diagnostic_context', diagnostics.capture_context())
 
             self.queries.append(query)
             self.cached_queries[(execution_context.workspace_uuid, query_uuid)] = query
diff --git a/src/langbot/pkg/pipeline/preproc/preproc.py b/src/langbot/pkg/pipeline/preproc/preproc.py
index 67db3b1a4..2a0878170 100644
--- a/src/langbot/pkg/pipeline/preproc/preproc.py
+++ b/src/langbot/pkg/pipeline/preproc/preproc.py
@@ -108,6 +108,12 @@ class PreProcessor(stage.PipelineStage):
         )
         catalog = ResourcePolicyProjector.filter_tools(catalog, policy)
         ToolManager.bind_query_tool_sources(query, catalog)
+        try:
+            from ...telemetry import resources as telemetry_resources
+
+            telemetry_resources.note_pipeline(tools=catalog, tools_all=policy.allow_all_tools)
+        except Exception:
+            pass
         return ToolManager.tools_from_catalog(catalog)
 
     def _runner_accepts_multimodal_input(self, descriptor: RunnerDescriptor | None) -> bool:
@@ -393,6 +399,18 @@ class PreProcessor(stage.PipelineStage):
             descriptor,
             runner_config,
         )
+        # Telemetry: the knowledge bases this lane may retrieve from, plus the
+        # schema-declared selector fields that selected them.
+        try:
+            from ...telemetry import resources as telemetry_resources
+
+            telemetry_resources.note_pipeline(
+                knowledge_bases=query.variables['_knowledge_base_uuids'],
+                runner_id=descriptor.id if descriptor is not None else '',
+                selectors=config_schema.extract_selected_resources(descriptor, runner_config),
+            )
+        except Exception:
+            pass
 
         # Emit PromptPreProcessing before the runner receives the query.
 
@@ -427,4 +445,10 @@ class PreProcessor(stage.PipelineStage):
                 bound_skills = extensions_prefs['skills']
 
             query.variables['_pipeline_bound_skills'] = bound_skills
+            try:
+                from ...telemetry import resources as telemetry_resources
+
+                telemetry_resources.note_pipeline(skills=bound_skills)
+            except Exception:
+                pass
         return entities.StageProcessResult(result_type=entities.ResultType.CONTINUE, new_query=query)
diff --git a/src/langbot/pkg/pipeline/process/handlers/chat.py b/src/langbot/pkg/pipeline/process/handlers/chat.py
index 2fc7b0f82..5226d56b0 100644
--- a/src/langbot/pkg/pipeline/process/handlers/chat.py
+++ b/src/langbot/pkg/pipeline/process/handlers/chat.py
@@ -1,22 +1,21 @@
 from __future__ import annotations
 
+from ....i18n import pipeline_notice
+
 import uuid
 import typing
 import traceback
-import time
-from datetime import datetime
 
 
 from .. import handler
+from ..stream_results import coalesce_stream_results
 from ... import entities
 from ... import plugin_diagnostics
 
 import langbot_plugin.api.entities.events as events
 from ....agent.runner.config_resolver import RunnerConfigResolver
 from ....agent.runner import config_schema
-from ....utils import constants, runner as runner_utils
-from ....telemetry import features as telemetry_features
-from ....telemetry.identity import workspace_identity
+from ....utils import runner as runner_utils
 import langbot_plugin.api.entities.builtin.provider.session as provider_session
 import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
 import langbot_plugin.api.entities.builtin.provider.message as provider_message
@@ -115,9 +114,6 @@ class ChatMessageHandler(handler.MessageHandler):
             text_length = 0
             runner = None
             try:
-                # Mark start time for telemetry
-                start_ts = time.time()
-
                 try_claim_steering = getattr(
                     self.ap.agent_run_orchestrator,
                     'try_claim_steering_from_query',
@@ -139,7 +135,10 @@ class ChatMessageHandler(handler.MessageHandler):
 
                 # Use AgentRunOrchestrator to run the agent
                 # This replaces direct runner lookup and PluginRunnerWrapper
-                async for result in self.ap.agent_run_orchestrator.run_from_query(query):
+                results = self.ap.agent_run_orchestrator.run_from_query(query)
+                if is_stream:
+                    results = coalesce_stream_results(results)
+                async for result in results:
                     has_result = True
                     self._check_response_size(result)
 
@@ -156,6 +155,11 @@ class ChatMessageHandler(handler.MessageHandler):
                         )
                         result.attachments = attachments
 
+                    if is_stream and result.tool_calls:
+                        # A model round ending in tool calls does not end the
+                        # reply. Keep the native stream open for the answer.
+                        result = result.model_copy(update={'is_final': False})
+
                     result.resp_message_id = str(resp_message_id)
 
                     # For streaming mode, pop previous response before adding new chunk
@@ -199,6 +203,11 @@ class ChatMessageHandler(handler.MessageHandler):
 
                     if is_stream:
                         yield entities.StageProcessResult(result_type=entities.ResultType.CONTINUE, new_query=query)
+                        if result.is_final and not result.tool_calls:
+                            # Native adapters release their card on completion.
+                            # A distinct later message needs a fresh card identity.
+                            is_create_card = False
+                            resp_message_id = uuid.uuid4()
 
                 # Log final summary after streaming completes
                 if is_stream:
@@ -231,14 +240,20 @@ class ChatMessageHandler(handler.MessageHandler):
 
                 # Handle specific runner errors with appropriate messages
                 if isinstance(e, RunnerNotFoundError):
-                    user_notice = f'Agent runner not found: {e.runner_id}'
+                    user_notice = pipeline_notice(
+                        query, 'runner-not-found-hint', 'Agent runner not found: {runner}', runner=e.runner_id
+                    )
                 elif isinstance(e, RunnerNotAuthorizedError):
-                    user_notice = 'Agent runner not authorized for this pipeline'
+                    user_notice = pipeline_notice(
+                        query, 'runner-not-authorized-hint', 'Agent runner not authorized for this pipeline.'
+                    )
                 elif isinstance(e, RunnerExecutionError):
                     if e.retryable:
-                        user_notice = 'Agent runner temporarily unavailable. Please try again.'
+                        user_notice = pipeline_notice(
+                            query, 'runner-unavailable-hint', 'Agent runner temporarily unavailable. Please try again.'
+                        )
                     else:
-                        user_notice = 'Agent runner execution failed.'
+                        user_notice = pipeline_notice(query, 'failure-hint', 'Agent runner execution failed.')
                 else:
                     # Use existing exception handling
                     exception_handling = query.pipeline_config['output']['misc'].get('exception-handling', 'show-hint')
@@ -246,7 +261,7 @@ class ChatMessageHandler(handler.MessageHandler):
                     if exception_handling == 'show-error':
                         user_notice = f'{e}'
                     elif exception_handling == 'show-hint':
-                        user_notice = query.pipeline_config['output']['misc'].get('failure-hint', 'Request failed.')
+                        user_notice = pipeline_notice(query, 'failure-hint', 'Request failed.')
                     else:  # hide
                         user_notice = None
 
@@ -258,69 +273,48 @@ class ChatMessageHandler(handler.MessageHandler):
                     debug_notice=traceback.format_exc(),
                 )
             finally:
-                # Telemetry reporting
+                # Telemetry: the per-execution record is built when the owning
+                # trace closes, so only attach the execution-scoped fields this
+                # lane knows here.
                 try:
-                    end_ts = time.time()
-                    duration_ms = None
-                    if 'start_ts' in locals():
-                        duration_ms = int((end_ts - start_ts) * 1000)
+                    from ....telemetry.trace import current as current_trace
 
+                    state = current_trace()
+                    if state is not None:
+                        adapter_name = query.adapter.__class__.__name__ if hasattr(query, 'adapter') else ''
+                        runner_name = self.ap.agent_run_orchestrator.resolve_runner_id_for_telemetry(query)
+                        state.adapter = state.adapter or adapter_name
+                        state.runner = state.runner or (runner_name or '')
+                        state.runner_category = state.runner_category or runner_utils.get_runner_category_from_runner(
+                            runner_name, None, query.pipeline_config
+                        )
+                        model_name = ''
+                        try:
+                            if getattr(query, 'use_llm_model_uuid', None):
+                                m = await self.ap.model_mgr.get_model_by_uuid(
+                                    get_query_execution_context(query),
+                                    query.use_llm_model_uuid,
+                                )
+                                if m and getattr(m, 'model_entity', None):
+                                    model_name = getattr(m.model_entity, 'name', '') or ''
+                        except Exception:
+                            model_name = ''
+                        state.model_name = state.model_name or model_name
+                        if state.pipeline_plugins is None:
+                            state.pipeline_plugins = query.variables.get('_pipeline_bound_plugins', None)
+                except Exception as ex:
+                    self.ap.logger.warning(f'Failed to attach execution telemetry fields: {ex}')
+
+                # Trigger survey events on successful non-WebSocket responses
+                try:
                     adapter_name = query.adapter.__class__.__name__ if hasattr(query, 'adapter') else None
-
-                    # Use orchestrator to resolve runner ID for telemetry
-                    runner_name = self.ap.agent_run_orchestrator.resolve_runner_id_for_telemetry(query)
-
-                    # Model name if available
-                    model_name = None
-                    try:
-                        if getattr(query, 'use_llm_model_uuid', None):
-                            m = await self.ap.model_mgr.get_model_by_uuid(
-                                get_query_execution_context(query),
-                                query.use_llm_model_uuid,
-                            )
-                            if m and getattr(m, 'model_entity', None):
-                                model_name = getattr(m.model_entity, 'name', None)
-                    except Exception:
-                        model_name = None
-
-                    pipeline_plugins = query.variables.get('_pipeline_bound_plugins', None)
-
-                    runner_category = runner_utils.get_runner_category_from_runner(
-                        runner_name, None, query.pipeline_config
-                    )
-
-                    # Feature usage collected during query processing (tool calls,
-                    # knowledge base usage, sandbox executions, activated skills, ...)
-                    features = telemetry_features.collect_features(query)
-
-                    payload = {
-                        'event_type': 'query',
-                        'query_id': query.query_id,
-                        'adapter': adapter_name,
-                        'runner': runner_name,
-                        'runner_category': runner_category,
-                        'duration_ms': duration_ms,
-                        'model_name': model_name,
-                        'version': constants.semantic_version,
-                        **workspace_identity(get_query_execution_context(query)),
-                        'runtime_instance_id': constants.instance_id,
-                        'edition': constants.edition,
-                        'pipeline_plugins': pipeline_plugins,
-                        'features': features,
-                        'error': locals().get('error_info', None),
-                        'timestamp': datetime.utcnow().isoformat(),
-                    }
-
-                    await self.ap.telemetry.start_send_task(payload)
-
-                    # Trigger survey events on successful non-WebSocket responses
                     if not locals().get('error_info') and adapter_name and 'WebSocket' not in adapter_name:
                         if self.ap.survey:
                             await self.ap.survey.trigger_event('first_bot_response_success')
                             # Counts toward the bot_response_success_100 milestone event
                             await self.ap.survey.record_bot_response_success()
                 except Exception as ex:
-                    self.ap.logger.warning(f'Failed to send telemetry: {ex}')
+                    self.ap.logger.warning(f'Failed to trigger survey event: {ex}')
 
     async def _ensure_conversation_for_history(
         self,
diff --git a/src/langbot/pkg/pipeline/process/stream_results.py b/src/langbot/pkg/pipeline/process/stream_results.py
new file mode 100644
index 000000000..35e0dd09a
--- /dev/null
+++ b/src/langbot/pkg/pipeline/process/stream_results.py
@@ -0,0 +1,55 @@
+"""Coalesce a runner's terminal delta and completed snapshot before delivery."""
+
+from collections.abc import AsyncIterable, AsyncIterator
+
+from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+
+
+async def coalesce_stream_results(
+    results: AsyncIterable[Message | MessageChunk],
+) -> AsyncIterator[Message | MessageChunk]:
+    pending: MessageChunk | None = None
+    snapshot = ''
+    async for result in results:
+        if pending is not None:
+            # Completion snapshots can add attachments and usage to the final
+            # delta. Deliver that richer result once, before closing the card.
+            content = pending.all_content if pending.all_content is not None else pending.content
+            # Multi-round runners stream a cumulative transcript, but complete
+            # with only the last model round. Keep the transcript on the same
+            # card while taking attachments and metadata from the completion.
+            final_round = (
+                isinstance(content, str)
+                and isinstance(result.content, str)
+                and bool(result.content)
+                and content.endswith(result.content)
+            )
+            same_message = (
+                isinstance(result, Message)
+                and result.role == pending.role
+                and (result.tool_calls or []) == (pending.tool_calls or [])
+                and (not content or result.content == content or final_round)
+            )
+            if same_message and final_round and content != result.content:
+                result = result.model_copy(update={'content': content})
+            elif not same_message:
+                yield pending
+            pending = None
+            snapshot = ''
+
+        if isinstance(result, MessageChunk) and isinstance(result.content, (str, type(None))):
+            # Runner output is already cumulative (unlike raw provider deltas).
+            # Explicit all_content also supports runners forwarding raw deltas.
+            # Never guess delta semantics from a missing all_content field.
+            snapshot = result.all_content if result.all_content is not None else (result.content or snapshot)
+            result = result.model_copy(update={'all_content': snapshot})
+        elif isinstance(result, Message):
+            snapshot = ''
+
+        if isinstance(result, MessageChunk) and result.is_final:
+            pending = result
+        else:
+            yield result
+
+    if pending is not None:
+        yield pending
diff --git a/src/langbot/pkg/pipeline/ratelimit/ratelimit.py b/src/langbot/pkg/pipeline/ratelimit/ratelimit.py
index cab62b8d5..8ee1bb639 100644
--- a/src/langbot/pkg/pipeline/ratelimit/ratelimit.py
+++ b/src/langbot/pkg/pipeline/ratelimit/ratelimit.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ...i18n import pipeline_notice
+
 import typing
 
 from .. import entities, stage
@@ -62,7 +64,7 @@ class RateLimit(stage.PipelineStage):
                     result_type=entities.ResultType.INTERRUPT,
                     new_query=query,
                     console_notice=f'根据限速规则忽略 {query.launcher_type.value}:{query.launcher_id} 消息',
-                    user_notice='请求数超过限速器设定值，已丢弃本消息。',
+                    user_notice=pipeline_notice(query, 'rate-limit-hint', 'Too many requests. Please try again later.'),
                 )
         elif stage_inst_name == 'ReleaseRateLimitOccupancy':
             await self.algo.release_access(
diff --git a/src/langbot/pkg/pipeline/wrapper/wrapper.py b/src/langbot/pkg/pipeline/wrapper/wrapper.py
index 9ea275e1d..2dcda3a05 100644
--- a/src/langbot/pkg/pipeline/wrapper/wrapper.py
+++ b/src/langbot/pkg/pipeline/wrapper/wrapper.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ...i18n import pipeline_notice
+
 import typing
 
 from .. import entities
@@ -175,7 +177,9 @@ class ResponseWrapper(stage.PipelineStage):
                     if result.tool_calls is not None and len(result.tool_calls) > 0:  # 有函数调用
                         function_names = [tc.function.name for tc in result.tool_calls]
 
-                        reply_text = f'Call {".".join(function_names)}...'
+                        reply_text = pipeline_notice(
+                            query, 'tool-call-hint', 'Calling {tools}...', tools=', '.join(function_names)
+                        )
 
                         # Only surface the tool-call notice when the pipeline option
                         # is enabled. Emitting it unconditionally would become the
diff --git a/src/langbot/pkg/platform/adapter_names.py b/src/langbot/pkg/platform/adapter_names.py
index 448182af6..be5c1f56b 100644
--- a/src/langbot/pkg/platform/adapter_names.py
+++ b/src/langbot/pkg/platform/adapter_names.py
@@ -2,7 +2,6 @@
 
 OMNI_ADAPTER_NAMES = frozenset(
     {
-        'aiocqhttp',
         'telegram',
         'discord',
         'qqofficial',
@@ -20,6 +19,8 @@ OMNI_ADAPTER_NAMES = frozenset(
 
 def canonical_adapter_name(name: str) -> str:
     """Accept previous built-in IDs without renaming legacy or custom adapters."""
+    if name in ('aiocqhttp-omni', 'aiocqhttp-eba'):
+        return 'onebot-v11'
     if name.endswith('-eba') and name[:-4] in OMNI_ADAPTER_NAMES:
         return f'{name[:-4]}-omni'
     return name
diff --git a/src/langbot/pkg/platform/adapters/aiocqhttp/adapter.py b/src/langbot/pkg/platform/adapters/aiocqhttp/adapter.py
index ae51f6c9b..6840d7131 100644
--- a/src/langbot/pkg/platform/adapters/aiocqhttp/adapter.py
+++ b/src/langbot/pkg/platform/adapters/aiocqhttp/adapter.py
@@ -1,9 +1,13 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import traceback
+import hmac
+from urllib.parse import urlsplit
+
+from websockets.asyncio.client import connect
+from langbot.pkg.core.task_boundary import create_detached_task
+from langbot.pkg.platform.adapters.aiocqhttp.transport import OneBotClient, OneBotConnection
 import typing
 
 import aiocqhttp
@@ -22,7 +26,10 @@ from langbot_plugin.api.entities.builtin.platform.errors import NotSupportedErro
 
 
 class AiocqhttpAdapter(AiocqhttpAPIMixin, abstract_platform_adapter.AbstractPlatformAdapter):
-    bot: aiocqhttp.CQHttp = pydantic.Field(exclude=True)
+    bot: typing.Any = pydantic.Field(exclude=True)
+    _runtime: typing.Any = pydantic.PrivateAttr(default=None)
+    _stopped: asyncio.Event = pydantic.PrivateAttr(default_factory=asyncio.Event)
+    _connection_task: typing.Any = pydantic.PrivateAttr(default=None)
 
     message_converter: AiocqhttpMessageConverter = AiocqhttpMessageConverter()
     event_converter: AiocqhttpEventConverter = AiocqhttpEventConverter()
@@ -38,24 +45,64 @@ class AiocqhttpAdapter(AiocqhttpAPIMixin, abstract_platform_adapter.AbstractPlat
         arbitrary_types_allowed = True
 
     def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger):
-        run_config = dict(config)
+        config = dict(config)
+        config.setdefault('connection_mode', 'reverse_ws')
+        if not isinstance(config.get('access-token'), str) or not config['access-token'].strip():
+            raise ValueError('OneBot requires a non-empty access token')
+        if config['connection_mode'] not in ('reverse_ws', 'forward_ws'):
+            raise ValueError('Unsupported OneBot connection mode')
+        if config['connection_mode'] == 'forward_ws':
+            url = urlsplit(config.get('ws_url', ''))
+            if url.scheme not in ('ws', 'wss') or not url.hostname or url.username or url.password or url.fragment:
+                raise ValueError('OneBot ws_url must be a ws:// or wss:// endpoint without embedded credentials')
+        super().__init__(config=config, logger=logger, bot=OneBotClient(), bot_account_id='', listeners={})
 
-        async def shutdown_trigger_placeholder():
-            while True:
-                await asyncio.sleep(1)
+    def bind_runtime(self, runtime):
+        self._runtime = runtime
 
-        run_config['shutdown_trigger'] = shutdown_trigger_placeholder
-        access_token = run_config.pop('access-token', '') or None
-        bot = aiocqhttp.CQHttp(access_token=access_token)
+    async def _validate_connection(self):
+        if self._stopped.is_set() or self._runtime is None:
+            raise RuntimeError('OneBot runtime is inactive')
+        current = await self._runtime.ap.platform_mgr.resolve_public_bot(self._runtime.bot_entity.uuid)
+        if current is not self._runtime or not current.enable:
+            raise RuntimeError('OneBot runtime placement changed')
 
-        super().__init__(
-            config=run_config,
-            logger=logger,
-            bot=bot,
-            bot_account_id='',
-            listeners={},
+    def authenticate(self, headers):
+        token = str(self.config.get('access-token', ''))
+        supplied = headers.get('Authorization', '')
+        return bool(token) and hmac.compare_digest(supplied.encode(), ('Bearer ' + token).encode())
+
+    async def serve_connection(self, send, receive, close, self_id=''):
+        await self._validate_connection()
+        if self.bot.connection is not None:
+            raise RuntimeError('OneBot already has an active connection')
+        expected = str(self.config.get('self_id', '') or '')
+        if expected and self_id and expected != self_id:
+            raise ValueError('OneBot account mismatch')
+        connection = OneBotConnection(
+            send,
+            receive,
+            close,
+            self._receive_event,
+            self._validate_connection,
+            self_id=expected or self_id,
         )
-        self._register_native_handlers()
+        self.bot.connection = connection
+        self.bot_account_id = expected or self_id
+        self._lookup.clear()
+        task = create_detached_task(connection.run(), workspace_uuid=self._runtime.workspace_uuid)
+        self._connection_task = task
+        try:
+            await task
+        finally:
+            await connection.close()
+            if self.bot.connection is connection:
+                self.bot.connection = None
+                self._connection_task = None
+                self._lookup.clear()
+
+    async def _receive_event(self, payload):
+        await self._handle_native_event(aiocqhttp.Event.from_payload(payload))
 
     def get_supported_events(self) -> list[str]:
         return [
@@ -96,7 +143,6 @@ class AiocqhttpAdapter(AiocqhttpAPIMixin, abstract_platform_adapter.AbstractPlat
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -124,34 +170,51 @@ class AiocqhttpAdapter(AiocqhttpAPIMixin, abstract_platform_adapter.AbstractPlat
             self.listeners.pop(event_type, None)
 
     async def run_async(self):
-        await self.bot._server_app.run_task(**self.config)
+        if not self.config.get('access-token'):
+            raise ValueError('OneBot requires a non-empty access token')
+        if self.config['connection_mode'] == 'reverse_ws':
+            await self._stopped.wait()
+            return
+        delay = 1
+        while not self._stopped.is_set():
+            await self._validate_connection()
+            try:
+                async with connect(
+                    self.config['ws_url'],
+                    additional_headers={'Authorization': 'Bearer ' + self.config['access-token']},
+                    open_timeout=10,
+                    close_timeout=5,
+                    ping_interval=20,
+                    ping_timeout=20,
+                    max_size=8 * 1024 * 1024,
+                    proxy=None,
+                ) as socket:
+                    delay = 1
+                    await self.serve_connection(socket.send, socket.recv, socket.close)
+            except asyncio.CancelledError:
+                raise
+            except Exception as exc:
+                await self.logger.warning(f'OneBot connection lost ({type(exc).__name__}); reconnecting')
+            try:
+                await asyncio.wait_for(self._stopped.wait(), timeout=delay)
+            except TimeoutError:
+                pass
+            delay = min(delay * 2, 30)
 
     async def kill(self) -> bool:
+        self._stopped.set()
+        task = self._connection_task
+        if task is not None:
+            task.cancel()
+            await asyncio.gather(task, return_exceptions=True)
         self._lookup.clear()
-        return False
+        return True
 
-    def _register_native_handlers(self):
-        @self.bot.on_message()
-        async def on_message(event: aiocqhttp.Event):
-            await self._handle_native_event(event)
-
-        @self.bot.on_notice()
-        async def on_notice(event: aiocqhttp.Event):
-            await self._handle_native_event(event)
-
-        @self.bot.on_request()
-        async def on_request(event: aiocqhttp.Event):
-            await self._handle_native_event(event)
-
-        @self.bot.on_websocket_connection
-        async def on_websocket_connection(event: aiocqhttp.Event):
-            self.bot_account_id = str(getattr(event, 'self_id', '') or self.bot_account_id)
-            await self.logger.info(f'WebSocket connection established, bot id: {self.bot_account_id}')
-            await self._dispatch_native_event(event)
-
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: aiocqhttp.Event):
         self.bot_account_id = str(getattr(event, 'self_id', '') or self.bot_account_id)
+        # Connection metadata belongs to the adapter, never business event routing.
+        if getattr(event, 'type', None) == 'meta_event':
+            return
         if getattr(event, 'type', None) == 'message' and str(getattr(event, 'user_id', '')) == self.bot_account_id:
             return
         try:
@@ -167,14 +230,12 @@ class AiocqhttpAdapter(AiocqhttpAPIMixin, abstract_platform_adapter.AbstractPlat
         except Exception:
             await self.logger.error(f'Error in aiocqhttp native event: {traceback.format_exc()}')
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _dispatch_native_event(self, event: aiocqhttp.Event):
         eba_event = await self.event_converter.target2yiri(event, self.bot, self.bot_account_id, self._lookup)
         if eba_event:
             await self._dispatch_eba_event(eba_event)
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/aiocqhttp/api_impl.py b/src/langbot/pkg/platform/adapters/aiocqhttp/api_impl.py
index 75c4b49ba..f4c315909 100644
--- a/src/langbot/pkg/platform/adapters/aiocqhttp/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/aiocqhttp/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 import aiocqhttp
@@ -17,7 +15,6 @@ from langbot_plugin.api.entities.builtin.platform.errors import NotSupportedErro
 class AiocqhttpAPIMixin:
     bot: aiocqhttp.CQHttp
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -38,7 +35,6 @@ class AiocqhttpAPIMixin:
             raise ValueError(f'Unsupported aiocqhttp target_type: {target_type}')
         return platform_events.MessageResult(message_id=(raw or {}).get('message_id'), raw=raw or {})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -53,7 +49,6 @@ class AiocqhttpAPIMixin:
         raw = await self.bot.send(message_source.source_platform_object, aiocq_msg)
         return platform_events.MessageResult(message_id=(raw or {}).get('message_id'), raw=raw or {})
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -62,7 +57,6 @@ class AiocqhttpAPIMixin:
     ) -> None:
         await self.bot.delete_msg(message_id=int(message_id))
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -81,7 +75,6 @@ class AiocqhttpAPIMixin:
             raise ValueError(f'Unsupported aiocqhttp to_chat_type: {to_chat_type}')
         return platform_events.MessageResult(message_id=(raw or {}).get('message_id'), raw=raw or {})
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -107,7 +100,6 @@ class AiocqhttpAPIMixin:
         )
         return await AiocqhttpEventConverter.message_to_eba(event, self.bot)
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         raw = await self.bot.get_group_info(group_id=int(group_id))
         return platform_entities.UserGroup(
@@ -116,7 +108,6 @@ class AiocqhttpAPIMixin:
             member_count=raw.get('member_count'),
         )
 
-    @diagnostics.observe('api', 'get_group_list', source='platform', stage='accepted')
     async def get_group_list(self) -> list[platform_entities.UserGroup]:
         raw_list = await self.bot.get_group_list()
         return [
@@ -128,7 +119,6 @@ class AiocqhttpAPIMixin:
             for item in raw_list
         ]
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -136,7 +126,6 @@ class AiocqhttpAPIMixin:
         raw_list = await self.bot.get_group_member_list(group_id=int(group_id))
         return [self._member_to_entity(item, group_id) for item in raw_list]
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -145,11 +134,9 @@ class AiocqhttpAPIMixin:
         raw = await self.bot.get_group_member_info(group_id=int(group_id), user_id=int(user_id), no_cache=True)
         return self._member_to_entity(raw, group_id)
 
-    @diagnostics.observe('api', 'set_group_name', source='platform', stage='accepted')
     async def set_group_name(self, group_id: typing.Union[int, str], name: str) -> None:
         await self.bot.set_group_name(group_id=int(group_id), group_name=name)
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(
         self,
         group_id: typing.Union[int, str],
@@ -158,19 +145,15 @@ class AiocqhttpAPIMixin:
     ) -> None:
         await self.bot.set_group_ban(group_id=int(group_id), user_id=int(user_id), duration=int(duration))
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]) -> None:
         await self.bot.set_group_ban(group_id=int(group_id), user_id=int(user_id), duration=0)
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]) -> None:
         await self.bot.set_group_kick(group_id=int(group_id), user_id=int(user_id), reject_add_request=False)
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(self, group_id: typing.Union[int, str]) -> None:
         await self.bot.set_group_leave(group_id=int(group_id), is_dismiss=False)
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         raw = await self.bot.get_stranger_info(user_id=int(user_id), no_cache=True)
         return platform_entities.User(
@@ -179,7 +162,6 @@ class AiocqhttpAPIMixin:
             avatar_url=raw.get('avatar_url'),
         )
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         raw_list = await self.bot.get_friend_list()
         return [
@@ -191,7 +173,6 @@ class AiocqhttpAPIMixin:
             for item in raw_list
         ]
 
-    @diagnostics.observe('api', 'approve_friend_request', source='platform', stage='accepted')
     async def approve_friend_request(
         self,
         request_id: typing.Union[int, str],
@@ -200,15 +181,12 @@ class AiocqhttpAPIMixin:
     ) -> None:
         await self.bot.set_friend_add_request(flag=str(request_id), approve=approve, remark=remark or '')
 
-    @diagnostics.observe('api', 'approve_group_invite', source='platform', stage='accepted')
     async def approve_group_invite(self, request_id: typing.Union[int, str], approve: bool = True) -> None:
         await self.bot.set_group_add_request(flag=str(request_id), sub_type='invite', approve=approve, reason='')
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
diff --git a/src/langbot/pkg/platform/adapters/aiocqhttp/event_converter.py b/src/langbot/pkg/platform/adapters/aiocqhttp/event_converter.py
index cf809c513..295f4ef18 100644
--- a/src/langbot/pkg/platform/adapters/aiocqhttp/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/aiocqhttp/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 import aiocqhttp
@@ -23,7 +21,6 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return getattr(event, 'source_platform_object', None)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(
         event: aiocqhttp.Event,
         bot: aiocqhttp.CQHttp | None = None,
@@ -37,12 +34,9 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
             return AiocqhttpEventConverter.notice_to_eba(event, bot_user_id)
         if event_type == 'request':
             return AiocqhttpEventConverter.request_to_eba(event)
-        if event_type == 'meta_event':
-            return AiocqhttpEventConverter.platform_specific(event, f'meta.{getattr(event, "detail_type", "")}')
         return None
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         event: aiocqhttp.Event,
         bot: aiocqhttp.CQHttp | None = None,
@@ -94,7 +88,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
             )
         return platform_events.MessageReceivedEvent(
             type='message.received',
-            adapter_name='aiocqhttp-omni',
+            adapter_name='onebot-v11',
             message_id=getattr(event, 'message_id', ''),
             message_chain=message_chain,
             sender=sender,
@@ -115,13 +109,19 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if notice_type in ('group_recall', 'friend_recall'):
             return platform_events.MessageDeletedEvent(
                 type='message.deleted',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 message_id=getattr(event, 'message_id', ''),
-                operator=AiocqhttpEventConverter.user(getattr(event, 'operator_id', None)),
+                operator=AiocqhttpEventConverter.user(
+                    getattr(event, 'user_id', None)
+                    if notice_type == 'friend_recall'
+                    else getattr(event, 'operator_id', None)
+                ),
                 chat_type=platform_entities.ChatType.GROUP
                 if notice_type == 'group_recall'
                 else platform_entities.ChatType.PRIVATE,
-                chat_id=getattr(event, 'group_id', getattr(event, 'user_id', '')),
+                chat_id=getattr(event, 'group_id', '')
+                if notice_type == 'group_recall'
+                else getattr(event, 'user_id', ''),
                 group=AiocqhttpEventConverter.group_from_event(event) if notice_type == 'group_recall' else None,
                 timestamp=float(getattr(event, 'time', 0) or 0),
                 source_platform_object=event,
@@ -133,7 +133,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
             if AiocqhttpEventConverter._is_bot_user(getattr(event, 'user_id', None), bot_user_id, event):
                 return platform_events.BotInvitedToGroupEvent(
                     type='bot.invited_to_group',
-                    adapter_name='aiocqhttp-omni',
+                    adapter_name='onebot-v11',
                     group=group,
                     inviter=AiocqhttpEventConverter.user(inviter_id) if inviter_id else None,
                     timestamp=float(getattr(event, 'time', 0) or 0),
@@ -141,7 +141,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
                 )
             return platform_events.MemberJoinedEvent(
                 type='group.member_joined',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 group=group,
                 member=user,
                 inviter=AiocqhttpEventConverter.user(inviter_id) if inviter_id else None,
@@ -155,7 +155,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
             if AiocqhttpEventConverter._is_bot_user(getattr(event, 'user_id', None), bot_user_id, event):
                 return platform_events.BotRemovedFromGroupEvent(
                     type='bot.removed_from_group',
-                    adapter_name='aiocqhttp-omni',
+                    adapter_name='onebot-v11',
                     group=group,
                     operator=operator,
                     timestamp=float(getattr(event, 'time', 0) or 0),
@@ -163,7 +163,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
                 )
             return platform_events.MemberLeftEvent(
                 type='group.member_left',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 group=group,
                 member=AiocqhttpEventConverter.user(getattr(event, 'user_id', '')),
                 is_kicked=getattr(event, 'sub_type', '') in ('kick', 'kick_me'),
@@ -179,7 +179,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
                 event_cls = platform_events.BotMutedEvent if duration > 0 else platform_events.BotUnmutedEvent
                 kwargs: dict[str, typing.Any] = {
                     'type': 'bot.muted' if duration > 0 else 'bot.unmuted',
-                    'adapter_name': 'aiocqhttp-omni',
+                    'adapter_name': 'onebot-v11',
                     'group': group,
                     'operator': operator,
                     'timestamp': float(getattr(event, 'time', 0) or 0),
@@ -191,7 +191,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
             if duration > 0:
                 return platform_events.MemberBannedEvent(
                     type='group.member_banned',
-                    adapter_name='aiocqhttp-omni',
+                    adapter_name='onebot-v11',
                     group=group,
                     member=AiocqhttpEventConverter.user(getattr(event, 'user_id', '')),
                     operator=operator,
@@ -202,7 +202,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if notice_type == 'friend_add':
             return platform_events.FriendAddedEvent(
                 type='friend.added',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 user=AiocqhttpEventConverter.user(getattr(event, 'user_id', '')),
                 timestamp=float(getattr(event, 'time', 0) or 0),
                 source_platform_object=event,
@@ -215,7 +215,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if request_type == 'friend':
             return platform_events.FriendRequestReceivedEvent(
                 type='friend.request_received',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 request_id=getattr(event, 'flag', ''),
                 user=AiocqhttpEventConverter.user(getattr(event, 'user_id', '')),
                 message=getattr(event, 'comment', None),
@@ -225,7 +225,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if request_type == 'group' and getattr(event, 'sub_type', '') == 'invite':
             return platform_events.BotInvitedToGroupEvent(
                 type='bot.invited_to_group',
-                adapter_name='aiocqhttp-omni',
+                adapter_name='onebot-v11',
                 group=AiocqhttpEventConverter.group_from_event(event),
                 inviter=AiocqhttpEventConverter.user(getattr(event, 'user_id', '')),
                 request_id=getattr(event, 'flag', ''),
@@ -262,7 +262,7 @@ class AiocqhttpEventConverter(abstract_platform_adapter.AbstractEventConverter):
     def platform_specific(event: aiocqhttp.Event, action: str) -> platform_events.PlatformSpecificEvent:
         return platform_events.PlatformSpecificEvent(
             type='platform.specific',
-            adapter_name='aiocqhttp-omni',
+            adapter_name='onebot-v11',
             action=action,
             data={key: value for key, value in dict(event).items() if key not in {'message'}},
             timestamp=float(getattr(event, 'time', 0) or 0),
diff --git a/src/langbot/pkg/platform/adapters/aiocqhttp/manifest.yaml b/src/langbot/pkg/platform/adapters/aiocqhttp/manifest.yaml
index 6911881c9..3b5dec81b 100644
--- a/src/langbot/pkg/platform/adapters/aiocqhttp/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/aiocqhttp/manifest.yaml
@@ -2,7 +2,7 @@ apiVersion: v1
 kind: MessagePlatformAdapter
 
 metadata:
-  name: aiocqhttp-omni
+  name: onebot-v11
   label:
     en_US: OneBot v11
     zh_Hans: OneBot v11
@@ -15,48 +15,66 @@ metadata:
 
 spec:
   categories:
+    - popular
     - protocol
   help_links:
     zh: https://langbot.app/docs/zh/usage/platforms/qq/aiocqhttp/napcat
     en: https://langbot.app/docs/en/usage/platforms/qq/aiocqhttp/napcat
     ja: https://langbot.app/docs/ja/usage/platforms/qq/aiocqhttp/napcat
   config:
-    - name: host
-      label:
-        en_US: Host
-        zh_Hans: 主机
-        zh_Hant: 主機
-      description:
-        en_US: The host that OneBot v11 listens on for reverse WebSocket connections. Unless you know what you're doing, use 0.0.0.0
-        zh_Hans: OneBot v11 反向 WebSocket 监听主机，除非你知道自己在做什么，否则请写 0.0.0.0
-        zh_Hant: OneBot v11 反向 WebSocket 監聽主機，除非你知道自己在做什麼，否則請填 0.0.0.0
-      type: string
+    - name: connection_mode
+      label: {en_US: Connection mode, zh_Hans: 连接模式, zh_Hant: 連線模式, ja_JP: 接続方式}
+      type: select
       required: true
-      default: 0.0.0.0
-    - name: port
-      label:
-        en_US: Port
-        zh_Hans: 端口
-        zh_Hant: 連接埠
+      default: reverse_ws
+      options:
+        - name: reverse_ws
+          label: {en_US: 'Reverse WS — protocol endpoint connects to LangBot', zh_Hans: '反向 WS — 协议端连接 LangBot', zh_Hant: '反向 WS — 協定端連接 LangBot', ja_JP: '逆方向 WS — プロトコル側から LangBot に接続'}
+        - name: forward_ws
+          label: {en_US: 'Forward WS — LangBot connects to protocol endpoint', zh_Hans: '正向 WS — LangBot 连接协议端', zh_Hant: '正向 WS — LangBot 連接協定端', ja_JP: '順方向 WS — LangBot からプロトコル側に接続'}
       description:
-        en_US: Reverse WebSocket listen port
-        zh_Hans: 反向 WebSocket 监听端口
-        zh_Hant: 反向 WebSocket 監聽連接埠
-      type: integer
-      required: true
-      default: 2280
-    - name: access-token
-      label:
-        en_US: Access Token
-        zh_Hans: 访问令牌
-        zh_Hant: 存取令牌
+        en_US: Reverse WS uses LangBot's main service port. Forward WS requires a protocol endpoint reachable from LangBot. HTTP is not supported.
+        zh_Hans: 反向 WS 由协议端连接 LangBot 主服务端口，无需额外监听端口；正向 WS 由 LangBot 连接可访问的协议端。不支持 HTTP。
+        zh_Hant: 反向 WS 由協定端連接 LangBot 主服務連接埠；正向 WS 由 LangBot 連接可存取的協定端。不支援 HTTP。
+        ja_JP: 逆方向 WS は LangBot のメインポートを使用。順方向 WS は LangBot から到達可能なプロトコル側に接続します。HTTP は非対応です。
+    - name: __system.onebot_ws_url
+      label: {en_US: Protocol endpoint connection URL, zh_Hans: 协议端连接地址, zh_Hant: 協定端連線位址, ja_JP: プロトコル側の接続先 URL}
       description:
-        en_US: Custom connection token for the protocol endpoint. Leave empty if the endpoint has no token configured
-        zh_Hans: 自定义的协议端连接令牌；若协议端未设置，则不填
-        zh_Hant: 自訂的協定端連線令牌；若協定端未設定，則不填
+        en_US: Save the bot, then copy this URL into the protocol endpoint's WebSocket client (Universal). Old host/port settings are ignored; update the protocol endpoint URL.
+        zh_Hans: 保存机器人后，将此地址填入协议端的 WebSocket 客户端（Universal）配置。旧 host/port 配置不再生效，请同步修改协议端地址。
+        zh_Hant: 儲存後將此位址填入協定端 WebSocket 用戶端（Universal）。舊 host/port 不再生效，請同步更新協定端位址。
+        ja_JP: 保存後、この URL をプロトコル側の WebSocket クライアント（Universal）に設定。旧 host/port は無効です。接続先を更新してください。
       type: string
       required: false
-      default: ""
+      default: ''
+      show_if: {field: connection_mode, operator: eq, value: reverse_ws}
+    - name: ws_url
+      label: {en_US: Protocol endpoint WebSocket URL, zh_Hans: 协议端 WebSocket 地址, zh_Hant: 協定端 WebSocket 位址, ja_JP: プロトコル側 WebSocket URL}
+      description:
+        en_US: LangBot connects to this ws:// or wss:// Universal endpoint. It must be reachable from LangBot.
+        zh_Hans: LangBot 将主动连接此 ws:// 或 wss:// 地址，使用协议端同时提供事件和 API 的 Universal 端点。请确保 LangBot 能访问它。
+        zh_Hant: LangBot 主動連接此 ws:// 或 wss:// Universal 位址，請確保可存取。
+        ja_JP: LangBot がこの ws:// または wss:// Universal URL に接続します。到達可能な URL を指定してください。
+      type: string
+      required: true
+      default: ''
+      show_if: {field: connection_mode, operator: eq, value: forward_ws}
+    - name: access-token
+      label: {en_US: Access token, zh_Hans: 访问令牌, zh_Hant: 存取權杖, ja_JP: アクセストークン}
+      description:
+        en_US: Required in both modes. Configure the same secret token on LangBot and the protocol endpoint. Sent using Authorization Bearer.
+        zh_Hans: 两种模式均必填。在 LangBot 与协议端设置相同的保密令牌，连接时使用 Authorization Bearer 验证。
+        zh_Hant: 兩種模式皆必填。在 LangBot 與協定端設定相同的保密權杖。
+        ja_JP: 両方式で必須。LangBot とプロトコル側に同じ秘密トークンを設定してください。
+      type: string
+      required: true
+      default: ''
+    - name: self_id
+      label: {en_US: Expected bot account (optional), zh_Hans: 预期机器人账号（可选）, zh_Hant: 預期機器人帳號（選填）, ja_JP: ボットアカウント（任意）}
+      description: {en_US: Reject connections or events from a different account., zh_Hans: 拒绝其他账号的连接或事件。, zh_Hant: 拒絕其他帳號的連線或事件。, ja_JP: 異なるアカウントの接続やイベントを拒否します。}
+      type: string
+      required: false
+      default: ''
 
   supported_events:
     - message.received
diff --git a/src/langbot/pkg/platform/adapters/aiocqhttp/transport.py b/src/langbot/pkg/platform/adapters/aiocqhttp/transport.py
new file mode 100644
index 000000000..36e09c3d7
--- /dev/null
+++ b/src/langbot/pkg/platform/adapters/aiocqhttp/transport.py
@@ -0,0 +1,125 @@
+"""OneBot v11 duplex transport, independent of an HTTP server."""
+
+from __future__ import annotations
+
+import asyncio
+import functools
+import json
+import uuid
+
+from aiocqhttp.exceptions import ActionFailed, ApiNotAvailable
+
+
+class OneBotConnection:
+    def __init__(self, send, receive, close, on_event, validate, *, self_id='', timeout=30):
+        self.send = send
+        self.receive = receive
+        self.close_socket = close
+        self.on_event = on_event
+        self.validate = validate
+        self.self_id = self_id
+        self.timeout = timeout
+        self.pending = {}
+        self.events = asyncio.Queue(maxsize=256)
+        self.closed = False
+        self.send_lock = asyncio.Lock()
+
+    async def call_action(self, action, **params):
+        await self.validate()
+        if self.closed:
+            raise ApiNotAvailable('OneBot connection closed')
+        if len(self.pending) >= 256:
+            raise ApiNotAvailable('Too many pending OneBot actions')
+        echo = uuid.uuid4().hex
+        future = asyncio.get_running_loop().create_future()
+        self.pending[echo] = future
+        try:
+            async with asyncio.timeout(self.timeout):
+                async with self.send_lock:
+                    await self.validate()
+                    await self.send(json.dumps({'action': action, 'params': params, 'echo': echo}))
+                result = await future
+            if (result.get('status'), result.get('retcode')) not in (('ok', 0), ('async', 1)):
+                raise ActionFailed(result)
+            return result.get('data')
+        finally:
+            self.pending.pop(echo, None)
+            if not future.done():
+                future.cancel()
+
+    async def _read(self):
+        while True:
+            raw = await self.receive()
+            if len(raw) > 8 * 1024 * 1024:
+                raise ValueError('OneBot frame too large')
+            data = json.loads(raw)
+            if not isinstance(data, dict):
+                raise ValueError('OneBot frame must be an object')
+            await self.validate()
+            if 'post_type' in data:
+                account = str(data.get('self_id', ''))
+                if not account or (self.self_id and account != self.self_id):
+                    raise ValueError('OneBot account mismatch')
+                self.self_id = account
+                # Never block responses behind an event that itself calls an API.
+                self.events.put_nowait(data)
+            else:
+                echo = data.get('echo')
+                future = self.pending.get(echo) if isinstance(echo, str) else None
+                if future is not None and not future.done():
+                    future.set_result(data)
+
+    async def _consume(self):
+        while True:
+            data = await self.events.get()
+            await self.validate()
+            await self.on_event(data)
+
+    async def _watch_placement(self):
+        while True:
+            await asyncio.sleep(5)
+            await self.validate()
+
+    async def run(self):
+        tasks = [asyncio.create_task(coro()) for coro in (self._read, self._consume, self._watch_placement)]
+        try:
+            done, _ = await asyncio.wait(tasks, return_when=asyncio.FIRST_COMPLETED)
+            for task in done:
+                task.result()
+        finally:
+            for task in tasks:
+                task.cancel()
+            await asyncio.gather(*tasks, return_exceptions=True)
+            await self.close()
+
+    async def close(self):
+        if self.closed:
+            return
+        self.closed = True
+        for future in self.pending.values():
+            if not future.done():
+                future.set_exception(ApiNotAvailable('OneBot connection closed'))
+        self.pending.clear()
+        await self.close_socket()
+
+
+class OneBotClient:
+    """Keep the existing converter/API call surface without CQHttp's server."""
+
+    def __init__(self):
+        self.connection = None
+
+    async def call_action(self, action, **params):
+        connection = self.connection
+        if connection is None:
+            raise ApiNotAvailable('OneBot protocol endpoint is not connected')
+        return await connection.call_action(action, **params)
+
+    def __getattr__(self, action):
+        if action.startswith('_'):
+            raise AttributeError(action)
+        return functools.partial(self.call_action, action)
+
+    async def send(self, event, message, **kwargs):
+        params = {key: event[key] for key in ('message_type', 'user_id', 'group_id', 'discuss_id') if key in event}
+        return await self.call_action('send_msg', **(params | {'message': message} | kwargs))
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/adapter.py b/src/langbot/pkg/platform/adapters/dingtalk/adapter.py
index 486421470..60a14087d 100644
--- a/src/langbot/pkg/platform/adapters/dingtalk/adapter.py
+++ b/src/langbot/pkg/platform/adapters/dingtalk/adapter.py
@@ -1,14 +1,15 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
+import time
 import traceback
 import typing
 
 import dingtalk_stream
 import pydantic
 
-from langbot.libs.dingtalk_api.api import DingTalkClient
+from .subscriptions import DingTalkSubscriptionHandler
+
+from langbot.libs.dingtalk_api.api import DingTalkClient, is_stream_card_configured
 from langbot.libs.dingtalk_api.dingtalkevent import DingTalkEvent
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
 import langbot_plugin.api.definition.abstract.platform.event_logger as abstract_platform_logger
@@ -41,6 +42,7 @@ class DingTalkCardCallbackHandler(dingtalk_stream.CallbackHandler):
             {
                 'conversation_type': 'CardCallback',
                 'Type': 'card_callback',
+                'timestamp': float(message.headers.time or 0) / 1000,
                 'CardCallback': {
                     'extension': callback.extension,
                     'corp_id': callback.corp_id,
@@ -77,7 +79,7 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         arbitrary_types_allowed = True
 
     def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger):
-        required_keys = ['client_id', 'client_secret', 'robot_name', 'robot_code']
+        required_keys = ['client_id', 'client_secret']
         missing_keys = [key for key in required_keys if key not in config]
         if missing_keys:
             raise Exception('钉钉缺少相关配置项，请查看文档或联系管理员')
@@ -85,8 +87,8 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         bot = DingTalkClient(
             client_id=config['client_id'],
             client_secret=config['client_secret'],
-            robot_name=config['robot_name'],
-            robot_code=config['robot_code'],
+            robot_name=config.get('robot_name', ''),
+            robot_code=config.get('robot_code', ''),
             markdown_card=config.get('markdown_card', True),
             logger=logger,
         )
@@ -94,7 +96,7 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
             config=config,
             logger=logger,
             card_instance_id_dict={},
-            bot_account_id=config['robot_name'],
+            bot_account_id=config['client_id'],
             bot=bot,
             listeners={},
             _message_cache={},
@@ -108,6 +110,10 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         return [
             'message.received',
             'feedback.received',
+            'group.member_joined',
+            'group.member_left',
+            'group.info_updated',
+            'dingtalk.card_action',
             'platform.specific',
         ]
 
@@ -135,7 +141,6 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
     def _plain_message(text: str) -> platform_message.MessageChain:
         return platform_message.MessageChain([platform_message.Plain(text=text)])
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -152,7 +157,6 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
             raise ValueError(f'Unsupported dingtalk target_type: {target_type}')
         return platform_events.MessageResult(raw=raw if isinstance(raw, dict) else {'result': raw})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -162,14 +166,28 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         assert isinstance(message_source.source_platform_object, DingTalkEvent)
         incoming_message = message_source.source_platform_object.incoming_message
         markdown_enabled = self.config.get('markdown_card', False)
-        content, at = await DingTalkMessageConverter.yiri2target(message, markdown_enabled)
+        text_message = platform_message.MessageChain(
+            [part for part in message if not isinstance(part, (platform_message.Image, platform_message.File))]
+        )
+        content, at = await DingTalkMessageConverter.yiri2target(text_message, markdown_enabled)
         raw = await self.bot.send_message(content, incoming_message, at)
+        await self._send_attachments(message, incoming_message)
         return platform_events.MessageResult(
             message_id=getattr(incoming_message, 'message_id', None),
             raw=raw if isinstance(raw, dict) else {'result': raw},
         )
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
+    async def _send_attachments(self, message, incoming_message):
+        failures = []
+        for part in message:
+            if isinstance(part, (platform_message.Image, platform_message.File)):
+                try:
+                    await self.bot.send_attachment(part, incoming_message)
+                except Exception as exc:
+                    failures.append(str(exc))
+        if failures:
+            raise RuntimeError('; '.join(failures))
+
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -179,12 +197,19 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         is_final: bool = False,
     ):
         message_id = bot_message.resp_message_id
-        msg_seq = bot_message.msg_sequence
-        if (msg_seq - 1) % 8 != 0 and not is_final:
+        now = time.monotonic()
+        last_updates = getattr(self, '_card_last_updates', None)
+        if last_updates is None:
+            self._card_last_updates = last_updates = {}
+        if not is_final and now - last_updates.get(message_id, 0) < 0.35:
             return
+        last_updates[message_id] = now
 
         markdown_enabled = self.config.get('markdown_card', False)
-        content, _ = await DingTalkMessageConverter.yiri2target(message, markdown_enabled)
+        text_message = platform_message.MessageChain(
+            [part for part in message if not isinstance(part, (platform_message.Image, platform_message.File))]
+        )
+        content, _ = await DingTalkMessageConverter.yiri2target(text_message, markdown_enabled)
         card_instance, card_instance_id = self.card_instance_id_dict[message_id]
         if not content and bot_message.content:
             content = bot_message.content
@@ -192,14 +217,17 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
             await self.bot.send_card_message(card_instance, card_instance_id, content, is_final)
         if is_final and bot_message.tool_calls is None:
             self.card_instance_id_dict.pop(message_id)
+            last_updates.pop(message_id, None)
+            await self._send_attachments(message, message_source.source_platform_object.incoming_message)
 
-    @diagnostics.observe('api', 'create_message_card', source='platform', stage='accepted')
     async def create_message_card(self, message_id, event):
         while len(self.card_instance_id_dict) >= 1000:
-            self.card_instance_id_dict.pop(next(iter(self.card_instance_id_dict)), None)
-        card_template_id = self.config['card_template_id']
+            oldest_id = next(iter(self.card_instance_id_dict))
+            self.card_instance_id_dict.pop(oldest_id, None)
+            getattr(self, '_card_last_updates', {}).pop(oldest_id, None)
+        card_template_id = self.config['card_template_id'].strip()
         incoming_message = event.source_platform_object.incoming_message
-        card_auto_layout = self.config.get('card_auto_layout', False)
+        card_auto_layout = self.config.get('card_auto_layout', True)
         card_instance, card_instance_id = await self.bot.create_and_card(
             card_template_id,
             incoming_message,
@@ -209,9 +237,8 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         return True
 
     async def is_stream_output_supported(self) -> bool:
-        return bool(self.config.get('enable-stream-reply', False))
+        return is_stream_card_configured(self.config)
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         if action == 'interaction.request' and action in self.get_supported_apis():
             return await send_interaction(self, params)
@@ -260,6 +287,7 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
         async def on_message(event: DingTalkEvent):
             await self._handle_native_event(event)
 
+        self.bot.client.register_all_event_handler(DingTalkSubscriptionHandler(self))
         self.bot.on_message('FriendMessage')(on_message)
         self.bot.on_message('GroupMessage')(on_message)
         self.bot.client.register_callback_handler(
@@ -267,9 +295,9 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
             DingTalkCardCallbackHandler(self),
         )
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: DingTalkEvent):
         try:
+            self.bot_account_id = getattr(event.incoming_message, 'chatbot_user_id', None) or self.bot_account_id
             interaction_event = interaction_event_from_native(event, self.interaction_callback_contexts)
             if interaction_event is not None:
                 await self._dispatch_eba_event(interaction_event)
@@ -278,14 +306,16 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
                 'DingTalk event received: '
                 f'conversation={event.conversation}, message_id={getattr(event.incoming_message, "message_id", None)}'
             )
-            if platform_events.FriendMessage in self.listeners or platform_events.GroupMessage in self.listeners:
-                legacy_event = await self.event_converter.target2legacy(event, self.config['robot_name'])
+            if event.conversation in {'FriendMessage', 'GroupMessage'} and (
+                platform_events.FriendMessage in self.listeners or platform_events.GroupMessage in self.listeners
+            ):
+                legacy_event = await self.event_converter.target2legacy(event, self.bot_account_id)
                 if legacy_event:
                     callback = self.listeners.get(type(legacy_event))
                     if callback:
                         await callback(legacy_event, self)
 
-            eba_event = await self.event_converter.target2yiri(event, self.config['robot_name'])
+            eba_event = await self.event_converter.target2yiri(event, self.bot_account_id)
             if eba_event:
                 self._cache_event(eba_event)
                 await self._dispatch_eba_event(eba_event)
@@ -293,7 +323,6 @@ class DingTalkAdapter(DingTalkAPIMixin, abstract_platform_adapter.AbstractPlatfo
             await self.logger.error(f'Error in dingtalk native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/api_impl.py b/src/langbot/pkg/platform/adapters/dingtalk/api_impl.py
index 358e7a29d..6db28ef4d 100644
--- a/src/langbot/pkg/platform/adapters/dingtalk/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/dingtalk/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.dingtalk_api.api import DingTalkClient
@@ -16,7 +14,6 @@ class DingTalkAPIMixin:
     _user_cache: dict[str, platform_entities.User]
     _group_cache: dict[str, platform_entities.UserGroup]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -28,22 +25,18 @@ class DingTalkAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         return self._group_cache.get(str(group_id)) or platform_entities.UserGroup(id=group_id, name='')
 
-    @diagnostics.observe('api', 'get_group_list', source='platform', stage='accepted')
     async def get_group_list(self) -> list[platform_entities.UserGroup]:
         return list(self._group_cache.values())
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
     ) -> list[platform_entities.UserGroupMember]:
         raise NotSupportedError('get_group_member_list')
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -59,18 +52,14 @@ class DingTalkAPIMixin:
             display_name=user.nickname,
         )
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         return self._user_cache.get(str(user_id)) or platform_entities.User(id=user_id, nickname='')
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         return await self.bot.get_file_url(file_id)
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/event_converter.py b/src/langbot/pkg/platform/adapters/dingtalk/event_converter.py
index b3b23da2c..dd7798b2d 100644
--- a/src/langbot/pkg/platform/adapters/dingtalk/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/dingtalk/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.dingtalk_api.dingtalkevent import DingTalkEvent
@@ -18,32 +16,41 @@ class DingTalkEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return getattr(event, 'source_platform_object', None)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
-    async def target2yiri(event: DingTalkEvent, bot_name: str) -> platform_events.Event | None:
+    async def target2yiri(event: DingTalkEvent, bot_id: str) -> platform_events.Event | None:
         if event.conversation in {'FriendMessage', 'GroupMessage'}:
-            return await DingTalkEventConverter.message_to_eba(event, bot_name)
+            return await DingTalkEventConverter.message_to_eba(event, bot_id)
         if event.conversation == 'CardCallback':
             feedback = DingTalkEventConverter.card_callback_to_feedback(event)
             if feedback is not None:
                 return feedback
-            return DingTalkEventConverter.platform_specific(event, 'card.callback')
+            callback = event.get('CardCallback') or {}
+            content = callback.get('content') or {}
+            return platform_events.DingTalkCardActionEvent(
+                adapter_name=ADAPTER_NAME,
+                user=platform_entities.User(id=str(callback.get('user_id') or '')),
+                space_id=str(callback.get('space_id') or ''),
+                card_instance_id=str(callback.get('card_instance_id') or ''),
+                action=str(content.get('action') or content.get('actionKey') or ''),
+                data=callback,
+                timestamp=float(event.get('timestamp') or 0),
+                source_platform_object=event,
+            )
         return DingTalkEventConverter.platform_specific(event, f'message.{event.conversation or "unknown"}')
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         event: DingTalkEvent,
-        bot_name: str,
+        bot_id: str,
     ) -> platform_events.FriendMessage | platform_events.GroupMessage | None:
-        eba_event = await DingTalkEventConverter.message_to_eba(event, bot_name)
+        eba_event = await DingTalkEventConverter.message_to_eba(event, bot_id)
         if eba_event:
             return eba_event.to_legacy_event()
         return None
 
     @staticmethod
-    async def message_to_eba(event: DingTalkEvent, bot_name: str) -> platform_events.MessageReceivedEvent:
+    async def message_to_eba(event: DingTalkEvent, bot_id: str) -> platform_events.MessageReceivedEvent:
         incoming_message = event.incoming_message
-        message_chain = await DingTalkMessageConverter.target2yiri(event, bot_name)
+        message_chain = await DingTalkMessageConverter.target2yiri(event, bot_id)
         sender = DingTalkEventConverter.user_from_event(event)
         chat_type = platform_entities.ChatType.PRIVATE
         chat_id = getattr(incoming_message, 'sender_staff_id', '')
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/manifest.yaml b/src/langbot/pkg/platform/adapters/dingtalk/manifest.yaml
index 8e74f21d1..11a0b0e1f 100644
--- a/src/langbot/pkg/platform/adapters/dingtalk/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/dingtalk/manifest.yaml
@@ -15,6 +15,7 @@ metadata:
 
 spec:
   categories:
+    - popular
     - china
   help_links:
     zh: https://langbot.app/docs/zh/usage/platforms/dingtalk
@@ -27,36 +28,12 @@ spec:
         zh_Hans: 一键创建应用
         zh_Hant: 一鍵建立應用
       description:
-        en_US: 'Scan QR code with DingTalk to automatically create an app and fill in credentials. Note: Robot Code cannot be
-          obtained automatically, you need to copy it from the DingTalk Developer Backend manually.'
-        zh_Hans: 使用钉钉扫码自动创建应用并填写凭据。注意：机器人代码无法自动获取，需前往钉钉开发者后台手动复制。
-        zh_Hant: 使用釘釘掃碼自動建立應用並填寫憑證。注意：機器人代碼無法自動取得，需前往釘釘開發者後台手動複製。
+        en_US: Scan with DingTalk to create an app and automatically fill in credentials. The bot identity is resolved automatically.
+        zh_Hans: 使用钉钉扫码创建应用并自动填写凭据，机器人身份会自动识别。
+        zh_Hant: 使用釘釘掃碼建立應用並自動填寫憑證，機器人身分會自動識別。
       type: qr-code-login
       login_platform: dingtalk
       required: false
-    - name: human_input_card_template_download
-      label:
-        en_US: Download Human Input Card Template
-        zh_Hans: 下载人工输入卡片模板
-        zh_Hant: 下載人工輸入卡片範本
-      description:
-        en_US: Download the template, import it in DingTalk Card Platform, then enter its template ID below to enable native interactions.
-        zh_Hans: 下载模板并导入钉钉卡片平台，再将模板 ID 填入下方以启用原生交互。
-        zh_Hant: 下載範本並匯入釘釘卡片平台，再將範本 ID 填入下方以啟用原生互動。
-      type: download-link
-      required: false
-      default: ''
-      url: /api/v1/platform/adapters/dingtalk/human-input-card-template
-      download_filename: dingtalk_human_input_card.json
-      help_links:
-        zh: https://open-dev.dingtalk.com/fe/card
-        en: https://open-dev.dingtalk.com/fe/card
-        ja: https://open-dev.dingtalk.com/fe/card
-      help_label:
-        en_US: Import Guide
-        zh_Hans: 导入指引
-        zh_Hant: 匯入指引
-        ja_JP: インポート手順
     - name: client_id
       label:
         en_US: Client ID
@@ -73,75 +50,111 @@ spec:
       type: string
       required: true
       default: ""
-    - name: robot_code
-      label:
-        en_US: Robot Code
-        zh_Hans: 机器人代码
-        zh_Hant: 機器人代碼
-      type: string
-      required: true
-      default: ""
-    - name: robot_name
-      label:
-        en_US: Robot Name
-        zh_Hans: 机器人名称
-        zh_Hant: 機器人名稱
-      type: string
-      required: true
-      default: ""
     - name: markdown_card
       label:
-        en_US: Markdown Card
-        zh_Hans: 是否使用 Markdown 卡片
-        zh_Hant: 是否使用 Markdown 卡片
+        en_US: Use Card Replies
+        zh_Hans: 使用卡片回复
+        zh_Hant: 使用卡片回覆
+      description:
+        en_US: Controls Markdown and streaming card replies. Disable to use plain text while keeping your template settings. When enabled, a configured template automatically enables streaming.
+        zh_Hans: 控制 Markdown 和流式卡片回复。关闭后使用纯文本，保留模板配置；开启且已配置模板时自动使用流式回复。
+        zh_Hant: 控制 Markdown 和串流卡片回覆。關閉後使用純文字，保留範本設定；開啟且已設定範本時自動使用串流回覆。
       type: boolean
       required: false
       default: true
-    - name: enable-stream-reply
-      label:
-        en_US: Enable Stream Reply Mode
-        zh_Hans: 启用钉钉卡片流式回复模式
-        zh_Hant: 啟用釘釘卡片串流回覆模式
-      description:
-        en_US: Used for streaming replies from pipelines or plugins. If enabled, the bot will use DingTalk card streaming replies.
-        zh_Hans: 用于流水线和插件主动发起的流式回复。如果启用，将使用钉钉卡片流式方式来回复内容
-        zh_Hant: 用於流水線和插件主動發起的串流回覆。如果啟用，將使用釘釘卡片串流方式來回覆內容
-      type: boolean
-      required: true
-      default: false
-    - name: card_auto_layout
-      label:
-        en_US: Card Auto Layout
-        zh_Hans: 卡片宽屏自动布局
-        zh_Hant: 卡片寬螢幕自動佈局
-      type: boolean
-      required: false
-      default: false
     - name: card_template_id
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
       label:
         en_US: Card Template ID
         zh_Hans: 卡片模板ID
         zh_Hant: 卡片範本ID
+      description:
+        en_US: Template for streaming text replies. Save and publish it in DingTalk Card Platform, then paste its ID here to automatically enable streaming while card replies are enabled. Leave empty for ordinary replies. Separate from the human input template.
+        zh_Hans: 用于流式文字回复。在钉钉卡片平台保存并发布模板后，开启卡片回复后，填写 ID 即自动启用流式回复，留空使用普通回复；与人工输入卡片模板分开配置。
+        zh_Hant: 用於串流文字回覆。在釘釘卡片平台儲存並發布範本後，開啟卡片回覆後，填寫 ID 即自動啟用串流回覆，留空使用一般回覆；與人工輸入卡片範本分開設定。
       type: string
-      required: true
-      default: "填写你的卡片template_id"
-
+      required: false
+      default: ""
+    - name: stream_card_template_download
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
+      label:
+        en_US: Download Streaming Reply Card Template
+        zh_Hans: 下载流式回复卡片模板
+        zh_Hant: 下載串流回覆卡片範本
+      description:
+        en_US: Displays reasoning and the answer separately. Import through More > Import Card Template in the DingTalk AI card editor, save and publish, then paste its ID into Card Template ID above. Respects the pipeline thinking removal setting.
+        zh_Hans: 分开展示思考与正文。在钉钉 AI 卡片编辑器中选择「更多 → 导入卡片模板」，保存并发布后，将 ID 填入上方「卡片模板 ID」。是否保留思考仍由流水线设置控制。
+        zh_Hant: 分開展示思考與正文。在釘釘 AI 卡片編輯器中選擇「更多 → 匯入卡片範本」，儲存並發布後，將 ID 填入上方「卡片範本 ID」。是否保留思考仍由流水線設定控制。
+      type: download-link
+      required: false
+      default: ''
+      url: /api/v1/platform/adapters/dingtalk/stream-card-template
+      download_filename: dingtalk_stream_card.json
+    - name: card_auto_layout
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
+      label:
+        en_US: Card Auto Layout
+        zh_Hans: 卡片宽屏自动布局
+        zh_Hant: 卡片寬螢幕自動佈局
+      description:
+        en_US: Enabled by default. Allows supported cards to adapt to wider windows. The LangBot streaming reply template includes the required variables; you can disable this for a narrower layout.
+        zh_Hans: 默认开启，让支持的卡片适应较宽的聊天窗口。LangBot 流式回复模板已包含所需变量；如需较窄布局，可关闭此项。
+        zh_Hant: 預設開啟，讓支援的卡片適應較寬的聊天視窗。LangBot 串流回覆範本已包含所需變數；如需較窄版面，可關閉此項。
+      type: boolean
+      required: false
+      default: true
     - name: human_input_card_template_id
       label:
         en_US: Human Input Card Template ID
         zh_Hans: 人工输入卡片模板 ID
         zh_Hant: 人工輸入卡片範本 ID
       description:
-        en_US: Import the bundled `src/langbot/templates/dingtalk_human_input_card.json`, then enter its DingTalk template ID here. It contains the `content`, `btns`, `input_*`, `select_*`, and `index_o` variables required for native interactions.
-        zh_Hans: 可选。模板需包含 `content`、`btns`、`input_*`、`select_*` 和 `index_o` 变量以支持原生交互。
-        zh_Hant: 可選。範本需包含 `content`、`btns`、`input_*`、`select_*` 和 `index_o` 變數以支援原生互動。
+        en_US: Optional. Used when a workflow or plugin requests user input, a choice, or confirmation before continuing (for example, a Dify Human Input node). Ordinary chat does not require it. Import the template below and enter its published ID here; streaming text replies use the separate Card Template ID.
+        zh_Hans: 可选。当工作流或插件需要用户填写信息、选择选项或确认后再继续执行时使用，例如 Dify 的人工输入节点。普通聊天无需配置。请导入下方模板，将发布后的 ID 填在这里；流式文字回复使用单独的「卡片模板 ID」。
+        zh_Hant: 可選。當工作流或外掛需要使用者填寫資訊、選擇選項或確認後再繼續執行時使用，例如 Dify 的人工輸入節點。一般聊天無需設定。請匯入下方範本，將發布後的 ID 填在這裡；串流文字回覆使用獨立的「卡片範本 ID」。
       type: string
       required: false
       default: ""
+    - name: human_input_card_template_download
+      label:
+        en_US: Download Human Input Card Template
+        zh_Hans: 下载人工输入卡片模板
+        zh_Hant: 下載人工輸入卡片範本
+      description:
+        en_US: Download the human input template. Create an AI card in DingTalk Card Platform, choose More > Import Card Template in the editor, then save and publish. Paste its ID into the field above.
+        zh_Hans: 下载人工输入模板后，在钉钉卡片平台新建 AI 卡片，在编辑器中选择「更多 → 导入卡片模板」，保存并发布，将 ID 填入上方字段。
+        zh_Hant: 下載人工輸入範本後，在釘釘卡片平台新建 AI 卡片，在編輯器中選擇「更多 → 匯入卡片範本」，儲存並發布，將 ID 填入上方欄位。
+      type: download-link
+      required: false
+      default: ''
+      url: /api/v1/platform/adapters/dingtalk/human-input-card-template
+      download_filename: dingtalk_human_input_card.json
+      help_links:
+        zh: https://langbot.app/docs/zh/usage/platforms/dingtalk#导入人工输入卡片模板
+        en: https://open-dev.dingtalk.com/fe/card
+        ja: https://open-dev.dingtalk.com/fe/card
+      help_label:
+        en_US: Import Guide
+        zh_Hans: 导入指引
+        zh_Hant: 匯入指引
+        ja_JP: インポート手順
 
   supported_events:
     - message.received
     - feedback.received
+    - group.member_joined
+    - group.member_left
+    - group.info_updated
+    - dingtalk.card_action
     - platform.specific
 
   supported_apis:
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/message_converter.py b/src/langbot/pkg/platform/adapters/dingtalk/message_converter.py
index 92bd44d6b..253eed8d1 100644
--- a/src/langbot/pkg/platform/adapters/dingtalk/message_converter.py
+++ b/src/langbot/pkg/platform/adapters/dingtalk/message_converter.py
@@ -86,7 +86,7 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
         return content, at
 
     @staticmethod
-    async def target2yiri(event: DingTalkEvent, bot_name: str) -> platform_message.MessageChain:
+    async def target2yiri(event: DingTalkEvent, bot_id: str) -> platform_message.MessageChain:
         incoming_message = event.incoming_message
         components: list[platform_message.MessageComponent] = [
             platform_message.Source(
@@ -95,15 +95,19 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             )
         ]
 
-        for at_user in getattr(incoming_message, 'at_users', []) or []:
-            if getattr(at_user, 'dingtalk_id', None) == getattr(incoming_message, 'chatbot_user_id', None):
-                components.append(platform_message.At(target=bot_name, display=bot_name))
+        chatbot_id = getattr(incoming_message, 'chatbot_user_id', None)
+        mentioned = getattr(incoming_message, 'is_in_at_list', False) or any(
+            chatbot_id and getattr(at_user, 'dingtalk_id', None) == chatbot_id
+            for at_user in getattr(incoming_message, 'at_users', []) or []
+        )
+        if mentioned:
+            components.append(platform_message.At(target=bot_id))
 
         rich_content = event.rich_content
         if rich_content:
             for element in rich_content.get('Elements') or []:
                 if element.get('Type') == 'text':
-                    text = DingTalkMessageConverter._strip_bot_mention(element.get('Content', ''), bot_name)
+                    text = element.get('Content', '')
                     if text.strip():
                         components.append(platform_message.Plain(text=text))
                 elif element.get('Type') == 'image' and element.get('Picture'):
@@ -112,7 +116,7 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             if event.content and event.type != 'audio':
                 components.append(
                     platform_message.Plain(
-                        text=DingTalkMessageConverter._strip_bot_mention(event.content, bot_name),
+                        text=event.content,
                     )
                 )
             if event.picture:
@@ -160,10 +164,6 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             origin=platform_message.MessageChain(origin_components),
         )
 
-    @staticmethod
-    def _strip_bot_mention(text: str, bot_name: str) -> str:
-        return text.replace('@' + bot_name, '')
-
     @staticmethod
     def _message_time(incoming_message: typing.Any) -> datetime.datetime:
         value = getattr(incoming_message, 'create_at', None)
diff --git a/src/langbot/pkg/platform/adapters/dingtalk/subscriptions.py b/src/langbot/pkg/platform/adapters/dingtalk/subscriptions.py
new file mode 100644
index 000000000..3ebabed8d
--- /dev/null
+++ b/src/langbot/pkg/platform/adapters/dingtalk/subscriptions.py
@@ -0,0 +1,91 @@
+"""Translate DingTalk Stream EVENT envelopes, keeping unknown payloads intact."""
+
+import asyncio
+from collections import OrderedDict
+
+import dingtalk_stream
+from langbot_plugin.api.entities.builtin.platform import entities, events
+
+from .types import ADAPTER_NAME
+
+
+def subscription_events(message):
+    data = message.data
+    headers = message.headers
+    action = headers.event_type or data.get('EventType') or data.get('eventType') or ''
+    raw = {'headers': vars(headers).copy(), 'data': data}
+    stamp = (
+        headers.event_born_time
+        or data.get('TimeStamp')
+        or data.get('EventTime')
+        or data.get('timestamp')
+        or headers.time
+        or 0
+    )
+    stamp = float(stamp)
+    common = dict(
+        adapter_name=ADAPTER_NAME,
+        timestamp=stamp / 1000 if stamp > 10_000_000_000 else stamp,
+        source_platform_object=raw,
+    )
+    group_id = (
+        data.get('OpenConversationId') or data.get('openConversationId') or data.get('chatId') or data.get('ChatId')
+    )
+    group = entities.UserGroup(id=str(group_id or ''), name=str(data.get('Title') or data.get('title') or ''))
+    owner_id = data.get('Owner') or data.get('owner')
+    if owner_id:
+        group.owner_id = str(owner_id)
+    operator_id = data.get('Operator') or data.get('operator')
+    operator = entities.User(id=str(operator_id)) if operator_id else None
+    members = data.get('UserId') or data.get('userId') or []
+    if isinstance(members, str):
+        members = [members]
+    if group_id and members and action in {'chat_add_member', 'chat_remove_member', 'chat_quit'}:
+        for member_id in dict.fromkeys(members):
+            member = entities.User(id=str(member_id))
+            if action == 'chat_add_member':
+                yield events.MemberJoinedEvent(group=group, member=member, inviter=operator, **common)
+            else:
+                yield events.MemberLeftEvent(
+                    group=group,
+                    member=member,
+                    operator=operator,
+                    is_kicked=action == 'chat_remove_member',
+                    **common,
+                )
+    elif group_id and action in {'chat_update_title', 'chat_update_owner'}:
+        yield events.GroupInfoUpdatedEvent(
+            group=group,
+            operator=operator,
+            changed_fields=['name' if action == 'chat_update_title' else 'owner_id'],
+            **common,
+        )
+    else:
+        yield events.PlatformSpecificEvent(action=action or 'unknown', data=raw, **common)
+
+
+class DingTalkSubscriptionHandler(dingtalk_stream.EventHandler):
+    def __init__(self, adapter):
+        super().__init__()
+        self.adapter = adapter
+        self.completed = OrderedDict()
+        self.lock = asyncio.Lock()
+
+    async def process(self, message):
+        # Serialize retries; mark each fan-out event only after successful dispatch.
+        async with self.lock:
+            try:
+                identity = message.headers.event_id or message.headers.message_id
+                for index, event in enumerate(subscription_events(message)):
+                    key = (message.headers.event_corp_id, identity, index)
+                    if identity and key in self.completed:
+                        continue
+                    await self.adapter._dispatch_eba_event(event)
+                    if identity:
+                        self.completed[key] = None
+                        while len(self.completed) > 4096:
+                            self.completed.popitem(last=False)
+                return dingtalk_stream.AckMessage.STATUS_OK, 'OK'
+            except Exception:
+                self.logger.exception('DingTalk subscription dispatch failed')
+                return 500, 'Event dispatch failed'
diff --git a/src/langbot/pkg/platform/adapters/discord/adapter.py b/src/langbot/pkg/platform/adapters/discord/adapter.py
index 73da2f1ab..8ace25ba7 100644
--- a/src/langbot/pkg/platform/adapters/discord/adapter.py
+++ b/src/langbot/pkg/platform/adapters/discord/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import os
 import time
 import traceback
@@ -47,30 +45,10 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
         adapter_self = self
 
         class LangBotDiscordClient(discord.Client):
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_ready(self: discord.Client):
                 adapter_self.bot_account_id = str(self.user.id) if self.user else ''
                 await adapter_self.logger.info(f'Discord adapter running as {self.user}')
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_message(self: discord.Client, message: discord.Message):
                 if self.user and message.author.id == self.user.id:
                     return
@@ -94,16 +72,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                 except Exception:
                     await adapter_self.logger.error(f'Error in discord on_message: {traceback.format_exc()}')
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_interaction(self: discord.Client, interaction: discord.Interaction):
                 custom_id = (interaction.data or {}).get('custom_id') if isinstance(interaction.data, dict) else None
                 try:
@@ -123,46 +91,16 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                 except Exception:
                     await adapter_self.logger.error(f'Error in Discord interaction callback: {traceback.format_exc()}')
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_message_edit(self: discord.Client, before: discord.Message, after: discord.Message):
                 await adapter_self._dispatch_gateway_tuple(
                     'message_edit', (before, after), self.user.id if self.user else None
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_message_delete(self: discord.Client, message: discord.Message):
                 await adapter_self._dispatch_gateway_tuple(
                     'message_delete', message, self.user.id if self.user else None
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_raw_message_delete(self: discord.Client, payload: discord.RawMessageDeleteEvent):
                 await adapter_self._dispatch_gateway_tuple(
                     'raw_message_delete',
@@ -170,16 +108,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                     self.user.id if self.user else None,
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_reaction_add(
                 self: discord.Client, reaction: discord.Reaction, user: discord.User | discord.Member
             ):
@@ -189,16 +117,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                     'reaction_add', (reaction, user), self.user.id if self.user else None
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_reaction_remove(
                 self: discord.Client, reaction: discord.Reaction, user: discord.User | discord.Member
             ):
@@ -208,16 +126,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                     'reaction_remove', (reaction, user), self.user.id if self.user else None
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_raw_reaction_add(self: discord.Client, payload: discord.RawReactionActionEvent):
                 if self.user and payload.user_id == self.user.id:
                     return
@@ -227,16 +135,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                     self.user.id if self.user else None,
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_raw_reaction_remove(self: discord.Client, payload: discord.RawReactionActionEvent):
                 if self.user and payload.user_id == self.user.id:
                     return
@@ -246,55 +144,15 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
                     self.user.id if self.user else None,
                 )
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_member_join(self: discord.Client, member: discord.Member):
                 await adapter_self._dispatch_gateway_tuple('member_join', member, self.user.id if self.user else None)
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_member_remove(self: discord.Client, member: discord.Member):
                 await adapter_self._dispatch_gateway_tuple('member_remove', member, self.user.id if self.user else None)
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_guild_join(self: discord.Client, guild: discord.Guild):
                 await adapter_self._dispatch_gateway_tuple('guild_join', guild, self.user.id if self.user else None)
 
-            @diagnostics.observe(
-                'event',
-                'platform.native_callback',
-                source='platform',
-                stage='convert',
-                ap=lambda: getattr(logger, 'ap', None),
-                fields=lambda b: {
-                    'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-                },
-            )
             async def on_guild_remove(self: discord.Client, guild: discord.Guild):
                 await adapter_self._dispatch_gateway_tuple('guild_remove', guild, self.user.id if self.user else None)
 
@@ -352,7 +210,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
     def get_interaction_capabilities(self) -> dict[str, typing.Any]:
         return interaction_delivery_capabilities()
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
         content, files = await self.message_converter.yiri2target(message)
         channel = await self._get_channel(target_id)
@@ -362,7 +219,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
         sent = await channel.send(**kwargs)
         return platform_events.MessageResult(message_id=sent.id, raw={'message_id': sent.id})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -380,7 +236,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
         sent = await message_source.source_platform_object.channel.send(**kwargs)
         return platform_events.MessageResult(message_id=sent.id, raw={'message_id': sent.id})
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _dispatch_gateway_tuple(self, kind: str, payload, bot_user_id: int | None):
         try:
             event = await self.event_converter.target2yiri((kind, payload), bot_user_id)
@@ -390,7 +245,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
             await self.logger.error(f'Error in discord {kind}: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
@@ -415,7 +269,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
     ):
         self.listeners.pop(event_type, None)
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         if action == 'interaction.request':
             return await send_interaction(self, params)
@@ -437,7 +290,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
     async def is_stream_output_supported(self) -> bool:
         return True
 
-    @diagnostics.observe('api', 'create_message_card', source='platform', stage='accepted')
     async def create_message_card(self, message_id: str, event: platform_events.MessageEvent) -> bool:
         """Set up a stream context for progressive editing.
 
@@ -457,7 +309,6 @@ class DiscordAdapter(DiscordAPIMixin, abstract_platform_adapter.AbstractPlatform
         }
         return True
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
diff --git a/src/langbot/pkg/platform/adapters/discord/api_impl.py b/src/langbot/pkg/platform/adapters/discord/api_impl.py
index 4c4ee9ecc..767ae9a27 100644
--- a/src/langbot/pkg/platform/adapters/discord/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/discord/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import datetime
 import typing
 
@@ -17,7 +15,6 @@ from langbot_plugin.api.entities.builtin.platform import message as platform_mes
 class DiscordAPIMixin:
     bot: discord.Client
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -34,7 +31,6 @@ class DiscordAPIMixin:
             return
         await message.edit(content=content)
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -45,7 +41,6 @@ class DiscordAPIMixin:
         message = await channel.fetch_message(int(message_id))
         await message.delete()
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -61,12 +56,10 @@ class DiscordAPIMixin:
         sent = await to_channel.send(content=message.content, files=files)
         return platform_events.MessageResult(message_id=sent.id, raw={'message_id': sent.id})
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         guild = await self._get_guild(group_id)
         return DiscordEventConverter.group_from_guild(guild)
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -75,7 +68,6 @@ class DiscordAPIMixin:
         members = guild.members or [member async for member in guild.fetch_members(limit=None)]
         return [self._member_to_entity(member) for member in members]
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -85,22 +77,18 @@ class DiscordAPIMixin:
         member = guild.get_member(int(user_id)) or await guild.fetch_member(int(user_id))
         return self._member_to_entity(member)
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         user = self.bot.get_user(int(user_id)) or await self.bot.fetch_user(int(user_id))
         return DiscordEventConverter.user_from_author(user)
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         from langbot_plugin.api.entities.builtin.platform.errors import NotSupportedError
 
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         return file_id
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(
         self,
         group_id: typing.Union[int, str],
@@ -114,7 +102,6 @@ class DiscordAPIMixin:
             until = datetime.datetime.now(datetime.UTC) + datetime.timedelta(seconds=duration)
         await member.timeout(until, reason='LangBot Omni mute_member')
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(
         self,
         group_id: typing.Union[int, str],
@@ -124,7 +111,6 @@ class DiscordAPIMixin:
         member = guild.get_member(int(user_id)) or await guild.fetch_member(int(user_id))
         await member.timeout(None, reason='LangBot Omni unmute_member')
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(
         self,
         group_id: typing.Union[int, str],
@@ -134,7 +120,6 @@ class DiscordAPIMixin:
         member = guild.get_member(int(user_id)) or await guild.fetch_member(int(user_id))
         await member.kick(reason='LangBot Omni kick_member')
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(self, group_id: typing.Union[int, str]) -> None:
         guild = await self._get_guild(group_id)
         await guild.leave()
diff --git a/src/langbot/pkg/platform/adapters/discord/event_converter.py b/src/langbot/pkg/platform/adapters/discord/event_converter.py
index b590b2f50..d34839725 100644
--- a/src/langbot/pkg/platform/adapters/discord/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/discord/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 import discord
@@ -18,7 +16,6 @@ class DiscordEventConverter(abstract_platform_adapter.AbstractEventConverter):
         raise NotImplementedError
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(event: typing.Any, bot_user_id: int | None = None) -> platform_events.Event | None:
         if isinstance(event, discord.Message):
             return await DiscordEventConverter.message_to_eba(event)
@@ -241,7 +238,6 @@ class DiscordEventConverter(abstract_platform_adapter.AbstractEventConverter):
         )
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(message: discord.Message) -> platform_events.FriendMessage | platform_events.GroupMessage:
         message_chain = await DiscordMessageConverter.target2yiri(message)
         if isinstance(message.channel, discord.DMChannel):
diff --git a/src/langbot/pkg/platform/adapters/kook/adapter.py b/src/langbot/pkg/platform/adapters/kook/adapter.py
index 198d1d186..b89bcb3f9 100644
--- a/src/langbot/pkg/platform/adapters/kook/adapter.py
+++ b/src/langbot/pkg/platform/adapters/kook/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 from langbot.pkg.platform.sources.kook import _decode_gateway_message
 
 import asyncio
@@ -101,7 +99,6 @@ class KookAdapter(KookAPIMixin, BasePlatformAdapter):
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -171,7 +168,6 @@ class KookAdapter(KookAPIMixin, BasePlatformAdapter):
         self.session_id = str(data.get('session_id') or '')
         await self.logger.info(f'KOOK WebSocket HELLO received, session_id: {self.session_id}')
 
-    @diagnostics.observe('event', 'platform.receive', source='platform', stage='accepted')
     async def _handle_event(self, data: dict, sn: int):
         self.current_sn = max(self.current_sn, sn)
 
@@ -196,13 +192,10 @@ class KookAdapter(KookAPIMixin, BasePlatformAdapter):
             if eba_event:
                 self._cache_event(eba_event)
                 await self._dispatch_eba_event(eba_event)
-        except Exception as exc:
-            diagnostics.annotate(error=exc)
-            diagnostics.set_outcome('failed')
+        except Exception:
             await self.logger.error(f'Error handling KOOK event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/kook/api_impl.py b/src/langbot/pkg/platform/adapters/kook/api_impl.py
index 6170479b0..1d48b36d8 100644
--- a/src/langbot/pkg/platform/adapters/kook/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/kook/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.pkg.platform.adapters.kook.message_converter import KookMessageConverter
@@ -16,7 +14,6 @@ class KookAPIMixin:
     _user_cache: dict[str, platform_entities.User]
     _group_cache: dict[str, platform_entities.UserGroup]
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -37,7 +34,6 @@ class KookAPIMixin:
         data = raw.get('data') or {}
         return platform_events.MessageResult(message_id=data.get('msg_id'), raw=raw)
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -77,7 +73,6 @@ class KookAPIMixin:
         data = raw.get('data') or {}
         return platform_events.MessageResult(message_id=data.get('msg_id'), raw=raw)
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -89,7 +84,6 @@ class KookAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         cached = self._group_cache.get(str(group_id))
         if cached:
@@ -102,18 +96,15 @@ class KookAPIMixin:
             member_count=data.get('user_count'),
         )
 
-    @diagnostics.observe('api', 'get_group_list', source='platform', stage='accepted')
     async def get_group_list(self) -> list[platform_entities.UserGroup]:
         return list(self._group_cache.values())
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
     ) -> list[platform_entities.UserGroupMember]:
         raise NotSupportedError('get_group_member_list')
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -137,7 +128,6 @@ class KookAPIMixin:
             display_name=user.nickname,
         )
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         cached = self._user_cache.get(str(user_id))
         if cached:
@@ -152,22 +142,18 @@ class KookAPIMixin:
             is_bot=bool(data.get('bot', False)),
         )
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         data = {'file': file_data}
         raw = await self._request('POST', '/asset/create', data=data, filename=filename)
         result = raw.get('data') or {}
         return str(result.get('url') or result.get('id') or '')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         return file_id
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -177,7 +163,6 @@ class KookAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -187,7 +172,6 @@ class KookAPIMixin:
         endpoint = '/message/delete' if str(chat_type).lower() in {'group', 'channel'} else '/direct-message/delete'
         await self._request('POST', endpoint, json={'msg_id': str(message_id)})
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -201,7 +185,6 @@ class KookAPIMixin:
             raise NotSupportedError('forward_message:message_not_cached')
         return await self.send_message(to_chat_type, str(to_chat_id), cached.message_chain)
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(
         self,
         group_id: typing.Union[int, str],
@@ -210,7 +193,6 @@ class KookAPIMixin:
     ) -> None:
         raise NotSupportedError('mute_member')
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(
         self,
         group_id: typing.Union[int, str],
@@ -218,7 +200,6 @@ class KookAPIMixin:
     ) -> None:
         raise NotSupportedError('unmute_member')
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(
         self,
         group_id: typing.Union[int, str],
@@ -226,6 +207,5 @@ class KookAPIMixin:
     ) -> None:
         raise NotSupportedError('kick_member')
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(self, group_id: typing.Union[int, str]) -> None:
         raise NotSupportedError('leave_group')
diff --git a/src/langbot/pkg/platform/adapters/kook/event_converter.py b/src/langbot/pkg/platform/adapters/kook/event_converter.py
index 34a2bbd19..5c7f8f4f2 100644
--- a/src/langbot/pkg/platform/adapters/kook/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/kook/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import time
 
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
@@ -18,7 +16,6 @@ class KookEventConverter(abstract_platform_adapter.AbstractEventConverter):
         raise NotImplementedError
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(kook_event: dict, bot_account_id: str = '') -> platform_events.Event | None:
         event_type = int(kook_event.get('type', 0) or 0)
         channel_type = kook_event.get('channel_type')
@@ -58,7 +55,6 @@ class KookEventConverter(abstract_platform_adapter.AbstractEventConverter):
         )
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         kook_event: dict, bot_account_id: str = ''
     ) -> platform_events.FriendMessage | platform_events.GroupMessage:
diff --git a/src/langbot/pkg/platform/adapters/lark/adapter.py b/src/langbot/pkg/platform/adapters/lark/adapter.py
index 6e75d79a1..aa04ae8a7 100644
--- a/src/langbot/pkg/platform/adapters/lark/adapter.py
+++ b/src/langbot/pkg/platform/adapters/lark/adapter.py
@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
+from langbot.pkg.i18n.adapters import adapter_t, localized_adapter
 
 from langbot.pkg.platform.sources.lark import (
     LarkAdapter as LegacyLarkAdapter,
@@ -48,6 +48,8 @@ from lark_oapi.api.im.v1 import (
     CreateMessageResponse,
     EventMessage,
     EventSender,
+    GetChatRequest,
+    GetMessageRequest,
     P2ImMessageReceiveV1,
     P2ImMessageReceiveV1Data,
     ReplyMessageRequest,
@@ -58,6 +60,8 @@ import lark_oapi.ws.exception
 import pydantic
 import quart
 
+from .stream_card import build_card, streaming_text
+
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
 import langbot_plugin.api.definition.abstract.platform.event_logger as abstract_platform_logger
 from langbot.pkg.platform.adapters.lark.api_impl import LarkAPIMixin
@@ -142,13 +146,15 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
     _message_cache: dict[str, platform_events.MessageReceivedEvent] = pydantic.PrivateAttr(default_factory=dict)
     _user_cache: dict[str, platform_entities.User] = pydantic.PrivateAttr(default_factory=dict)
     _group_cache: dict[str, platform_entities.UserGroup] = pydantic.PrivateAttr(default_factory=dict)
+    _bot_open_ids: dict[str, str] = pydantic.PrivateAttr(default_factory=dict)
+    _bot_identity_lock: asyncio.Lock = pydantic.PrivateAttr(default_factory=asyncio.Lock)
     _monitoring_mapping_ttl: int = 600
 
     class Config:
         arbitrary_types_allowed = True
 
     def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger, **kwargs):
-        required_keys = ['app_id', 'app_secret', 'bot_name']
+        required_keys = ['app_id', 'app_secret']
         missing_keys = [key for key in required_keys if not config.get(key)]
         if missing_keys:
             raise ValueError(f'Lark missing required config: {", ".join(missing_keys)}')
@@ -167,7 +173,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             config=config,
             logger=logger,
             lark_tenant_key=config.get('lark_tenant_key', ''),
-            bot_account_id=config['bot_name'],
+            bot_account_id=config['app_id'],
             bot=bot,
             api_client=api_client,
             quart_app=quart.Quart(__name__),
@@ -188,16 +194,6 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
         self.request_app_ticket()
 
     def _build_event_handler(self):
-        @diagnostics.observe(
-            'event',
-            'platform.native_callback',
-            source='platform',
-            stage='convert',
-            ap=lambda: getattr(logger, 'ap', None),
-            fields=lambda b: {
-                'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-            },
-        )
         async def on_message(event: lark_oapi.im.v1.P2ImMessageReceiveV1):
             await self._handle_message_event(event)
 
@@ -207,17 +203,32 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
         def sync_on_card_action(event):
             return self._handle_card_action_sync(event)
 
-        return (
+        def sync_on_bot_added(event):
+            self._submit_coro(self._handle_bot_added_event(event))
+
+        def sync_on_standard_event(event):
+            self._submit_coro(self._handle_standard_event(json.loads(lark_oapi.JSON.marshal(event))))
+
+        builder = (
             lark_oapi.EventDispatcherHandler.builder('', '')
             .register_p2_im_message_receive_v1(sync_on_message)
+            .register_p2_im_chat_member_bot_added_v1(sync_on_bot_added)
             .register_p2_card_action_trigger(sync_on_card_action)
-            .build()
         )
+        for registration in LarkEventConverter.STANDARD_EVENT_CALLBACKS.values():
+            getattr(builder, registration)(sync_on_standard_event)
+        return builder.build()
 
     def get_supported_events(self) -> list[str]:
         return [
             'message.received',
             'bot.invited_to_group',
+            'bot.removed_from_group',
+            'group.member_joined',
+            'group.member_left',
+            'group.info_updated',
+            'message.deleted',
+            'message.reaction',
             'platform.specific',
         ]
 
@@ -338,7 +349,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             self.request_tenant_access_token(tenant_key)
         return self.tenant_access_tokens.get(tenant_key, {}).get('token')
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
+    @localized_adapter
     async def send_message(
         self,
         target_type: str,
@@ -372,7 +383,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             message_id=message_ids[-1] if message_ids else '', raw={'message_ids': message_ids}
         )
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
+    @localized_adapter
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -449,7 +460,6 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             while len(self.pending_monitoring_msg) > 1000:
                 self.pending_monitoring_msg.pop(next(iter(self.pending_monitoring_msg)), None)
 
-    @diagnostics.observe('api', 'create_message_card', source='platform', stage='accepted')
     async def create_message_card(self, message_id, event) -> bool:
         card_id = await self.create_card_id(message_id)
         content = {'type': 'card', 'data': {'card_id': card_id, 'template_variable': {'content': 'Thinking...'}}}
@@ -486,28 +496,13 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             self.card_sequence_dict.pop(old_card, None)
             self.card_last_update_dict.pop(old_card, None)
             self.closed_streaming_cards.discard(old_card)
-        card_data = {
-            'schema': '2.0',
-            'config': {
-                'update_multi': True,
-                'streaming_mode': True,
-                'streaming_config': {
-                    'print_step': {'default': 1},
-                    'print_frequency_ms': {'default': 70},
-                    'print_strategy': 'fast',
-                },
-            },
-            'body': {
-                'direction': 'vertical',
-                'elements': [{'tag': 'markdown', 'content': '', 'element_id': 'streaming_txt'}],
-            },
-        }
+        card_data = build_card(streaming=True)
         request = (
             CreateCardRequest.builder()
             .request_body(CreateCardRequestBody.builder().type('card_json').data(json.dumps(card_data)).build())
             .build()
         )
-        response: CreateCardResponse = self.api_client.cardkit.v1.card.create(request)
+        response: CreateCardResponse = await self.api_client.cardkit.v1.card.acreate(request)
         if not response.success():
             raise RuntimeError(f'Lark create_card failed: {response.code} {response.msg}')
         card_id = str(response.data.card_id)
@@ -527,16 +522,9 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
     def _streaming_mode_closed(response: ContentCardElementResponse) -> bool:
         return response.code == 300309 or 'streaming mode is closed' in str(response.msg).lower()
 
-    async def _replace_streaming_card(self, card_id: str, content: str) -> None:
+    async def _replace_streaming_card(self, card_id: str, content: str, *, finished: bool = True) -> None:
         sequence = self._next_card_sequence(card_id)
-        card_data = {
-            'schema': '2.0',
-            'config': {'update_multi': True},
-            'body': {
-                'direction': 'vertical',
-                'elements': [{'tag': 'markdown', 'content': content}],
-            },
-        }
+        card_data = build_card(content, finished=finished)
         request = (
             UpdateCardRequest.builder()
             .card_id(card_id)
@@ -554,7 +542,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             raise RuntimeError(f'Lark card update failed: {response.code} {response.msg}')
         self.closed_streaming_cards.add(card_id)
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
+    @localized_adapter
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -580,14 +568,16 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
                 for paragraph in text_elements
             )
         if (is_final and not bot_message.tool_calls) or card_id in self.closed_streaming_cards:
-            await self._replace_streaming_card(card_id, content)
+            await self._replace_streaming_card(card_id, content, finished=is_final and not bot_message.tool_calls)
         else:
             sequence = self._next_card_sequence(card_id)
             request = (
                 ContentCardElementRequest.builder()
                 .card_id(card_id)
                 .element_id('streaming_txt')
-                .request_body(ContentCardElementRequestBody.builder().content(content).sequence(sequence).build())
+                .request_body(
+                    ContentCardElementRequestBody.builder().content(streaming_text(content)).sequence(sequence).build()
+                )
                 .build()
             )
             response: ContentCardElementResponse = await self.api_client.cardkit.v1.card_element.acontent(
@@ -595,7 +585,9 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             )
             if not response.success():
                 if self._streaming_mode_closed(response):
-                    await self._replace_streaming_card(card_id, content)
+                    await self._replace_streaming_card(
+                        card_id, content, finished=is_final and not bot_message.tool_calls
+                    )
                 else:
                     raise RuntimeError(f'Lark card_element update failed: {response.code} {response.msg}')
         self.card_last_update_dict[card_id] = now
@@ -605,7 +597,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             self.card_last_update_dict.pop(card_id, None)
             self.closed_streaming_cards.discard(card_id)
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
+    @localized_adapter
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         if action == 'interaction.request':
             return await send_interaction(self, params)
@@ -646,6 +638,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             return f'{event.group.id}_{thread_id}'
         return None
 
+    @localized_adapter
     async def handle_unified_webhook(self, bot_uuid: str, path: str, request):
         try:
             data = await request.json
@@ -657,6 +650,9 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             if event_type == 'app_ticket':
                 self.app_ticket = self._webhook_event(data).get('app_ticket')
                 return {'code': 200, 'message': 'ok'}
+            if event_type in LarkEventConverter.STANDARD_EVENT_CALLBACKS:
+                await self._handle_standard_event(data)
+                return {'code': 200, 'message': 'ok'}
             if event_type == 'im.message.receive_v1':
                 p2v1 = P2ImMessageReceiveV1()
                 p2v1.header = self._webhook_header(data)
@@ -669,11 +665,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
                 await self._handle_message_event(p2v1)
                 return {'code': 200, 'message': 'ok'}
             if event_type == 'im.chat.member.bot.added_v1':
-                raw_event = self._webhook_event(data)
-                header = self._webhook_header(data)
-                chat_id = raw_event.get('chat_id', '')
-                await self._send_bot_added_welcome(chat_id, getattr(header, 'tenant_key', None))
-                await self._dispatch_eba_event(LarkEventConverter.bot_invited_to_group(data, chat_id))
+                await self._handle_bot_added_event(lark_oapi.im.v1.P2ImChatMemberBotAddedV1(data))
                 return {'code': 200, 'message': 'ok'}
             if event_type == 'card.action.trigger':
                 interaction_event = interaction_event_from_webhook(data)
@@ -683,7 +675,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
                 feedback_event = self._feedback_event_from_webhook(data)
                 if feedback_event and platform_events.FeedbackEvent in self.listeners:
                     await self.listeners[platform_events.FeedbackEvent](feedback_event, self)
-                return {'toast': {'type': 'success', 'content': '感谢您的反馈'}}
+                return {'toast': {'type': 'success', 'content': adapter_t(self, 'Thank you for your feedback')}}
             await self._dispatch_eba_event(LarkEventConverter.platform_specific(data, event_type, data))
             return {'code': 200, 'message': 'ok'}
         except Exception:
@@ -704,8 +696,14 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
     def _webhook_header(self, data: dict):
         return type('LarkWebhookHeader', (), data.get('header', {}))()
 
+    @localized_adapter
     async def run_async(self):
         self.event_loop = asyncio.get_running_loop()
+        if self.config.get('app_type', 'self') != 'isv':
+            try:
+                await self._get_bot_open_id(None)
+            except Exception as exc:
+                await self.logger.warning(f'Lark bot identity lookup failed; will retry on incoming message: {exc}')
         if not self.config.get('enable-webhook', False):
             try:
                 await self.bot._connect()
@@ -750,14 +748,118 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
     async def is_muted(self, group_id: int | None = None) -> bool:
         return False
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
+    async def _standard_message_context(self, raw: dict) -> dict:
+        data = raw['event']
+        cached = self._message_cache.get(data['message_id'])
+        if cached:
+            return {'chat_type': cached.chat_type, 'chat_id': cached.chat_id, 'group': cached.group}
+        tenant_key = (raw.get('header') or {}).get('tenant_key')
+        option = await asyncio.to_thread(self.request_option, tenant_key)
+        chat_id = data.get('chat_id')
+        if not chat_id:
+            request = GetMessageRequest.builder().message_id(data['message_id']).build()
+            response = await self.api_client.im.v1.message.aget(request, option)
+            if not response.success() or not response.data or not response.data.items:
+                raise RuntimeError(f'Lark message context lookup failed: {response.code} {response.msg}')
+            chat_id = response.data.items[0].chat_id
+        request = GetChatRequest.builder().chat_id(chat_id).build()
+        response = await self.api_client.im.v1.chat.aget(request, option)
+        if not response.success() or not response.data:
+            raise RuntimeError(f'Lark chat context lookup failed: {response.code} {response.msg}')
+        mode = response.data.chat_mode
+        if mode not in {'group', 'p2p'}:
+            raise RuntimeError(f'Lark chat context has unknown chat mode: {mode}')
+        is_group = mode == 'group'
+        return {
+            'chat_type': platform_entities.ChatType.GROUP if is_group else platform_entities.ChatType.PRIVATE,
+            'chat_id': chat_id,
+            'group': platform_entities.UserGroup(id=chat_id, name=response.data.name or '') if is_group else None,
+        }
+
+    async def _handle_standard_event(self, raw: dict):
+        event_type = (raw.get('header') or {}).get('event_type', 'unknown')
+        await self.logger.info(f'Lark event received: {event_type}')
+        try:
+            await self._process_standard_event(raw)
+        except Exception:
+            await self.logger.error(f'Lark event processing failed ({event_type}): {traceback.format_exc()}')
+            raise
+
+    async def _process_standard_event(self, raw: dict):
+        event_type = raw['header']['event_type']
+        context = None
+        if event_type.startswith('im.message.'):
+            try:
+                context = await self._standard_message_context(raw)
+            except Exception as exc:
+                # Preserve the event without falsely classifying an unknown conversation as a private chat.
+                await self.logger.warning(f'Lark {event_type}: {exc}; forwarding original platform event')
+                await self._dispatch_eba_event(LarkEventConverter.platform_specific(raw, event_type, raw))
+                return
+        events = LarkEventConverter.standard_events(raw, context)
+        if not events:
+            await self.logger.warning(f'Lark {event_type}: no standard events could be extracted')
+        for event in events:
+            if isinstance(event, (platform_events.GroupInfoUpdatedEvent, platform_events.BotRemovedFromGroupEvent)):
+                self._group_cache.pop(str(event.group.id), None)
+            await self._dispatch_eba_event(event)
+        await self.logger.info(f'Lark event dispatched: {event_type}, count={len(events)}')
+        if event_type == 'im.message.recalled_v1':
+            self._message_cache.pop(raw['event']['message_id'], None)
+
+    async def _handle_bot_added_event(self, event: lark_oapi.im.v1.P2ImChatMemberBotAddedV1):
+        chat_id = event.event.chat_id
+        operator = event.event.operator_id
+        operator_id = getattr(operator, 'open_id', None) or getattr(operator, 'user_id', None)
+        await self._dispatch_eba_event(LarkEventConverter.bot_invited_to_group(event, chat_id, operator_id))
+        try:
+            await self._send_bot_added_welcome(chat_id, getattr(event.header, 'tenant_key', None))
+        except Exception:
+            await self.logger.warning(f'Lark bot_added_welcome failed: {traceback.format_exc()}')
+
+    async def _get_bot_open_id(self, tenant_key: str | None) -> str:
+        key = tenant_key or ''
+        async with self._bot_identity_lock:
+            if key in self._bot_open_ids:
+                return self._bot_open_ids[key]
+            request = (
+                lark_oapi.BaseRequest.builder()
+                .http_method(lark_oapi.HttpMethod.GET)
+                .uri('/open-apis/bot/v3/info')
+                .token_types({lark_oapi.AccessTokenType.TENANT})
+                .build()
+            )
+            option = await asyncio.to_thread(self.request_option, tenant_key)
+            response = await self.api_client.arequest(request, option)
+            if not response.success():
+                raise RuntimeError(f'Lark bot identity lookup failed: {response.code} {response.msg}')
+            bot_info = json.loads(response.raw.content).get('bot') or {}
+            open_id = bot_info.get('open_id') or bot_info.get('openid')
+            if not isinstance(open_id, str) or not open_id:
+                raise RuntimeError('Lark bot identity response is missing open_id')
+            # Keep a stable adapter identity across tenant-specific callbacks.
+            if not self._bot_open_ids:
+                self.bot_account_id = open_id
+            self._bot_open_ids[key] = open_id
+            return open_id
+
+    def _normalize_bot_mention(self, event, open_id: str) -> None:
+        if event is not None:
+            for component in event.message_chain:
+                if isinstance(component, platform_message.At) and str(component.target) == open_id:
+                    component.target = self.bot_account_id
+
     async def _handle_message_event(self, event: lark_oapi.im.v1.P2ImMessageReceiveV1):
         try:
+            tenant_key = getattr(getattr(event, 'header', None), 'tenant_key', None)
+            open_id = await self._get_bot_open_id(tenant_key if self.config.get('app_type', 'self') == 'isv' else None)
             if platform_events.FriendMessage in self.listeners or platform_events.GroupMessage in self.listeners:
                 legacy_event = await self.event_converter.target2legacy(event, self.api_client)
+                self._normalize_bot_mention(legacy_event, open_id)
                 if legacy_event and type(legacy_event) in self.listeners:
                     await self.listeners[type(legacy_event)](legacy_event, self)
             eba_event = await self.event_converter.target2yiri(event, self.api_client)
+            self._normalize_bot_mention(eba_event, open_id)
             if eba_event:
                 self._cache_event(eba_event)
                 await self._dispatch_eba_event(eba_event)
@@ -765,7 +867,6 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             await self.logger.error(f'Error in lark message event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.Event):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
@@ -799,12 +900,13 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
             self._submit_coro(self.listeners[platform_events.FeedbackEvent](feedback_event, self))
         from lark_oapi.event.callback.model.p2_card_action_trigger import P2CardActionTriggerResponse
 
-        return P2CardActionTriggerResponse({'toast': {'type': 'success', 'content': '感谢您的反馈'}})
+        return P2CardActionTriggerResponse(
+            {'toast': {'type': 'success', 'content': adapter_t(self, 'Thank you for your feedback')}}
+        )
 
-    @staticmethod
-    def _interaction_action_response(event: platform_events.PlatformSpecificEvent) -> dict[str, typing.Any]:
+    def _interaction_action_response(self, event: platform_events.PlatformSpecificEvent) -> dict[str, typing.Any]:
         response: dict[str, typing.Any] = {
-            'toast': {'type': 'success', 'content': 'Submitted / 已提交'},
+            'toast': {'type': 'success', 'content': adapter_t(self, 'Submitted')},
         }
         if not event.data.get('cardkit'):
             response['card'] = {
@@ -814,7 +916,7 @@ class LarkAdapter(LarkAPIMixin, abstract_platform_adapter.AbstractPlatformAdapte
                     'elements': [
                         {
                             'tag': 'div',
-                            'text': {'tag': 'lark_md', 'content': '**Submitted / 已提交**'},
+                            'text': {'tag': 'lark_md', 'content': '**' + adapter_t(self, 'Submitted') + '**'},
                         }
                     ],
                 },
diff --git a/src/langbot/pkg/platform/adapters/lark/api_impl.py b/src/langbot/pkg/platform/adapters/lark/api_impl.py
index fb122a74a..ae2ff23d5 100644
--- a/src/langbot/pkg/platform/adapters/lark/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/lark/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from lark_oapi.api.im.v1 import GetChatRequest, GetMessageRequest
@@ -19,7 +17,6 @@ class LarkAPIMixin:
     _user_cache: dict[str, platform_entities.User]
     _group_cache: dict[str, platform_entities.UserGroup]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -55,7 +52,6 @@ class LarkAPIMixin:
         self._message_cache[str(message_id)] = event
         return event
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         cached = self._group_cache.get(str(group_id))
         if cached:
@@ -75,7 +71,6 @@ class LarkAPIMixin:
         self._group_cache[str(group.id)] = group
         return group
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -84,14 +79,12 @@ class LarkAPIMixin:
         user = self._user_cache.get(str(user_id)) or platform_entities.User(id=user_id)
         return platform_entities.UserGroupMember(user=user, group_id=group_id, role=platform_entities.MemberRole.MEMBER)
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         cached = self._user_cache.get(str(user_id))
         if cached:
             return cached
         return platform_entities.User(id=user_id)
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         if str(file_id).startswith('file://'):
             return str(file_id)
diff --git a/src/langbot/pkg/platform/adapters/lark/event_converter.py b/src/langbot/pkg/platform/adapters/lark/event_converter.py
index 6cfd590c7..b2736009d 100644
--- a/src/langbot/pkg/platform/adapters/lark/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/lark/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import time
 import typing
 
@@ -17,6 +15,103 @@ from langbot_plugin.api.entities.builtin.platform import message as platform_mes
 
 
 class LarkEventConverter(abstract_platform_adapter.AbstractEventConverter):
+    STANDARD_EVENT_CALLBACKS: typing.ClassVar[dict[str, str]] = {
+        'im.chat.member.bot.deleted_v1': 'register_p2_im_chat_member_bot_deleted_v1',
+        'im.chat.member.user.added_v1': 'register_p2_im_chat_member_user_added_v1',
+        'im.chat.member.user.deleted_v1': 'register_p2_im_chat_member_user_deleted_v1',
+        'im.chat.updated_v1': 'register_p2_im_chat_updated_v1',
+        'im.message.recalled_v1': 'register_p2_im_message_recalled_v1',
+        'im.message.reaction.created_v1': 'register_p2_im_message_reaction_created_v1',
+        'im.message.reaction.deleted_v1': 'register_p2_im_message_reaction_deleted_v1',
+    }
+
+    @staticmethod
+    def _event_user(identity: dict | None, name: str = '') -> platform_entities.User | None:
+        identity = identity or {}
+        user_id = identity.get('open_id') or identity.get('user_id') or identity.get('union_id')
+        return platform_entities.User(id=user_id, nickname=name) if user_id else None
+
+    @classmethod
+    def standard_events(cls, raw: dict, message_context: dict | None = None) -> list[platform_events.EBAEvent]:
+        data = raw.get('event') or {}
+        header = raw.get('header') or {}
+        event_type = header.get('event_type', '')
+        common = {
+            'adapter_name': ADAPTER_NAME,
+            'timestamp': cls._timestamp(data.get('action_time') or data.get('recall_time') or header.get('create_time'))
+            or time.time(),
+            'source_platform_object': raw,
+        }
+        group = platform_entities.UserGroup(id=data.get('chat_id') or '', name=data.get('name') or '')
+        operator = cls._event_user(data.get('operator_id'))
+        if event_type == 'im.chat.member.bot.deleted_v1':
+            return [platform_events.BotRemovedFromGroupEvent(group=group, operator=operator, **common)]
+        if event_type in {'im.chat.member.user.added_v1', 'im.chat.member.user.deleted_v1'}:
+            events = []
+            for user in data.get('users') or []:
+                member = cls._event_user(user.get('user_id'), user.get('name') or '')
+                if member is None:
+                    continue
+                if event_type == 'im.chat.member.user.added_v1':
+                    inviter = operator if operator and operator.id != member.id else None
+                    events.append(
+                        platform_events.MemberJoinedEvent(
+                            group=group,
+                            member=member,
+                            inviter=inviter,
+                            join_type='invite' if inviter else None,
+                            **common,
+                        )
+                    )
+                else:
+                    events.append(
+                        platform_events.MemberLeftEvent(
+                            group=group,
+                            member=member,
+                            operator=operator,
+                            is_kicked=operator is not None and operator.id != member.id,
+                            **common,
+                        )
+                    )
+            return events
+        if event_type == 'im.chat.updated_v1':
+            after = data.get('after_change') or {}
+            before = data.get('before_change') or {}
+            changed = [key for key in before.keys() | after.keys() if before.get(key) != after.get(key)]
+            if data.get('moderator_list'):
+                changed.append('moderator_list')
+            group.name = after.get('name') or ''
+            group.description = after.get('description')
+            group.avatar_url = after.get('avatar')
+            owner = cls._event_user(after.get('owner_id'))
+            group.owner_id = owner.id if owner else None
+            return [
+                platform_events.GroupInfoUpdatedEvent(
+                    group=group,
+                    operator=operator,
+                    changed_fields=sorted(changed),
+                    **common,
+                )
+            ]
+        context = message_context or {}
+        if event_type == 'im.message.recalled_v1':
+            return [platform_events.MessageDeletedEvent(message_id=data['message_id'], **context, **common)]
+        if event_type in {'im.message.reaction.created_v1', 'im.message.reaction.deleted_v1'}:
+            user = cls._event_user(data.get('user_id'))
+            if user is None:
+                user = platform_entities.User(id=data.get('app_id') or '', is_bot=data.get('operator_type') == 'app')
+            return [
+                platform_events.MessageReactionEvent(
+                    message_id=data['message_id'],
+                    user=user,
+                    reaction=(data.get('reaction_type') or {}).get('emoji_type') or '',
+                    is_add=event_type == 'im.message.reaction.created_v1',
+                    **context,
+                    **common,
+                )
+            ]
+        return []
+
     _processed_thread_quote_cache: typing.ClassVar[dict[str, float]] = {}
     _processed_thread_quote_cache_max_size: typing.ClassVar[int] = 4096
     _processed_thread_quote_cache_ttl_seconds: typing.ClassVar[int] = 86400
@@ -26,7 +121,6 @@ class LarkEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return getattr(event, 'source_platform_object', None)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(
         event: lark_oapi.im.v1.P2ImMessageReceiveV1,
         api_client: lark_oapi.Client,
@@ -34,7 +128,6 @@ class LarkEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return await LarkEventConverter.message_to_eba(event, api_client)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         event: lark_oapi.im.v1.P2ImMessageReceiveV1,
         api_client: lark_oapi.Client,
diff --git a/src/langbot/pkg/platform/adapters/lark/interaction.py b/src/langbot/pkg/platform/adapters/lark/interaction.py
index 9c9873125..8896f6316 100644
--- a/src/langbot/pkg/platform/adapters/lark/interaction.py
+++ b/src/langbot/pkg/platform/adapters/lark/interaction.py
@@ -2,6 +2,8 @@
 
 from __future__ import annotations
 
+from langbot.pkg.i18n import t
+
 import json
 import time
 import typing
@@ -61,8 +63,7 @@ def _field_form_elements(
     component_name = 'lbi_field_0'
     label = str(field.get('label') or field_id)
     placeholder = str(
-        field.get('placeholder')
-        or ('Select an option / 请选择' if field_type == 'select' else 'Enter a value / 请输入')
+        field.get('placeholder') or (t('Select an option') if field_type == 'select' else t('Enter a value'))
     )
     required = bool(field.get('required'))
     if field_type == 'select':
@@ -122,7 +123,7 @@ def _field_form_elements(
     submit_button = {
         'tag': 'button',
         'name': 'lbi_submit',
-        'text': {'tag': 'plain_text', 'content': 'Submit / 提交'},
+        'text': {'tag': 'plain_text', 'content': t('Submit')},
         'type': 'primary',
         'width': 'fill',
         'form_action_type': 'submit',
diff --git a/src/langbot/pkg/platform/adapters/lark/manifest.yaml b/src/langbot/pkg/platform/adapters/lark/manifest.yaml
index 5d80781ca..af722d3cf 100644
--- a/src/langbot/pkg/platform/adapters/lark/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/lark/manifest.yaml
@@ -108,20 +108,6 @@ spec:
       type: string
       required: true
       default: ""
-    - name: bot_name
-      label:
-        en_US: Bot Name
-        zh_Hans: 机器人名称
-        zh_Hant: 機器人名稱
-        ja_JP: ボット名
-      description:
-        en_US: Must match the Lark bot name so group mentions can be recognized.
-        zh_Hans: 必须与飞书机器人名称一致，否则机器人将无法在群内正常识别 @。
-        zh_Hant: 必須與飛書機器人名稱一致，否則機器人將無法在群組內正常識別 @。
-        ja_JP: グループメンションを認識するには Lark のボット名と一致する必要があります。
-      type: string
-      required: true
-      default: ""
     - name: enable-webhook
       label:
         en_US: Enable Webhook Mode
@@ -216,6 +202,12 @@ spec:
   supported_events:
     - message.received
     - bot.invited_to_group
+    - bot.removed_from_group
+    - group.member_joined
+    - group.member_left
+    - group.info_updated
+    - message.deleted
+    - message.reaction
     - platform.specific
 
   supported_apis:
diff --git a/src/langbot/pkg/platform/adapters/lark/message_converter.py b/src/langbot/pkg/platform/adapters/lark/message_converter.py
index 117740d01..510246cf8 100644
--- a/src/langbot/pkg/platform/adapters/lark/message_converter.py
+++ b/src/langbot/pkg/platform/adapters/lark/message_converter.py
@@ -241,6 +241,13 @@ class LarkMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
                     flattened.append(ele)
                 elif isinstance(ele, list):
                     flattened.extend(item for item in ele if isinstance(item, dict))
+            for index, ele in enumerate(flattened):
+                if ele.get('tag') == 'at':
+                    mentions = LarkMessageConverter._split_text_mentions(
+                        ele.get('user_id') or '', getattr(message, 'mentions', []) or []
+                    )
+                    if len(mentions) == 1 and mentions[0].get('tag') == 'at':
+                        flattened[index] = {**ele, **mentions[0]}
             return flattened
         if message.message_type == 'image':
             return [{'tag': 'img', 'image_key': content.get('image_key', ''), 'style': []}]
@@ -275,10 +282,15 @@ class LarkMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
                 result.append({'tag': 'text', 'text': text[pos : match.start()], 'style': []})
             mention = mention_by_key.get(match.group(0))
             if mention:
+                identity = getattr(mention, 'id', None)
+                if isinstance(identity, dict):
+                    identity = identity.get('open_id') or identity.get('user_id')
+                elif identity is not None and not isinstance(identity, str):
+                    identity = getattr(identity, 'open_id', None) or getattr(identity, 'user_id', None)
                 result.append(
                     {
                         'tag': 'at',
-                        'user_id': getattr(mention, 'id', None)
+                        'user_id': identity
                         or getattr(mention, 'open_id', None)
                         or getattr(mention, 'user_id', None)
                         or getattr(mention, 'key', match.group(0)),
diff --git a/src/langbot/pkg/platform/adapters/lark/stream_card.py b/src/langbot/pkg/platform/adapters/lark/stream_card.py
new file mode 100644
index 000000000..cd5adbd56
--- /dev/null
+++ b/src/langbot/pkg/platform/adapters/lark/stream_card.py
@@ -0,0 +1,99 @@
+"""Compact streaming previews and a final card with collapsible reasoning."""
+
+import re
+
+from langbot.pkg.i18n import t
+
+
+def split_content(content: str) -> tuple[str, str]:
+    reasoning, answer = [], []
+    inside, offset = False, 0
+    for marker in re.finditer(r'</?think>', content, re.IGNORECASE):
+        (reasoning if inside else answer).append(content[offset : marker.start()])
+        inside = marker.group().lower() == '<think>'
+        offset = marker.end()
+    tail = content[offset:]
+    for size in range(min(7, len(tail)), 0, -1):
+        if any(tag.startswith(tail[-size:].lower()) for tag in ('<think>', '</think>')):
+            tail = tail[:-size]
+            break
+    (reasoning if inside else answer).append(tail)
+    return tuple('\n\n'.join(part.strip() for part in pieces if part.strip()) for pieces in (reasoning, answer))
+
+
+def streaming_text(content: str) -> str:
+    reasoning, answer = split_content(content)
+    if not reasoning:
+        return answer
+    # Keep live cards compact; the final card retains the complete reasoning.
+    preview = reasoning[-360:]
+    if len(reasoning) > 360:
+        preview = '…' + preview
+    quote = '> ' + preview.replace('\n', '\n> ')
+    return '**' + t('Thinking') + '**\n\n' + quote + ('\n\n---\n\n' + answer if answer else '')
+
+
+def build_card(content: str = '', *, streaming: bool = False, finished: bool = True) -> dict:
+    reasoning, answer = split_content(content)
+    elements = []
+    if streaming:
+        elements.append({'tag': 'markdown', 'content': streaming_text(content), 'element_id': 'streaming_txt'})
+    else:
+        if reasoning:
+            elements.append(
+                {
+                    'tag': 'collapsible_panel',
+                    'expanded': not finished,
+                    'header': {'title': {'tag': 'plain_text', 'content': t('Thinking')}},
+                    'elements': [{'tag': 'markdown', 'content': reasoning}],
+                }
+            )
+        if answer:
+            elements.append({'tag': 'markdown', 'content': answer})
+    elements.append(
+        {
+            'tag': 'column_set',
+            'horizontal_spacing': '8px',
+            'columns': [
+                {
+                    'tag': 'column',
+                    'width': 'weighted',
+                    'weight': 1,
+                    'vertical_align': 'center',
+                    'elements': [{'tag': 'markdown', 'content': t('AI-generated'), 'text_size': 'notation'}],
+                },
+                {
+                    'tag': 'column',
+                    'width': 'auto',
+                    'vertical_align': 'center',
+                    'elements': [
+                        {
+                            'tag': 'markdown',
+                            'content': '[LangBot](https://langbot.app?utm_source=feishu&utm_medium=bot_card&utm_campaign=langbot)',
+                            'text_size': 'notation',
+                            'text_align': 'right',
+                            'icon': {
+                                'tag': 'custom_icon',
+                                'img_key': 'img_v3_02p3_05c65d5d-9bad-440a-a2fb-c89571bfd5bg',
+                            },
+                        }
+                    ],
+                },
+            ],
+        }
+    )
+    config = {'update_multi': True}
+    if streaming:
+        config.update(
+            streaming_mode=True,
+            streaming_config={
+                'print_step': {'default': 3},
+                'print_frequency_ms': {'default': 40},
+                'print_strategy': 'fast',
+            },
+        )
+    return {
+        'schema': '2.0',
+        'config': config,
+        'body': {'direction': 'vertical', 'padding': '12px 12px 12px 12px', 'elements': elements},
+    }
diff --git a/src/langbot/pkg/platform/adapters/officialaccount/adapter.py b/src/langbot/pkg/platform/adapters/officialaccount/adapter.py
index cd1a27077..f4f362b4a 100644
--- a/src/langbot/pkg/platform/adapters/officialaccount/adapter.py
+++ b/src/langbot/pkg/platform/adapters/officialaccount/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import traceback
 import typing
@@ -58,13 +56,14 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
         }
         if mode == 'drop':
             bot = OAClient(**common_kwargs)
-        elif mode == 'passive':
+        elif mode in ('passive', 'auto'):
             bot = OAClientForLongerResponse(
                 **common_kwargs,
                 LoadingMessage=config.get('LoadingMessage', ''),
+                customer_service=mode == 'auto',
             )
         else:
-            raise KeyError('OfficialAccount Mode must be "drop" or "passive"')
+            raise KeyError('OfficialAccount Mode must be "drop", "passive" or "auto"')
 
         super().__init__(
             config=config,
@@ -96,7 +95,6 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -105,7 +103,6 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
     ) -> platform_events.MessageResult:
         raise NotSupportedError('send_message:official_account_requires_inbound_webhook_reply')
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -116,13 +113,13 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
         if not isinstance(source, OAEvent):
             raise ValueError('OfficialAccount reply_message requires an OAEvent source object')
         content = await OfficialAccountMessageConverter.yiri2target(message)
-        if self.config.get('Mode') == 'passive':
-            await self.bot.set_message(source.user_id, source.message_id, content)
+        if self.config.get('Mode') in ('passive', 'auto'):
+            queued = await self.bot.set_message(source.user_id, source.message_id, content)
         else:
             await self.bot.set_message(source.message_id, content)
-        return platform_events.MessageResult(message_id=source.message_id, raw={'queued': True})
+            queued = True
+        return platform_events.MessageResult(message_id=source.message_id, raw={'queued': queued})
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -175,7 +172,6 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
         for msg_type in ('text', 'image', 'voice', 'event'):
             self.bot.on_message(msg_type)(self._handle_native_event)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: OAEvent):
         self.bot_account_id = event.receiver_id or self.bot_account_id
         try:
@@ -192,7 +188,6 @@ class OfficialAccountAdapter(OfficialAccountAPIMixin, abstract_platform_adapter.
             await self.logger.error(f'Error in officialaccount native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/officialaccount/api_impl.py b/src/langbot/pkg/platform/adapters/officialaccount/api_impl.py
index 4468cb660..377986c64 100644
--- a/src/langbot/pkg/platform/adapters/officialaccount/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/officialaccount/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot_plugin.api.entities.builtin.platform import entities as platform_entities
@@ -14,7 +12,6 @@ class OfficialAccountAPIMixin:
     _message_cache: dict[str, platform_events.MessageReceivedEvent]
     _user_cache: dict[str, platform_entities.User]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -26,18 +23,15 @@ class OfficialAccountAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         user = self._user_cache.get(str(user_id))
         if user is None:
             raise NotSupportedError('get_user_info:not_cached')
         return user
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -47,7 +41,6 @@ class OfficialAccountAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -56,7 +49,6 @@ class OfficialAccountAPIMixin:
     ) -> None:
         raise NotSupportedError('delete_message')
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -67,30 +59,24 @@ class OfficialAccountAPIMixin:
     ) -> platform_events.MessageResult:
         raise NotSupportedError('forward_message')
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         raise NotSupportedError('get_group_info')
 
-    @diagnostics.observe('api', 'get_group_list', source='platform', stage='accepted')
     async def get_group_list(self) -> list[platform_entities.UserGroup]:
         raise NotSupportedError('get_group_list')
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
     ) -> list[platform_entities.UserGroupMember]:
         raise NotSupportedError('get_group_member_list')
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
diff --git a/src/langbot/pkg/platform/adapters/officialaccount/event_converter.py b/src/langbot/pkg/platform/adapters/officialaccount/event_converter.py
index dec3a2834..5523a55f8 100644
--- a/src/langbot/pkg/platform/adapters/officialaccount/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/officialaccount/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import time
 import typing
 
@@ -18,7 +16,6 @@ class OfficialAccountEventConverter(abstract_platform_adapter.AbstractEventConve
     async def yiri2target(event: platform_events.Event) -> typing.Any:
         return getattr(event, 'source_platform_object', None)
 
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(self, event: OAEvent) -> platform_events.FriendMessage | None:
         eba_event = await self.target2yiri(event)
         if not isinstance(eba_event, platform_events.MessageReceivedEvent):
@@ -34,7 +31,6 @@ class OfficialAccountEventConverter(abstract_platform_adapter.AbstractEventConve
             source_platform_object=event,
         )
 
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(self, event: OAEvent) -> platform_events.Event | None:
         if event.type in {'text', 'image', 'voice'}:
             return await self.message_to_eba(event)
diff --git a/src/langbot/pkg/platform/adapters/officialaccount/manifest.yaml b/src/langbot/pkg/platform/adapters/officialaccount/manifest.yaml
index 033c00811..2751023fd 100644
--- a/src/langbot/pkg/platform/adapters/officialaccount/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/officialaccount/manifest.yaml
@@ -34,6 +34,22 @@ spec:
       type: array[string]
       required: false
       default: []
+    - name: AppID
+      label:
+        en_US: App ID
+        zh_Hans: 应用 ID
+        zh_Hant: 應用 ID
+      type: string
+      required: true
+      default: ""
+    - name: AppSecret
+      label:
+        en_US: App Secret
+        zh_Hans: 应用密钥
+        zh_Hant: 應用密鑰
+      type: string
+      required: true
+      default: ""
     - name: webhook_url
       label:
         en_US: Webhook Callback URL
@@ -62,32 +78,32 @@ spec:
       type: string
       required: true
       default: ""
-    - name: AppID
-      label:
-        en_US: App ID
-        zh_Hans: 应用 ID
-        zh_Hant: 應用 ID
-      type: string
-      required: true
-      default: ""
-    - name: AppSecret
-      label:
-        en_US: App Secret
-        zh_Hans: 应用密钥
-        zh_Hant: 應用密鑰
-      type: string
-      required: true
-      default: ""
     - name: Mode
       label:
         en_US: Mode
         zh_Hans: 接入模式
         zh_Hant: 接入模式
       description:
-        en_US: "drop replies within the current callback; passive returns a loading message first and queues the real reply for the user's next message."
-        zh_Hans: "drop 会在当前回调内等待回复；passive 会先返回加载提示，并将真实回复排队到用户下一条消息。"
-        zh_Hant: "drop 會在目前回調內等待回覆；passive 會先回傳載入提示，並將真實回覆排隊到使用者下一則訊息。"
-      type: string
+        en_US: "Auto uses customer service messaging when available, falling back to next-message retrieval for permission, window or quota limits. Passive always waits for the next message. Drop requires a reply within 5 seconds."
+        zh_Hans: "自动模式在生成完成后通过客服接口回复；无权限、超出回复窗口或额度不足时，改为下一条消息领取。passive 始终由下一条消息领取；drop 需要在 5 秒内完成回复。客服接口通常需企业主体认证，以微信实际权限为准。"
+        zh_Hant: "自動模式在生成完成後透過客服介面回覆；無權限、超出回覆時限或額度不足時，改為下一則訊息領取。passive 始終由下一則訊息領取；drop 需要在 5 秒內完成回覆。客服介面通常需企業主體認證，以微信實際權限為準。"
+      type: select
+      options:
+        - name: auto
+          label:
+            en_US: Automatic reply with fallback (auto)
+            zh_Hans: 自动回复，受限时转为领取（auto）
+            zh_Hant: 自動回覆，受限時轉為領取（auto）
+        - name: drop
+          label:
+            en_US: Reply in current callback (drop)
+            zh_Hans: 当前回调内回复（drop）
+            zh_Hant: 目前回調內回覆（drop）
+        - name: passive
+          label:
+            en_US: Queue reply for next message (passive)
+            zh_Hans: 下一条消息时返回回复（passive）
+            zh_Hant: 下一則訊息時傳回回覆（passive）
       required: true
       default: "drop"
     - name: LoadingMessage
diff --git a/src/langbot/pkg/platform/adapters/qqofficial/adapter.py b/src/langbot/pkg/platform/adapters/qqofficial/adapter.py
index d57cca2da..7703a5935 100644
--- a/src/langbot/pkg/platform/adapters/qqofficial/adapter.py
+++ b/src/langbot/pkg/platform/adapters/qqofficial/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import time
 import traceback
@@ -134,7 +132,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
     def _plain_message(text: str) -> platform_message.MessageChain:
         return platform_message.MessageChain([platform_message.Plain(text=text)])
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -146,7 +143,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
         )
         return platform_events.MessageResult(raw={'results': raw})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -165,7 +161,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
         )
         return platform_events.MessageResult(message_id=source.d_id or source.id, raw={'results': raw})
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         if action == 'interaction.request':
             return await send_interaction(self, params)
@@ -263,7 +258,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
     async def is_stream_output_supported(self) -> bool:
         return bool(self.config.get('enable-stream-reply') or self.config.get('enable_stream_reply'))
 
-    @diagnostics.observe('api', 'create_message_card', source='platform', stage='accepted')
     async def create_message_card(self, message_id: str, event: platform_events.MessageEvent) -> bool:
         source = event.source_platform_object
         if not isinstance(source, QQOfficialEvent) or source.t != 'C2C_MESSAGE_CREATE':
@@ -283,7 +277,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
         self._stream_ctx_ts[message_id] = time.time()
         return True
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -371,7 +364,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
             if event is not None:
                 await self._dispatch_eba_event(event)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: QQOfficialEvent):
         self.bot_account_id = self.config.get('appid', self.bot_account_id)
         try:
@@ -388,7 +380,6 @@ class QQOfficialAdapter(QQOfficialAPIMixin, abstract_platform_adapter.AbstractPl
             await self.logger.error(f'Error in qqofficial native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/qqofficial/api_impl.py b/src/langbot/pkg/platform/adapters/qqofficial/api_impl.py
index e64570138..59ed22956 100644
--- a/src/langbot/pkg/platform/adapters/qqofficial/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/qqofficial/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.pkg.platform.adapters.qqofficial.errors import NotSupportedError
@@ -16,7 +14,6 @@ class QQOfficialAPIMixin:
     _group_cache: dict[str, platform_entities.UserGroup]
     _member_cache: dict[tuple[str, str], platform_entities.UserGroupMember]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -28,25 +25,21 @@ class QQOfficialAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         user = self._user_cache.get(str(user_id))
         if user is None:
             raise NotSupportedError('get_user_info:not_cached')
         return user
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         group = self._group_cache.get(str(group_id))
         if group is None:
             raise NotSupportedError('get_group_info:not_cached')
         return group
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -57,7 +50,6 @@ class QQOfficialAPIMixin:
             raise NotSupportedError('get_group_member_info:not_cached')
         return member
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -66,7 +58,6 @@ class QQOfficialAPIMixin:
             member for (cached_group_id, _), member in self._member_cache.items() if cached_group_id == str(group_id)
         ]
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -76,7 +67,6 @@ class QQOfficialAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -85,7 +75,6 @@ class QQOfficialAPIMixin:
     ) -> None:
         raise NotSupportedError('delete_message')
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -96,26 +85,20 @@ class QQOfficialAPIMixin:
     ) -> platform_events.MessageResult:
         raise NotSupportedError('forward_message')
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str], duration: int = 0):
         raise NotSupportedError('mute_member')
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]):
         raise NotSupportedError('unmute_member')
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]):
         raise NotSupportedError('kick_member')
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(self, group_id: typing.Union[int, str]):
         raise NotSupportedError('leave_group')
diff --git a/src/langbot/pkg/platform/adapters/qqofficial/event_converter.py b/src/langbot/pkg/platform/adapters/qqofficial/event_converter.py
index 130061ce5..66b550418 100644
--- a/src/langbot/pkg/platform/adapters/qqofficial/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/qqofficial/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import datetime
 import time
 import typing
@@ -33,7 +31,6 @@ class QQOfficialEventConverter(abstract_platform_adapter.AbstractEventConverter)
     async def yiri2target(event: platform_events.Event) -> typing.Any:
         return getattr(event, 'source_platform_object', None)
 
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         self, event: QQOfficialEvent
     ) -> platform_events.FriendMessage | platform_events.GroupMessage | None:
@@ -68,7 +65,6 @@ class QQOfficialEventConverter(abstract_platform_adapter.AbstractEventConverter)
             source_platform_object=event,
         )
 
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(self, event: QQOfficialEvent) -> platform_events.Event:
         if event.t in MESSAGE_EVENT_TYPES:
             return await self.message_to_eba(event)
diff --git a/src/langbot/pkg/platform/adapters/slack/adapter.py b/src/langbot/pkg/platform/adapters/slack/adapter.py
index fd5a5dfbf..81b1c94ca 100644
--- a/src/langbot/pkg/platform/adapters/slack/adapter.py
+++ b/src/langbot/pkg/platform/adapters/slack/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import traceback
 import typing
@@ -10,6 +8,7 @@ import pydantic
 
 from langbot.libs.slack_api.api import SlackClient
 from langbot.libs.slack_api.slackevent import SlackEvent
+from langbot.libs.slack_api.media import send_chain
 from langbot.pkg.platform.adapters.slack.api_impl import SlackAPIMixin
 from langbot.pkg.platform.adapters.slack.errors import NotSupportedError
 from langbot.pkg.platform.adapters.slack.event_converter import SlackEventConverter
@@ -34,7 +33,7 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
         typing.Type[platform_events.Event],
         typing.Callable[[platform_events.Event, abstract_platform_adapter.AbstractMessagePlatformAdapter], None],
     ] = {}
-    _message_cache: dict[str, platform_events.MessageReceivedEvent] = {}
+    _message_cache: dict[tuple[str, str], platform_events.MessageReceivedEvent] = {}
     _user_cache: dict[str, platform_entities.User] = {}
     _group_cache: dict[str, platform_entities.UserGroup] = {}
     _member_cache: dict[tuple[str, str], platform_entities.UserGroupMember] = {}
@@ -43,14 +42,14 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
         arbitrary_types_allowed = True
 
     def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger):
-        required_keys = ['bot_token', 'signing_secret']
+        required_keys = ['bot_token', 'app_token' if config.get('socket_mode', False) else 'signing_secret']
         missing_keys = [key for key in required_keys if not config.get(key)]
         if missing_keys:
             raise Exception(f'Slack Omni adapter missing config: {missing_keys}')
 
         bot = SlackClient(
             bot_token=config['bot_token'],
-            signing_secret=config['signing_secret'],
+            signing_secret=config.get('signing_secret', ''),
             logger=logger,
             unified_mode=True,
         )
@@ -75,6 +74,14 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
     def get_supported_events(self) -> list[str]:
         return [
             'message.received',
+            'message.edited',
+            'message.deleted',
+            'message.reaction',
+            'group.member_joined',
+            'group.member_left',
+            'group.info_updated',
+            'bot.invited_to_group',
+            'bot.removed_from_group',
             'platform.specific',
         ]
 
@@ -92,18 +99,15 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
         target_id: str,
         message: platform_message.MessageChain,
     ) -> platform_events.MessageResult:
-        content = await SlackMessageConverter.yiri2target(message)
-        raw = await self._send_text(str(target_type), str(target_id), content)
+        raw = await send_chain(self.bot, self._normalize_target_type(str(target_type)), str(target_id), message)
         return platform_events.MessageResult(raw=raw)
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -115,10 +119,16 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             raise ValueError('Slack reply_message requires a SlackEvent source object')
         target_type = 'channel' if source.type == 'channel' else 'person'
         target_id = source.channel_id if source.type == 'channel' else source.user_id
-        raw = await self._send_text(target_type, target_id, await SlackMessageConverter.yiri2target(message))
-        return platform_events.MessageResult(message_id=source.message_id, raw=raw)
+        thread_ts = source.get('event', {}).get('thread_ts') or (source.message_id if quote_origin else None)
+        if thread_ts:
+            target_type, target_id = 'channel', source.channel_id
+        raw = await send_chain(self.bot, target_type, target_id, message, thread_ts)
+        responses = raw.get('responses', [])
+        sent = responses[-1] if responses else {}
+        return platform_events.MessageResult(
+            message_id=sent.get('ts') or sent.get('message', {}).get('ts', ''), raw=raw
+        )
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -149,11 +159,18 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
         return await self.bot.handle_unified_webhook(request)
 
     async def run_async(self):
+        identity = await self.bot.client.auth_test()
+        self.bot_account_id = identity.get('user_id', '')
+        self.bot.bot_user_id = self.bot_account_id
+        if self.config.get('socket_mode', False):
+            await self.bot.run_socket(self.config['app_token'])
+            return
         await self.logger.info('Slack Omni adapter running in unified webhook mode')
         while True:
             await asyncio.sleep(1)
 
     async def kill(self) -> bool:
+        await self.bot.close_socket()
         self._message_cache.clear()
         self._user_cache.clear()
         self._group_cache.clear()
@@ -164,12 +181,30 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
         return False
 
     def _register_native_handlers(self):
-        for msg_type in ('im', 'channel'):
+        for msg_type in ('im', 'channel', 'event'):
             self.bot.on_message(msg_type)(self._handle_native_event)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: SlackEvent):
         try:
+            self.event_converter.bot_user_id = getattr(self.bot, 'bot_user_id', None) or self.bot_account_id
+            raw = event.get('event', {})
+            channel = raw.get('channel') or raw.get('item', {}).get('channel', '')
+            if (
+                isinstance(channel, str)
+                and channel.startswith('D')
+                and channel not in self.event_converter.private_chats
+                and (
+                    raw.get('subtype') in {'message_changed', 'message_deleted'}
+                    or raw.get('type') in {'reaction_added', 'reaction_removed'}
+                )
+            ):
+                try:
+                    conversation = await self.bot.client.conversations_info(channel=channel)
+                    self.event_converter.remember_private_chat(channel, conversation.get('channel', {}).get('user', ''))
+                except Exception:
+                    await self.logger.warning('Slack DM participant lookup failed; using channel ID for this event')
+            if event.get('event', {}).get('type') in {'app_uninstalled', 'tokens_revoked'}:
+                await self.logger.warning(f'Slack application lifecycle event: {event["event"]["type"]}')
             if platform_events.FriendMessage in self.listeners or platform_events.GroupMessage in self.listeners:
                 legacy_event = await self.event_converter.target2legacy(event)
                 if legacy_event and type(legacy_event) in self.listeners:
@@ -183,7 +218,6 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             await self.logger.error(f'Error in slack native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
@@ -193,7 +227,7 @@ class SlackAdapter(SlackAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
     def _cache_event(self, event: platform_events.Event):
         if not isinstance(event, platform_events.MessageReceivedEvent):
             return
-        self._message_cache[str(event.message_id)] = event
+        self._message_cache[(str(event.chat_id), str(event.message_id))] = event
         self._user_cache[str(event.sender.id)] = event.sender
         if event.group:
             self._group_cache[str(event.group.id)] = event.group
diff --git a/src/langbot/pkg/platform/adapters/slack/api_impl.py b/src/langbot/pkg/platform/adapters/slack/api_impl.py
index 269444415..13fe35096 100644
--- a/src/langbot/pkg/platform/adapters/slack/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/slack/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.pkg.platform.adapters.slack.errors import NotSupportedError
@@ -11,46 +9,40 @@ from langbot_plugin.api.entities.builtin.platform import message as platform_mes
 
 
 class SlackAPIMixin:
-    _message_cache: dict[str, platform_events.MessageReceivedEvent]
+    _message_cache: dict[tuple[str, str], platform_events.MessageReceivedEvent]
     _user_cache: dict[str, platform_entities.User]
     _group_cache: dict[str, platform_entities.UserGroup]
     _member_cache: dict[tuple[str, str], platform_entities.UserGroupMember]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
         chat_id: typing.Union[int, str],
         message_id: typing.Union[int, str],
     ) -> platform_events.MessageReceivedEvent:
-        event = self._message_cache.get(str(message_id))
+        event = self._message_cache.get((str(chat_id), str(message_id)))
         if event is None:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         user = self._user_cache.get(str(user_id))
         if user is None:
             raise NotSupportedError('get_user_info:not_cached')
         return user
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         group = self._group_cache.get(str(group_id))
         if group is None:
             raise NotSupportedError('get_group_info:not_cached')
         return group
 
-    @diagnostics.observe('api', 'get_group_list', source='platform', stage='accepted')
     async def get_group_list(self) -> list[platform_entities.UserGroup]:
         return list(self._group_cache.values())
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -59,7 +51,6 @@ class SlackAPIMixin:
             member for (cached_group_id, _), member in self._member_cache.items() if cached_group_id == str(group_id)
         ]
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -70,7 +61,6 @@ class SlackAPIMixin:
             raise NotSupportedError('get_group_member_info:not_cached')
         return member
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -80,7 +70,6 @@ class SlackAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -89,7 +78,6 @@ class SlackAPIMixin:
     ) -> None:
         raise NotSupportedError('delete_message')
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -100,10 +88,8 @@ class SlackAPIMixin:
     ) -> platform_events.MessageResult:
         raise NotSupportedError('forward_message')
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
diff --git a/src/langbot/pkg/platform/adapters/slack/event_converter.py b/src/langbot/pkg/platform/adapters/slack/event_converter.py
index 8e7d4cc3c..2dae3491e 100644
--- a/src/langbot/pkg/platform/adapters/slack/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/slack/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import time
 import typing
 
@@ -14,14 +12,21 @@ from langbot_plugin.api.entities.builtin.platform import events as platform_even
 
 
 class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
-    def __init__(self, bot_token: str = ''):
+    def __init__(self, bot_token: str = '', bot_user_id: str = ''):
         self.bot_token = bot_token
+        self.bot_user_id = bot_user_id
+        self.private_chats: dict[str, str] = {}
+
+    def remember_private_chat(self, channel_id: str, user_id: str):
+        if channel_id and user_id:
+            self.private_chats[channel_id] = user_id
+            while len(self.private_chats) > 4096:
+                self.private_chats.pop(next(iter(self.private_chats)))
 
     @staticmethod
     async def yiri2target(event: platform_events.Event) -> typing.Any:
         return getattr(event, 'source_platform_object', None)
 
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         self, event: SlackEvent
     ) -> platform_events.FriendMessage | platform_events.GroupMessage | None:
@@ -56,11 +61,94 @@ class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
             source_platform_object=event,
         )
 
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(self, event: SlackEvent) -> platform_events.Event:
-        if event.type in {'im', 'channel'}:
+        raw = event.get('event', {})
+        kind = raw.get('type', '')
+        subtype = raw.get('subtype', '')
+        common = dict(
+            adapter_name=ADAPTER_NAME,
+            timestamp=_timestamp_value(event),
+            source_platform_object=SlackEvent(_public_payload(dict(event))),
+        )
+        channel = raw.get('channel', '') or raw.get('item', {}).get('channel', '')
+        channel_info = channel if isinstance(channel, dict) else {'id': channel}
+        channel_id = channel_info.get('id', '')
+        group = platform_entities.UserGroup(id=channel_id, name=channel_info.get('name', channel_id))
+
+        def user(uid):
+            return platform_entities.User(id=uid, nickname=uid) if uid else None
+
+        private = raw.get('channel_type') == 'im' or str(channel_id).startswith('D')
+        chat = dict(
+            chat_type=platform_entities.ChatType.PRIVATE if private else platform_entities.ChatType.GROUP,
+            chat_id=self.private_chats.get(channel_id, channel_id) if private else channel_id,
+            group=None if private else group,
+        )
+        if kind == 'message' and subtype == 'message_changed':
+            message = raw.get('message', {})
+            updated = SlackEvent({**event, 'event': {**raw, **message}})
+            return platform_events.MessageEditedEvent(
+                **common,
+                **chat,
+                message_id=message.get('ts', ''),
+                editor=user(message.get('edited', {}).get('user') or message.get('user')) or user('unknown'),
+                new_content=await SlackMessageConverter.target2yiri(updated, self.bot_token),
+            )
+        if kind == 'message' and subtype == 'message_deleted':
+            return platform_events.MessageDeletedEvent(**common, **chat, message_id=raw.get('deleted_ts', ''))
+        if kind in {'reaction_added', 'reaction_removed'} and raw.get('item', {}).get('type') == 'message':
+            return platform_events.MessageReactionEvent(
+                **common,
+                **chat,
+                message_id=raw['item'].get('ts', ''),
+                user=user(raw.get('user')) or user('unknown'),
+                reaction=raw.get('reaction', ''),
+                is_add=kind == 'reaction_added',
+            )
+        if kind in {'member_joined_channel', 'member_left_channel'}:
+            member = user(raw.get('user')) or user('unknown')
+            inviter = user(raw.get('inviter'))
+            if self.bot_user_id and raw.get('user') == self.bot_user_id:
+                if kind == 'member_joined_channel' and inviter:
+                    return platform_events.BotInvitedToGroupEvent(**common, group=group, inviter=inviter)
+                if kind == 'member_left_channel':
+                    return platform_events.BotRemovedFromGroupEvent(**common, group=group)
+            if kind == 'member_joined_channel':
+                return platform_events.MemberJoinedEvent(
+                    **common,
+                    group=group,
+                    member=member,
+                    inviter=inviter,
+                    join_type='invite' if inviter else 'direct',
+                )
+            return platform_events.MemberLeftEvent(**common, group=group, member=member)
+        fields = {
+            'channel_rename': 'name',
+            'group_rename': 'name',
+            'channel_archive': 'archived',
+            'channel_unarchive': 'archived',
+            'group_archive': 'archived',
+            'group_unarchive': 'archived',
+            'channel_topic': 'topic',
+            'group_topic': 'topic',
+            'channel_purpose': 'purpose',
+            'group_purpose': 'purpose',
+        }
+        change = subtype if kind == 'message' else kind
+        if change in fields:
+            return platform_events.GroupInfoUpdatedEvent(
+                **common,
+                group=group,
+                operator=user(raw.get('user')),
+                changed_fields=[fields[change]],
+            )
+        if (
+            kind in {'message', 'app_mention'}
+            and subtype in {'', 'file_share', 'thread_broadcast'}
+            and event.type in {'im', 'channel'}
+        ):
             return await self.message_to_eba(event)
-        return self.platform_specific(event, f'slack.{event.type or "unknown"}')
+        return self.platform_specific(event, f'slack.{subtype or kind or event.get("type") or event.type or "unknown"}')
 
     async def message_to_eba(self, event: SlackEvent) -> platform_events.MessageReceivedEvent:
         sender_id = event.user_id or ''
@@ -71,6 +159,8 @@ class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
         chat_type = platform_entities.ChatType.PRIVATE
         chat_id = sender_id
         group = None
+        if event.type == 'im':
+            self.remember_private_chat(event.channel_id, sender_id)
         if event.type == 'channel':
             chat_type = platform_entities.ChatType.GROUP
             chat_id = event.channel_id or ''
@@ -86,7 +176,7 @@ class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
             chat_id=chat_id or '',
             group=group,
             timestamp=_timestamp_value(event),
-            source_platform_object=event,
+            source_platform_object=SlackEvent(_public_payload(dict(event))),
         )
 
     @staticmethod
@@ -95,15 +185,28 @@ class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
             type='platform.specific',
             adapter_name=ADAPTER_NAME,
             action=action,
-            data=dict(event),
+            data=_public_payload(dict(event)),
             timestamp=_timestamp_value(event),
-            source_platform_object=event,
+            source_platform_object=SlackEvent(_public_payload(dict(event))),
         )
 
 
 def _timestamp_value(event: SlackEvent) -> float:
-    raw_ts = event.get('event', {}).get('ts') or event.get('event', {}).get('event_ts')
+    raw_ts = event.get('event', {}).get('event_ts') or event.get('event', {}).get('ts') or event.get('event_time')
     try:
         return float(raw_ts)
     except (TypeError, ValueError):
         return time.time()
+
+
+def _public_payload(value):
+    """Keep callback credentials out of event data and monitoring records."""
+    if isinstance(value, dict):
+        return {
+            key: _public_payload(item)
+            for key, item in value.items()
+            if key not in {'token', 'response_url', 'response_urls'}
+        }
+    if isinstance(value, list):
+        return [_public_payload(item) for item in value]
+    return value
diff --git a/src/langbot/pkg/platform/adapters/slack/manifest.yaml b/src/langbot/pkg/platform/adapters/slack/manifest.yaml
index b5163f55e..7f59b510d 100644
--- a/src/langbot/pkg/platform/adapters/slack/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/slack/manifest.yaml
@@ -8,9 +8,9 @@ metadata:
     zh_Hans: Slack
     zh_Hant: Slack
   description:
-    en_US: Slack adapter with event-driven orchestration support, using LangBot's unified webhook endpoint.
-    zh_Hans: Slack 适配器，支持事件驱动编排，通过 LangBot 统一 Webhook 接收 Slack 事件订阅消息。
-    zh_Hant: Slack 適配器，支援事件驅動編排，透過 LangBot 統一 Webhook 接收 Slack 事件訂閱訊息。
+    en_US: Slack adapter with event-driven orchestration, supporting Webhook and Socket Mode.
+    zh_Hans: Slack 适配器，支持事件驱动编排，可使用 Webhook 或无需公网地址的长连接。
+    zh_Hant: Slack 適配器，支援事件驅動編排，可使用 Webhook 或無需公開網址的長連線。
   icon: slack.png
 
 spec:
@@ -22,7 +22,26 @@ spec:
     en: https://langbot.app/docs/en/usage/platforms/slack
     ja: https://langbot.app/docs/ja/usage/platforms/slack
   config:
+    - name: quick_setup
+      label: {en_US: Quick Slack setup, zh_Hans: Slack 快速配置, zh_Hant: Slack 快速設定, ja_JP: Slack クイック設定}
+      description: {en_US: Create and configure an app using App Configuration Tokens., zh_Hans: 填入配置令牌，自动创建应用并配置权限，授权后填入凭据。, zh_Hant: 填入設定權杖，自動建立應用程式並設定權限，授權後填入憑證。, ja_JP: 設定トークンでアプリを作成し、承認後に認証情報を入力します。}
+      type: qr-code-login
+      login_platform: slack
+      required: false
+    - name: socket_mode
+      label: {en_US: Socket Mode, zh_Hans: 长连接模式, zh_Hant: 長連線模式, ja_JP: Socket Mode}
+      type: boolean
+      default: false
+      required: false
+    - name: app_token
+      label: {en_US: App-Level Token, zh_Hans: App-Level Token, zh_Hant: App-Level Token, ja_JP: App-Level Token}
+      description: {en_US: Generate an xapp- token with connections:write in Basic Information., zh_Hans: 在 Basic Information 中生成带 connections:write 权限的 xapp- 令牌。, zh_Hant: 在 Basic Information 產生具有 connections:write 權限的 xapp- 權杖。, ja_JP: Basic Information で connections:write 権限の xapp- トークンを生成してください。}
+      type: string
+      required: false
+      default: ''
+      show_if: {field: socket_mode, operator: eq, value: true}
     - name: webhook_url
+      show_if: {field: socket_mode, operator: eq, value: false}
       label:
         en_US: Webhook Callback URL
         zh_Hans: Webhook 回调地址
@@ -43,16 +62,25 @@ spec:
       required: true
       default: ""
     - name: signing_secret
+      show_if: {field: socket_mode, operator: eq, value: false}
       label:
         en_US: Signing Secret
         zh_Hans: 签名密钥
         zh_Hant: 簽名密鑰
       type: string
-      required: true
+      required: false
       default: ""
 
   supported_events:
     - message.received
+    - message.edited
+    - message.deleted
+    - message.reaction
+    - group.member_joined
+    - group.member_left
+    - group.info_updated
+    - bot.invited_to_group
+    - bot.removed_from_group
     - platform.specific
 
   supported_apis:
diff --git a/src/langbot/pkg/platform/adapters/slack/message_converter.py b/src/langbot/pkg/platform/adapters/slack/message_converter.py
index 07a246dae..20c6f82d4 100644
--- a/src/langbot/pkg/platform/adapters/slack/message_converter.py
+++ b/src/langbot/pkg/platform/adapters/slack/message_converter.py
@@ -4,6 +4,7 @@ import datetime
 
 from langbot.libs.slack_api.slackevent import SlackEvent
 from langbot.pkg.utils import image
+from langbot.libs.slack_api.media import download_media
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
 from langbot_plugin.api.entities.builtin.platform import message as platform_message
 
@@ -55,15 +56,24 @@ class SlackMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
         if event.type == 'channel':
             components.append(platform_message.At(target='SlackBot'))
 
-        if event.pic_url:
-            try:
+        for file in event.get('event', {}).get('files', []):
+            url = file.get('url_private_download') or file.get('url_private')
+            if not url:
+                raise ValueError('Slack attachment is missing its download URL')
+            data, mime = await download_media(url, bot_token)
+            mime = file.get('mimetype') or mime
+            encoded = await image.encode_base64(data)
+            if mime.startswith('image/') and mime != 'image/svg+xml':
+                components.append(platform_message.Image(base64=f'data:{mime};base64,{encoded}'))
+            else:
                 components.append(
-                    platform_message.Image(
-                        url=event.pic_url, base64=await image.get_slack_image_to_base64(event.pic_url, bot_token)
+                    platform_message.File(
+                        id=file.get('id', ''),
+                        name=file.get('name') or file.get('title') or 'attachment',
+                        size=len(data),
+                        base64=encoded,
                     )
                 )
-            except Exception:
-                components.append(platform_message.Image(url=event.pic_url))
 
         if event.text:
             components.append(platform_message.Plain(text=event.text))
diff --git a/src/langbot/pkg/platform/adapters/slack/platform_api.py b/src/langbot/pkg/platform/adapters/slack/platform_api.py
index e1a80bf04..1e6d37d6c 100644
--- a/src/langbot/pkg/platform/adapters/slack/platform_api.py
+++ b/src/langbot/pkg/platform/adapters/slack/platform_api.py
@@ -5,7 +5,8 @@ import typing
 
 async def get_mode(adapter, params: dict) -> dict:
     return {
-        'webhook': True,
+        'webhook': not adapter.config.get('socket_mode', False),
+        'socket_mode': adapter.config.get('socket_mode', False),
         'bot_account_id': adapter.bot_account_id,
     }
 
diff --git a/src/langbot/pkg/platform/adapters/telegram/adapter.py b/src/langbot/pkg/platform/adapters/telegram/adapter.py
index 0e4ae9c08..5d6519267 100644
--- a/src/langbot/pkg/platform/adapters/telegram/adapter.py
+++ b/src/langbot/pkg/platform/adapters/telegram/adapter.py
@@ -6,9 +6,6 @@ Preserves all existing functionality (messaging, streaming output, markdown card
 
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-from langbot.pkg.telemetry.adapter_diagnostics import record_api_result
-
 import typing
 import traceback
 
@@ -71,16 +68,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         arbitrary_types_allowed = True
 
     def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger):
-        @diagnostics.observe(
-            'event',
-            'platform.native_callback',
-            source='platform',
-            stage='convert',
-            ap=lambda: getattr(logger, 'ap', None),
-            fields=lambda b: {
-                'workspace_uuid': getattr(getattr(logger, 'execution_context', None), 'workspace_uuid', '')
-            },
-        )
         async def telegram_callback(update: Update, context: ContextTypes.DEFAULT_TYPE):
             if (
                 not update.message
@@ -162,7 +149,7 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         application.add_handler(
             MessageReactionHandler(
                 telegram_callback,
-                MessageReactionHandler.MESSAGE_REACTION,
+                message_reaction_types=MessageReactionHandler.MESSAGE_REACTION,
             )
         )
 
@@ -219,7 +206,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
     # ---- Message Send / Reply (preserving original logic) ----
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
         components = await TelegramMessageConverter.yiri2target(message, self.bot)
 
@@ -227,6 +213,7 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         chat_id: int | str = int(chat_id_str) if chat_id_str.lstrip('-').isdigit() else chat_id_str
         message_thread_id = int(thread_id_str) if thread_id_str and thread_id_str.isdigit() else None
 
+        sent_messages = []
         for component in components:
             component_type = component.get('type')
             args = {'chat_id': chat_id}
@@ -239,22 +226,25 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
                     text = telegramify_markdown.markdownify(content=text)
                     args['parse_mode'] = 'MarkdownV2'
                 args['text'] = text
-                record_api_result(await self.bot.send_message(**args))
+                sent_messages.append(await self.bot.send_message(**args))
             elif component_type == 'photo':
                 photo = component.get('photo')
                 if photo is None:
                     continue
-                args['photo'] = telegram.InputFile(photo)
-                record_api_result(await self.bot.send_photo(**args))
+                sent_messages.append(await self._send_photo_or_document(photo, args))
             elif component_type == 'document':
                 doc = component.get('document')
                 if doc is None:
                     continue
                 filename = component.get('filename', 'file')
                 args['document'] = telegram.InputFile(doc, filename=filename)
-                record_api_result(await self.bot.send_document(**args))
+                sent_messages.append(await self.bot.send_document(**args))
+
+        return platform_events.MessageResult(
+            message_id=sent_messages[-1].message_id if sent_messages else None,
+            raw={'message_ids': [sent.message_id for sent in sent_messages]},
+        )
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -264,6 +254,7 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         assert isinstance(message_source.source_platform_object, Update)
         components = await TelegramMessageConverter.yiri2target(message, self.bot)
 
+        sent_messages = []
         for component in components:
             component_type = component.get('type')
             args = {
@@ -286,20 +277,34 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
                 if self.config['markdown_card'] is True:
                     args['parse_mode'] = 'MarkdownV2'
                 args['text'] = content
-                record_api_result(await self.bot.send_message(**args))
+                sent_messages.append(await self.bot.send_message(**args))
             elif component_type == 'photo':
                 photo = component.get('photo')
                 if photo is None:
                     continue
-                args['photo'] = telegram.InputFile(photo)
-                record_api_result(await self.bot.send_photo(**args))
+                sent_messages.append(await self._send_photo_or_document(photo, args))
             elif component_type == 'document':
                 doc = component.get('document')
                 if doc is None:
                     continue
                 filename = component.get('filename', 'file')
                 args['document'] = telegram.InputFile(doc, filename=filename)
-                record_api_result(await self.bot.send_document(**args))
+                sent_messages.append(await self.bot.send_document(**args))
+
+        return platform_events.MessageResult(
+            message_id=sent_messages[-1].message_id if sent_messages else None,
+            raw={'message_ids': [sent.message_id for sent in sent_messages]},
+        )
+
+    async def _send_photo_or_document(self, photo: bytes, args: dict):
+        try:
+            return await self.bot.send_photo(photo=telegram.InputFile(photo), **args)
+        except telegram.error.BadRequest as exc:
+            if 'image_process_failed' not in str(exc).lower() and 'photo_invalid_dimensions' not in str(exc).lower():
+                raise
+            # Preserve rejected image bytes without retrying already-sent parts.
+            await self.logger.warning('Telegram rejected photo encoding; sending the image as a document')
+            return await self.bot.send_document(document=telegram.InputFile(photo, filename='image'), **args)
 
     # ---- Streaming Output (preserving original logic) ----
 
@@ -318,6 +323,14 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
     _MAX_STREAM_STATES: typing.ClassVar[int] = 1000
 
+    async def _delete_group_stream_message(self, chat_mode: str, chat_id: int, stream_id: int | None):
+        if chat_mode == 'private' or stream_id is None:
+            return
+        try:
+            await self.bot.delete_message(chat_id=chat_id, message_id=stream_id)
+        except telegram.error.TelegramError:
+            pass
+
     def _cap_stream_states(self) -> None:
         while len(self.msg_stream_id) > self._MAX_STREAM_STATES:
             self.msg_stream_id.pop(next(iter(self.msg_stream_id)), None)
@@ -332,7 +345,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         cleaned = text.replace('\u200b', '').replace('\u200c', '').replace('\u200d', '').replace('\ufeff', '').strip()
         return cleaned == ''
 
-    @diagnostics.observe('api', 'create_message_card', source='platform', stage='accepted')
     async def create_message_card(self, message_id, event):
         assert isinstance(event.source_platform_object, Update)
         update = event.source_platform_object
@@ -347,7 +359,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
         return True
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -367,6 +378,15 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
         if message_id not in self.msg_stream_id:
             return
 
+        # Media cannot be edited into a text stream. Deliver terminal attachments
+        # through the regular reply path, including attachment-only completions.
+        if is_final and not bot_message.tool_calls:
+            attachments = platform_message.MessageChain(
+                [part for part in message if isinstance(part, (platform_message.Image, platform_message.File))]
+            )
+            if attachments:
+                await self.reply_message(message_source, attachments, quote_origin=quote_origin)
+
         stream_state = self.msg_stream_id[message_id]
         chat_mode, stream_id = stream_state[:2]
         has_visible_content = len(stream_state) > 2 and stream_state[2]
@@ -374,6 +394,8 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
         if not components or components[0]['type'] != 'text':
             if is_final and bot_message.tool_calls is None:
+                if not has_visible_content:
+                    await self._delete_group_stream_message(chat_mode, chat_id, stream_id)
                 self.msg_stream_id.pop(message_id)
             return
 
@@ -483,7 +505,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
         """Dispatch once, preferring the most specific registered listener."""
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
@@ -510,7 +531,6 @@ class TelegramAdapter(TelegramAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
     # ---- Pass-through API ----
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(
         self,
         action: str,
diff --git a/src/langbot/pkg/platform/adapters/telegram/api_impl.py b/src/langbot/pkg/platform/adapters/telegram/api_impl.py
index e6e053485..f058d54e6 100644
--- a/src/langbot/pkg/platform/adapters/telegram/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/telegram/api_impl.py
@@ -5,9 +5,6 @@ Implements optional API methods defined in AbstractPlatformAdapter.
 
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-from langbot.pkg.telemetry.adapter_diagnostics import record_api_result
-
 import typing
 
 import telegram
@@ -28,7 +25,6 @@ class TelegramAPIMixin:
 
     bot: telegram.Bot
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -53,10 +49,9 @@ class TelegramAPIMixin:
                 }
                 if self.config.get('markdown_card', False):
                     args['parse_mode'] = 'MarkdownV2'
-                record_api_result(await self.bot.edit_message_text(**args), edited_content=new_content)
+                await self.bot.edit_message_text(**args)
                 return
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -64,9 +59,8 @@ class TelegramAPIMixin:
         message_id: typing.Union[int, str],
     ) -> None:
         """Delete / recall a message."""
-        record_api_result(await self.bot.delete_message(chat_id=chat_id, message_id=message_id))
+        await self.bot.delete_message(chat_id=chat_id, message_id=message_id)
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -86,7 +80,6 @@ class TelegramAPIMixin:
             raw={'message_id': result.message_id},
         )
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(
         self,
         group_id: typing.Union[int, str],
@@ -107,7 +100,6 @@ class TelegramAPIMixin:
         except Exception:
             return None
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -142,7 +134,6 @@ class TelegramAPIMixin:
             )
         return members
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -169,7 +160,6 @@ class TelegramAPIMixin:
             display_name=member.custom_title if hasattr(member, 'custom_title') else None,
         )
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(
         self,
         user_id: typing.Union[int, str],
@@ -182,7 +172,6 @@ class TelegramAPIMixin:
             username=chat.username,
         )
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(
         self,
         file_data: bytes,
@@ -197,7 +186,6 @@ class TelegramAPIMixin:
 
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(
         self,
         file_id: str,
@@ -206,7 +194,6 @@ class TelegramAPIMixin:
         file = await self.bot.get_file(file_id)
         return file.file_path
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(
         self,
         group_id: typing.Union[int, str],
@@ -224,9 +211,8 @@ class TelegramAPIMixin:
         }
         if duration > 0:
             kwargs['until_date'] = datetime.datetime.now(datetime.timezone.utc) + datetime.timedelta(seconds=duration)
-        record_api_result(await self.bot.restrict_chat_member(**kwargs))
+        await self.bot.restrict_chat_member(**kwargs)
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(
         self,
         group_id: typing.Union[int, str],
@@ -244,27 +230,23 @@ class TelegramAPIMixin:
             can_send_video_notes=True,
             can_send_voice_notes=True,
         )
-        record_api_result(
-            await self.bot.restrict_chat_member(
-                chat_id=group_id,
-                user_id=user_id,
-                permissions=permissions,
-            )
+        await self.bot.restrict_chat_member(
+            chat_id=group_id,
+            user_id=user_id,
+            permissions=permissions,
         )
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(
         self,
         group_id: typing.Union[int, str],
         user_id: typing.Union[int, str],
     ) -> None:
         """Kick a member from the group."""
-        record_api_result(await self.bot.ban_chat_member(chat_id=group_id, user_id=user_id))
+        await self.bot.ban_chat_member(chat_id=group_id, user_id=user_id)
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(
         self,
         group_id: typing.Union[int, str],
     ) -> None:
         """Make the bot leave a group."""
-        record_api_result(await self.bot.leave_chat(chat_id=group_id))
+        await self.bot.leave_chat(chat_id=group_id)
diff --git a/src/langbot/pkg/platform/adapters/telegram/event_converter.py b/src/langbot/pkg/platform/adapters/telegram/event_converter.py
index f62fb2946..30baf66da 100644
--- a/src/langbot/pkg/platform/adapters/telegram/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/telegram/event_converter.py
@@ -5,9 +5,8 @@ Converts all Telegram Update types to unified EBA events, not just messages.
 
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
+import datetime
 
 import telegram
 from telegram import Update
@@ -19,6 +18,13 @@ import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platf
 from langbot.pkg.platform.adapters.telegram.message_converter import TelegramMessageConverter
 
 
+def _restriction_duration(until_date: datetime.datetime | None, event_date: datetime.datetime) -> int | None:
+    """Telegram uses Unix epoch zero for an indefinite restriction."""
+    if until_date is None or until_date.timestamp() <= 0:
+        return None
+    return max(0, int(until_date.timestamp() - event_date.timestamp()))
+
+
 def _make_user(tg_user: telegram.User) -> platform_entities.User:
     """Convert a Telegram User to a unified User entity."""
     return platform_entities.User(
@@ -56,7 +62,6 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return None
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(
         update: Update,
         bot: telegram.Bot,
@@ -232,9 +237,7 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if new_status == 'restricted' and cm.new_chat_member:
             restricted = cm.new_chat_member
             if hasattr(restricted, 'can_send_messages') and not restricted.can_send_messages:
-                duration = None
-                if hasattr(restricted, 'until_date') and restricted.until_date:
-                    duration = int(restricted.until_date.timestamp() - time.time())
+                duration = _restriction_duration(restricted.until_date, cm.date)
                 return platform_events.MemberBannedEvent(
                     type='group.member_banned',
                     timestamp=cm.date.timestamp() if cm.date else time.time(),
@@ -286,7 +289,7 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
             )
 
         # Bot removed from group
-        if old_status in ('member', 'administrator', 'creator') and new_status in ('left', 'kicked'):
+        if old_status in ('member', 'administrator', 'creator', 'restricted') and new_status in ('left', 'kicked'):
             return platform_events.BotRemovedFromGroupEvent(
                 type='bot.removed_from_group',
                 timestamp=mcm.date.timestamp() if mcm.date else time.time(),
@@ -300,9 +303,7 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
         if new_status == 'restricted' and mcm.new_chat_member:
             restricted = mcm.new_chat_member
             if hasattr(restricted, 'can_send_messages') and not restricted.can_send_messages:
-                duration = None
-                if hasattr(restricted, 'until_date') and restricted.until_date:
-                    duration = int(restricted.until_date.timestamp() - time.time())
+                duration = _restriction_duration(restricted.until_date, mcm.date)
                 return platform_events.BotMutedEvent(
                     type='bot.muted',
                     timestamp=mcm.date.timestamp() if mcm.date else time.time(),
@@ -313,7 +314,11 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
                     source_platform_object=update,
                 )
 
-        if old_status == 'restricted' and new_status in ('member', 'administrator') and mcm.new_chat_member:
+        was_muted = old_status == 'restricted' and not mcm.old_chat_member.can_send_messages
+        can_send = new_status in ('member', 'administrator') or (
+            new_status == 'restricted' and mcm.new_chat_member.can_send_messages
+        )
+        if was_muted and can_send:
             return platform_events.BotUnmutedEvent(
                 type='bot.unmuted',
                 timestamp=mcm.date.timestamp() if mcm.date else time.time(),
@@ -344,14 +349,18 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
         reaction = update.message_reaction
         chat = reaction.chat
 
-        # Extract newly added emojis
-        new_emojis = []
-        if reaction.new_reaction:
-            for r in reaction.new_reaction:
-                if hasattr(r, 'emoji'):
-                    new_emojis.append(r.emoji)
-                elif hasattr(r, 'custom_emoji_id'):
-                    new_emojis.append(str(r.custom_emoji_id))
+        def emojis(items):
+            return [
+                str(r.emoji if hasattr(r, 'emoji') else r.custom_emoji_id)
+                for r in items
+                if hasattr(r, 'emoji') or hasattr(r, 'custom_emoji_id')
+            ]
+
+        old_emojis = emojis(reaction.old_reaction or [])
+        new_emojis = emojis(reaction.new_reaction or [])
+        added = [emoji for emoji in new_emojis if emoji not in old_emojis]
+        removed = [emoji for emoji in old_emojis if emoji not in new_emojis]
+        changed = added or removed
 
         user = platform_entities.User(id='')
         if reaction.user:
@@ -366,8 +375,8 @@ class TelegramEventConverter(abstract_platform_adapter.AbstractEventConverter):
             adapter_name='telegram-omni',
             message_id=reaction.message_id,
             user=user,
-            reaction=new_emojis[0] if new_emojis else '',
-            is_add=len(new_emojis) > 0,
+            reaction=changed[0] if changed else '',
+            is_add=bool(added),
             chat_type=ct,
             chat_id=chat.id,
             group=group,
@@ -387,13 +396,6 @@ class LegacyEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return event.source_platform_object
 
     @staticmethod
-    @diagnostics.observe(
-        'event',
-        'platform.target2yiri',
-        source='platform',
-        stage='convert',
-        fields=lambda _: {'attributes': {'adapter_evidence': False}},
-    )
     async def target2yiri(event: Update, bot: telegram.Bot, bot_account_id: str):
         """Convert to legacy format (FriendMessage / GroupMessage)."""
         import langbot_plugin.api.entities.builtin.platform.events as legacy_events
diff --git a/src/langbot/pkg/platform/adapters/wecom/adapter.py b/src/langbot/pkg/platform/adapters/wecom/adapter.py
index 447980891..7096e8d80 100644
--- a/src/langbot/pkg/platform/adapters/wecom/adapter.py
+++ b/src/langbot/pkg/platform/adapters/wecom/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import traceback
 import typing
@@ -92,7 +90,6 @@ class WecomAdapter(WecomAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -109,7 +106,6 @@ class WecomAdapter(WecomAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             raw_results.append(await self._send_content(user_id, agent_id, content))
         return platform_events.MessageResult(raw={'results': raw_results})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -125,7 +121,6 @@ class WecomAdapter(WecomAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             raw_results.append(await self._send_content(wecom_event.user_id, int(wecom_event.agent_id), content))
         return platform_events.MessageResult(message_id=wecom_event.message_id, raw={'results': raw_results})
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -179,7 +174,6 @@ class WecomAdapter(WecomAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
         self.bot.on_message('text')(on_message)
         self.bot.on_message('image')(on_message)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: WecomEvent):
         self.bot_account_id = event.receiver_id or self.bot_account_id
         try:
@@ -198,7 +192,6 @@ class WecomAdapter(WecomAPIMixin, abstract_platform_adapter.AbstractPlatformAdap
             await self.logger.error(f'Error in wecom native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/wecom/api_impl.py b/src/langbot/pkg/platform/adapters/wecom/api_impl.py
index b768a1259..07ec2784d 100644
--- a/src/langbot/pkg/platform/adapters/wecom/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/wecom/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.wecom_api.api import WecomClient
@@ -16,7 +14,6 @@ class WecomAPIMixin:
     _message_cache: dict[str, platform_events.MessageReceivedEvent]
     _user_cache: dict[str, platform_entities.User]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -28,7 +25,6 @@ class WecomAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         cached = self._user_cache.get(str(user_id))
         if cached is not None:
@@ -40,30 +36,24 @@ class WecomAPIMixin:
             username=info.get('alias') or info.get('userid') or None,
         )
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         raise NotSupportedError('get_group_info')
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
     ) -> list[platform_entities.UserGroupMember]:
         raise NotSupportedError('get_group_member_list')
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -71,7 +61,6 @@ class WecomAPIMixin:
     ) -> platform_entities.UserGroupMember:
         raise NotSupportedError('get_group_member_info')
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -81,7 +70,6 @@ class WecomAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
diff --git a/src/langbot/pkg/platform/adapters/wecom/event_converter.py b/src/langbot/pkg/platform/adapters/wecom/event_converter.py
index 41d301ae6..94c43fd78 100644
--- a/src/langbot/pkg/platform/adapters/wecom/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/wecom/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.wecom_api.api import WecomClient
@@ -19,7 +17,6 @@ class WecomEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return getattr(event, 'source_platform_object', None)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(event: WecomEvent, bot: WecomClient | None = None) -> platform_events.FriendMessage | None:
         eba_event = await WecomEventConverter.target2yiri(event, bot)
         if hasattr(eba_event, 'to_legacy_event'):
@@ -39,7 +36,6 @@ class WecomEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return None
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(event: WecomEvent, bot: WecomClient | None = None) -> platform_events.Event | None:
         if event.type in {'text', 'image'}:
             return await WecomEventConverter.message_to_eba(event, bot)
diff --git a/src/langbot/pkg/platform/adapters/wecom/manifest.yaml b/src/langbot/pkg/platform/adapters/wecom/manifest.yaml
index fd63e8348..a5b11284e 100644
--- a/src/langbot/pkg/platform/adapters/wecom/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/wecom/manifest.yaml
@@ -8,9 +8,9 @@ metadata:
     zh_Hans: 企业微信
     zh_Hant: 企業微信
   description:
-    en_US: WeCom application message adapter with event-driven orchestration support
-    zh_Hans: 企业微信内部应用消息适配器，支持事件驱动编排
-    zh_Hant: 企業微信內部應用訊息適配器，支援事件驅動編排
+    en_US: Serve employees through a custom WeCom app
+    zh_Hans: 通过企业自建应用，为内部员工提供服务
+    zh_Hant: 透過企業自建應用，為內部員工提供服務
   icon: wecom.png
 
 spec:
diff --git a/src/langbot/pkg/platform/adapters/wecombot/adapter.py b/src/langbot/pkg/platform/adapters/wecombot/adapter.py
index ce23bab27..90ee727e7 100644
--- a/src/langbot/pkg/platform/adapters/wecombot/adapter.py
+++ b/src/langbot/pkg/platform/adapters/wecombot/adapter.py
@@ -1,10 +1,9 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 from langbot.pkg.platform.sources.wecombot import WecomBotAdapter as LegacyWecomBotAdapter
 
 import asyncio
+import base64
 import time
 import traceback
 import typing
@@ -49,6 +48,7 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
     _group_cache: dict[str, platform_entities.UserGroup] = {}
     _member_cache: dict[tuple[str, str], platform_entities.UserGroupMember] = {}
     _stream_to_monitoring_msg: dict[str, tuple[str, float]] = {}
+    _business_event_ids: dict[tuple[str, str], float] = {}
     _STREAM_MAPPING_TTL: int = 600
 
     class Config:
@@ -105,6 +105,8 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
         return [
             'message.received',
             'feedback.received',
+            'wecombot.enter_chat',
+            'wecombot.template_card_event',
             'platform.specific',
         ]
 
@@ -133,9 +135,30 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
     _join_text_components = staticmethod(LegacyWecomBotAdapter._join_text_components)
     _iter_media_components = staticmethod(LegacyWecomBotAdapter._iter_media_components)
-    _send_media = staticmethod(LegacyWecomBotAdapter._send_media)
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
+    @staticmethod
+    async def _send_media(bot, req_id: str, item: dict) -> bool:
+        """Require upload and delivery acknowledgements instead of dropping attachments."""
+        kind = item['type']
+        payload = item.get('base64') or ''
+        if payload.startswith('data:'):
+            payload = payload.split(',', 1)[-1]
+        try:
+            data = base64.b64decode(payload, validate=True)
+        except ValueError as exc:
+            raise ValueError(f'Invalid {kind} attachment data') from exc
+        if not data:
+            raise ValueError(f'{kind} attachment has no data to upload')
+        result = await bot.upload_media(data, item.get('name') or f'attachment.{kind}', media_type=kind)
+        media_id = result.get('media_id') if isinstance(result, dict) else None
+        if not media_id:
+            raise RuntimeError(f'WeCom {kind} upload failed: no media ID returned')
+        ack = await getattr(bot, f'reply_{kind}')(req_id, media_id)
+        if ack is None or ack.get('errcode', 0) != 0:
+            code = ack.get('errcode') if isinstance(ack, dict) else 'no acknowledgement'
+            raise RuntimeError(f'WeCom {kind} delivery failed: {code}')
+        return True
+
     async def send_message(
         self,
         target_type: str,
@@ -148,10 +171,37 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
             raise NotSupportedError(f'send_message:{target_type}')
         items = await WecomBotMessageConverter.yiri2target(message)
         content = self._join_text_components(items)
+        if list(self._iter_media_components(items)):
+            raise NotSupportedError('send_message:media requires an originating message callback')
         raw = await self.bot.send_message(str(target_id), content)
         return platform_events.MessageResult(raw={'result': raw})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
+    async def reply_to_event(
+        self, *, context: dict, target_type: str, target_id: str, message: platform_message.MessageChain
+    ) -> platform_events.MessageResult:
+        """Bind Agent replies to the originating callback and its loading stream."""
+        message_id = ((context.get('delivery') or {}).get('reply_target') or {}).get('message_id')
+        if context.get('event_type') == 'message.received' and message_id:
+            source = await self.get_message(target_type, target_id, message_id)
+            return await self.reply_message(source, message)
+        if context.get('event_type') == 'wecombot.enter_chat':
+            data = context.get('data') or {}
+            raw_event = data.get('data') or {}
+            if time.time() - float(data.get('timestamp') or 0) >= 5:
+                raise ValueError('WeCom welcome replies must complete within 5 seconds of entering the chat')
+            items = await WecomBotMessageConverter.yiri2target(message)
+            if list(self._iter_media_components(items)):
+                raise NotSupportedError('welcome_reply:media')
+            content = self._join_text_components(items)
+            callback_id = data.get('message_id') if self.config.get('enable-webhook') else raw_event.get('req_id')
+            if not callback_id:
+                raise ValueError('WeCom welcome reply is missing its callback ID')
+            raw = await self.bot.reply_welcome(callback_id, content)
+            if raw is None:
+                raise RuntimeError('WeCom welcome reply was not acknowledged')
+            return platform_events.MessageResult(raw={'result': raw})
+        return await self.send_message(target_type, target_id, message)
+
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -162,18 +212,24 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
         if not isinstance(event, WecomBotEvent):
             raise ValueError('WeComBot reply_message requires a WecomBotEvent source object')
         items = await WecomBotMessageConverter.yiri2target(message)
+        if self.config.get('enable-webhook') and list(self._iter_media_components(items)):
+            raise NotSupportedError('reply_media:webhook_mode; use WebSocket mode to deliver files')
         content = self._join_text_components(items)
         raw = None
+        from .stream_text import format_stream_text
+
+        content = format_stream_text(content)
         if not self.config.get('enable-webhook', False) and event.get('req_id'):
             if content:
-                raw = await self.bot.reply_text(event.get('req_id'), content)
+                finished = await self.bot.push_stream_chunk(event.message_id, content, is_final=True)
+                if not finished:
+                    raw = await self.bot.reply_text(event.get('req_id'), content)
             for item in self._iter_media_components(items):
                 await self._send_media(self.bot, event.get('req_id'), item)
         else:
             raw = await self.bot.set_message(event.message_id, content)
         return platform_events.MessageResult(message_id=event.message_id, raw={'result': raw})
 
-    @diagnostics.observe('api', 'reply_message_chunk', source='platform', stage='accepted')
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -186,7 +242,11 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
         if not isinstance(event, WecomBotEvent):
             raise ValueError('WeComBot reply_message_chunk requires a WecomBotEvent source object')
         items = await WecomBotMessageConverter.yiri2target(message)
-        content = self._join_text_components(items)
+        if is_final and self.config.get('enable-webhook') and list(self._iter_media_components(items)):
+            raise NotSupportedError('reply_media:webhook_mode; use WebSocket mode to deliver files')
+        from .stream_text import format_stream_text
+
+        content = format_stream_text(self._join_text_components(items))
         success = await self.bot.push_stream_chunk(event.message_id, content, is_final=is_final)
         if not success and is_final and not self.config.get('enable-webhook', False) and event.get('req_id'):
             await self.bot.reply_text(event.get('req_id'), content)
@@ -201,7 +261,6 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
     async def is_stream_output_supported(self) -> bool:
         return self.config.get('enable-stream-reply', True)
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         if action == 'interaction.request' and 'interaction.request' in self.get_supported_apis():
             return await send_interaction(self, params)
@@ -277,17 +336,18 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
             self.bot.on_message('event')(self._handle_native_event)
             self.bot.on_message('template_card_event')(self._handle_interaction_event)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_interaction_event(self, event: WecomBotEvent):
         try:
             interaction_event = interaction_event_from_native(event)
             if interaction_event is not None:
                 await self._dispatch_eba_event(interaction_event)
+            else:
+                await self._handle_native_event(event)
         except Exception:
             await self.logger.error(f'Error in WeComBot interaction callback: {traceback.format_exc()}')
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: WecomBotEvent):
+        key = None
         try:
             if platform_events.FriendMessage in self.listeners or platform_events.GroupMessage in self.listeners:
                 legacy_event = await self.event_converter.target2legacy(event)
@@ -296,9 +356,24 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
 
             eba_event = await self.event_converter.target2yiri(event)
             if eba_event:
+                if eba_event.type in {'wecombot.enter_chat', 'wecombot.template_card_event'}:
+                    message_id = eba_event.message_id
+                    now = time.monotonic()
+                    self._business_event_ids = {
+                        key: seen for key, seen in self._business_event_ids.items() if now - seen < 600
+                    }
+                    key = (eba_event.type, message_id)
+                    if message_id and key in self._business_event_ids:
+                        return
+                    if message_id:
+                        self._business_event_ids[key] = now
+                        if len(self._business_event_ids) > 2048:
+                            self._business_event_ids.pop(next(iter(self._business_event_ids)))
                 self._cache_event(eba_event)
                 await self._dispatch_eba_event(eba_event)
         except Exception:
+            if key is not None:
+                self._business_event_ids.pop(key, None)
             await self.logger.error(f'Error in wecombot native event: {traceback.format_exc()}')
 
     async def _handle_feedback(self, **kwargs):
@@ -312,7 +387,6 @@ class WecomBotAdapter(WecomBotAPIMixin, abstract_platform_adapter.AbstractPlatfo
             await self.logger.error(f'Error in wecombot feedback event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/wecombot/api_impl.py b/src/langbot/pkg/platform/adapters/wecombot/api_impl.py
index cbd34ba38..15758fa5b 100644
--- a/src/langbot/pkg/platform/adapters/wecombot/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/wecombot/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot_plugin.api.entities.builtin.platform import entities as platform_entities
@@ -16,7 +14,6 @@ class WecomBotAPIMixin:
     _group_cache: dict[str, platform_entities.UserGroup]
     _member_cache: dict[tuple[str, str], platform_entities.UserGroupMember]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -28,25 +25,21 @@ class WecomBotAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         cached = self._user_cache.get(str(user_id))
         if cached is None:
             raise NotSupportedError('get_user_info:not_cached')
         return cached
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         cached = self._group_cache.get(str(group_id))
         if cached is None:
             raise NotSupportedError('get_group_info:not_cached')
         return cached
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -57,7 +50,6 @@ class WecomBotAPIMixin:
             raise NotSupportedError('get_group_member_info:not_cached')
         return cached
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
@@ -66,15 +58,12 @@ class WecomBotAPIMixin:
             member for (cached_group_id, _), member in self._member_cache.items() if cached_group_id == str(group_id)
         ]
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -84,7 +73,6 @@ class WecomBotAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
@@ -93,7 +81,6 @@ class WecomBotAPIMixin:
     ) -> None:
         raise NotSupportedError('delete_message')
 
-    @diagnostics.observe('api', 'forward_message', source='platform', stage='accepted')
     async def forward_message(
         self,
         from_chat_type: str,
@@ -104,18 +91,14 @@ class WecomBotAPIMixin:
     ) -> platform_events.MessageResult:
         raise NotSupportedError('forward_message')
 
-    @diagnostics.observe('api', 'mute_member', source='platform', stage='accepted')
     async def mute_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str], duration: int = 0):
         raise NotSupportedError('mute_member')
 
-    @diagnostics.observe('api', 'unmute_member', source='platform', stage='accepted')
     async def unmute_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]):
         raise NotSupportedError('unmute_member')
 
-    @diagnostics.observe('api', 'kick_member', source='platform', stage='accepted')
     async def kick_member(self, group_id: typing.Union[int, str], user_id: typing.Union[int, str]):
         raise NotSupportedError('kick_member')
 
-    @diagnostics.observe('api', 'leave_group', source='platform', stage='accepted')
     async def leave_group(self, group_id: typing.Union[int, str]):
         raise NotSupportedError('leave_group')
diff --git a/src/langbot/pkg/platform/adapters/wecombot/event_converter.py b/src/langbot/pkg/platform/adapters/wecombot/event_converter.py
index 207cfd71c..a99508813 100644
--- a/src/langbot/pkg/platform/adapters/wecombot/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/wecombot/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import time
 import typing
 
@@ -21,7 +19,6 @@ class WecomBotEventConverter(abstract_platform_adapter.AbstractEventConverter):
     async def yiri2target(event: platform_events.Event) -> typing.Any:
         return getattr(event, 'source_platform_object', None)
 
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         self, event: WecomBotEvent
     ) -> platform_events.FriendMessage | platform_events.GroupMessage | None:
@@ -52,8 +49,37 @@ class WecomBotEventConverter(abstract_platform_adapter.AbstractEventConverter):
             source_platform_object=event,
         )
 
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(self, event: WecomBotEvent) -> platform_events.Event:
+        from langbot.libs.wecom_ai_bot_api.api import (
+            extract_wecom_event_type,
+            extract_template_card_event_payload,
+            extract_template_card_action,
+        )
+
+        event_type = extract_wecom_event_type(event) or event.get('eventtype')
+        if event_type in {'enter_chat', 'template_card_event'}:
+            group = event.get('chattype', event.type) == 'group'
+            fields = dict(
+                adapter_name=ADAPTER_NAME,
+                user=platform_entities.User(id=event.userid, nickname=event.username),
+                chat_id=str(event.chatid if group else event.userid),
+                chat_type=platform_entities.ChatType.GROUP if group else platform_entities.ChatType.PRIVATE,
+                message_id=event.message_id or '',
+                timestamp=float(event.get('create_time') or time.time()),
+                data=dict(event),
+                source_platform_object=event,
+            )
+            if event_type == 'enter_chat':
+                return platform_events.WecomBotEnterChatEvent(**fields)
+            card = extract_template_card_event_payload(event)
+            task_id, event_key, card_type = extract_template_card_action(card)
+            return platform_events.WecomBotTemplateCardEvent(
+                **fields,
+                task_id=task_id,
+                event_key=event_key,
+                card_type=card_type,
+                selected_items=card.get('selected_items') or card.get('SelectedItems') or [],
+            )
         if event.type in {'single', 'group'} and event.msgtype != 'event':
             return await self.message_to_eba(event)
         return self.platform_specific(
diff --git a/src/langbot/pkg/platform/adapters/wecombot/manifest.yaml b/src/langbot/pkg/platform/adapters/wecombot/manifest.yaml
index 2b722142f..d5b7c1eeb 100644
--- a/src/langbot/pkg/platform/adapters/wecombot/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/wecombot/manifest.yaml
@@ -8,9 +8,9 @@ metadata:
     zh_Hans: 企业微信智能机器人
     zh_Hant: 企業微信智慧機器人
   description:
-    en_US: WeCom AI Bot adapter with event-driven orchestration support
-    zh_Hans: 企业微信智能机器人适配器，支持事件驱动编排、长连接和 Webhook 两种接入方式
-    zh_Hant: 企業微信智慧機器人適配器，支援事件驅動編排、長連線和 Webhook 兩種接入方式
+    en_US: Answer questions and assist teams in WeCom chats and groups
+    zh_Hans: 在企业微信单聊和群聊中提供智能问答与协作
+    zh_Hant: 在企業微信單聊和群聊中提供智慧問答與協作
   icon: wecombot.png
 
 spec:
@@ -27,10 +27,9 @@ spec:
         zh_Hans: 一键创建机器人
         zh_Hant: 一鍵建立機器人
       description:
-        en_US: 'Scan QR code with WeCom to automatically create a bot and fill in BotId and Secret. Note: Robot Name needs to
-          be filled in manually.'
-        zh_Hans: 使用企业微信扫码自动创建机器人并填写 BotId 和 Secret。注意：机器人名称需手动填写。
-        zh_Hant: 使用企業微信掃碼自動建立機器人並填寫 BotId 和 Secret。注意：機器人名稱需手動填寫。
+        en_US: Scan with WeCom to create a bot and fill in BotId and Secret automatically.
+        zh_Hans: 使用企业微信扫码自动创建机器人并填写 BotId 和 Secret。
+        zh_Hant: 使用企業微信掃碼自動建立機器人並填寫 BotId 和 Secret。
       type: qr-code-login
       login_platform: wecombot
       required: false
@@ -60,8 +59,12 @@ spec:
         en_US: Robot Name
         zh_Hans: 机器人名称
         zh_Hant: 機器人名稱
+      description:
+        en_US: Optional. Only removes matching @name text in group messages; bot mentions are identified by BotId.
+        zh_Hans: 可选。仅用于清理群消息中的 @名称 文本，机器人提及通过 BotId 识别。
+        zh_Hant: 選填。僅用於清理群訊息中的 @名稱 文字，機器人提及透過 BotId 識別。
       type: string
-      required: true
+      required: false
       default: ""
     - name: enable-webhook
       label:
@@ -103,6 +106,10 @@ spec:
       type: string
       required: false
       default: ""
+      show_if:
+        field: enable-webhook
+        operator: neq
+        value: true
     - name: Corpid
       label:
         en_US: Corpid
@@ -115,6 +122,10 @@ spec:
       type: string
       required: false
       default: ""
+      show_if:
+        field: enable-webhook
+        operator: eq
+        value: true
     - name: Token
       label:
         en_US: Token
@@ -127,18 +138,26 @@ spec:
       type: string
       required: false
       default: ""
+      show_if:
+        field: enable-webhook
+        operator: eq
+        value: true
     - name: EncodingAESKey
       label:
         en_US: EncodingAESKey
         zh_Hans: 消息加解密密钥 (EncodingAESKey)
         zh_Hant: 訊息加解密密鑰 (EncodingAESKey)
       description:
-        en_US: Required for webhook mode. Optional for WebSocket mode when encrypted files need to be decrypted.
-        zh_Hans: Webhook 模式必填。WebSocket 模式下如需解密文件则填写。
-        zh_Hant: Webhook 模式必填。WebSocket 模式下如需解密檔案則填寫。
+        en_US: Required for webhook mode. Long connections use the decryption key included in each message.
+        zh_Hans: Webhook 模式必填。长连接模式使用消息自带的解密密钥。
+        zh_Hant: Webhook 模式必填。長連線模式使用訊息自帶的解密金鑰。
       type: string
       required: false
       default: ""
+      show_if:
+        field: enable-webhook
+        operator: eq
+        value: true
     - name: enable-stream-reply
       label:
         en_US: Enable Stream Reply
@@ -155,6 +174,8 @@ spec:
   supported_events:
     - message.received
     - feedback.received
+    - wecombot.enter_chat
+    - wecombot.template_card_event
     - platform.specific
 
   supported_apis:
diff --git a/src/langbot/pkg/platform/adapters/wecombot/message_converter.py b/src/langbot/pkg/platform/adapters/wecombot/message_converter.py
index fca3e1f5d..95ce52202 100644
--- a/src/langbot/pkg/platform/adapters/wecombot/message_converter.py
+++ b/src/langbot/pkg/platform/adapters/wecombot/message_converter.py
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import datetime
+import re
 
 from langbot.libs.wecom_ai_bot_api.wecombotevent import WecomBotEvent
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
@@ -46,12 +47,14 @@ class WecomBotMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             platform_message.Source(id=event.message_id, time=datetime.datetime.now()),
         ]
         if event.type == 'group' and event.ai_bot_id:
+            # Group callbacks identify the mentioned bot by ID, independently
+            # of its display name or the optional text-cleanup configuration.
             components.append(platform_message.At(target=event.ai_bot_id))
 
         if event.content:
             content = event.content
-            if bot_name:
-                content = content.replace(f'@{bot_name}', '').strip()
+            if event.type == 'group' and bot_name:
+                content = re.sub(rf'(?<!\S)@{re.escape(bot_name)}(?=\s|$)', '', content).strip()
             if content:
                 components.append(platform_message.Plain(text=content))
 
diff --git a/src/langbot/pkg/platform/adapters/wecombot/stream_text.py b/src/langbot/pkg/platform/adapters/wecombot/stream_text.py
new file mode 100644
index 000000000..90cc46a09
--- /dev/null
+++ b/src/langbot/pkg/platform/adapters/wecombot/stream_text.py
@@ -0,0 +1,31 @@
+"""Render multi-round reasoning as one WeCom thinking section."""
+
+import re
+
+
+def format_stream_text(content: str) -> str:
+    if content.strip().lower() in {'<', '<t', '<th', '<thi', '<thin', '<think'}:
+        return ''
+    if '<think' not in content.lower() and '</think' not in content.lower():
+        return content
+    # Every invocation receives the full snapshot, so split tags may simply
+    # wait for the next snapshot instead of leaking into the visible answer.
+    content = re.sub(r'<(?:/?t(?:h(?:i(?:n(?:k)?)?)?)?|/?)$', '', content, flags=re.IGNORECASE)
+    tokens = re.split(r'(</?think\s*>)', content, flags=re.IGNORECASE)
+    thoughts: list[str] = []
+    answers: list[str] = []
+    thinking = False
+    for token in tokens:
+        if re.fullmatch(r'<think\s*>', token, flags=re.IGNORECASE):
+            thinking = True
+        elif re.fullmatch(r'</think\s*>', token, flags=re.IGNORECASE):
+            thinking = False
+        elif token.strip():
+            (thoughts if thinking else answers).append(token.strip())
+    if not thoughts:
+        return '\n\n'.join(answers)
+    reasoning = '\n\n'.join(thoughts)
+    answer = '\n\n'.join(answers)
+    if thinking and not answer:
+        return f'<think>\n{reasoning}'
+    return f'<think>\n{reasoning}\n</think>' + (f'\n\n{answer}' if answer else '')
diff --git a/src/langbot/pkg/platform/adapters/wecomcs/adapter.py b/src/langbot/pkg/platform/adapters/wecomcs/adapter.py
index 26f132581..f82d1b58a 100644
--- a/src/langbot/pkg/platform/adapters/wecomcs/adapter.py
+++ b/src/langbot/pkg/platform/adapters/wecomcs/adapter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import time
 import traceback
@@ -81,6 +79,12 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
     def get_supported_events(self) -> list[str]:
         return [
             'message.received',
+            'message.deleted',
+            'wecomcs.enter_session',
+            'wecomcs.msg_send_fail',
+            'wecomcs.servicer_status_change',
+            'wecomcs.session_status_change',
+            'wecomcs.reject_customer_msg_switch_change',
             'platform.specific',
         ]
 
@@ -94,7 +98,6 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
             'call_platform_api',
         ]
 
-    @diagnostics.observe('api', 'send_message', source='platform', stage='accepted')
     async def send_message(
         self,
         target_type: str,
@@ -113,7 +116,36 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
             )
         return platform_events.MessageResult(raw={'results': raw_results})
 
-    @diagnostics.observe('api', 'reply_message', source='platform', stage='accepted')
+    async def reply_to_event(
+        self, *, context: dict, target_type: str, target_id: str, message: platform_message.MessageChain
+    ) -> platform_events.MessageResult:
+        """Use event credentials for welcome and session-status replies."""
+        event_type = context.get('event_type')
+        if event_type not in ('wecomcs.enter_session', 'wecomcs.session_status_change'):
+            return await self.send_message(target_type, target_id, message)
+        data = context.get('data') or {}
+        code_key = 'welcome_code' if event_type == 'wecomcs.enter_session' else 'msg_code'
+        code = data.get(code_key) or (data.get('data') or {}).get(code_key)
+        if not code:
+            raise ValueError(
+                f'WeCom CS did not provide {code_key} for this event; an event response is not available. '
+                'Entering a session does not open the ordinary messaging window. '
+                'Ask the customer to send a message before sending an ordinary reply.'
+            )
+        timestamp = data.get('timestamp')
+        if event_type == 'wecomcs.enter_session' and timestamp and time.time() - float(timestamp) >= 20:
+            raise ValueError(
+                'WeCom CS welcome reply expired: welcome_code is valid for only 20 seconds after entry. '
+                'Use a fast, fixed welcome response; do not wait for a long AI run or retry this event.'
+            )
+        if not message.root or any(not isinstance(item, platform_message.Plain) for item in message.root):
+            raise ValueError('WeCom CS event replies currently support a single text message only')
+        text = ''.join(item.text for item in message.root)
+        if not text.strip():
+            raise ValueError('WeCom CS event reply text must not be empty')
+        result = await self.bot.send_event_text(str(code), self._make_outbound_msgid(), text)
+        return platform_events.MessageResult(message_id=result.get('msgid'), raw=result)
+
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -136,7 +168,6 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
             )
         return platform_events.MessageResult(message_id=wecom_event.message_id, raw={'results': raw_results})
 
-    @diagnostics.observe('api', 'call_platform_api', source='platform', stage='accepted')
     async def call_platform_api(self, action: str, params: dict = {}) -> dict:
         handler = PLATFORM_API_MAP.get(action)
         if handler is None:
@@ -188,10 +219,9 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
         async def on_message(event: WecomCSEvent):
             await self._handle_native_event(event)
 
-        for msg_type in ('text', 'image', 'file', 'voice'):
+        for msg_type in ('text', 'image', 'file', 'voice', 'event'):
             self.bot.on_message(msg_type)(on_message)
 
-    @diagnostics.observe('event', 'platform.native_receive', source='platform', stage='convert')
     async def _handle_native_event(self, event: WecomCSEvent):
         self.bot_account_id = event.receiver_id or self.bot_account_id
         try:
@@ -214,7 +244,6 @@ class WecomCSAdapter(WecomCSAPIMixin, abstract_platform_adapter.AbstractPlatform
             await self.logger.error(f'Error in wecomcs native event: {traceback.format_exc()}')
 
     async def _dispatch_eba_event(self, event: platform_events.EBAEvent):
-        diagnostics.adapter_event_received(self, event)
         for event_type in (type(event), platform_events.EBAEvent, platform_events.Event):
             callback = self.listeners.get(event_type)
             if callback:
diff --git a/src/langbot/pkg/platform/adapters/wecomcs/api_impl.py b/src/langbot/pkg/platform/adapters/wecomcs/api_impl.py
index d14565b2a..3fca63275 100644
--- a/src/langbot/pkg/platform/adapters/wecomcs/api_impl.py
+++ b/src/langbot/pkg/platform/adapters/wecomcs/api_impl.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.wecom_customer_service_api.api import WecomCSClient
@@ -16,7 +14,6 @@ class WecomCSAPIMixin:
     _message_cache: dict[str, platform_events.MessageReceivedEvent]
     _user_cache: dict[str, platform_entities.User]
 
-    @diagnostics.observe('api', 'get_message', source='platform', stage='accepted')
     async def get_message(
         self,
         chat_type: str,
@@ -28,7 +25,6 @@ class WecomCSAPIMixin:
             raise NotSupportedError('get_message:message_not_cached')
         return event
 
-    @diagnostics.observe('api', 'get_user_info', source='platform', stage='accepted')
     async def get_user_info(self, user_id: typing.Union[int, str]) -> platform_entities.User:
         cached = self._user_cache.get(str(user_id))
         if cached is not None:
@@ -43,30 +39,24 @@ class WecomCSAPIMixin:
             username=info.get('external_userid') or None,
         )
 
-    @diagnostics.observe('api', 'get_friend_list', source='platform', stage='accepted')
     async def get_friend_list(self) -> list[platform_entities.User]:
         return list(self._user_cache.values())
 
-    @diagnostics.observe('api', 'upload_file', source='platform', stage='accepted')
     async def upload_file(self, file_data: bytes, filename: str) -> str:
         raise NotSupportedError('upload_file')
 
-    @diagnostics.observe('api', 'get_file_url', source='platform', stage='accepted')
     async def get_file_url(self, file_id: str) -> str:
         raise NotSupportedError('get_file_url')
 
-    @diagnostics.observe('api', 'get_group_info', source='platform', stage='accepted')
     async def get_group_info(self, group_id: typing.Union[int, str]) -> platform_entities.UserGroup:
         raise NotSupportedError('get_group_info')
 
-    @diagnostics.observe('api', 'get_group_member_list', source='platform', stage='accepted')
     async def get_group_member_list(
         self,
         group_id: typing.Union[int, str],
     ) -> list[platform_entities.UserGroupMember]:
         raise NotSupportedError('get_group_member_list')
 
-    @diagnostics.observe('api', 'get_group_member_info', source='platform', stage='accepted')
     async def get_group_member_info(
         self,
         group_id: typing.Union[int, str],
@@ -74,7 +64,6 @@ class WecomCSAPIMixin:
     ) -> platform_entities.UserGroupMember:
         raise NotSupportedError('get_group_member_info')
 
-    @diagnostics.observe('api', 'edit_message', source='platform', stage='accepted')
     async def edit_message(
         self,
         chat_type: str,
@@ -84,7 +73,6 @@ class WecomCSAPIMixin:
     ) -> None:
         raise NotSupportedError('edit_message')
 
-    @diagnostics.observe('api', 'delete_message', source='platform', stage='accepted')
     async def delete_message(
         self,
         chat_type: str,
diff --git a/src/langbot/pkg/platform/adapters/wecomcs/event_converter.py b/src/langbot/pkg/platform/adapters/wecomcs/event_converter.py
index 64201513b..4c3d76eb2 100644
--- a/src/langbot/pkg/platform/adapters/wecomcs/event_converter.py
+++ b/src/langbot/pkg/platform/adapters/wecomcs/event_converter.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 
 from langbot.libs.wecom_customer_service_api.api import WecomCSClient
@@ -19,7 +17,6 @@ class WecomCSEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return getattr(event, 'source_platform_object', None)
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2legacy', source='platform', stage='convert')
     async def target2legacy(
         event: WecomCSEvent, bot: WecomCSClient | None = None
     ) -> platform_events.FriendMessage | None:
@@ -29,10 +26,44 @@ class WecomCSEventConverter(abstract_platform_adapter.AbstractEventConverter):
         return None
 
     @staticmethod
-    @diagnostics.observe('event', 'platform.target2yiri', source='platform', stage='convert')
     async def target2yiri(event: WecomCSEvent, bot: WecomCSClient | None = None) -> platform_events.Event | None:
         if event.type in {'text', 'image', 'file', 'voice'}:
             return await WecomCSEventConverter.message_to_eba(event, bot)
+        if event.type == 'event':
+            data = event.get('event') or {}
+            event_type = data.get('event_type', 'unknown')
+            customer_id = data.get('external_userid') or ''
+            common = {
+                'adapter_name': ADAPTER_NAME,
+                'timestamp': float(event.timestamp or 0),
+                'source_platform_object': event,
+                'chat_id': make_private_chat_id(customer_id, data.get('open_kfid')) if customer_id else '',
+            }
+            if event_type in {'user_recall_msg', 'servicer_recall_msg'}:
+                operator_id = data.get('servicer_userid') if event_type == 'servicer_recall_msg' else customer_id
+                return platform_events.MessageDeletedEvent(
+                    message_id=data.get('recall_msgid') or '',
+                    operator=platform_entities.User(id=operator_id) if operator_id else None,
+                    **common,
+                )
+            event_classes = {
+                'enter_session': platform_events.WecomCSEnterSessionEvent,
+                'msg_send_fail': platform_events.WecomCSMessageSendFailedEvent,
+                'servicer_status_change': platform_events.WecomCSServicerStatusChangedEvent,
+                'session_status_change': platform_events.WecomCSSessionStatusChangedEvent,
+                'reject_customer_msg_switch_change': platform_events.WecomCSRejectCustomerMessageChangedEvent,
+            }
+            event_class = event_classes.get(event_type)
+            if event_class:
+                fields = {
+                    key: value for key, value in data.items() if key in event_class.model_fields and value is not None
+                }
+                fields.update(common)
+                fields['data'] = dict(data)
+                actor_id = data.get('servicer_userid') or customer_id
+                fields['user'] = platform_entities.User(id=actor_id) if actor_id else None
+                return event_class(**fields)
+            return WecomCSEventConverter.platform_specific(event, f'wecomcs.{event_type}')
         return WecomCSEventConverter.platform_specific(event, f'wecomcs.{event.type or "unknown"}')
 
     @staticmethod
diff --git a/src/langbot/pkg/platform/adapters/wecomcs/manifest.yaml b/src/langbot/pkg/platform/adapters/wecomcs/manifest.yaml
index d1f04c7cc..3c8821675 100644
--- a/src/langbot/pkg/platform/adapters/wecomcs/manifest.yaml
+++ b/src/langbot/pkg/platform/adapters/wecomcs/manifest.yaml
@@ -8,9 +8,9 @@ metadata:
     zh_Hans: 企业微信客服
     zh_Hant: 企業微信客服
   description:
-    en_US: WeCom Customer Service adapter with event-driven orchestration support
-    zh_Hans: 企业微信客服适配器，支持事件驱动编排，通过统一 Webhook 接收客服会话消息
-    zh_Hant: 企業微信客服適配器，支援事件驅動編排，透過統一 Webhook 接收客服會話訊息
+    en_US: Support customers on WeChat through WeChat Customer Service
+    zh_Hans: 通过微信客服，为微信用户提供咨询与售后服务
+    zh_Hant: 透過微信客服，為微信使用者提供諮詢與售後服務
   icon: wecom.png
 
 spec:
@@ -93,6 +93,12 @@ spec:
 
   supported_events:
     - message.received
+    - message.deleted
+    - wecomcs.enter_session
+    - wecomcs.msg_send_fail
+    - wecomcs.servicer_status_change
+    - wecomcs.session_status_change
+    - wecomcs.reject_customer_msg_switch_change
     - platform.specific
 
   supported_apis:
diff --git a/src/langbot/pkg/platform/botmgr.py b/src/langbot/pkg/platform/botmgr.py
index e1c819fda..bb3beb0e8 100644
--- a/src/langbot/pkg/platform/botmgr.py
+++ b/src/langbot/pkg/platform/botmgr.py
@@ -1,11 +1,15 @@
 from __future__ import annotations
 
-from ..telemetry import diagnostics
+from ..cloud.plan_policy import require_workspace_platform
+
+from ..i18n import language_scope
+from ..i18n.workspace import get_workspace_language
 
 import asyncio
 import contextlib
 import dataclasses
 import functools
+import hashlib
 import json
 import re
 import time
@@ -87,6 +91,7 @@ class RuntimeBot:
     ):
         if not isinstance(execution_context, ExecutionContext):
             raise WorkspaceRequiredError('RuntimeBot requires an ExecutionContext')
+        self._ingress_dedup_lock = asyncio.Lock()
         if not execution_context.instance_uuid.strip() or not execution_context.workspace_uuid.strip():
             raise WorkspaceRequiredError('RuntimeBot requires an instance and Workspace')
         if execution_context.placement_generation <= 0:
@@ -114,6 +119,7 @@ class RuntimeBot:
     async def assert_execution_active(self) -> None:
         """Fail closed when this long-lived adapter belongs to a stale placement."""
 
+        await require_workspace_platform(self.ap, self.workspace_uuid, self.bot_entity.adapter)
         await self.ap.workspace_service.get_execution_binding(
             self.workspace_uuid,
             expected_generation=self.placement_generation,
@@ -160,7 +166,7 @@ class RuntimeBot:
 
     @classmethod
     def _is_message_event_type(cls, event_type: str) -> bool:
-        return cls._match_event_pattern(event_type, 'message.*')
+        return event_type == 'message.received'
 
     @classmethod
     def _agent_supports_event_type(
@@ -312,6 +318,12 @@ class RuntimeBot:
                 step['reason'] = 'Event type does not match binding event_pattern'
                 diagnostic_steps.append(step)
                 continue
+            if binding.get('target_type') == 'pipeline' and not cls._is_message_event_type(event_type):
+                step['failure_code'] = 'processor_incompatible'
+                step['reason'] = 'Pipeline targets only support message.received'
+                diagnostic_steps.append(step)
+                continue
+
             if not cls._match_event_filters(event, binding.get('filters')):
                 step['failure_code'] = 'filters_mismatch'
                 step['reason'] = 'Event data does not satisfy binding filters'
@@ -354,6 +366,20 @@ class RuntimeBot:
         """Return the selected event binding plus per-binding diagnostic steps."""
         return self._evaluate_eba_event_bindings(self._get_event_bindings(), event, event_type)
 
+    @staticmethod
+    def _route_ref(
+        binding: dict | None,
+        target_type: str | None = None,
+        target_uuid: str | None = None,
+    ) -> str:
+        """Code-defined route identity for telemetry; never a user-facing name."""
+        binding = binding or {}
+        kind = str(target_type or binding.get('target_type') or '').strip()
+        target = str(target_uuid or binding.get('target_uuid') or '').strip()
+        if not kind or not target:
+            return ''
+        return f'{kind}:{target}'[:160]
+
     async def _record_event_route_trace(
         self,
         *,
@@ -370,22 +396,6 @@ class RuntimeBot:
     ) -> dict[str, typing.Any]:
         """Record structured event routing state while preserving the human log."""
         binding = binding or {}
-        diagnostics.event(
-            self,
-            'route',
-            'route.primary',
-            {
-                'not_matched': 'skipped',
-                'discarded': 'skipped',
-                'matched': 'started',
-                'delivered': 'succeeded',
-                'failed': 'failed',
-            }.get(status, 'unknown'),
-            stage='dispatch',
-            platform_event_type=event_type,
-            processor_type=target_type or binding.get('target_type', ''),
-            reason_code=failure_code or status,
-        )
         metadata = {
             'kind': 'event_route_trace',
             'event_type': event_type,
@@ -398,22 +408,37 @@ class RuntimeBot:
             'reason': reason or text,
             'run_id': run_id,
         }
-        diagnostics.set_outcome(
-            {
-                'not_matched': 'skipped',
-                'discarded': 'skipped',
-                'matched': 'started',
-                'delivered': 'succeeded',
-                'failed': 'failed',
-            }.get(status, 'unknown'),
-            reason_code=failure_code or status,
-        )
         log_method = getattr(self.logger, level, self.logger.info)
         await log_method(text, metadata=metadata)
+        if status in {'delivered', 'failed'}:
+            from ..telemetry import trace as trace_mod
+            from ..telemetry.execution import record
+
+            state = trace_mod.current()
+            if status == 'failed' and state is not None and state.reportable and not trace_mod.reporting_suppressed():
+                # The root event can finish successfully after a delivery failed.
+                # Preserve that failure for closure reporting in failures mode.
+                state.mark_failure(failure_code or 'route_failed')
+            with trace_mod.scope(route_ref=self._route_ref(binding, target_type, target_uuid)):
+                # The routing decision is its own workflow step; the Pipeline or
+                # processor it delivers to hangs under this node.
+                with trace_mod.stage_scope() as node:
+                    record(
+                        getattr(self, 'ap', None),
+                        getattr(self, 'execution_context', None),
+                        family='event_route',
+                        operation=event_type,
+                        adapter=type(getattr(self, 'adapter', None)).__name__,
+                        mode=target_type if target_type in {'pipeline', 'agent', 'event_processor'} else 'none',
+                        outcome={'delivered': 'success', 'failed': 'failed'}.get(status, 'skipped'),
+                        node=node,
+                    )
         return metadata
 
     def get_pipeline_target_for_event_type(self, event_type: str = 'message.received') -> str | None:
         """Return the first Pipeline target configured for an event type."""
+        if not self._is_message_event_type(event_type):
+            return None
         matched: list[tuple[int, int, str]] = []
         for index, binding in enumerate(self._get_event_bindings()):
             if not binding.get('enabled', True):
@@ -441,10 +466,13 @@ class RuntimeBot:
             return {}
         if hasattr(model, 'model_dump'):
             try:
-                return model.model_dump(mode='json')
+                # Native adapter objects are for replying, not event logging.
+                # Exclude them before JSON serialization, since some (OAEvent,
+                # for example) cannot be serialized by Pydantic.
+                return model.model_dump(mode='json', exclude={'source_platform_object'})
             except TypeError:
                 try:
-                    return model.model_dump()
+                    return model.model_dump(exclude={'source_platform_object'})
                 except Exception:
                     return {}
             except Exception:
@@ -695,33 +723,20 @@ class RuntimeBot:
 
     @classmethod
     def _build_agent_input(cls, event: platform_events.EBAEvent) -> AgentInput:
-        text = None
+        text = platform_events.event_summary(event)
         contents: list[dict[str, typing.Any]] = []
 
         message_chain = getattr(event, 'message_chain', None)
         if message_chain:
-            text_parts: list[str] = []
             try:
                 for component in message_chain:
-                    if isinstance(component, platform_message.Plain):
-                        text_parts.append(component.text)
-                    elif isinstance(component, platform_message.Image):
+                    if isinstance(component, platform_message.Image):
                         if component.url:
                             contents.append({'type': 'image_url', 'image_url': {'url': component.url}})
                         elif component.base64:
                             contents.append({'type': 'image_base64', 'image_base64': component.base64})
             except TypeError:
-                text_parts.append(str(message_chain))
-            text = ''.join(text_parts) or str(message_chain)
-
-        if text is None:
-            feedback_content = getattr(event, 'feedback_content', None)
-            if feedback_content:
-                text = str(feedback_content)
-            elif getattr(event, 'action', None):
-                text = str(getattr(event, 'action'))
-            else:
-                text = str(getattr(event, 'type', 'event'))
+                pass
 
         if text:
             contents.insert(0, {'type': 'text', 'text': text})
@@ -736,12 +751,19 @@ class RuntimeBot:
         self,
         event: platform_events.EBAEvent,
         adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
+        execution_id: str | None = None,
     ) -> AgentEventEnvelope:
         event_type = getattr(event, 'type', None) or event.__class__.__name__
         event_time = getattr(event, 'timestamp', None) or time.time()
-        event_id = (
-            getattr(event, 'message_id', None) or getattr(event, 'feedback_id', None) or f'{event_type}:{uuid.uuid4()}'
-        )
+        if execution_id:
+            # Reuse the ingress execution identity so the run view, the journaled
+            # event and the uploaded chain all share one id.
+            envelope_event_id = str(execution_id)
+            prefix = f'platform:{self.bot_entity.uuid}:'
+            event_id = envelope_event_id[len(prefix) :] if envelope_event_id.startswith(prefix) else envelope_event_id
+        else:
+            event_id = self._platform_event_raw_id(event)
+            envelope_event_id = f'platform:{self.bot_entity.uuid}:{event_id}'
         target_type, target_id, target_metadata = self._infer_reply_target(event)
         supported_apis = self._get_adapter_supported_apis(adapter)
 
@@ -773,7 +795,7 @@ class RuntimeBot:
             delivery_data['interactions'] = interaction_capabilities
 
         return AgentEventEnvelope(
-            event_id=f'platform:{self.bot_entity.uuid}:{event_id}',
+            event_id=envelope_event_id,
             event_type=event_type,
             event_time=int(event_time) if isinstance(event_time, (int, float)) else None,
             source='platform',
@@ -863,38 +885,189 @@ class RuntimeBot:
             processor_id=agent.get('uuid'),
         )
 
-    @diagnostics.observe('event', 'platform.receive', source='platform', stage='dispatch')
+    def _platform_event_execution_id(self, event: platform_events.EBAEvent) -> str:
+        """Space-visible execution identity for one inbound platform event.
+
+        Matches the ``AgentEventEnvelope.event_id`` the same event produces, so
+        the stored chain and the Space run view share one identity.
+        """
+        return f'platform:{self.bot_entity.uuid}:{self._platform_event_raw_id(event)}'
+
+    @staticmethod
+    def _platform_event_raw_id(event: platform_events.EBAEvent) -> str:
+        event_type = getattr(event, 'type', None) or event.__class__.__name__
+        if isinstance(event, platform_events.MessageDeletedEvent) and event.message_id:
+            # A recall is distinct from the original message.
+            return f'{event_type}:{event.message_id}'
+        if getattr(event, 'adapter_name', None) == 'wecomcs-omni':
+            source = getattr(event, 'source_platform_object', None)
+            if isinstance(source, dict):
+                if source.get('msgid'):
+                    return str(source['msgid'])
+                # Native service events can lack msgid. Preserve their identity
+                # across sync retries and process restarts instead of using UUIDs.
+                payload = json.dumps(source, sort_keys=True, ensure_ascii=False, separators=(',', ':'))
+                return f'{event_type}:{hashlib.sha256(payload.encode()).hexdigest()}'
+        return str(
+            getattr(event, 'message_id', None) or getattr(event, 'feedback_id', None) or f'{event_type}:{uuid.uuid4()}'
+        )
+
     async def _handle_platform_event(
         self,
         event: platform_events.EBAEvent,
         adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
     ) -> None:
-        event.bot_uuid = self.bot_entity.uuid
-        await self._record_adapter_event(event, adapter)
+        # One inbound event owns one execution trace; every stage recorded while
+        # it is handled (routing, runner, platform API calls) joins that trace.
+        from ..telemetry import trace as trace_mod
+        from ..telemetry.execution import ingress
 
-        primary = (
-            self._handle_interaction_submission(event, adapter)
-            if isinstance(event, platform_events.PlatformSpecificEvent) and event.action == 'interaction.submitted'
-            else self._dispatch_eba_event_to_processor(event, adapter)
-        )
-        subscriptions = self._get_event_bindings_from_value(getattr(self.bot_entity, 'plugin_processors', []))
-        seen = set()
-        tasks = [primary]
-        for subscription in subscriptions:
-            processor_uuid = subscription.get('processor_uuid')
-            if not processor_uuid or processor_uuid in seen or not subscription.get('enabled', True):
-                continue
-            seen.add(processor_uuid)
-            tasks.append(self._dispatch_plugin_subscription(event, adapter, processor_uuid))
-        # Start all deliveries together. A slow or failed subscriber cannot block
-        # another subscriber or the primary route from receiving the event.
-        results = await asyncio.gather(*tasks, return_exceptions=True)
+        await self.assert_execution_active()
+        event.bot_uuid = self.bot_entity.uuid
+        execution_id = self._platform_event_execution_id(event)
+        if getattr(event, 'adapter_name', None) == 'wecomcs-omni':
+            from ..agent.runner.event_log_store import EventLogStore
+
+            # The sync API can replay history after an adapter restart. Reuse
+            # the durable ingress journal, including records from earlier runs.
+            async with self._ingress_dedup_lock:
+                store = EventLogStore(self.ap.persistence_mgr.get_db_engine())
+                if await store.get_event(execution_id):
+                    return
+                await self._persist_monitoring_ingress(event, adapter, execution_id)
+        with ingress(
+            getattr(self, 'ap', None),
+            'event_done',
+            getattr(self, 'execution_context', None),
+            execution_id,
+            require_route=True,
+        ):
+            # The inbound event is the root workflow step: everything recorded
+            # while handling it hangs under this node.
+            with trace_mod.stage_scope() as node:
+                await self._handle_platform_event_body(event, adapter, execution_id, node)
+
+    @staticmethod
+    def _summarize_event_delivery(results: list[typing.Any]) -> tuple[str, str]:
+        outcome = 'success'
+        error_detail = ''
         for result in results:
             if isinstance(result, BaseException):
-                await self.logger.error(f'Event delivery failed: {result}')
+                outcome = 'failed'
+                error_detail = error_detail or f'{type(result).__name__}: {result}'
+        return outcome, error_detail
 
-    @diagnostics.observe('route', 'route.subscription', source='platform', stage='dispatch')
-    async def _dispatch_plugin_subscription(self, event, adapter, processor_uuid):
+    def _record_platform_event(self, event, adapter, outcome: str, error_detail: str, node: str = '') -> None:
+        from ..telemetry.execution import record
+
+        record(
+            getattr(self, 'ap', None),
+            getattr(self, 'execution_context', None),
+            family='platform_event',
+            operation=str(getattr(event, 'type', None) or event.__class__.__name__),
+            adapter=adapter.__class__.__name__,
+            outcome=outcome,
+            error=error_detail,
+            node=node,
+        )
+
+    async def _persist_monitoring_ingress(self, event, adapter, execution_id, status='running', routes=None):
+        service = getattr(getattr(self, 'ap', None), 'monitoring_service', None)
+        if service is None or not execution_id:
+            return
+        try:
+            sender = getattr(event, 'sender', None) or getattr(event, 'user', None) or getattr(event, 'operator', None)
+            target_type, target_id, _ = self._infer_reply_target(event)
+            await service.record_ingress_event(
+                self.execution_context,
+                event_id=execution_id,
+                event_type=getattr(event, 'type', None) or event.__class__.__name__,
+                bot_id=self.bot_entity.uuid,
+                bot_name=self.bot_entity.name,
+                platform=getattr(self.bot_entity, 'adapter', None) or adapter.__class__.__name__,
+                payload={
+                    key: value
+                    for key, value in self._safe_model_dump(event).items()
+                    if key != 'source_platform_object' and not key.startswith('_')
+                },
+                actor_id=self._get_entity_id(sender),
+                actor_name=self._get_entity_name(sender),
+                conversation_id=f'{target_type}_{target_id}' if target_type and target_id else None,
+                status=status,
+                routes=routes,
+            )
+        except Exception as exc:
+            await self.logger.warning(f'Failed to persist monitoring ingress: {exc}')
+
+    async def _handle_platform_event_body(
+        self,
+        event: platform_events.EBAEvent,
+        adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
+        execution_id: str | None = None,
+        node: str = '',
+    ) -> None:
+        results: list[typing.Any] = []
+        try:
+            await self._record_adapter_event(event, adapter)
+            await self._persist_monitoring_ingress(event, adapter, execution_id)
+
+            primary_handler = (
+                self._handle_interaction_submission
+                if isinstance(event, platform_events.PlatformSpecificEvent) and event.action == 'interaction.submitted'
+                else self._dispatch_eba_event_to_processor
+            )
+            subscriptions = self._get_event_bindings_from_value(getattr(self.bot_entity, 'plugin_processors', []))
+            seen = set()
+            tasks = [self._run_event_delivery(primary_handler, event, adapter, execution_id=execution_id)]
+            for subscription in subscriptions:
+                processor_uuid = subscription.get('processor_uuid')
+                if not processor_uuid or processor_uuid in seen or not subscription.get('enabled', True):
+                    continue
+                seen.add(processor_uuid)
+                tasks.append(
+                    self._run_event_delivery(
+                        self._dispatch_plugin_subscription, event, adapter, processor_uuid, execution_id
+                    )
+                )
+            # Start all deliveries together. A slow or failed subscriber cannot block
+            # another subscriber or the primary route from receiving the event.
+            results = await asyncio.gather(*tasks, return_exceptions=True)
+        except BaseException as exc:
+            self._record_platform_event(event, adapter, 'failed', f'{type(exc).__name__}: {exc}', node)
+            await self._persist_monitoring_ingress(event, adapter, execution_id, 'failed', [exc])
+            raise
+
+        # Recorded after handling so the inbound stage carries the real outcome
+        # instead of a hardcoded success.
+        outcome, error_detail = self._summarize_event_delivery(results)
+        for result in results:
+            if isinstance(result, BaseException):
+                try:
+                    await self.logger.error(f'Event delivery failed: {result}')
+                except Exception:
+                    pass
+        self._record_platform_event(event, adapter, outcome, error_detail, node)
+        await self._persist_monitoring_ingress(
+            event,
+            adapter,
+            execution_id,
+            {'success': 'completed', 'skipped': 'ignored'}.get(outcome, 'failed'),
+            results,
+        )
+
+    async def _run_event_delivery(self, handler, *args, **kwargs):
+        # gather creates child tasks. Each delivery needs its own database scope;
+        # copying the listener's ContextVar does not transfer task ownership.
+        persistence = getattr(getattr(self, 'ap', None), 'persistence_mgr', None)
+        if getattr(getattr(persistence, 'mode', None), 'value', None) == 'cloud_runtime':
+            tenant_scope = getattr(persistence, 'tenant_scope', None)
+            if not callable(tenant_scope):
+                raise RuntimeError('Cloud event deliveries require a tenant scope')
+            async with tenant_scope(self.workspace_uuid):
+                return await handler(*args, **kwargs)
+        return await handler(*args, **kwargs)
+
+    async def _dispatch_plugin_subscription(self, event, adapter, processor_uuid, execution_id=None):
         event_type = event.type
         event_binding = {
             'id': f'plugin_processor:{processor_uuid}',
@@ -912,30 +1085,36 @@ class RuntimeBot:
             # Resolve the installed declaration each time, including after plugin updates.
             patterns = descriptor.supported_event_patterns
             if not patterns or not self._agent_supports_event_type(patterns, event_type):
-                diagnostics.set_outcome('skipped', reason_code='not_matched')
                 return
-            agent = {**agent, 'supported_event_patterns': patterns}
-            return await self._dispatch_eba_event_to_processor(event, adapter, event_binding, agent)
         except Exception as exc:
-            return await self._record_event_route_trace(
-                event_type=event_type,
-                status='failed',
-                level='error',
-                binding=event_binding,
-                target_type='event_processor',
-                target_uuid=processor_uuid,
-                failure_code='runner_failed',
-                reason=str(exc),
-                text=f'Plugin processor {processor_uuid} failed: {exc}',
-            )
+            from ..telemetry import trace as trace_mod
+
+            # No matching declaration was resolved: log locally without
+            # reporting a failed delivery for every unrelated incoming event.
+            with trace_mod.suppress_reporting():
+                return await self._record_event_route_trace(
+                    event_type=event_type,
+                    status='failed',
+                    level='error',
+                    binding=event_binding,
+                    target_type='event_processor',
+                    target_uuid=processor_uuid,
+                    failure_code='runner_failed',
+                    reason=str(exc),
+                    text=f'Plugin processor {processor_uuid} failed: {exc}',
+                )
+        agent = {**agent, 'supported_event_patterns': patterns}
+        return await self._dispatch_eba_event_to_processor(
+            event, adapter, event_binding, agent, execution_id=execution_id
+        )
 
-    @diagnostics.observe('route', 'route.dispatch', source='platform', stage='dispatch')
     async def _dispatch_eba_event_to_processor(
         self,
         event: platform_events.EBAEvent,
         adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
         event_binding: dict | None = None,
         agent: dict | None = None,
+        execution_id: str | None = None,
     ) -> dict[str, typing.Any]:
         event_type = getattr(event, 'type', None) or event.__class__.__name__
 
@@ -951,6 +1130,12 @@ class RuntimeBot:
             )
 
         target_type = event_binding.get('target_type')
+        from ..telemetry import trace as trace_mod
+
+        if target_type != 'discard':
+            # Matching precedes target lookup/execution, so genuine delivery
+            # failures remain observable even when the target is unavailable.
+            trace_mod.admit()
         await self._record_event_route_trace(
             event_type=event_type,
             status='matched',
@@ -961,12 +1146,15 @@ class RuntimeBot:
         )
         if target_type == 'discard':
             if isinstance(event, platform_events.MessageReceivedEvent):
-                await self._dispatch_eba_message_to_pipeline(
-                    event,
-                    adapter,
-                    pipeline_uuid=self.PIPELINE_DISCARD,
-                    routed_by_event_binding=True,
-                )
+                from ..telemetry import trace as trace_mod
+
+                with trace_mod.suppress_reporting(), trace_mod.scope(route_ref=self._route_ref(event_binding)):
+                    await self._dispatch_eba_message_to_pipeline(
+                        event,
+                        adapter,
+                        pipeline_uuid=self.PIPELINE_DISCARD,
+                        routed_by_event_binding=True,
+                    )
                 return await self._record_event_route_trace(
                     event_type=event_type,
                     status='discarded',
@@ -994,12 +1182,17 @@ class RuntimeBot:
                     reason='Pipeline targets only support message events',
                     text=f'Event {event_type} ignored Pipeline target for non-message event',
                 )
-            await self._dispatch_eba_message_to_pipeline(
-                event,
-                adapter,
-                pipeline_uuid=event_binding.get('target_uuid'),
-                routed_by_event_binding=True,
-            )
+            from ..telemetry import trace as trace_mod
+
+            with trace_mod.scope(
+                route_ref=self._route_ref(event_binding, target_type, event_binding.get('target_uuid'))
+            ):
+                await self._dispatch_eba_message_to_pipeline(
+                    event,
+                    adapter,
+                    pipeline_uuid=event_binding.get('target_uuid'),
+                    routed_by_event_binding=True,
+                )
             return await self._record_event_route_trace(
                 event_type=event_type,
                 status='delivered',
@@ -1075,21 +1268,34 @@ class RuntimeBot:
                 text=f'Event {event_type} target agent has no runner: {target_uuid}',
             )
 
-        envelope = self._eba_event_to_agent_envelope(event, adapter)
+        envelope = self._eba_event_to_agent_envelope(event, adapter, execution_id)
         if target_type == 'event_processor':
             envelope.data = event.model_dump(mode='json', exclude={'source_platform_object', 'legacy_event'})
+            if isinstance(event, platform_events.MessageReceivedEvent):
+                # Use the SDK wire format for Source timestamps and nested chains.
+                # Pydantic's enclosing event serializer bypasses model_dump overrides.
+                envelope.data['message_chain'] = event.message_chain.model_dump()
+        from ..telemetry import trace as trace_mod
+
         try:
-            async for _ in self.ap.agent_run_orchestrator.run(
-                envelope,
-                binding,
-                adapter_context={
-                    '_delivery_adapter': adapter,
-                    '_platform_event': event,
-                    '_execution_context': self.execution_context,
-                },
-            ):
-                # Results are journaled by the orchestrator; platform sends require explicit actions.
-                pass
+            with trace_mod.scope(route_ref=self._route_ref(event_binding, target_type, target_uuid)):
+                from .processing_indicator import processing_indicator
+
+                reply_type, reply_id = (None, None)
+                if isinstance(event, platform_events.MessageReceivedEvent):
+                    reply_type, reply_id, _ = self._infer_reply_target(event)
+                async with processing_indicator(adapter, reply_type, reply_id):
+                    async for _ in self.ap.agent_run_orchestrator.run(
+                        envelope,
+                        binding,
+                        adapter_context={
+                            '_delivery_adapter': adapter,
+                            '_platform_event': event,
+                            '_execution_context': self.execution_context,
+                        },
+                    ):
+                        # Results are journaled by the orchestrator; platform sends require explicit actions.
+                        pass
         except Exception:
             return await self._record_event_route_trace(
                 event_type=event_type,
@@ -1200,7 +1406,6 @@ class RuntimeBot:
         except Exception as e:
             await self.logger.error(f'Failed to record discarded message: {e}')
 
-    @diagnostics.observe('event', 'platform.legacy_receive', source='platform', stage='dispatch')
     async def _handle_legacy_message_event(
         self,
         event: platform_events.FriendMessage | platform_events.GroupMessage,
@@ -1286,11 +1491,11 @@ class RuntimeBot:
             execution_context=self.execution_context,
         )
 
-    @diagnostics.observe('interaction', 'interaction.submit', source='platform', stage='dispatch')
     async def _handle_interaction_submission(
         self,
         event: platform_events.PlatformSpecificEvent,
         adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
+        execution_id: str | None = None,
     ) -> None:
         """Consume a platform interaction callback and resume its original processor."""
         data = event.data if isinstance(event.data, dict) else {}
@@ -1316,13 +1521,18 @@ class RuntimeBot:
             conversation_id=conversation_id,
             actor_id=actor_id,
         )
+        if record['processor_type'] not in {'agent', 'pipeline'}:
+            raise ValueError(f'Unsupported interaction processor type: {record["processor_type"]}')
+        from ..telemetry import trace as trace_mod
+
+        # Only a validated callback resolving an original processor is eligible.
+        # Admit before acknowledgement and target lookup to preserve failures.
+        trace_mod.admit()
         await self.ap.agent_run_orchestrator.interaction_manager.acknowledge_submission(record, adapter)
 
         if record['processor_type'] == 'agent':
-            await self._resume_agent_interaction(record, event, adapter, actor_id)
+            await self._resume_agent_interaction(record, event, adapter, actor_id, execution_id)
             return
-        if record['processor_type'] != 'pipeline':
-            raise ValueError(f'Unsupported interaction processor type: {record["processor_type"]}')
 
         pipeline = await self.ap.pipeline_service.get_pipeline(
             self.execution_context,
@@ -1395,13 +1605,13 @@ class RuntimeBot:
             execution_context=self.execution_context,
         )
 
-    @diagnostics.observe('interaction', 'interaction.resume', source='platform', stage='resume')
     async def _resume_agent_interaction(
         self,
         record: dict[str, typing.Any],
         event: platform_events.PlatformSpecificEvent,
         adapter: abstract_platform_adapter.AbstractMessagePlatformAdapter,
         actor_id: str | None,
+        execution_id: str | None = None,
     ) -> None:
         """Resume the exact independent Agent that produced an interaction."""
         agent = await self.ap.agent_service.get_agent(
@@ -1454,7 +1664,7 @@ class RuntimeBot:
             delivery_data['interactions'] = interaction_capabilities
 
         envelope = AgentEventEnvelope(
-            event_id=f'interaction:{record["id"]}:{uuid.uuid4()}',
+            event_id=execution_id or f'interaction:{record["id"]}:{uuid.uuid4()}',
             event_type='interaction.submitted',
             event_time=int(event.timestamp) if event.timestamp else int(time.time()),
             source='platform',
@@ -1503,24 +1713,29 @@ class RuntimeBot:
         )
 
     async def initialize(self):
+        from ..telemetry.platform import observe_adapter
+
+        observe_adapter(self.ap, self.execution_context, self.adapter)
+
         def tenant_scoped_listener(listener):
             @functools.wraps(listener)
             async def wrapped(*args, **kwargs):
-                tenant_scope = getattr(self.ap.persistence_mgr, 'tenant_scope', None)
-                cloud_runtime = (
-                    getattr(
-                        getattr(self.ap.persistence_mgr, 'mode', None),
-                        'value',
-                        None,
+                with language_scope(await get_workspace_language(self.ap, self.workspace_uuid)):
+                    tenant_scope = getattr(self.ap.persistence_mgr, 'tenant_scope', None)
+                    cloud_runtime = (
+                        getattr(
+                            getattr(self.ap.persistence_mgr, 'mode', None),
+                            'value',
+                            None,
+                        )
+                        == 'cloud_runtime'
                     )
-                    == 'cloud_runtime'
-                )
-                if cloud_runtime:
-                    if not callable(tenant_scope):
-                        raise RuntimeError('Cloud platform callbacks require a tenant scope')
-                    async with tenant_scope(self.workspace_uuid):
-                        return await listener(*args, **kwargs)
-                return await listener(*args, **kwargs)
+                    if cloud_runtime:
+                        if not callable(tenant_scope):
+                            raise RuntimeError('Cloud platform callbacks require a tenant scope')
+                        async with tenant_scope(self.workspace_uuid):
+                            return await listener(*args, **kwargs)
+                    return await listener(*args, **kwargs)
 
             return wrapped
 
@@ -1550,8 +1765,6 @@ class RuntimeBot:
                 return
             await self._handle_platform_event(self._legacy_message_to_eba_event(event, adapter), adapter)
 
-        from ..telemetry.diagnostic_catalog import snapshot_bot
-
         get_supported_events = getattr(self.adapter, 'get_supported_events', None)
         supported_events: list[str] = []
         if callable(get_supported_events):
@@ -1623,8 +1836,6 @@ class RuntimeBot:
             platform_events.EBAEvent,
             tenant_scoped_listener(on_eba_event),
         )
-        # Registration is evidence of an installed listener, not a live connection.
-        snapshot_bot(self, listener_registered=True)
 
     async def run(self):
         async def exception_wrapper():
@@ -2071,6 +2282,10 @@ class PlatformManager:
         )
         binding = _binding
         for bot in result.all():
+            # Disabled drafts may not have the credentials needed to construct
+            # an adapter. They become runtime bots when explicitly enabled.
+            if not bot.enable:
+                continue
             try:
                 if binding is None:
                     binding = await self.ap.workspace_service.get_execution_binding(workspace_uuid)
@@ -2105,6 +2320,7 @@ class PlatformManager:
             bot_entity = persistence_bot.Bot(**bot_entity)
 
         execution_context = self._normalize_execution_context(context, bot_uuid=bot_entity.uuid)
+        await require_workspace_platform(self.ap, execution_context.workspace_uuid, bot_entity.adapter)
         if bot_entity.workspace_uuid != execution_context.workspace_uuid:
             raise WorkspaceRequiredError('Bot entity Workspace does not match its runtime context')
         if not _binding_validated:
@@ -2150,6 +2366,9 @@ class PlatformManager:
                 execution_context=execution_context,
             )
 
+            if hasattr(adapter_inst, 'bind_runtime'):
+                adapter_inst.bind_runtime(runtime_bot)
+
             await runtime_bot.initialize()
 
             bot_key = self._runtime_bot_key(runtime_bot)
@@ -2191,6 +2410,7 @@ class PlatformManager:
         )
         if bot is None or bot.placement_generation != execution_context.placement_generation:
             return None
+        await require_workspace_platform(self.ap, execution_context.workspace_uuid, bot.bot_entity.adapter)
         return bot
 
     async def resolve_public_bot(self, route_key: str) -> RuntimeBot | None:
diff --git a/src/langbot/pkg/platform/processing_indicator.py b/src/langbot/pkg/platform/processing_indicator.py
new file mode 100644
index 000000000..723efa356
--- /dev/null
+++ b/src/langbot/pkg/platform/processing_indicator.py
@@ -0,0 +1,13 @@
+"""Optional platform processing status around actual processor execution."""
+
+import contextlib
+
+
+@contextlib.asynccontextmanager
+async def processing_indicator(adapter, target_type, target_id):
+    factory = getattr(type(adapter), 'processing_indicator', None)
+    if callable(factory) and target_type and target_id is not None:
+        async with factory(adapter, target_type, str(target_id)):
+            yield
+    else:
+        yield
diff --git a/src/langbot/pkg/platform/sources/aiocqhttp.yaml b/src/langbot/pkg/platform/sources/aiocqhttp.yaml
index d9880bcfe..fdabf4969 100644
--- a/src/langbot/pkg/platform/sources/aiocqhttp.yaml
+++ b/src/langbot/pkg/platform/sources/aiocqhttp.yaml
@@ -14,6 +14,7 @@ metadata:
 spec:
   legacy: true
   categories:
+    - popular
     - protocol
   help_links:
     zh: https://langbot.app/docs/zh/usage/platforms/qq/aiocqhttp/napcat
@@ -59,4 +60,4 @@ spec:
 execution:
   python:
     path: ./aiocqhttp.py
-    attr: AiocqhttpAdapter
\ No newline at end of file
+    attr: AiocqhttpAdapter
diff --git a/src/langbot/pkg/platform/sources/dingtalk.py b/src/langbot/pkg/platform/sources/dingtalk.py
index 6d76ddd3d..a8fd2349d 100644
--- a/src/langbot/pkg/platform/sources/dingtalk.py
+++ b/src/langbot/pkg/platform/sources/dingtalk.py
@@ -1,3 +1,4 @@
+from langbot.libs.dingtalk_api.thinking import visible_think_markers
 import asyncio
 import json
 import pathlib
@@ -13,7 +14,7 @@ import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platf
 import langbot_plugin.api.entities.builtin.platform.events as platform_events
 import langbot_plugin.api.entities.builtin.platform.entities as platform_entities
 import langbot_plugin.api.entities.builtin.provider.session as provider_session
-from langbot.libs.dingtalk_api.api import DingTalkClient
+from langbot.libs.dingtalk_api.api import DingTalkClient, is_stream_card_configured
 import datetime
 from langbot.pkg.platform.logger import EventLogger
 from langbot.pkg.platform.human_input import format_human_input_text
@@ -65,15 +66,19 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             platform_message.Source(id=event.incoming_message.message_id, time=datetime.datetime.now())
         )
 
-        for atUser in event.incoming_message.at_users:
-            if atUser.dingtalk_id == event.incoming_message.chatbot_user_id:
-                yiri_msg_list.append(platform_message.At(target=bot_name))
+        chatbot_id = getattr(event.incoming_message, 'chatbot_user_id', None)
+        mentioned = getattr(event.incoming_message, 'is_in_at_list', False) or any(
+            chatbot_id and getattr(at_user, 'dingtalk_id', None) == chatbot_id
+            for at_user in getattr(event.incoming_message, 'at_users', []) or []
+        )
+        if mentioned:
+            yiri_msg_list.append(platform_message.At(target=bot_name))
 
         if event.rich_content:
             elements = event.rich_content.get('Elements')
             for element in elements:
                 if element.get('Type') == 'text':
-                    text = element.get('Content', '').replace('@' + bot_name, '')
+                    text = element.get('Content', '')
                     if text.strip():
                         yiri_msg_list.append(platform_message.Plain(text=text))
                 elif element.get('Type') == 'image' and element.get('Picture'):
@@ -82,7 +87,7 @@ class DingTalkMessageConverter(abstract_platform_adapter.AbstractMessageConverte
             # 回退到原有简单逻辑
             # 对于音频消息，content 来自 recognition 转写文字，在下方音频处理块中统一处理
             if event.content and event.type != 'audio':
-                text_content = event.content.replace('@' + bot_name, '')
+                text_content = event.content
                 yiri_msg_list.append(platform_message.Plain(text=text_content))
             if event.picture:
                 yiri_msg_list.append(platform_message.Image(base64=event.picture))
@@ -519,8 +524,6 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         required_keys = [
             'client_id',
             'client_secret',
-            'robot_name',
-            'robot_code',
         ]
         missing_keys = [key for key in required_keys if key not in config]
         if missing_keys:
@@ -528,12 +531,12 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         bot = DingTalkClient(
             client_id=config['client_id'],
             client_secret=config['client_secret'],
-            robot_name=config['robot_name'],
-            robot_code=config['robot_code'],
+            robot_name=config.get('robot_name', ''),
+            robot_code=config.get('robot_code', ''),
             markdown_card=config['markdown_card'],
             logger=logger,
         )
-        bot_account_id = config['robot_name']
+        bot_account_id = config['client_id']
         super().__init__(
             config=config,
             logger=logger,
@@ -610,8 +613,14 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         incoming_message = event.incoming_message
 
         markdown_enabled = self.config.get('markdown_card', False)
-        content, at = await DingTalkMessageConverter.yiri2target(message, markdown_enabled)
+        text_message = platform_message.MessageChain(
+            [part for part in message if not isinstance(part, (platform_message.Image, platform_message.File))]
+        )
+        content, at = await DingTalkMessageConverter.yiri2target(text_message, markdown_enabled)
         await self.bot.send_message(content, incoming_message, at)
+        for part in message:
+            if isinstance(part, (platform_message.Image, platform_message.File)):
+                await self.bot.send_attachment(part, incoming_message)
 
     async def reply_message_chunk(
         self,
@@ -638,7 +647,10 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
 
         if (msg_seq - 1) % 8 == 0 or is_final:
             markdown_enabled = self.config.get('markdown_card', False)
-            content, at = await DingTalkMessageConverter.yiri2target(message, markdown_enabled)
+            text_message = platform_message.MessageChain(
+                [part for part in message if not isinstance(part, (platform_message.Image, platform_message.File))]
+            )
+            content, at = await DingTalkMessageConverter.yiri2target(text_message, markdown_enabled)
             if not content and bot_message.content:
                 content = bot_message.content  # 兼容直接传入content的情况
 
@@ -707,6 +719,10 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                         pass
                 if bot_message.tool_calls is None:
                     self.card_instance_id_dict.pop(message_id, None)
+                    event = await DingTalkEventConverter.yiri2target(message_source)
+                    for part in message:
+                        if isinstance(part, (platform_message.Image, platform_message.File)):
+                            await self.bot.send_attachment(part, event.incoming_message)
 
     async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
         markdown_enabled = self.config.get('markdown_card', False)
@@ -717,15 +733,12 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             await self.bot.send_proactive_message_to_group(target_id, content)
 
     async def is_stream_output_supported(self) -> bool:
-        is_stream = False
-        if self.config.get('enable-stream-reply', None):
-            is_stream = True
-        return is_stream
+        return is_stream_card_configured(self.config)
 
     async def create_message_card(self, message_id, event):
         self._prune_card_state()
         form_template_id = (self.config.get('human_input_card_template_id') or '').strip()
-        legacy_template_id = self.config.get('card_template_id', '')
+        legacy_template_id = (self.config.get('card_template_id') or '').strip()
 
         # Synthetic events (button clicks): look up the card already in
         # active_turn_card so reply_message_chunk can stream to it.
@@ -776,7 +789,7 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
 
         # Legacy chat-card path (no form template).
         incoming_message = event.source_platform_object.incoming_message
-        card_auto_layout = self.config.get('card_auto_layout', False)
+        card_auto_layout = self.config.get('card_auto_layout', True)
         card_instance, card_instance_id = await self.bot.create_and_card(
             legacy_template_id, incoming_message, card_auto_layout=card_auto_layout
         )
@@ -812,8 +825,9 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
     ):
         async def on_message(event: DingTalkEvent):
             try:
+                self.bot_account_id = getattr(event.incoming_message, 'chatbot_user_id', None) or self.bot_account_id
                 return await callback(
-                    await self.event_converter.target2yiri(event, self.config['robot_name']),
+                    await self.event_converter.target2yiri(event, self.bot_account_id),
                     self,
                 )
             except Exception:
@@ -853,6 +867,8 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         if self.bot_avatar_media_id:
             params['bot_avatar'] = self.bot_avatar_media_id
         params.update(extra)
+        if isinstance(params.get('content'), str):
+            params['content'] = visible_think_markers(params['content'])
         return params
 
     async def kill(self) -> bool:
@@ -1196,7 +1212,7 @@ class DingTalkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             # Legacy chat-card template doesn't carry a `bot_avatar`
             # variable, so don't decorate the param map here.
             card_param_map = {'content': '', 'query': '...'}
-            card_data_config = {'autoLayout': self.config.get('card_auto_layout', False)}
+            card_data_config = {'autoLayout': self.config.get('card_auto_layout', True)}
         try:
             success = await self.bot.create_and_deliver_card(
                 card_template_id=template_id,
diff --git a/src/langbot/pkg/platform/sources/dingtalk.yaml b/src/langbot/pkg/platform/sources/dingtalk.yaml
index f8a7e6395..43392cf0b 100644
--- a/src/langbot/pkg/platform/sources/dingtalk.yaml
+++ b/src/langbot/pkg/platform/sources/dingtalk.yaml
@@ -14,6 +14,7 @@ metadata:
 spec:
   legacy: true
   categories:
+    - popular
     - china
   help_links:
     zh: https://langbot.app/docs/zh/usage/platforms/dingtalk
@@ -26,9 +27,9 @@ spec:
         zh_Hans: 一键创建应用
         zh_Hant: 一鍵建立應用
       description:
-        en_US: "Scan QR code with DingTalk to automatically create an app and fill in credentials. Note: Robot Code cannot be obtained automatically, you need to copy it from the DingTalk Developer Backend manually."
-        zh_Hans: "使用钉钉扫码自动创建应用并填写凭据。注意：机器人代码无法自动获取，需前往钉钉开发者后台手动复制。"
-        zh_Hant: "使用釘釘掃碼自動建立應用並填寫憑證。注意：機器人代碼無法自動取得，需前往釘釘開發者後台手動複製。"
+        en_US: Scan with DingTalk to create an app and automatically fill in credentials. The bot identity is resolved automatically.
+        zh_Hans: 使用钉钉扫码创建应用并自动填写凭据，机器人身份会自动识别。
+        zh_Hant: 使用釘釘掃碼建立應用並自動填寫憑證，機器人身分會自動識別。
       type: qr-code-login
       login_platform: dingtalk
       required: false
@@ -48,78 +49,96 @@ spec:
       type: string
       required: true
       default: ""
-    - name: robot_code
-      label:
-        en_US: Robot Code
-        zh_Hans: 机器人代码
-        zh_Hant: 機器人代碼
-      description:
-        en_US: "Required for image recognition, file upload and other features. Get it from DingTalk Developer Backend > Robot Configuration."
-        zh_Hans: "识图、上传文件等功能必填。请前往钉钉开发者后台 > 机器人配置中获取。"
-        zh_Hant: "識圖、上傳檔案等功能必填。請前往釘釘開發者後台 > 機器人設定中取得。"
-      type: string
-      required: true
-      default: ""
-    - name: robot_name
-      label:
-        en_US: Robot Name
-        zh_Hans: 机器人名称
-        zh_Hant: 機器人名稱
-      type: string
-      required: true
-      default: ""
     - name: markdown_card
       label:
-        en_US: Markdown Card
-        zh_Hans: 是否使用 Markdown 卡片
-        zh_Hant: 是否使用 Markdown 卡片
+        en_US: Use Card Replies
+        zh_Hans: 使用卡片回复
+        zh_Hant: 使用卡片回覆
+      description:
+        en_US: Controls Markdown and streaming card replies. Disable to use plain text while keeping your template settings. When enabled, a configured template automatically enables streaming.
+        zh_Hans: 控制 Markdown 和流式卡片回复。关闭后使用纯文本，保留模板配置；开启且已配置模板时自动使用流式回复。
+        zh_Hant: 控制 Markdown 和串流卡片回覆。關閉後使用純文字，保留範本設定；開啟且已設定範本時自動使用串流回覆。
       type: boolean
       required: false
       default: true
-    - name: enable-stream-reply
-      label:
-        en_US: Enable Stream Reply Mode
-        zh_Hans: 启用钉钉卡片流式回复模式
-        zh_Hant: 啟用釘釘卡片串流回覆模式
-      description:
-        en_US: Used for streaming replies from pipelines or plugins. If enabled, the bot will use the stream of lark reply mode
-        zh_Hans: 用于流水线和插件主动发起的流式回复。如果启用，将使用钉钉卡片流式方式来回复内容
-        zh_Hant: 用於流水線和插件主動發起的串流回覆。如果啟用，將使用釘釘卡片串流方式來回覆內容
-      type: boolean
-      required: true
-      default: false
-    - name: card_auto_layout
-      label:
-        en_US: Card Auto Layout
-        zh_Hans: 卡片宽屏自动布局
-        zh_Hant: 卡片寬螢幕自動佈局
-      type: boolean
-      required: false
-      default: false
     - name: card_template_id
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
       label:
         en_US: card template id
         zh_Hans: 卡片模板ID
         zh_Hant: 卡片範本ID
+      description:
+        en_US: Template for streaming text replies. Save and publish it in DingTalk Card Platform, then paste its ID here to automatically enable streaming while card replies are enabled. Leave empty for ordinary replies. Separate from the human input template.
+        zh_Hans: 用于流式文字回复。在钉钉卡片平台保存并发布模板后，开启卡片回复后，填写 ID 即自动启用流式回复，留空使用普通回复；与人工输入卡片模板分开配置。
+        zh_Hant: 用於串流文字回覆。在釘釘卡片平台儲存並發布範本後，開啟卡片回覆後，填寫 ID 即自動啟用串流回覆，留空使用一般回覆；與人工輸入卡片範本分開設定。
       type: string
-      required: true
-      default: "填写你的卡片template_id"
+      required: false
+      default: ""
+    - name: stream_card_template_download
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
+      label:
+        en_US: Download Streaming Reply Card Template
+        zh_Hans: 下载流式回复卡片模板
+        zh_Hant: 下載串流回覆卡片範本
+      description:
+        en_US: Displays reasoning and the answer separately. Import through More > Import Card Template in the DingTalk AI card editor, save and publish, then paste its ID into Card Template ID above. Respects the pipeline thinking removal setting.
+        zh_Hans: 分开展示思考与正文。在钉钉 AI 卡片编辑器中选择「更多 → 导入卡片模板」，保存并发布后，将 ID 填入上方「卡片模板 ID」。是否保留思考仍由流水线设置控制。
+        zh_Hant: 分開展示思考與正文。在釘釘 AI 卡片編輯器中選擇「更多 → 匯入卡片範本」，儲存並發布後，將 ID 填入上方「卡片範本 ID」。是否保留思考仍由流水線設定控制。
+      type: download-link
+      required: false
+      default: ''
+      url: /api/v1/platform/adapters/dingtalk/stream-card-template
+      download_filename: dingtalk_stream_card.json
+    - name: card_auto_layout
+      show_if:
+        field: markdown_card
+        operator: eq
+        value: true
+      label:
+        en_US: Card Auto Layout
+        zh_Hans: 卡片宽屏自动布局
+        zh_Hant: 卡片寬螢幕自動佈局
+      description:
+        en_US: Enabled by default. Allows supported cards to adapt to wider windows. The LangBot streaming reply template includes the required variables; you can disable this for a narrower layout.
+        zh_Hans: 默认开启，让支持的卡片适应较宽的聊天窗口。LangBot 流式回复模板已包含所需变量；如需较窄布局，可关闭此项。
+        zh_Hant: 預設開啟，讓支援的卡片適應較寬的聊天視窗。LangBot 串流回覆範本已包含所需變數；如需較窄版面，可關閉此項。
+      type: boolean
+      required: false
+      default: true
+    - name: human_input_card_template_id
+      label:
+        en_US: Human Input Card Template ID
+        zh_Hans: 人工输入卡片模板ID
+        zh_Hant: 人工輸入卡片範本ID
+      description:
+        en_US: Optional. Used when a workflow or plugin requests user input, a choice, or confirmation before continuing (for example, a Dify Human Input node). Ordinary chat does not require it. Import the template below and enter its published ID here; streaming text replies use the separate Card Template ID.
+        zh_Hans: 可选。当工作流或插件需要用户填写信息、选择选项或确认后再继续执行时使用，例如 Dify 的人工输入节点。普通聊天无需配置。请导入下方模板，将发布后的 ID 填在这里；流式文字回复使用单独的「卡片模板 ID」。
+        zh_Hant: 可選。當工作流或外掛需要使用者填寫資訊、選擇選項或確認後再繼續執行時使用，例如 Dify 的人工輸入節點。一般聊天無需設定。請匯入下方範本，將發布後的 ID 填在這裡；串流文字回覆使用獨立的「卡片範本 ID」。
+      type: string
+      required: false
+      default: ""
     - name: human_input_card_template_download
       label:
         en_US: Download Human Input Card Template
         zh_Hans: 下载人工输入卡片模板
         zh_Hant: 下載人工輸入卡片範本
       description:
-        en_US: "Used as the only card template ID for the whole conversation turn. Download the built-in template, then import the JSON in DingTalk Open Platform > Card Platform / Card Template Management. After DingTalk creates the template, copy its template ID into the field below. The template already wires `content` (MarkdownBlock) and `btns` (ButtonGroup). Leave empty to fall back to the legacy two-card behavior."
-        zh_Hans: "用作整个对话回合唯一卡片的模板 ID。先下载内置模板，再到钉钉开放平台 > 卡片平台 / 卡片模板管理中导入该 JSON；钉钉生成模板后，将模板 ID 填到这里。模板已预先连好 `content` (MarkdownBlock) 与 `btns` (ButtonGroup)。留空则降级为旧的双卡行为。"
-        zh_Hant: "用作整個對話回合唯一卡片的範本 ID。先下載內建範本，再到釘釘開放平台 > 卡片平台 / 卡片範本管理中匯入該 JSON；釘釘產生範本後，將範本 ID 填到這裡。範本已預先連好 `content` (MarkdownBlock) 與 `btns` (ButtonGroup)。留空則降級為舊的雙卡行為。"
+        en_US: Download the human input template. Create an AI card in DingTalk Card Platform, choose More > Import Card Template in the editor, then save and publish. Paste its ID into the field above.
+        zh_Hans: 下载人工输入模板后，在钉钉卡片平台新建 AI 卡片，在编辑器中选择「更多 → 导入卡片模板」，保存并发布，将 ID 填入上方字段。
+        zh_Hant: 下載人工輸入範本後，在釘釘卡片平台新建 AI 卡片，在編輯器中選擇「更多 → 匯入卡片範本」，儲存並發布，將 ID 填入上方欄位。
       type: download-link
       required: false
       default: ""
       url: /api/v1/platform/adapters/dingtalk/human-input-card-template
       download_filename: dingtalk_human_input_card.json
       help_links:
-        zh: https://open-dev.dingtalk.com/fe/card
+        zh: https://langbot.app/docs/zh/usage/platforms/dingtalk#导入人工输入卡片模板
         en: https://open-dev.dingtalk.com/fe/card
         ja: https://open-dev.dingtalk.com/fe/card
       help_label:
@@ -127,18 +146,6 @@ spec:
         zh_Hans: 导入指引
         zh_Hant: 匯入指引
         ja_JP: インポート手順
-    - name: human_input_card_template_id
-      label:
-        en_US: Human Input Card Template ID
-        zh_Hans: 人工输入卡片模板ID
-        zh_Hant: 人工輸入卡片範本ID
-      description:
-        en_US: "Paste the template ID generated after importing the human input card template."
-        zh_Hans: "填写导入人工输入卡片模板后生成的模板 ID。"
-        zh_Hant: "填寫匯入人工輸入卡片範本後產生的範本 ID。"
-      type: string
-      required: false
-      default: ""
 execution:
   python:
     path: ./dingtalk.py
diff --git a/src/langbot/pkg/platform/sources/discord.py b/src/langbot/pkg/platform/sources/discord.py
index e7119e391..1d9a29234 100644
--- a/src/langbot/pkg/platform/sources/discord.py
+++ b/src/langbot/pkg/platform/sources/discord.py
@@ -1,5 +1,8 @@
 from __future__ import annotations
 
+from langbot.pkg.i18n import t
+from langbot.pkg.i18n.adapters import adapter_t, localized_adapter
+
 import discord
 from discord import ui as discord_ui
 
@@ -219,11 +222,13 @@ class VoiceConnectionManager:
                 # 获取服务器和频道对象
                 guild = self.bot.get_guild(guild_id)
                 if not guild:
-                    raise VoiceConnectionError(f'无法找到服务器 {guild_id}', 'GUILD_NOT_FOUND', guild_id)
+                    raise VoiceConnectionError(t('Resource not found') + f': {guild_id}', 'GUILD_NOT_FOUND', guild_id)
 
                 channel = guild.get_channel(channel_id)
                 if not channel or not isinstance(channel, discord.VoiceChannel):
-                    raise VoiceConnectionError(f'无法找到语音频道 {channel_id}', 'CHANNEL_NOT_FOUND', guild_id)
+                    raise VoiceConnectionError(
+                        t('Resource not found') + f': {channel_id}', 'CHANNEL_NOT_FOUND', guild_id
+                    )
 
                 # 验证用户是否在语音频道中（如果提供了用户ID）
                 if user_id:
@@ -258,12 +263,14 @@ class VoiceConnectionManager:
                 return voice_client
 
             except discord.ClientException as e:
-                raise VoiceNetworkError(f'Discord 客户端错误: {str(e)}')
+                raise VoiceNetworkError(t('Voice connection failed') + f': {e}')
             except discord.opus.OpusNotLoaded as e:
-                raise VoiceConnectionError(f'Opus 编码器未加载: {str(e)}', 'OPUS_NOT_LOADED', guild_id)
+                raise VoiceConnectionError(
+                    t('Voice service is unavailable') + f': Opus: {e}', 'OPUS_NOT_LOADED', guild_id
+                )
             except Exception as e:
                 await self.logger.error(f'连接语音频道时发生未知错误: {str(e)}')
-                raise VoiceConnectionError(f'连接失败: {str(e)}', 'UNKNOWN_ERROR', guild_id)
+                raise VoiceConnectionError(t('Voice connection failed') + f': {e}', 'UNKNOWN_ERROR', guild_id)
 
     async def leave_voice_channel(self, guild_id: int) -> bool:
         """
@@ -494,11 +501,13 @@ class VoiceConnectionManager:
         """
         member = guild.get_member(user_id)
         if not member:
-            raise VoicePermissionError(f'无法找到用户 {user_id}', ['member_not_found'], user_id, channel.id)
+            raise VoicePermissionError(
+                t('Resource not found') + f': {user_id}', ['member_not_found'], user_id, channel.id
+            )
 
         if not member.voice or member.voice.channel != channel:
             raise VoicePermissionError(
-                f'用户 {member.display_name} 不在语音频道 {channel.name} 中',
+                t('User is not in the voice channel') + f': {member.display_name}, {channel.name}',
                 ['user_not_in_channel'],
                 user_id,
                 channel.id,
@@ -528,7 +537,7 @@ class VoiceConnectionManager:
 
         if missing_permissions:
             raise VoicePermissionError(
-                f'机器人在频道 {channel.name} 中缺少权限: {", ".join(missing_permissions)}',
+                t('Permission denied') + f': {channel.name}: {", ".join(missing_permissions)}',
                 missing_permissions,
                 channel_id=channel.id,
             )
@@ -843,7 +852,7 @@ class DiscordFormView(discord_ui.View):
         # silently — most Dify forms have ≤10 actions in practice.
         for idx, action in enumerate(actions[:25]):
             action_id = str(action.get('id') or '')
-            label = str(action.get('title') or action_id or f'Option {idx + 1}')
+            label = str(action.get('title') or action_id or adapter_t(adapter, 'Option {number}', number=idx + 1))
             style = self._STYLE_MAP.get(
                 str(action.get('button_style') or '').lower(),
                 discord.ButtonStyle.secondary,
@@ -966,6 +975,7 @@ class DiscordAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             self._pending_forms.pop(next(iter(self._pending_forms)), None)
 
     # Voice functionality methods
+    @localized_adapter
     async def join_voice_channel(self, guild_id: int, channel_id: int, user_id: int = None) -> discord.VoiceClient:
         """
         加入语音频道
@@ -990,7 +1000,7 @@ class DiscordAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             VoiceConnectionError: 其他连接错误
         """
         if not self.voice_manager:
-            raise VoiceConnectionError('语音管理器未初始化', 'MANAGER_NOT_READY')
+            raise VoiceConnectionError(t('Voice service is unavailable'), 'MANAGER_NOT_READY')
 
         return await self.voice_manager.join_voice_channel(guild_id, channel_id, user_id)
 
@@ -1358,7 +1368,7 @@ class DiscordAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                 )
             return
 
-        node_title = str(form_data.get('node_title') or 'Confirmation needed')
+        node_title = str(form_data.get('node_title') or adapter_t(self, 'Confirmation needed'))
         form_content = str(form_data.get('form_content') or '').strip()
 
         # Two paths:
diff --git a/src/langbot/pkg/platform/sources/http_bot.yaml b/src/langbot/pkg/platform/sources/http_bot.yaml
index 97392a444..5e13f437e 100644
--- a/src/langbot/pkg/platform/sources/http_bot.yaml
+++ b/src/langbot/pkg/platform/sources/http_bot.yaml
@@ -8,10 +8,14 @@ metadata:
     zh_Hant: HTTP 通用接入
     ja_JP: HTTP ボット
   description:
-    en_US: Integrate any backend over plain HTTP. Push messages in via a signed webhook, receive replies on a callback URL. Server-to-server, no long-lived connection. Preserves message aggregation (N->1) and multi-part replies (1->M).
-    zh_Hans: 通过 HTTP 接入任意后端系统。以签名 Webhook 推入消息，在回调地址接收回复。面向服务间集成，无需长连接。完整保留消息聚合（多条合一）与多段回复（一条问、多条回）能力。
-    zh_Hant: 透過 HTTP 接入任意後端系統。以簽名 Webhook 推入訊息，在回調地址接收回覆。面向服務間整合，無需長連線。完整保留訊息聚合（多條合一）與多段回覆（一條問、多條回）能力。
-    ja_JP: 任意のバックエンドを HTTP で接続。署名付き Webhook でメッセージを送信し、コールバック URL で返信を受信します。サーバー間連携、長時間接続不要。メッセージ集約（N→1）とマルチパート返信（1→M）に対応。
+    en_US: Connect backend systems via HTTP. Send messages through signed webhooks and receive replies via callbacks.
+    zh_Hans: 通过 HTTP 接入后端系统，以签名 Webhook 发送消息，通过回调接收回复。
+    zh_Hant: 透過 HTTP 接入後端系統，以簽名 Webhook 傳送訊息，透過回呼接收回覆。
+    ja_JP: HTTP でバックエンドに接続。署名付き Webhook でメッセージを送信し、コールバックで返信を受信します。
+    es_ES: Conecta sistemas mediante HTTP. Envía mensajes con webhooks firmados y recibe respuestas mediante callbacks.
+    ru_RU: Подключайте серверные системы по HTTP. Отправляйте сообщения через подписанные вебхуки и получайте ответы через обратные вызовы.
+    th_TH: เชื่อมต่อระบบผ่าน HTTP ส่งข้อความด้วย Webhook ที่ลงลายเซ็นและรับคำตอบผ่าน callback
+    vi_VN: Kết nối hệ thống qua HTTP. Gửi tin nhắn bằng webhook có chữ ký và nhận phản hồi qua callback.
   icon: http_bot.svg
 spec:
   categories:
diff --git a/src/langbot/pkg/platform/sources/lark.py b/src/langbot/pkg/platform/sources/lark.py
index b4c4dc827..c6c2c9455 100644
--- a/src/langbot/pkg/platform/sources/lark.py
+++ b/src/langbot/pkg/platform/sources/lark.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from langbot.pkg.i18n.adapters import adapter_t, localized_adapter
+
 import lark_oapi
 from lark_oapi.api.im.v1 import CreateImageRequest, CreateImageRequestBody, CreateFileRequest, CreateFileRequestBody
 import traceback
@@ -1240,7 +1242,9 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
 
                     from lark_oapi.event.callback.model.p2_card_action_trigger import P2CardActionTriggerResponse
 
-                    return P2CardActionTriggerResponse({'toast': {'type': 'success', 'content': '操作成功'}})
+                    return P2CardActionTriggerResponse(
+                        {'toast': {'type': 'success', 'content': adapter_t(self, 'Done')}}
+                    )
 
                 if action_value == '有帮助':
                     feedback_type = 1
@@ -1249,7 +1253,9 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                 else:
                     from lark_oapi.event.callback.model.p2_card_action_trigger import P2CardActionTriggerResponse
 
-                    return P2CardActionTriggerResponse({'toast': {'type': 'success', 'content': '操作成功'}})
+                    return P2CardActionTriggerResponse(
+                        {'toast': {'type': 'success', 'content': adapter_t(self, 'Done')}}
+                    )
 
                 operator = getattr(event.event, 'operator', None)
                 context = getattr(event.event, 'context', None)
@@ -1286,13 +1292,17 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
 
                 from lark_oapi.event.callback.model.p2_card_action_trigger import P2CardActionTriggerResponse
 
-                return P2CardActionTriggerResponse({'toast': {'type': 'success', 'content': '感谢您的反馈'}})
+                return P2CardActionTriggerResponse(
+                    {'toast': {'type': 'success', 'content': adapter_t(self, 'Thank you for your feedback')}}
+                )
             except Exception:
                 traceback.print_exc()
                 schedule_on_app_loop(self.logger.error(f'Error in lark card action callback: {traceback.format_exc()}'))
                 from lark_oapi.event.callback.model.p2_card_action_trigger import P2CardActionTriggerResponse
 
-                return P2CardActionTriggerResponse({'toast': {'type': 'error', 'content': '反馈处理失败'}})
+                return P2CardActionTriggerResponse(
+                    {'toast': {'type': 'error', 'content': adapter_t(self, 'Failed to process feedback')}}
+                )
 
         event_handler = (
             lark_oapi.EventDispatcherHandler.builder('', '')
@@ -1545,6 +1555,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                         return True
         return False
 
+    @localized_adapter
     async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
         text_elements, media_items = await self.message_converter.yiri2target(message, self.api_client)
 
@@ -1809,7 +1820,11 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                     'elements': [
                                         {
                                             'tag': 'markdown',
-                                            'content': '<font color="grey-600">以上内容由 AI 生成，仅供参考。更多详细、准确信息可点击引用链接查看</font>',
+                                            'content': '<font color="grey-600">'
+                                            + adapter_t(
+                                                self, 'AI-generated content. Refer to the cited sources for details.'
+                                            )
+                                            + '</font>',
                                             'text_align': 'left',
                                             'text_size': 'notation',
                                             'margin': '4px 0px 0px 0px',
@@ -1840,7 +1855,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                             'width': 'fill',
                                             'size': 'medium',
                                             'icon': {'tag': 'standard_icon', 'token': 'thumbsup_outlined'},
-                                            'hover_tips': {'tag': 'plain_text', 'content': '有帮助'},
+                                            'hover_tips': {'tag': 'plain_text', 'content': adapter_t(self, 'Helpful')},
                                             'behaviors': [{'type': 'callback', 'value': {'feedback': '有帮助'}}],
                                             'margin': '0px 0px 0px 0px',
                                         }
@@ -1864,7 +1879,10 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                             'width': 'default',
                                             'size': 'medium',
                                             'icon': {'tag': 'standard_icon', 'token': 'thumbdown_outlined'},
-                                            'hover_tips': {'tag': 'plain_text', 'content': '无帮助'},
+                                            'hover_tips': {
+                                                'tag': 'plain_text',
+                                                'content': adapter_t(self, 'Not helpful'),
+                                            },
                                             'behaviors': [{'type': 'callback', 'value': {'feedback': '无帮助'}}],
                                             'margin': '0px 0px 0px 0px',
                                         }
@@ -2064,6 +2082,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         )
         return new_card_id
 
+    @localized_adapter
     async def reply_message(
         self,
         message_source: platform_events.MessageEvent,
@@ -2174,6 +2193,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                     f'client.im.v1.message.reply ({media["msg_type"]}) failed, code: {response.code}, msg: {response.msg}, log_id: {response.get_log_id()}, resp: \n{json.dumps(json.loads(response.raw.content), indent=4, ensure_ascii=False)}'
                 )
 
+    @localized_adapter
     async def reply_message_chunk(
         self,
         message_source: platform_events.MessageEvent,
@@ -2491,7 +2511,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                         'tag': 'select_static',
                         'name': component_name,
                         'label': {'tag': 'plain_text', 'content': field_name},
-                        'placeholder': {'tag': 'plain_text', 'content': '请选择'},
+                        'placeholder': {'tag': 'plain_text', 'content': adapter_t(self, 'Select an option')},
                         'options': [
                             {
                                 'text': {'tag': 'plain_text', 'content': option},
@@ -2525,7 +2545,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                     'tag': 'input',
                     'name': component_name,
                     'label': {'tag': 'plain_text', 'content': field_name},
-                    'placeholder': {'tag': 'plain_text', 'content': '请输入'},
+                    'placeholder': {'tag': 'plain_text', 'content': adapter_t(self, 'Enter a value')},
                     'default_value': _dify_default_value(field),
                     'width': 'fill',
                     'required': False,
@@ -2820,7 +2840,11 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                 'elements': [
                                     {
                                         'tag': 'markdown',
-                                        'content': '<font color="grey-600">以上内容由 AI 生成，仅供参考。更多详细、准确信息可点击引用链接查看</font>',
+                                        'content': '<font color="grey-600">'
+                                        + adapter_t(
+                                            self, 'AI-generated content. Refer to the cited sources for details.'
+                                        )
+                                        + '</font>',
                                         'text_align': 'left',
                                         'text_size': 'notation',
                                         'margin': '4px 0px 0px 0px',
@@ -2855,7 +2879,10 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                                 'width': 'fill',
                                                 'size': 'medium',
                                                 'icon': {'tag': 'standard_icon', 'token': 'thumbsup_outlined'},
-                                                'hover_tips': {'tag': 'plain_text', 'content': '有帮助'},
+                                                'hover_tips': {
+                                                    'tag': 'plain_text',
+                                                    'content': adapter_t(self, 'Helpful'),
+                                                },
                                                 'behaviors': [{'type': 'callback', 'value': {'feedback': '有帮助'}}],
                                                 'margin': '0px 0px 0px 0px',
                                             }
@@ -2879,7 +2906,10 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                                                 'width': 'default',
                                                 'size': 'medium',
                                                 'icon': {'tag': 'standard_icon', 'token': 'thumbdown_outlined'},
-                                                'hover_tips': {'tag': 'plain_text', 'content': '无帮助'},
+                                                'hover_tips': {
+                                                    'tag': 'plain_text',
+                                                    'content': adapter_t(self, 'Not helpful'),
+                                                },
                                                 'behaviors': [{'type': 'callback', 'value': {'feedback': '无帮助'}}],
                                                 'margin': '0px 0px 0px 0px',
                                             }
@@ -2973,6 +3003,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         else:
             return data['type']
 
+    @localized_adapter
     async def handle_unified_webhook(self, bot_uuid: str, path: str, request):
         """处理统一 webhook 请求。
         Args:
@@ -3132,14 +3163,14 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                             },
                         )
 
-                        return {'toast': {'type': 'success', 'content': '操作成功'}}
+                        return {'toast': {'type': 'success', 'content': adapter_t(self, 'Done')}}
 
                     if action_value == '有帮助':
                         feedback_type = 1
                     elif action_value == '无帮助':
                         feedback_type = 2
                     else:
-                        return {'toast': {'type': 'success', 'content': '操作成功'}}
+                        return {'toast': {'type': 'success', 'content': adapter_t(self, 'Done')}}
 
                     user_id = operator.get('open_id') or operator.get('user_id')
                     open_chat_id = context_data.get('open_chat_id')
@@ -3171,10 +3202,10 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                     if platform_events.FeedbackEvent in self.listeners:
                         await self.listeners[platform_events.FeedbackEvent](feedback_event, self)
 
-                    return {'toast': {'type': 'success', 'content': '感谢您的反馈'}}
+                    return {'toast': {'type': 'success', 'content': adapter_t(self, 'Thank you for your feedback')}}
                 except Exception:
                     await self.logger.error(f'Error in lark card action callback: {traceback.format_exc()}')
-                    return {'toast': {'type': 'error', 'content': '反馈处理失败'}}
+                    return {'toast': {'type': 'error', 'content': adapter_t(self, 'Failed to process feedback')}}
 
             elif 'im.chat.member.bot.added_v1' == type:
                 try:
@@ -3227,6 +3258,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             await self.logger.error(f'Error in lark callback: {traceback.format_exc()}')
             return {'code': 500, 'message': 'error'}
 
+    @localized_adapter
     async def run_async(self):
         enable_webhook = self.config['enable-webhook']
 
diff --git a/src/langbot/pkg/platform/sources/line.yaml b/src/langbot/pkg/platform/sources/line.yaml
index 83f520286..b99ab5b2c 100644
--- a/src/langbot/pkg/platform/sources/line.yaml
+++ b/src/langbot/pkg/platform/sources/line.yaml
@@ -20,6 +20,7 @@ metadata:
   icon: line.png
 spec:
   categories:
+    - popular
     - global
   help_links:
     zh: https://langbot.app/docs/zh/usage/platforms/line
@@ -81,4 +82,4 @@ spec:
 execution:
   python:
     path: ./line.py
-    attr: LINEAdapter
\ No newline at end of file
+    attr: LINEAdapter
diff --git a/src/langbot/pkg/platform/sources/matrix.py b/src/langbot/pkg/platform/sources/matrix.py
index f0a9822e3..3fc311392 100644
--- a/src/langbot/pkg/platform/sources/matrix.py
+++ b/src/langbot/pkg/platform/sources/matrix.py
@@ -1,5 +1,8 @@
 from __future__ import annotations
 
+from langbot.pkg.i18n import t
+from langbot.pkg.i18n.adapters import localized_adapter
+
 import typing
 import asyncio
 import traceback
@@ -666,25 +669,26 @@ class MatrixAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         except Exception:
             await self.logger.error(f'[{bridge.user_id}] Bridge status check fatal error: {traceback.format_exc()}')
 
+    @localized_adapter
     async def _handle_relogin_command(self, room_id: str):
         """Handle !relogin command: logout then re-login all bridges."""
         if not self._bridges:
             await self.client.room_send(
                 room_id=room_id,
                 message_type='m.room.message',
-                content={'msgtype': 'm.text', 'body': '没有配置任何桥。'},
+                content={'msgtype': 'm.text', 'body': t('No bridges configured.')},
             )
             return
 
-        lines = ['开始重新登录所有桥...']
+        lines = [t('Signing in to all bridges again...')]
         for bridge in self._bridges:
             if not bridge.login_command or not bridge.dm_room_id:
-                lines.append(f'[{bridge.user_id}] 跳过（未配置登录命令或无DM房间）')
+                lines.append(t('Bridge {bridge}: skipped (login command or DM room missing).', bridge=bridge.user_id))
                 continue
 
             # Use configured logout command, fallback to deriving from login command
             logout_cmd = bridge.logout_command or bridge.login_command.replace('login', 'logout')
-            lines.append(f'[{bridge.user_id}] 发送 "{logout_cmd}"...')
+            lines.append(t('Bridge {bridge}: sending {command}...', bridge=bridge.user_id, command=logout_cmd))
 
             # Cancel existing tasks
             if bridge.login_task and not bridge.login_task.done():
@@ -700,14 +704,14 @@ class MatrixAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
                     content={'msgtype': 'm.text', 'body': logout_cmd},
                 )
             except Exception as e:
-                lines.append(f'[{bridge.user_id}] logout 发送失败: {e}')
+                lines.append(t('Bridge {bridge}: sign-out failed.', bridge=bridge.user_id) + f' {e}')
 
             await asyncio.sleep(2)
 
             # Reset state and restart login
             bridge.logged_in = False
             self._restart_bridge_login(bridge)
-            lines.append(f'[{bridge.user_id}] 已触发重新登录')
+            lines.append(t('Bridge {bridge}: sign-in started.', bridge=bridge.user_id))
 
         await self.client.room_send(
             room_id=room_id,
@@ -715,20 +719,21 @@ class MatrixAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
             content={'msgtype': 'm.text', 'body': '\n'.join(lines)},
         )
 
+    @localized_adapter
     async def _handle_status_command(self, room_id: str):
         """Handle !status command: show bridge states."""
         if not self._bridges:
             await self.client.room_send(
                 room_id=room_id,
                 message_type='m.room.message',
-                content={'msgtype': 'm.text', 'body': '没有配置任何桥。'},
+                content={'msgtype': 'm.text', 'body': t('No bridges configured.')},
             )
             return
 
-        lines = ['桥状态:']
+        lines = [t('Bridge status:')]
         for bridge in self._bridges:
-            status = '已登录 ✓' if bridge.logged_in else '未登录 ✗'
-            dm = bridge.dm_room_id or '无'
+            status = t('Signed in ✓') if bridge.logged_in else t('Not signed in ✗')
+            dm = bridge.dm_room_id or t('None')
             lines.append(f'• {bridge.user_id}: {status} (DM: {dm})')
         await self.client.room_send(
             room_id=room_id,
diff --git a/src/langbot/pkg/platform/sources/mattermost.yaml b/src/langbot/pkg/platform/sources/mattermost.yaml
index f46b125bb..6c958e42d 100644
--- a/src/langbot/pkg/platform/sources/mattermost.yaml
+++ b/src/langbot/pkg/platform/sources/mattermost.yaml
@@ -22,7 +22,6 @@ metadata:
 spec:
   categories:
     - global
-    - popular
   config:
     - name: server_url
       label:
diff --git a/src/langbot/pkg/platform/sources/openclaw_weixin.py b/src/langbot/pkg/platform/sources/openclaw_weixin.py
index 4726888b1..b4c9dc270 100644
--- a/src/langbot/pkg/platform/sources/openclaw_weixin.py
+++ b/src/langbot/pkg/platform/sources/openclaw_weixin.py
@@ -10,6 +10,7 @@ from __future__ import annotations
 
 import asyncio
 import base64
+import contextlib
 import os
 import traceback
 import typing
@@ -256,6 +257,8 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
     # context_token cache: from_user_id -> context_token
     _context_tokens: dict[str, str] = pydantic.PrivateAttr(default_factory=dict)
 
+    _typing_indicator: typing.Any = pydantic.PrivateAttr(default=None)
+
     _polling: bool = pydantic.PrivateAttr(default=False)
     _poll_task: typing.Optional[asyncio.Task] = pydantic.PrivateAttr(default=None)
     _bot_uuid: typing.Optional[str] = pydantic.PrivateAttr(default=None)
@@ -347,6 +350,18 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
         # Persist token to database so it survives restart
         await self._persist_config()
 
+    @contextlib.asynccontextmanager
+    async def processing_indicator(self, target_type, target_id):
+        if target_type not in ('person', 'friend'):
+            yield
+            return
+        from langbot.libs.openclaw_weixin_api.typing_indicator import TypingIndicator
+
+        if self._typing_indicator is None:
+            self._typing_indicator = TypingIndicator(self.client, lambda peer: self._context_tokens.get(peer, ''))
+        async with self._typing_indicator.processing(target_id):
+            yield
+
     async def send_message(
         self,
         target_type: str,
@@ -385,6 +400,7 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
                 await self.logger.error(
                     f'Failed to send component {type(component).__name__}: {traceback.format_exc()}'
                 )
+                raise
 
     async def reply_message(
         self,
@@ -478,6 +494,9 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
             token=self.config.get('token', token),
         )
         self.bot_account_id = self.config.get('account_id', 'openclaw-weixin')
+        from langbot.libs.openclaw_weixin_api.typing_indicator import TypingIndicator
+
+        self._typing_indicator = TypingIndicator(self.client, lambda peer: self._context_tokens.get(peer, ''))
         self._polling = True
 
         # Start the long-poll loop
@@ -488,6 +507,9 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
             await self._poll_task
         except asyncio.CancelledError:
             pass
+        finally:
+            if self._typing_indicator is not None:
+                await self._typing_indicator.close()
 
     async def _poll_loop(self):
         """Long-poll loop: call getUpdates continuously.
@@ -522,6 +544,9 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
                         await self.logger.error('OpenClaw WeChat session expired, attempting re-login...')
                         try:
                             await self._do_login()
+                            if self._typing_indicator is not None:
+                                await self._typing_indicator.close()
+                                self._typing_indicator = None
                             # Rebuild client with new credentials
                             self.client = OpenClawWeixinClient(
                                 base_url=self.config.get('base_url', DEFAULT_BASE_URL),
@@ -615,6 +640,8 @@ class OpenClawWeixinAdapter(abstract_platform_adapter.AbstractMessagePlatformAda
     async def kill(self) -> bool:
         """Stop the adapter."""
         self._polling = False
+        if self._typing_indicator is not None:
+            await self._typing_indicator.close()
         if self._poll_task and not self._poll_task.done():
             self._poll_task.cancel()
             try:
diff --git a/src/langbot/pkg/platform/sources/slack.py b/src/langbot/pkg/platform/sources/slack.py
index df18a0765..86481f9a1 100644
--- a/src/langbot/pkg/platform/sources/slack.py
+++ b/src/langbot/pkg/platform/sources/slack.py
@@ -6,6 +6,8 @@ import traceback
 import datetime
 
 from langbot.libs.slack_api.api import SlackClient
+from langbot.libs.slack_api.media import send_chain
+from langbot.pkg.platform.adapters.slack.message_converter import SlackMessageConverter as OmniMessageConverter
 import langbot_plugin.api.definition.abstract.platform.adapter as abstract_platform_adapter
 from langbot.libs.slack_api.slackevent import SlackEvent
 import langbot_plugin.api.entities.builtin.platform.events as platform_events
@@ -61,13 +63,9 @@ class SlackEventConverter(abstract_platform_adapter.AbstractEventConverter):
 
     @staticmethod
     async def target2yiri(event: SlackEvent, bot: SlackClient):
-        yiri_chain = await SlackMessageConverter.target2yiri(
-            message=event.text, message_id=event.message_id, pic_url=event.pic_url, bot=bot
-        )
+        yiri_chain = await OmniMessageConverter.target2yiri(event, bot.bot_token)
 
         if event.type == 'channel':
-            yiri_chain.insert(0, platform_message.At(target='SlackBot'))
-
             sender = platform_entities.GroupMember(
                 id=event.user_id,
                 member_name=str(event.sender_name),
@@ -102,14 +100,17 @@ class SlackAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
     def __init__(self, config: dict, logger: EventLogger):
         required_keys = [
             'bot_token',
-            'signing_secret',
+            'app_token' if config.get('socket_mode', False) else 'signing_secret',
         ]
-        missing_keys = [key for key in required_keys if key not in config]
+        missing_keys = [key for key in required_keys if not config.get(key)]
         if missing_keys:
             raise command_errors.ParamNotEnoughError('Slack机器人缺少相关配置项，请查看文档或联系管理员')
 
         bot = SlackClient(
-            bot_token=config['bot_token'], signing_secret=config['signing_secret'], logger=logger, unified_mode=True
+            bot_token=config['bot_token'],
+            signing_secret=config.get('signing_secret', ''),
+            logger=logger,
+            unified_mode=True,
         )
 
         super().__init__(
@@ -127,27 +128,13 @@ class SlackAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
     ):
         slack_event = await SlackEventConverter.yiri2target(message_source)
 
-        content_list = await SlackMessageConverter.yiri2target(message)
-
-        for content in content_list:
-            # Both text and image (URL) are sent as text messages
-            # Slack will auto-unfurl image URLs
-            message_content = content['content']
-            if slack_event.type == 'channel':
-                await self.bot.send_message_to_channel(message_content, slack_event.channel_id)
-            if slack_event.type == 'im':
-                await self.bot.send_message_to_one(message_content, slack_event.user_id)
+        target_type = 'channel' if slack_event.type == 'channel' else 'person'
+        target_id = slack_event.channel_id if target_type == 'channel' else slack_event.user_id
+        return await send_chain(self.bot, target_type, target_id, message)
 
     async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
-        content_list = await SlackMessageConverter.yiri2target(message)
-        for content in content_list:
-            # Both text and image (URL) are sent as text messages
-            # Slack will auto-unfurl image URLs
-            message_content = content['content']
-            if target_type == 'person':
-                await self.bot.send_message_to_one(message_content, target_id)
-            if target_type == 'group':
-                await self.bot.send_message_to_channel(message_content, target_id)
+        target_type = 'channel' if target_type in {'group', 'channel'} else target_type
+        return await send_chain(self.bot, target_type, target_id, message)
 
     def register_listener(
         self,
@@ -186,6 +173,10 @@ class SlackAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         return await self.bot.handle_unified_webhook(request)
 
     async def run_async(self):
+        if self.config.get('socket_mode', False):
+            await self.bot.run_socket(self.config['app_token'])
+            return
+
         # 统一 webhook 模式下，不启动独立的 Quart 应用
         # 保持运行但不启动独立端口
         async def keep_alive():
@@ -195,7 +186,8 @@ class SlackAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
         await keep_alive()
 
     async def kill(self) -> bool:
-        return False
+        await self.bot.close_socket()
+        return True
 
     async def unregister_listener(
         self,
diff --git a/src/langbot/pkg/platform/sources/slack.yaml b/src/langbot/pkg/platform/sources/slack.yaml
index 44fd4b337..f6bbd97e2 100644
--- a/src/langbot/pkg/platform/sources/slack.yaml
+++ b/src/langbot/pkg/platform/sources/slack.yaml
@@ -12,9 +12,9 @@ metadata:
     es_ES: Slack
   description:
     en_US: Slack Adapter
-    zh_Hans: Slack 适配器，需要公网地址以接收 Slack 消息推送，请查看文档了解使用方式
-    zh_Hant: Slack 適配器，需要公網地址以接收 Slack 訊息推送，請查看文件了解使用方式
-    ja_JP: Slack アダプター、Slackのメッセージプッシュを受信するためにパブリックURLが必要です。使用方法の詳細については、ドキュメントを参照してください。
+    zh_Hans: Slack 适配器，支持 Webhook 和无需公网地址的长连接，请查看文档了解使用方式
+    zh_Hant: Slack 適配器，支援 Webhook 或無需公開網址的 Socket Mode。
+    ja_JP: Slack アダプター。Webhook または公開 URL が不要な Socket Mode を利用できます。
     th_TH: อะแดปเตอร์ Slack ต้องการที่อยู่สาธารณะเพื่อรับการแจ้งเตือนข้อความจาก Slack โปรดดูเอกสารประกอบสำหรับวิธีการใช้งาน
     vi_VN: Bộ điều hợp Slack, cần địa chỉ công cộng để nhận thông báo tin nhắn từ Slack, vui lòng xem tài liệu để biết cách sử dụng
     es_ES: Adaptador de Slack, requiere una dirección pública para recibir notificaciones de mensajes de Slack, consulte la documentación para obtener instrucciones de uso
@@ -29,7 +29,26 @@ spec:
     en: https://langbot.app/docs/en/usage/platforms/slack
     ja: https://langbot.app/docs/ja/usage/platforms/slack
   config:
+    - name: quick_setup
+      label: {en_US: Quick Slack setup, zh_Hans: Slack 快速配置, zh_Hant: Slack 快速設定, ja_JP: Slack クイック設定}
+      description: {en_US: Create and configure an app using App Configuration Tokens., zh_Hans: 填入配置令牌，自动创建应用并配置权限，授权后填入凭据。, zh_Hant: 填入設定權杖，自動建立應用程式並設定權限，授權後填入憑證。, ja_JP: 設定トークンでアプリを作成し、承認後に認証情報を入力します。}
+      type: qr-code-login
+      login_platform: slack
+      required: false
+    - name: socket_mode
+      label: {en_US: Socket Mode, zh_Hans: 长连接模式, zh_Hant: 長連線模式, ja_JP: Socket Mode}
+      type: boolean
+      default: false
+      required: false
+    - name: app_token
+      label: {en_US: App-Level Token, zh_Hans: App-Level Token, zh_Hant: App-Level Token, ja_JP: App-Level Token}
+      description: {en_US: Generate an xapp- token with connections:write in Basic Information., zh_Hans: 在 Basic Information 中生成带 connections:write 权限的 xapp- 令牌。, zh_Hant: 在 Basic Information 產生具有 connections:write 權限的 xapp- 權杖。, ja_JP: Basic Information で connections:write 権限の xapp- トークンを生成してください。}
+      type: string
+      required: false
+      default: ''
+      show_if: {field: socket_mode, operator: eq, value: true}
     - name: webhook_url
+      show_if: {field: socket_mode, operator: eq, value: false}
       label:
         en_US: Webhook Callback URL
         zh_Hans: Webhook 回调地址
@@ -62,6 +81,7 @@ spec:
       required: true
       default: ""
     - name: signing_secret
+      show_if: {field: socket_mode, operator: eq, value: false}
       label:
         en_US: signing_secret
         zh_Hans: 密钥
@@ -71,7 +91,7 @@ spec:
         vi_VN: Khóa ký
         es_ES: Secreto de firma
       type: string
-      required: true
+      required: false
       default: ""
 execution:
   python:
diff --git a/src/langbot/pkg/plugin/agent_run_support.py b/src/langbot/pkg/plugin/agent_run_support.py
index 4bd770d20..a14d044d2 100644
--- a/src/langbot/pkg/plugin/agent_run_support.py
+++ b/src/langbot/pkg/plugin/agent_run_support.py
@@ -2,8 +2,6 @@
 
 from __future__ import annotations
 
-from ..telemetry import diagnostics
-
 from typing import Any, Union
 import json
 import time
@@ -303,8 +301,6 @@ async def _validate_agent_run_session(
         )
         if not available_apis.get(api_capability, False) and not has_admin_permission:
             return None, handler.ActionResponse.error(message=f'{api_name} access not authorized')
-
-    diagnostics.link_context(session.get('_diagnostic_context'))
     return session, None
 
 
diff --git a/src/langbot/pkg/plugin/certification.py b/src/langbot/pkg/plugin/certification.py
index 6a538980b..b6eabaf0f 100644
--- a/src/langbot/pkg/plugin/certification.py
+++ b/src/langbot/pkg/plugin/certification.py
@@ -18,6 +18,7 @@ from langbot_plugin.entities.io.context import PluginExecutionMode
 
 
 SHARED_RUNTIME_V1 = 'shared-runtime-v1'
+STATELESS_COMPONENT_MODEL_V1 = 'stateless-v1'
 DEDICATED_RUNTIME = 'dedicated'
 
 
@@ -42,12 +43,11 @@ class AdmissionDisposition(str, Enum):
 
 class AdmissionCode(str, Enum):
     SHARED_ELIGIBLE = 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE'
-    CLOUD_CERTIFICATE_REQUIRED = 'CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_REQUIRED'
+    UNSIGNED_DEDICATED = 'CERTIFIED_PLUGIN_UNSIGNED_DEDICATED'
     CLOUD_CERTIFICATE_INVALID = 'CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_INVALID'
     OSS_LEGACY_DEDICATED = 'CERTIFIED_PLUGIN_OSS_LEGACY_DEDICATED'
     OSS_FORCE_REQUIRED = 'CERTIFIED_PLUGIN_OSS_FORCE_REQUIRED'
     OSS_FORCED_DEDICATED = 'CERTIFIED_PLUGIN_OSS_FORCED_DEDICATED'
-    OSS_CERTIFIED_DEDICATED = 'CERTIFIED_PLUGIN_OSS_CERTIFIED_DEDICATED'
     OSS_UNTRUSTED_DEDICATED = 'CERTIFIED_PLUGIN_OSS_UNTRUSTED_DEDICATED'
 
 
@@ -66,11 +66,16 @@ class CertificateFacts:
 
     verification: CertificateVerification
     runtime_profile: str | None = None
+    component_model: str | None = None
     certificate_id: str | None = None
 
     @property
     def is_valid_shared_runtime(self) -> bool:
-        return self.verification is CertificateVerification.VALID and self.runtime_profile == SHARED_RUNTIME_V1
+        return (
+            self.verification is CertificateVerification.VALID
+            and self.runtime_profile == SHARED_RUNTIME_V1
+            and self.component_model == STATELESS_COMPONENT_MODEL_V1
+        )
 
     @property
     def is_declared(self) -> bool:
@@ -153,6 +158,7 @@ def verify_plugin_archive_certificate(
     verification = verify_archive(archive, key_ring.get)
     envelope = verification.envelope
     runtime_profile = envelope.shared_runtime if envelope is not None else None
+    component_model = envelope.component_model if envelope is not None else None
     # Record the issuer identity only when this instance actually resolved it
     # through the configured ring. When the ring is empty (for example a
     # self-hosted deployment that never configured
@@ -170,6 +176,7 @@ def verify_plugin_archive_certificate(
         certificate=CertificateFacts(
             verification=state,
             runtime_profile=runtime_profile,
+            component_model=component_model,
             certificate_id=certificate_id,
         ),
     )
@@ -188,39 +195,38 @@ def decide_plugin_admission(
     facts: PluginCertificationFacts,
     administrator_force: bool = False,
 ) -> PluginAdmissionDecision:
-    """Apply Cloud fail-closed and OSS administrator-force admission rules."""
+    """Only verified cross-tenant certificates grant shared placement."""
 
     mode = DeploymentMode(deployment)
     certificate = facts.certificate
-    if certificate.is_valid_shared_runtime:
+    if mode is DeploymentMode.CLOUD and certificate.is_valid_shared_runtime:
         return PluginAdmissionDecision(
             AdmissionDisposition.SHARED_ELIGIBLE,
             AdmissionCode.SHARED_ELIGIBLE,
             SHARED_RUNTIME_V1,
         )
 
-    if mode is DeploymentMode.CLOUD:
-        code = (
-            AdmissionCode.CLOUD_CERTIFICATE_REQUIRED
-            if certificate.verification is CertificateVerification.ABSENT
-            else AdmissionCode.CLOUD_CERTIFICATE_INVALID
-        )
-        return PluginAdmissionDecision(AdmissionDisposition.REJECTED, code, DEDICATED_RUNTIME)
-
     if certificate.verification is CertificateVerification.ABSENT:
         return PluginAdmissionDecision(
             AdmissionDisposition.DEDICATED_ALLOWED,
-            AdmissionCode.OSS_LEGACY_DEDICATED,
+            AdmissionCode.UNSIGNED_DEDICATED if mode is DeploymentMode.CLOUD else AdmissionCode.OSS_LEGACY_DEDICATED,
             DEDICATED_RUNTIME,
         )
 
-    if certificate.verification is CertificateVerification.VALID:
+    # OSS has no cross-tenant shared placement; a verified claim stays dedicated
+    # there without inventing a dedicated certification tier.
+    if mode is DeploymentMode.OSS and certificate.is_valid_shared_runtime:
         return PluginAdmissionDecision(
             AdmissionDisposition.DEDICATED_ALLOWED,
-            AdmissionCode.OSS_CERTIFIED_DEDICATED,
+            AdmissionCode.OSS_UNTRUSTED_DEDICATED,
             DEDICATED_RUNTIME,
         )
 
+    if mode is DeploymentMode.CLOUD:
+        return PluginAdmissionDecision(
+            AdmissionDisposition.REJECTED, AdmissionCode.CLOUD_CERTIFICATE_INVALID, DEDICATED_RUNTIME
+        )
+
     # A self-hosted deployment that has not configured the issuer key ring cannot
     # verify a marketplace archive's declaration. Falling back to the dedicated
     # runtime (rather than blocking the install) is not a privilege escalation:
@@ -299,6 +305,7 @@ def execution_mode_for_persisted_installation(
         'artifact_digest': artifact_digest,
         'verification': CertificateVerification.VALID.value,
         'certificate_runtime_profile': SHARED_RUNTIME_V1,
+        'certificate_component_model': STATELESS_COMPONENT_MODEL_V1,
         'runtime_profile': SHARED_RUNTIME_V1,
         'admission_code': AdmissionCode.SHARED_ELIGIBLE.value,
     }
diff --git a/src/langbot/pkg/plugin/connector.py b/src/langbot/pkg/plugin/connector.py
index a7d140539..d0a531355 100644
--- a/src/langbot/pkg/plugin/connector.py
+++ b/src/langbot/pkg/plugin/connector.py
@@ -1,11 +1,10 @@
 # For connect to plugin runtime.
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import asyncio
 import contextlib
 import contextvars
+import datetime
 import hashlib
 import json
 import math
@@ -23,6 +22,7 @@ from langbot_plugin.api.entities.builtin.pipeline.query import provider_session
 
 from ..core import app
 from . import handler
+from . import runtime_ops
 from .errors import (
     PluginRuntimeNotConnectedError,
     PluginInstallationFailedError,
@@ -56,7 +56,9 @@ from langbot_plugin.api.entities.builtin.command import (
     errors as command_errors,
 )
 from langbot_plugin.runtime.plugin.mgr import PluginInstallSource
+from langbot_plugin.runtime.io.handler import SHARED_WORKER_FILE_STORAGE_DIR
 from langbot_plugin.runtime.security import (
+    PLUGIN_FILE_STORAGE_DIR_ENV,
     PLUGIN_RUNTIME_CONTROL_TOKEN_ENV,
     PLUGIN_RUNTIME_CONTROL_TOKEN_HEADER,
     validate_runtime_secret,
@@ -99,6 +101,12 @@ _HEARTBEAT_FAILURE_THRESHOLD = 3
 _RECONNECT_MAX_DELAY_SEC = 60.0
 
 
+def _utc_now_iso() -> str:
+    """Return the current UTC time as a second-precision RFC 3339 string."""
+
+    return datetime.datetime.now(datetime.timezone.utc).isoformat(timespec='seconds').replace('+00:00', 'Z')
+
+
 async def _read_httpx_response_limited(
     response: httpx.Response,
     *,
@@ -225,6 +233,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
         self.runtime_profile: typing.Literal['oss_dev', 'shared'] = (
             'shared' if getattr(getattr(ap, 'deployment', None), 'mode', 'oss') == 'cloud' else 'oss_dev'
         )
+        self._align_plugin_file_transfer_root()
         self.runtime_identity: RuntimeIdentity | None = None
         self._runtime_id = self._build_runtime_id()
         self.worker_policy: PluginWorkerPolicy | None = None
@@ -241,6 +250,32 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
         self._reconnect_task: asyncio.Task | None = None
         self._generation = 0
         self._connected = asyncio.Event()
+        self.runtime_ops_reporter = runtime_ops.RuntimeOpsReporter(ap)
+        self._reconcile_summary: dict[str, Any] = {
+            'last_started_at': None,
+            'last_duration_ms': None,
+            'last_ok': None,
+            'failed_installations': 0,
+            'missing_artifacts': 0,
+        }
+
+    def _align_plugin_file_transfer_root(self) -> None:
+        """Use the Runtime's plugin file-transfer root when running shared.
+
+        The SDK binds its transfer root to one directory inode with mode 0700 and
+        picks the default per process: a shared worker uses
+        ``SHARED_WORKER_FILE_STORAGE_DIR``, every other process falls back to
+        ``data/temp/lbp``. A host talking to a shared Runtime must use the same root,
+        otherwise plugin icons and assets die with "Invalid file transfer
+        capability" (the icon route then answers 500/404 instead of the image).
+        An explicit operator value always wins.
+        """
+
+        if self.runtime_profile != 'shared':
+            return
+        if os.environ.get(PLUGIN_FILE_STORAGE_DIR_ENV):
+            return
+        os.environ[PLUGIN_FILE_STORAGE_DIR_ENV] = SHARED_WORKER_FILE_STORAGE_DIR
 
     @staticmethod
     def _build_runtime_id() -> str:
@@ -299,6 +334,24 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
         ):
             if field_name in PluginWorkerPolicy.model_fields and field_name in worker:
                 policy_data[field_name] = worker.get(field_name)
+        if self.runtime_profile == 'shared':
+            for mode in ('dedicated', 'shared'):
+                overrides = worker.get(mode)
+                if isinstance(overrides, dict):
+                    overrides = dict(overrides)
+                    for name, value in overrides.items():
+                        # Optional leaves start as null, so env loading preserves
+                        # strings. Parse only these trusted numeric config fields.
+                        if isinstance(value, str) and name in (
+                            'max_cpus',
+                            'max_memory_mb',
+                            'max_pids',
+                            'max_open_files',
+                            'max_file_size_mb',
+                        ):
+                            overrides[name] = float(value) if name == 'max_cpus' else int(value)
+                if overrides is not None:
+                    policy_data[mode] = overrides
         return PluginWorkerPolicy.model_validate(policy_data)
 
     def _control_headers(self, *, allow_generate: bool) -> dict[str, str]:
@@ -345,8 +398,9 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             artifact_digest=setting.artifact_digest,
         )
 
-    @staticmethod
-    def _execution_mode_from_setting(setting: persistence_plugin.PluginSetting) -> PluginExecutionMode:
+    def _execution_mode_from_setting(self, setting: persistence_plugin.PluginSetting) -> PluginExecutionMode:
+        if getattr(getattr(self.ap, 'deployment', None), 'mode', 'oss') != 'cloud':
+            return PluginExecutionMode.DEDICATED
         return execution_mode_for_persisted_installation(
             artifact_digest=setting.artifact_digest,
             install_info=setting.install_info,
@@ -619,6 +673,10 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             'installation_uuid': installation_uuid,
             'error_code': error_code,
             'message': message,
+            # When the failure actually happened. The ops reporter must publish this
+            # instead of its own sampling time, otherwise every sample re-stamps the
+            # same failure with the newest report time.
+            'failed_at': _utc_now_iso(),
         }
         self._installation_failures[installation_uuid] = failure
         self.ap.logger.error(
@@ -660,9 +718,15 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                 'installation_uuid': installation_uuid,
                 'error_code': error_code,
                 'message': message,
+                # See _raise_apply_failure: never let the reporter substitute its own
+                # sampling time for the moment the failure was observed.
+                'failed_at': _utc_now_iso(),
             }
             failures[installation_uuid] = failure
-            if self._installation_failures.get(installation_uuid) != failure:
+            previous = self._installation_failures.get(installation_uuid) or {}
+            if {key: value for key, value in previous.items() if key != 'failed_at'} != {
+                key: value for key, value in failure.items() if key != 'failed_at'
+            }:
                 self.ap.logger.error(
                     'Plugin installation %s failed during reconcile [%s]: %s',
                     installation_uuid,
@@ -675,6 +739,21 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                 self._installation_failures.pop(installation_uuid, None)
         self._installation_failures.update(failures)
 
+    def _record_reconcile_outcome(self, started_at: str, duration_ms: float, result: dict[str, Any]) -> None:
+        """Expose the last reconcile outcome for the runtime ops sample."""
+
+        failed = result.get('failed_installations')
+        missing = result.get('missing_artifacts')
+        failed_installations = len(failed) if isinstance(failed, list) else 0
+        missing_artifacts = len(missing) if isinstance(missing, list) else 0
+        self._reconcile_summary = {
+            'last_started_at': started_at,
+            'last_duration_ms': int(max(duration_ms, 0.0)),
+            'last_ok': failed_installations == 0 and missing_artifacts == 0,
+            'failed_installations': failed_installations,
+            'missing_artifacts': missing_artifacts,
+        }
+
     async def _repair_reconcile_missing_artifacts(
         self,
         desired_states: dict[str, PluginInstallationDesiredState],
@@ -719,7 +798,6 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                         }
                     )
 
-    @diagnostics.observe('lifecycle', 'runtime.prepare_connected_runtime', source='runtime', stage='execute')
     async def _prepare_connected_runtime(self) -> None:
         """Handshake follow-up: pin OSS compatibility, then replay authority."""
 
@@ -756,12 +834,19 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
         reconcile_timeout_seconds = max(
             300.0, self._runtime_connect_timeout(self.ap.instance_config.data.get('plugin', {}))
         )
+        reconcile_started_at = _utc_now_iso()
+        reconcile_started_monotonic = time.monotonic()
         result = await runtime_handler.reconcile_plugin_installations(
             tuple(self._known_desired_states.values()),
             timeout=reconcile_timeout_seconds,
         )
         await self._repair_reconcile_missing_artifacts(self._known_desired_states, result)
         self._record_reconcile_failures(self._known_desired_states, result)
+        self._record_reconcile_outcome(
+            reconcile_started_at,
+            (time.monotonic() - reconcile_started_monotonic) * 1000.0,
+            result,
+        )
 
     async def reconcile_projected_workspaces(
         self,
@@ -777,6 +862,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
 
         runtime_handler = self._runtime_handler()
         started_at = time.monotonic()
+        started_at_iso = _utc_now_iso()
         async with self._state_lock:
             all_states: dict[str, PluginInstallationDesiredState] = {}
             workspace_installations: dict[str, set[str]] = {}
@@ -803,6 +889,11 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             )
             await self._repair_reconcile_missing_artifacts(all_states, result)
             self._record_reconcile_failures(all_states, result)
+            self._record_reconcile_outcome(
+                started_at_iso,
+                (time.monotonic() - started_at) * 1000.0,
+                result,
+            )
             for installation_uuid, previous in tuple(self._known_desired_states.items()):
                 if installation_uuid not in all_states:
                     runtime_handler.unregister_installation_binding(previous.binding)
@@ -905,7 +996,6 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                     self.schedule_reconnect()
                     failures = 0
 
-    @diagnostics.observe('lifecycle', 'runtime.initialize', source='runtime', stage='execute')
     async def initialize(self):
         if not self.is_enable_plugin:
             self.ap.logger.info('Plugin system is disabled.')
@@ -1071,6 +1161,8 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             if self.heartbeat_task is None or self.heartbeat_task.done():
                 self.heartbeat_task = asyncio.create_task(self.heartbeat_loop())
 
+            self.runtime_ops_reporter.start()
+
     def schedule_reconnect(self) -> None:
         if self._closing or not self.is_enable_plugin:
             return
@@ -1133,6 +1225,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             await asyncio.gather(self.heartbeat_task, return_exceptions=True)
             self.heartbeat_task = None
         await self._stop_transport()
+        await self.runtime_ops_reporter.stop()
         await self._close_managed_subprocess()
 
     @staticmethod
@@ -1809,13 +1902,13 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             'normalized_digest': facts.artifact_digest,
             'verification': facts.certificate.verification.value,
             'certificate_runtime_profile': facts.certificate.runtime_profile,
+            'certificate_component_model': facts.certificate.component_model,
             'certificate_id': facts.certificate.certificate_id,
             'runtime_profile': decision.runtime_profile,
             'admission_code': decision.code.value,
         }
         return {**install_info, '_certification': certification_info}, verified
 
-    @diagnostics.observe('lifecycle', 'runtime.install_plugin', source='runtime', stage='execute')
     async def install_plugin(
         self,
         install_source: PluginInstallSource,
@@ -1982,7 +2075,6 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
             task_context.set_current_action('plugin updated' if operation == 'upgrade' else 'plugin installed')
             task_context.metadata['progress_percent'] = 100
 
-    @diagnostics.observe('lifecycle', 'runtime.upgrade_plugin', source='runtime', stage='execute')
     async def upgrade_plugin(
         self,
         plugin_author: str,
@@ -2009,7 +2101,6 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
         )
         return {}
 
-    @diagnostics.observe('lifecycle', 'runtime.delete_plugin', source='runtime', stage='execute')
     async def delete_plugin(
         self,
         plugin_author: str,
diff --git a/src/langbot/pkg/plugin/handler.py b/src/langbot/pkg/plugin/handler.py
index 9df92e778..12fade67b 100644
--- a/src/langbot/pkg/plugin/handler.py
+++ b/src/langbot/pkg/plugin/handler.py
@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from ..telemetry import diagnostics
+from contextlib import aclosing
 
 import asyncio
 import inspect
@@ -443,13 +443,19 @@ def _resolve_remove_think(data: dict[str, Any], query: Any | None) -> bool:
     return False
 
 
-def _merge_model_extra_args(model: Any, call_extra_args: Any) -> dict[str, Any]:
+def _merge_model_extra_args(model: Any, call_extra_args: Any, *, llm_policy: bool = False) -> dict[str, Any]:
     """Merge persisted model extra_args with action-level overrides."""
     merged: dict[str, Any] = {}
 
     model_extra_args = getattr(getattr(model, 'model_entity', None), 'extra_args', None)
     if isinstance(model_extra_args, dict):
         merged.update(model_extra_args)
+    if llm_policy:
+        # RuntimeProvider resolves model and call policy layers independently.
+        # Flattening them here makes a model request_policy.retries override a
+        # caller's legacy num_retries=0 (and similarly hides timeout overrides).
+        for key in ('request_policy', 'num_retries', 'max_retries', 'timeout'):
+            merged.pop(key, None)
     if isinstance(call_extra_args, dict):
         merged.update(call_extra_args)
 
@@ -646,8 +652,6 @@ class RuntimeConnectionHandler(handler.Handler):
         payload fields can never select the scope. A transaction-free boundary
         avoids reserving one pooled connection across provider and network waits.
         """
-
-        diagnostics.annotate(workspace_uuid=action_context.workspace_uuid)
         persistence_mgr = getattr(self.ap, 'persistence_mgr', None)
         if persistence_mgr is None:
             yield
@@ -700,9 +704,7 @@ class RuntimeConnectionHandler(handler.Handler):
                             async for response in responses:
                                 yield response
 
-                self.actions[action_name] = diagnostics.observe(
-                    'api', 'host.' + action_name, source='plugin', ap=self.ap
-                )(secured_stream_action)
+                self.actions[action_name] = secured_stream_action
                 continue
 
             async def secured_action(
@@ -731,9 +733,7 @@ class RuntimeConnectionHandler(handler.Handler):
                         response = await response
                     return response
 
-            self.actions[action_name] = diagnostics.observe('api', 'host.' + action_name, source='plugin', ap=self.ap)(
-                secured_action
-            )
+            self.actions[action_name] = secured_action
 
     async def _get_plugin_setting(
         self,
@@ -1490,7 +1490,9 @@ class RuntimeConnectionHandler(handler.Handler):
 
             funcs_obj = [resource_tool.LLMTool.model_validate({**func, 'func': _placeholder_func}) for func in funcs]
             query = _resolve_action_query(data, session, self.ap, action_context)
-            effective_extra_args = _merge_model_extra_args(llm_model, extra_args)
+            if query is None:
+                query = _build_plugin_action_query(action_context)
+            effective_extra_args = _merge_model_extra_args(llm_model, extra_args, llm_policy=True)
             remove_think = _resolve_remove_think(data, query)
             effective_funcs = funcs_obj if 'func_call' in (llm_model.model_entity.abilities or []) else []
 
@@ -1586,26 +1588,31 @@ class RuntimeConnectionHandler(handler.Handler):
 
             funcs_obj = [resource_tool.LLMTool.model_validate({**func, 'func': _placeholder_func}) for func in funcs]
             query = _resolve_action_query(data, session, self.ap, action_context)
-            effective_extra_args = _merge_model_extra_args(llm_model, extra_args)
+            if query is None:
+                query = _build_plugin_action_query(action_context)
+            effective_extra_args = _merge_model_extra_args(llm_model, extra_args, llm_policy=True)
             remove_think = _resolve_remove_think(data, query)
             effective_funcs = funcs_obj if 'func_call' in (llm_model.model_entity.abilities or []) else []
 
-            async for chunk in llm_model.provider.invoke_llm_stream(
-                query=query,
-                model=llm_model,
-                messages=messages_obj,
-                funcs=effective_funcs,
-                extra_args=effective_extra_args,
-                remove_think=remove_think,
-                execution_context=execution_context,
-            ):
-                if chunk is None:
-                    continue
-                yield handler.ActionResponse.success(
-                    data={
-                        'chunk': chunk.model_dump(),
-                    },
+            async with aclosing(
+                llm_model.provider.invoke_llm_stream(
+                    query=query,
+                    model=llm_model,
+                    messages=messages_obj,
+                    funcs=effective_funcs,
+                    extra_args=effective_extra_args,
+                    remove_think=remove_think,
+                    execution_context=execution_context,
                 )
+            ) as owned_stream:
+                async for chunk in owned_stream:
+                    if chunk is None:
+                        continue
+                    yield handler.ActionResponse.success(
+                        data={
+                            'chunk': chunk.model_dump(),
+                        },
+                    )
             usage = _pop_query_llm_usage(query)
             if usage is not None:
                 yield handler.ActionResponse.success(
@@ -2808,7 +2815,7 @@ class RuntimeConnectionHandler(handler.Handler):
             return await self.call_action(
                 LangBotToRuntimeAction.APPLY_PLUGIN_INSTALLATION,
                 request.model_dump(exclude_none=True),
-                timeout=120,
+                timeout=600,  # dependency install can be slow
             )
 
     async def remove_plugin_installation(
@@ -2832,7 +2839,7 @@ class RuntimeConnectionHandler(handler.Handler):
                 'install_source': install_source,
                 'install_info': install_info,
             },
-            timeout=120,
+            timeout=600,  # dependency install can be slow
         )
 
         async for ret in gen:
@@ -2846,7 +2853,7 @@ class RuntimeConnectionHandler(handler.Handler):
                 'plugin_author': plugin_author,
                 'plugin_name': plugin_name,
             },
-            timeout=120,
+            timeout=600,  # dependency install can be slow
         )
 
         async for ret in gen:
diff --git a/src/langbot/pkg/plugin/runtime_ops.py b/src/langbot/pkg/plugin/runtime_ops.py
new file mode 100644
index 000000000..0c96820a0
--- /dev/null
+++ b/src/langbot/pkg/plugin/runtime_ops.py
@@ -0,0 +1,390 @@
+"""Cloud Edition runtime ops sample reporter.
+
+Aggregates this instance's plugin Runtime health (probed from the Runtime
+``/healthz`` endpoint) together with the connector's own desired-state and
+failure registries, then pushes one compact sample to the Space control
+plane where the internal ops dashboard stores and renders it.
+
+This is a Cloud-only ops channel, deliberately separate from product
+telemetry so ``space.disable_telemetry`` cannot blind operations. Sending is
+best-effort: a failed probe or upload never raises out of the reporter.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import contextlib
+import datetime
+import math
+import os
+import typing
+from urllib.parse import urlsplit, urlunsplit
+
+import httpx
+
+from ..utils import constants, httpclient
+
+_CONTROL_PLANE_TOKEN_ENV = 'LANGBOT_SPACE_CONTROL_PLANE_TOKEN'
+_RUNTIME_OPS_PATH = '/api/v1/cloud/v2/internal/runtime-ops'
+_DEFAULT_INTERVAL_SECONDS = 30.0
+_DEFAULT_TIMEOUT_SECONDS = 10.0
+_DEFAULT_MAX_FAILURES = 5
+_MAX_FAILURE_MESSAGE_CHARS = 512
+
+
+def _utc_now_iso() -> str:
+    """Return the current UTC time as a second-precision RFC 3339 string."""
+
+    return datetime.datetime.now(datetime.timezone.utc).isoformat(timespec='seconds').replace('+00:00', 'Z')
+
+
+def _coerce_bool(value: typing.Any, default: bool) -> bool:
+    return value if isinstance(value, bool) else default
+
+
+def _coerce_positive_float(value: typing.Any, default: float) -> float:
+    if isinstance(value, bool) or not isinstance(value, (int, float)):
+        return default
+    value = float(value)
+    if not math.isfinite(value) or value <= 0:
+        return default
+    return value
+
+
+def _coerce_positive_int(value: typing.Any, default: int) -> int:
+    if isinstance(value, bool) or not isinstance(value, int) or value <= 0:
+        return default
+    return value
+
+
+def _sanitize_failure_message(value: typing.Any) -> str:
+    """Flatten and bound one failure message for the wire.
+
+    Mirrors the connector's own sanitization so ops never receives control
+    characters, unbounded payloads, or newline-injected log content.
+    """
+
+    candidate = str(value or '').replace('\r', ' ').replace('\n', ' ').strip()
+    return candidate[:_MAX_FAILURE_MESSAGE_CHARS]
+
+
+class RuntimeOpsReporter:
+    """Collect and push one runtime ops sample per interval.
+
+    The reporter never raises out of its send loop; failures are counted and
+    surfaced through :meth:`get_stats`.
+    """
+
+    def __init__(self, ap: typing.Any):
+        self.ap = ap
+        instance_data = getattr(getattr(ap, 'instance_config', None), 'data', None)
+        if not isinstance(instance_data, dict):
+            instance_data = {}
+        plugin_config = instance_data.get('plugin', {})
+        if not isinstance(plugin_config, dict):
+            plugin_config = {}
+        config = plugin_config.get('runtime_ops', {})
+        if not isinstance(config, dict):
+            config = {}
+        self.config = config
+
+        deployment_mode = getattr(getattr(ap, 'deployment', None), 'mode', 'oss')
+        cloud_edition = deployment_mode == 'cloud'
+        # Cloud-only feature: an explicit flag may turn it off, but a
+        # non-cloud deployment is never enabled.
+        self.enabled = cloud_edition and _coerce_bool(config.get('enabled'), True)
+        self.interval_seconds = _coerce_positive_float(config.get('interval_seconds'), _DEFAULT_INTERVAL_SECONDS)
+        self.timeout_seconds = _coerce_positive_float(config.get('timeout_seconds'), _DEFAULT_TIMEOUT_SECONDS)
+        self.max_failures = _coerce_positive_int(config.get('max_failures'), _DEFAULT_MAX_FAILURES)
+
+        self.health_url = self._derive_health_url(plugin_config.get('runtime_ws_url'))
+        self.space_url = str(instance_data.get('space', {}).get('url', '') or '').rstrip('/')
+        self._token = str(os.environ.get(_CONTROL_PLANE_TOKEN_ENV, '') or '').strip()
+
+        if self.enabled and not self.health_url:
+            self._log_info('Plugin runtime ops reporter is disabled: no plugin runtime WebSocket URL is configured')
+            self.enabled = False
+        if self.enabled and not self.space_url:
+            self._log_info('Plugin runtime ops reporter is disabled: no Space URL is configured')
+            self.enabled = False
+
+        self._task: asyncio.Task | None = None
+        self._client: httpx.AsyncClient | None = None
+        self._sent_total = 0
+        self._dropped_total = 0
+        self._last_error: str | None = None
+
+    def _log_info(self, message: str) -> None:
+        logger = getattr(self.ap, 'logger', None)
+        info = getattr(logger, 'info', None)
+        if callable(info):
+            info(message)
+
+    @staticmethod
+    def _derive_health_url(runtime_ws_url: typing.Any) -> str | None:
+        """Derive the Runtime ``/healthz`` URL from the control WebSocket URL."""
+
+        if not isinstance(runtime_ws_url, str):
+            return None
+        candidate = runtime_ws_url.strip()
+        if not candidate:
+            return None
+        try:
+            parts = urlsplit(candidate)
+            port = parts.port
+        except ValueError:
+            return None
+        if parts.scheme not in ('ws', 'wss') or not parts.hostname:
+            return None
+        scheme = 'https' if parts.scheme == 'wss' else 'http'
+        netloc = parts.hostname if port is None else f'{parts.hostname}:{port}'
+        return urlunsplit((scheme, netloc, '/healthz', '', ''))
+
+    def start(self) -> None:
+        """Start the background sample loop once the control link is up."""
+
+        if not self.enabled:
+            return
+        if self._task is not None and not self._task.done():
+            return
+        self._task = asyncio.create_task(self._run())
+
+    async def stop(self) -> None:
+        """Cancel the sample loop and release the HTTP client."""
+
+        task = self._task
+        self._task = None
+        if task is not None and task is not asyncio.current_task():
+            task.cancel()
+            await asyncio.gather(task, return_exceptions=True)
+        client = self._client
+        self._client = None
+        if client is not None and not client.is_closed:
+            with contextlib.suppress(Exception):
+                await client.aclose()
+
+    def get_stats(self) -> dict[str, typing.Any]:
+        """Return in-memory counters for the instance liveness surface."""
+
+        return {
+            'sent_total': self._sent_total,
+            'dropped_total': self._dropped_total,
+            'last_error': self._last_error,
+        }
+
+    @contextlib.asynccontextmanager
+    async def _client_context(self):
+        if self._client is None or self._client.is_closed:
+            self._client = httpx.AsyncClient(
+                timeout=httpx.Timeout(self.timeout_seconds),
+                event_hooks=httpclient.httpx_response_limit_hooks(),
+            )
+        yield self._client
+
+    async def _run(self) -> None:
+        while True:
+            try:
+                await self._collect_and_send()
+            except asyncio.CancelledError:
+                raise
+            except Exception as exc:  # never let the ops channel crash the connector
+                self._last_error = _sanitize_failure_message(f'sample cycle failed: {exc}')
+            await asyncio.sleep(self.interval_seconds)
+
+    async def _collect_and_send(self) -> None:
+        health = await self._fetch_health()
+        if health is None:
+            self._dropped_total += 1
+            return
+        await self._post(self.build_payload(health))
+
+    async def _fetch_health(self) -> dict[str, typing.Any] | None:
+        """Probe the Runtime ``/healthz`` endpoint, returning its verbatim JSON."""
+
+        try:
+            async with self._client_context() as client:
+                response = await asyncio.wait_for(
+                    client.get(self.health_url),
+                    timeout=self.timeout_seconds,
+                )
+            if response.status_code >= 400:
+                self._last_error = f'health probe returned status {response.status_code}'
+                return None
+            payload = await httpclient.parse_json_response(response)
+            if not isinstance(payload, dict):
+                self._last_error = 'health probe returned a non-object payload'
+                return None
+            self._last_error = None
+            return payload
+        except asyncio.TimeoutError:
+            self._last_error = 'health probe timed out'
+            return None
+        except Exception as exc:
+            self._last_error = _sanitize_failure_message(f'health probe failed: {exc}')
+            return None
+
+    async def _post(self, payload: dict[str, typing.Any]) -> None:
+        """Best-effort POST of one sample; never raises."""
+
+        url = f'{self.space_url}{_RUNTIME_OPS_PATH}'
+        headers = {'Authorization': f'Bearer {self._token}'}
+        try:
+            async with self._client_context() as client:
+                response = await asyncio.wait_for(
+                    client.post(url, json=payload, headers=headers),
+                    timeout=self.timeout_seconds,
+                )
+            if response.status_code >= 400:
+                body = await httpclient.response_text(response, max_chars=200)
+                self._dropped_total += 1
+                self._last_error = f'POST {url} returned status {response.status_code} - {body}'
+                return
+            self._sent_total += 1
+            self._last_error = None
+        except asyncio.TimeoutError:
+            self._dropped_total += 1
+            self._last_error = f'POST {url} timed out'
+        except Exception as exc:
+            self._dropped_total += 1
+            self._last_error = _sanitize_failure_message(f'POST {url} failed: {exc}')
+
+    def build_payload(self, health: dict[str, typing.Any]) -> dict[str, typing.Any]:
+        """Assemble one ops sample in the frozen Core -> Space contract shape."""
+
+        connector = getattr(self.ap, 'plugin_connector', None)
+        instance_data = getattr(getattr(self.ap, 'instance_config', None), 'data', None)
+        if not isinstance(instance_data, dict):
+            instance_data = {}
+        system_config = instance_data.get('system', {})
+        if not isinstance(system_config, dict):
+            system_config = {}
+        instance_uuid = str(system_config.get('instance_id', '') or '') or str(constants.instance_id or '')
+        edition = str(system_config.get('edition', 'oss') or 'oss')
+        return {
+            'event': 'runtime_ops_sample',
+            'kind': 'aggregate',
+            'instance_uuid': instance_uuid,
+            'generated_at': _utc_now_iso(),
+            'core': {
+                'version': constants.semantic_version,
+                'edition': edition,
+                'connected': self._runtime_connected(connector),
+                'runtime_profile': getattr(connector, 'runtime_profile', 'oss_dev'),
+                'worker_policy': self._worker_policy(connector),
+                'desired_states': self._desired_states(connector),
+                'reconcile': self._reconcile_summary(connector),
+            },
+            'health': health,
+            'failures': self._failures(connector),
+        }
+
+    @staticmethod
+    def _runtime_connected(connector: typing.Any) -> bool:
+        available = getattr(connector, '_runtime_available', None)
+        if not callable(available):
+            return False
+        try:
+            return bool(available())
+        except Exception:
+            return False
+
+    @staticmethod
+    def _worker_policy(connector: typing.Any) -> dict[str, typing.Any] | None:
+        policy = getattr(connector, 'worker_policy', None)
+        if policy is None:
+            return None
+        model_dump = getattr(policy, 'model_dump', None)
+        if callable(model_dump):
+            return model_dump()
+        if isinstance(policy, dict):
+            return dict(policy)
+        return None
+
+    @staticmethod
+    def _desired_states(connector: typing.Any) -> dict[str, typing.Any]:
+        states = getattr(connector, '_known_desired_states', None)
+        if not isinstance(states, dict):
+            states = {}
+        enabled = 0
+        by_mode: dict[str, int] = {}
+        for state in states.values():
+            if bool(getattr(state, 'enabled', False)):
+                enabled += 1
+            mode = getattr(getattr(state, 'execution_mode', None), 'value', None)
+            mode = str(mode) if isinstance(mode, str) and mode else 'dedicated'
+            by_mode[mode] = by_mode.get(mode, 0) + 1
+        return {'total': len(states), 'enabled': enabled, 'by_mode': by_mode}
+
+    @staticmethod
+    def _reconcile_summary(connector: typing.Any) -> dict[str, typing.Any]:
+        summary = getattr(connector, '_reconcile_summary', None)
+        if isinstance(summary, dict):
+            return dict(summary)
+        return {
+            'last_started_at': None,
+            'last_duration_ms': None,
+            'last_ok': None,
+            'failed_installations': 0,
+            'missing_artifacts': 0,
+        }
+
+    def _failures(self, connector: typing.Any) -> list[dict[str, typing.Any]]:
+        records = getattr(connector, '_installation_failures', None)
+        if not isinstance(records, dict):
+            return []
+        states = getattr(connector, '_known_desired_states', None)
+        if not isinstance(states, dict):
+            states = {}
+        identities = self._plugin_identities(connector)
+        observed_at = _utc_now_iso()
+        failures: list[dict[str, typing.Any]] = []
+        for installation_uuid in list(records)[: self.max_failures]:
+            record = records.get(installation_uuid)
+            if not isinstance(record, dict):
+                continue
+            state = states.get(installation_uuid)
+            binding = getattr(state, 'binding', None)
+            mode = getattr(getattr(state, 'execution_mode', None), 'value', None)
+            mode = str(mode) if isinstance(mode, str) and mode else 'dedicated'
+            plugin_author, plugin_name = identities.get(str(installation_uuid), ('', ''))
+            failures.append(
+                {
+                    'workspace_uuid': str(getattr(binding, 'workspace_uuid', '') or ''),
+                    'installation_uuid': str(installation_uuid),
+                    'plugin_author': plugin_author,
+                    'plugin_name': plugin_name,
+                    'execution_mode': mode,
+                    'error_code': str(record.get('error_code', '') or ''),
+                    'message': _sanitize_failure_message(record.get('message')),
+                    # The moment the failure was observed by the connector. Falling back to
+                    # the sampling time would make a stale failure look freshly broken on
+                    # every report.
+                    'observed_at': str(record.get('failed_at') or observed_at),
+                }
+            )
+        return failures
+
+    @staticmethod
+    def _plugin_identities(connector: typing.Any) -> dict[str, tuple[str, str]]:
+        """Read the connector handler's installation identity registry.
+
+        The Runtime connection handler binds every desired installation to
+        its plugin identity; the connector's failure records themselves only
+        carry the installation UUID, so the reporter joins them here.
+        """
+
+        handler = getattr(connector, 'handler', None)
+        bindings = getattr(handler, '_installation_bindings', None)
+        if not isinstance(bindings, dict):
+            return {}
+        identities: dict[str, tuple[str, str]] = {}
+        for installation_uuid, entry in bindings.items():
+            try:
+                identity = entry[1]
+            except (TypeError, IndexError, KeyError):
+                continue
+            identities[str(installation_uuid)] = (
+                str(getattr(identity, 'plugin_author', '') or ''),
+                str(getattr(identity, 'plugin_name', '') or ''),
+            )
+        return identities
diff --git a/src/langbot/pkg/provider/modelmgr/codex_auth.py b/src/langbot/pkg/provider/modelmgr/codex_auth.py
index 571a67b40..1f6b618ec 100644
--- a/src/langbot/pkg/provider/modelmgr/codex_auth.py
+++ b/src/langbot/pkg/provider/modelmgr/codex_auth.py
@@ -23,6 +23,7 @@ from ...api.http.context import PrincipalType, RequestContext
 from ...api.http.authz import Permission, has_permission
 from ...api.http.service.tenant import require_workspace_uuid
 from ...workspace.errors import WorkspaceNotFoundError
+from ...i18n import t
 
 REQUESTER = 'openai-codex'
 BASE_URL = 'https://chatgpt.com/backend-api/codex'
@@ -77,7 +78,7 @@ def _tokens(data: dict, previous: dict | None = None) -> dict:
         or not math.isfinite(expires_at)
         or expires_at <= time.time()
     ):
-        raise ValueError('ChatGPT returned an incomplete authorization. Please sign in again.')
+        raise ValueError(t('ChatGPT returned an incomplete authorization. Please sign in again.'))
     return {
         'access_token': access,
         'refresh_token': refresh,
@@ -99,7 +100,7 @@ class CodexAuth:
         try:
             return await self.ap.persistence_mgr.execute_async(statement)
         except Exception:
-            raise ValueError('ChatGPT credential storage failed. Please retry.') from None
+            raise ValueError(t('ChatGPT credential storage failed. Please retry.')) from None
 
     async def _read(self, workspace: str, provider: str) -> dict | None:
         result = await self._execute(sa.select(CodexCredential).where(*self._where(workspace, provider)))
@@ -107,7 +108,7 @@ class CodexAuth:
             row = result.first()
             return dict(row._mapping) if row is not None else None
         except Exception:
-            raise ValueError('ChatGPT credential storage failed. Please retry.') from None
+            raise ValueError(t('ChatGPT credential storage failed. Please retry.')) from None
 
     async def _provider(self, context, provider: str, *, user: bool = False) -> str:
         workspace = require_workspace_uuid(context)
@@ -117,7 +118,7 @@ class CodexAuth:
             or not context.account_uuid
             or not has_permission(context, Permission.PROVIDER_SECRET_MANAGE)
         ):
-            raise ValueError('ChatGPT authorization requires an authorized workspace user')
+            raise ValueError(t('ChatGPT authorization requires an authorized workspace user'))
         result = await self._execute(
             sa.select(ModelProvider.requester).where(
                 ModelProvider.workspace_uuid == workspace, ModelProvider.uuid == provider
@@ -147,9 +148,9 @@ class CodexAuth:
             if result.rowcount == 1:
                 break
             if await self._read(workspace, provider) is None:
-                raise ValueError(LOGIN_REQUIRED)
+                raise ValueError(t(LOGIN_REQUIRED))
             if time.monotonic() >= deadline:
-                raise ValueError('ChatGPT authorization is busy. Please retry shortly.')
+                raise ValueError(t('ChatGPT authorization is busy. Please retry shortly.'))
             await asyncio.sleep(0.1)
         try:
             yield owner
@@ -171,7 +172,7 @@ class CodexAuth:
             .values(payload=payload, version=CodexCredential.version + 1)
         )
         if result.rowcount != 1:
-            raise ValueError('ChatGPT authorization was cancelled or replaced. Please retry.')
+            raise ValueError(t('ChatGPT authorization was cancelled or replaced. Please retry.'))
 
     async def _post(self, path: str, *, data=None, json_body=None) -> httpx.Response:
         try:
@@ -186,7 +187,7 @@ class CodexAuth:
                     25,
                 )
         except (httpx.HTTPError, TimeoutError):
-            raise ValueError('ChatGPT authorization network error. Please retry.') from None
+            raise ValueError(t('ChatGPT authorization network error. Please retry.')) from None
 
     @staticmethod
     def _json(response: httpx.Response) -> dict:
@@ -196,7 +197,7 @@ class CodexAuth:
                 raise ValueError
             return value
         except ValueError:
-            raise ValueError('ChatGPT returned an invalid authorization response') from None
+            raise ValueError(t('ChatGPT returned an invalid authorization response')) from None
 
     async def status(self, context, provider: str) -> dict:
         workspace = await self._provider(context, provider, user=True)
@@ -215,7 +216,9 @@ class CodexAuth:
         async with self._lease(workspace, provider) as owner:
             response = await self._post('/api/accounts/deviceauth/usercode', json_body={'client_id': CLIENT_ID})
             if response.status_code != 200:
-                raise ValueError('Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.')
+                raise ValueError(
+                    t('Unable to start ChatGPT device login. Enable device code login in ChatGPT settings.')
+                )
             data = self._json(response)
             try:
                 code = data.get('user_code') or data['usercode']
@@ -224,7 +227,7 @@ class CodexAuth:
                 if not isinstance(code, str) or not isinstance(device, str) or not code or not device:
                     raise ValueError
             except (KeyError, ValueError, TypeError):
-                raise ValueError('ChatGPT returned an invalid device code') from None
+                raise ValueError(t('ChatGPT returned an invalid device code')) from None
             now = time.time()
             try:
                 expiry = data.get('expires_at')
@@ -241,7 +244,7 @@ class CodexAuth:
                     raise ValueError
                 expires_at = min(now + 900, expires_at)
             except (ValueError, TypeError):
-                raise ValueError('ChatGPT returned an invalid device code expiry') from None
+                raise ValueError(t('ChatGPT returned an invalid device code expiry')) from None
             pending = {
                 'authorization_id': secrets.token_urlsafe(32),
                 'user_code': code,
@@ -306,12 +309,12 @@ class CodexAuth:
             if response.status_code != 200:
                 payload.pop('pending', None)
                 await self._save(workspace, provider, owner, payload)
-                raise ValueError('ChatGPT device authorization failed. Please start again.')
+                raise ValueError(t('ChatGPT device authorization failed. Please start again.'))
             data = self._json(response)
             if not data.get('authorization_code') or not data.get('code_verifier'):
                 payload.pop('pending', None)
                 await self._save(workspace, provider, owner, payload)
-                raise ValueError('ChatGPT returned an incomplete device authorization')
+                raise ValueError(t('ChatGPT returned an incomplete device authorization'))
             # Keep an attempt tombstone so cancel can preempt exchange, but never replay a code.
             pending['consumed'] = True
             await self._save(workspace, provider, owner, payload)
@@ -326,7 +329,7 @@ class CodexAuth:
                 },
             )
             if response.status_code != 200:
-                raise ValueError('ChatGPT token exchange failed. Please start sign-in again.')
+                raise ValueError(t('ChatGPT token exchange failed. Please start sign-in again.'))
             tokens = _tokens(self._json(response))
             await self._save(
                 workspace,
@@ -384,7 +387,7 @@ class CodexAuth:
         payload = row['payload'] if row else {}
         tokens = payload.get('tokens')
         if not tokens or payload.get('invalid'):
-            raise ValueError(LOGIN_REQUIRED)
+            raise ValueError(t(LOGIN_REQUIRED))
         if tokens['expires_at'] > time.time() + 120 and tokens['access_token'] != rejected_token:
             return tokens
         async with self._lease(workspace, provider, refresh=True) as owner:
@@ -392,7 +395,7 @@ class CodexAuth:
             payload = dict(row['payload'])
             tokens = payload.get('tokens')
             if not tokens or payload.get('invalid'):
-                raise ValueError(LOGIN_REQUIRED)
+                raise ValueError(t(LOGIN_REQUIRED))
             if tokens['expires_at'] > time.time() + 120 and tokens['access_token'] != rejected_token:
                 return tokens
             response = await self._post(
@@ -411,9 +414,9 @@ class CodexAuth:
                 payload.pop('tokens', None)
                 payload.pop('completed', None)
                 await self._save(workspace, provider, owner, payload)
-                raise ValueError(LOGIN_REQUIRED)
+                raise ValueError(t(LOGIN_REQUIRED))
             if response.status_code != 200:
-                raise ValueError('ChatGPT token refresh temporarily failed. Please retry.')
+                raise ValueError(t('ChatGPT token refresh temporarily failed. Please retry.'))
             refreshed = _tokens(self._json(response), tokens)
             payload['tokens'] = refreshed
             await self._save(workspace, provider, owner, payload)
diff --git a/src/langbot/pkg/provider/modelmgr/codex_errors.py b/src/langbot/pkg/provider/modelmgr/codex_errors.py
index a7ceb0301..78141afb7 100644
--- a/src/langbot/pkg/provider/modelmgr/codex_errors.py
+++ b/src/langbot/pkg/provider/modelmgr/codex_errors.py
@@ -1,10 +1,20 @@
 """Explicitly safe Codex failures; never construct messages from upstream bodies."""
 
+from ...i18n import t
+from .errors import RequesterError
+
 
 class CodexProviderError(ValueError):
     """A known provider failure safe to expose at the HTTP boundary."""
 
     def __init__(self, message: str, status_code: int = 502, error_code: str = 'codex_upstream_failure'):
-        super().__init__(message)
+        ValueError.__init__(self, t(message))
+        self.code = error_code
+        self.retryable = status_code in (500, 502, 503, 504)
+        self.retry_after = None
+        self.request_id = None
+        self.phase = None
         self.status_code = status_code
         self.error_code = error_code
+
+    as_dict = RequesterError.as_dict
diff --git a/src/langbot/pkg/provider/modelmgr/errors.py b/src/langbot/pkg/provider/modelmgr/errors.py
index dc3b35b6f..6df4c297b 100644
--- a/src/langbot/pkg/provider/modelmgr/errors.py
+++ b/src/langbot/pkg/provider/modelmgr/errors.py
@@ -1,5 +1,31 @@
+from ...i18n import t
+
+
 class RequesterError(Exception):
     """Base class for all Requester errors."""
 
-    def __init__(self, message: str):
-        super().__init__('模型请求失败: ' + message)
+    def __init__(
+        self,
+        message: str,
+        *,
+        code: str = 'provider_error',
+        status_code: int | None = None,
+        retryable: bool = False,
+        retry_after: float | None = None,
+        request_id: str | None = None,
+        phase: str | None = None,
+    ):
+        super().__init__(t('Model request failed') + ': ' + message)
+        self.code = code
+        self.status_code = status_code
+        self.retryable = retryable
+        self.retry_after = retry_after
+        self.request_id = request_id
+        self.phase = phase
+
+    def as_dict(self) -> dict:
+        """Machine-readable diagnostics without upstream bodies or credentials."""
+        return {
+            key: getattr(self, key)
+            for key in ('code', 'status_code', 'retryable', 'retry_after', 'request_id', 'phase')
+        }
diff --git a/src/langbot/pkg/provider/modelmgr/http_pool.py b/src/langbot/pkg/provider/modelmgr/http_pool.py
new file mode 100644
index 000000000..a29efefd9
--- /dev/null
+++ b/src/langbot/pkg/provider/modelmgr/http_pool.py
@@ -0,0 +1,114 @@
+"""Requester-local connections, retired only after their active leases finish."""
+
+from contextlib import asynccontextmanager
+import asyncio
+import re
+
+import httpx
+
+from .errors import RequesterError
+
+
+class _CompletedSSEStream(httpx.AsyncByteStream):
+    """Finish bounded HTTP framing before SDKs stop at the SSE DONE record."""
+
+    _done = re.compile(rb'\ndata: ?\[DONE\]\r?\n\r?\n')
+
+    def __init__(self, stream):
+        self.stream = stream
+
+    async def __aiter__(self):
+        iterator = self.stream.__aiter__()
+        tail = b'\n'
+        async for chunk in iterator:
+            probe = tail + chunk
+            if self._done.search(probe):
+                # SDK parsers stop at [DONE] before httpcore observes EOF. This
+                # otherwise discards the connection on every successful stream.
+                # Do not drain on cancellation, truncation, or merely a finish
+                # reason: providers can still send usage before the DONE record.
+                try:
+                    async with asyncio.timeout(0.05):
+                        drained = 0
+                        async for extra in iterator:
+                            drained += len(extra)
+                            if drained >= 65536:
+                                break
+                except (TimeoutError, httpx.HTTPError):
+                    pass
+                yield chunk
+                return
+            tail = probe[-32:]
+            yield chunk
+
+    async def aclose(self):
+        await self.stream.aclose()
+
+
+async def _finish_sse_body(response):
+    if response.headers.get('content-type', '').split(';')[0].strip() == 'text/event-stream':
+        response.stream = _CompletedSSEStream(response.stream)
+
+
+class RequesterHTTPPool:
+    def __init__(self, *, follow_redirects=False, drain_sse=False):
+        self.client = None
+        self.active = 0
+        self.retired = False
+        self.follow_redirects = follow_redirects
+        self.drain_sse = drain_sse
+        self._creation_lock = asyncio.Lock()
+
+    @asynccontextmanager
+    async def lease(self):
+        if self.retired:
+            raise RequesterError('Provider has been retired', code='provider_retired')
+        if self.client is None:
+            async with self._creation_lock:
+                if self.retired:
+                    raise RequesterError('Provider has been retired', code='provider_retired')
+                if self.client is None:
+                    # HTTPX otherwise reloads the same CA bundle for its direct
+                    # transport and each environment proxy mount. Load it once,
+                    # off the event loop, preserving HTTPX's TLS/env semantics.
+                    # Cancellation here leaves no open HTTP client or sockets.
+                    ssl_context = await asyncio.to_thread(httpx.create_ssl_context)
+                    if self.retired:
+                        raise RequesterError('Provider has been retired', code='provider_retired')
+                    # Auth and deadlines remain request-local. No pool or TLS
+                    # context is shared across runtime providers.
+                    self.client = httpx.AsyncClient(
+                        verify=ssl_context,
+                        follow_redirects=self.follow_redirects,
+                        limits=httpx.Limits(max_connections=100, max_keepalive_connections=20, keepalive_expiry=30),
+                        event_hooks={'response': [_finish_sse_body]} if self.drain_sse else None,
+                    )
+        self.active += 1
+        try:
+            yield self.client
+        finally:
+            self.active -= 1
+            if self.retired and not self.active:
+                await self._close()
+
+    async def _close(self):
+        client, self.client = self.client, None
+        if client is not None:
+            await client.aclose()
+
+    async def aclose(self):
+        self.retired = True
+        if not self.active:
+            await self._close()
+
+
+@asynccontextmanager
+async def requester_http_client(owner):
+    pool = getattr(owner, '_http_pool', None)
+    if pool is not None:
+        async with pool.lease() as client:
+            yield client
+    else:
+        # Preserve direct/uninitialized requester use by older extensions/tests.
+        async with httpx.AsyncClient(follow_redirects=False) as client:
+            yield client
diff --git a/src/langbot/pkg/provider/modelmgr/modelmgr.py b/src/langbot/pkg/provider/modelmgr/modelmgr.py
index e86f1761b..e13464a9c 100644
--- a/src/langbot/pkg/provider/modelmgr/modelmgr.py
+++ b/src/langbot/pkg/provider/modelmgr/modelmgr.py
@@ -743,6 +743,7 @@ class ModelManager:
         config = {
             'base_url': provider_entity.base_url,
             'requester_name': provider_entity.requester,
+            'request_policy': self.ap.instance_config.data.get('llm_request_policy', {}),
         }
 
         if provider_entity.requester == 'openai-codex':
diff --git a/src/langbot/pkg/provider/modelmgr/prompt_caching.py b/src/langbot/pkg/provider/modelmgr/prompt_caching.py
new file mode 100644
index 000000000..5ec55277f
--- /dev/null
+++ b/src/langbot/pkg/provider/modelmgr/prompt_caching.py
@@ -0,0 +1,82 @@
+"""Opt-in upstream prompt caching; never cache generated answers locally."""
+
+import copy
+import hashlib
+
+from .errors import RequesterError
+
+
+def _has_breakpoint(value):
+    if isinstance(value, dict):
+        return 'cache_control' in value or any(_has_breakpoint(item) for item in value.values())
+    if isinstance(value, list):
+        return any(_has_breakpoint(item) for item in value)
+    return False
+
+
+def apply_prompt_caching(args, state, *, provider, family):
+    config = args.pop('prompt_caching', None)
+    if config is None:
+        return
+    if not isinstance(config, dict) or set(config) - {'mode', 'retention', 'ttl'}:
+        raise RequesterError('Invalid prompt_caching fields', code='invalid_policy')
+    mode = config.get('mode', 'provider_default')
+    if mode not in ('provider_default', 'openai', 'anthropic'):
+        raise RequesterError('Unsupported prompt_caching mode', code='invalid_policy')
+    if mode == 'provider_default':
+        if set(config) - {'mode'}:
+            raise RequesterError('Cache settings require an explicit mode', code='invalid_policy')
+        return
+    if mode == 'openai':
+        if provider not in ('openai', 'openrouter') or family not in ('openai', 'openrouter'):
+            raise RequesterError('OpenAI caching requires an OpenAI-compatible model', code='invalid_policy')
+        if 'ttl' in config or config.get('retention') not in (None, 'in_memory', '24h'):
+            raise RequesterError('Invalid OpenAI cache retention', code='invalid_policy')
+        body = copy.deepcopy(args.get('extra_body') or {})
+        if not isinstance(body, dict):
+            raise RequesterError('extra_body must be an object', code='invalid_policy')
+        # Bypass LiteLLM's model-specific parameter filter for explicit opt-in
+        # caching fields as well as generated hints. The SDK's extra_body wins
+        # when the same native field appears in both locations.
+        for key in ('prompt_cache_key', 'prompt_cache_retention', 'prompt_cache_options'):
+            if key in args:
+                body.setdefault(key, args.pop(key))
+        # Explicit provider-native settings win. A key is a routing hint only;
+        # it neither enables caching nor guarantees a hit.
+        if (
+            state
+            and state.conversation_id
+            and not any(key in body for key in ('prompt_cache_key', 'prompt_cache_options'))
+        ):
+            scope = f'{state.conversation_id}:{args["model"]}'
+            body['prompt_cache_key'] = 'lb-' + hashlib.sha256(scope.encode()).hexdigest()
+        if config.get('retention') and not any(
+            key in location for key in ('prompt_cache_retention', 'prompt_cache_options') for location in (args, body)
+        ):
+            body['prompt_cache_retention'] = config['retention']
+        if body:
+            args['extra_body'] = body
+        return
+    if family != 'anthropic':
+        raise RequesterError('Anthropic caching requires an Anthropic model', code='invalid_policy')
+    if 'retention' in config or config.get('ttl', '5m') not in ('5m', '1h'):
+        raise RequesterError('Invalid Anthropic cache TTL', code='invalid_policy')
+    # Existing manual or automatic breakpoints are authoritative. Do not exceed
+    # a provider's breakpoint limit or change its TTL ordering.
+    if _has_breakpoint(args):
+        return
+    messages = copy.deepcopy(args['messages'])
+    boundary = None
+    for message in messages:
+        if message.get('role') not in ('system', 'developer'):
+            break
+        content = message.get('content')
+        if isinstance(content, str) and content:
+            message['content'] = [{'type': 'text', 'text': content}]
+        if isinstance(message.get('content'), list):
+            for part in message['content']:
+                if isinstance(part, dict) and part.get('type') == 'text' and part.get('text'):
+                    boundary = part
+    if boundary is not None:
+        boundary['cache_control'] = {'type': 'ephemeral', 'ttl': config.get('ttl', '5m')}
+        args['messages'] = messages
diff --git a/src/langbot/pkg/provider/modelmgr/reasoning.py b/src/langbot/pkg/provider/modelmgr/reasoning.py
index ecc6c38e2..dead45620 100644
--- a/src/langbot/pkg/provider/modelmgr/reasoning.py
+++ b/src/langbot/pkg/provider/modelmgr/reasoning.py
@@ -48,20 +48,27 @@ _CONFLICTING_EXTRA_BODY_ARGS = {
 }
 
 
-def normalize_reasoning_config(value: typing.Any) -> dict[str, str]:
+def normalize_reasoning_config(value: typing.Any) -> dict[str, typing.Any]:
     """Return the canonical model reasoning configuration."""
     if value is None:
         return dict(DEFAULT_REASONING_CONFIG)
     if not isinstance(value, dict):
         raise ValueError('reasoning_config must be an object')
 
-    unknown_fields = set(value) - {'level'}
+    unknown_fields = set(value) - {'level', 'budget_tokens'}
     if unknown_fields:
         raise ValueError(f'Unsupported reasoning_config fields: {", ".join(sorted(unknown_fields))}')
 
-    level = value.get('level', 'provider_default')
+    budget = value.get('budget_tokens')
+    level = value.get('level', 'enabled' if budget is not None else 'provider_default')
     if level not in REASONING_LEVELS:
         raise ValueError(f'Unsupported reasoning level: {level}')
+    if budget is not None:
+        if type(budget) is not int or budget <= 0:
+            raise ValueError('budget_tokens must be a positive integer')
+        if level != 'enabled':
+            raise ValueError('budget_tokens requires level enabled; do not combine a token budget with an effort level')
+        return {'level': level, 'budget_tokens': budget}
     return {'level': typing.cast(str, level)}
 
 
@@ -69,7 +76,7 @@ def validate_reasoning_config(
     value: typing.Any,
     abilities: typing.Iterable[str] | None,
     extra_args: typing.Any,
-) -> dict[str, str]:
+) -> dict[str, typing.Any]:
     """Validate a model-facing reasoning config and conflicting raw arguments."""
     config = normalize_reasoning_config(value)
     if config['level'] == 'provider_default':
@@ -101,7 +108,13 @@ def validate_reasoning_capabilities(
     model_name: str,
 ) -> None:
     """Ensure an explicit reasoning level can be honored by the requester."""
-    level = normalize_reasoning_config(config)['level']
+    config = normalize_reasoning_config(config)
+    level = config['level']
+    if 'budget_tokens' in config:
+        limits = capabilities.get('budget_tokens')
+        if not isinstance(limits, dict) or not limits['min'] <= config['budget_tokens'] <= limits['max']:
+            raise ValueError(f'Unsupported reasoning token budget for model {model_name}: {config["budget_tokens"]}')
+        return
     if level == 'provider_default':
         return
 
@@ -154,3 +167,17 @@ def model_with_reasoning_level(
     scoped_model = copy.copy(model)
     scoped_model.reasoning_config_override = copy.deepcopy(config)
     return scoped_model
+
+
+def resolve_reasoning_config(model, extra_args=None):
+    """Runner policy wins; a call object replaces, rather than merges, a model policy."""
+    value = getattr(model.model_entity, 'reasoning_config', None)
+    if value is not None and not isinstance(value, (dict, str, int, float, bool, list, tuple)):
+        value = None
+    for args in (model.model_entity.extra_args, extra_args):
+        if isinstance(args, dict) and 'reasoning_config' in args:
+            value = args['reasoning_config']
+    override = getattr(model, 'reasoning_config_override', None)
+    if override is not None:
+        value = override
+    return normalize_reasoning_config(value)
diff --git a/src/langbot/pkg/provider/modelmgr/request_policy.py b/src/langbot/pkg/provider/modelmgr/request_policy.py
new file mode 100644
index 000000000..02438d8bd
--- /dev/null
+++ b/src/langbot/pkg/provider/modelmgr/request_policy.py
@@ -0,0 +1,173 @@
+"""Host-owned LLM deadlines, retries and additive diagnostics.
+
+No SDK wire shape changes: request_policy is a reserved model/call extra argument.
+Context-local state lets adapters report raw events without sharing mutable state.
+"""
+
+from __future__ import annotations
+
+import asyncio
+from contextvars import ContextVar
+from dataclasses import dataclass, fields
+from email.utils import parsedate_to_datetime
+import hashlib
+import json
+import math
+import random
+import time
+
+from .errors import RequesterError
+from .codex_errors import CodexProviderError
+
+STRUCTURED_ERRORS = (RequesterError, CodexProviderError)
+
+
+@dataclass
+class RequestPolicy:
+    retries: int = 0
+    total_timeout: float = 300
+    first_event_timeout: float = 120
+    stream_idle_timeout: float = 120
+    connect_timeout: float = 10
+    read_timeout: float = 120
+    write_timeout: float = 30
+    pool_timeout: float = 10
+    backoff_base: float = 0.5
+    backoff_max: float = 8
+    attribution: bool = True
+    conversation_header: str | None = None
+
+    @classmethod
+    def resolve(cls, *layers):
+        values = {}
+        names = {field.name for field in fields(cls)}
+        for layer in layers:
+            if not isinstance(layer, dict):
+                continue
+            # Preserve the legacy per-model/call retry knob, including zero.
+            if 'max_retries' in layer:
+                values['retries'] = layer['max_retries']
+            if 'num_retries' in layer:
+                values['retries'] = layer['num_retries']
+            if isinstance(layer.get('timeout'), (int, float)):
+                values['read_timeout'] = layer['timeout']
+                values['first_event_timeout'] = layer['timeout']
+                values['stream_idle_timeout'] = layer['timeout']
+            policy = layer.get('request_policy', {})
+            if not isinstance(policy, dict) or set(policy) - names:
+                raise RequesterError('Invalid request_policy fields', code='invalid_policy')
+            values.update(policy)
+        policy = cls(**values)
+        for field in fields(cls):
+            value = getattr(policy, field.name)
+            if field.name == 'conversation_header':
+                if value not in (None, 'x-session-id', 'x-litellm-session-id', 'x-langbot-conversation-id'):
+                    raise RequesterError('Unsupported conversation header', code='invalid_policy')
+            elif field.name == 'attribution':
+                if not isinstance(value, bool):
+                    raise RequesterError('attribution must be boolean', code='invalid_policy')
+            elif field.name == 'retries':
+                if type(value) is not int or not 0 <= value <= 5:
+                    raise RequesterError('retries must be an integer from 0 to 5', code='invalid_policy')
+            elif (
+                isinstance(value, bool) or not isinstance(value, (int, float)) or not math.isfinite(value) or value <= 0
+            ):
+                raise RequesterError(f'{field.name} must be finite and positive', code='invalid_policy')
+        return policy
+
+
+class CallState:
+    def __init__(self, policy, conversation_id=None):
+        self.policy = policy
+        self.conversation_id = conversation_id
+        self.start = time.perf_counter()
+        self.attempt_start = self.start
+        self.metrics = {'attempts': 0, 'retry_wait_ms': 0}
+        self.raw_started = False
+        self.error = None
+        self.usage = None
+        self.prepared = {}
+
+    def mark(self, name):
+        self.metrics.setdefault(name, round((time.perf_counter() - self.start) * 1000, 3))
+
+    async def wait(self, operation, phase, limit):
+        remaining = self.policy.total_timeout - (time.perf_counter() - self.start)
+        if phase == 'first_event':
+            limit = min(limit, self.policy.first_event_timeout - (time.perf_counter() - self.attempt_start))
+        token = current_call.set(self)
+        try:
+            # timeout(0) only cancels at a scheduling point. Buffered SDK
+            # responses can complete synchronously, so enforce expired budgets
+            # before starting another operation as well.
+            if min(limit, remaining) <= 0:
+                raise TimeoutError
+            async with asyncio.timeout(max(0, min(limit, remaining))):
+                return await operation()
+        except TimeoutError as exc:
+            raise RequesterError(
+                'LLM request timed out', code='timeout', retryable=True, phase='total' if remaining <= limit else phase
+            ) from exc
+        finally:
+            current_call.reset(token)
+
+    async def retry(self, error):
+        self.error = error.as_dict() if isinstance(error, STRUCTURED_ERRORS) else {'code': 'provider_error'}
+        self.metrics.setdefault('attempt_errors', []).append(dict(self.error))
+        if not isinstance(error, STRUCTURED_ERRORS) or not error.retryable or self.raw_started:
+            return False
+        if self.metrics['attempts'] > self.policy.retries:
+            return False
+        delay = random.uniform(
+            0, min(self.policy.backoff_max, self.policy.backoff_base * 2 ** (self.metrics['attempts'] - 1))
+        )
+        delay = max(delay, error.retry_after or 0)
+        remaining = self.policy.total_timeout - (time.perf_counter() - self.start)
+        if delay >= remaining:
+            return False
+        started = time.perf_counter()
+        await asyncio.sleep(delay)
+        self.metrics['retry_wait_ms'] += round((time.perf_counter() - started) * 1000, 3)
+        return True
+
+
+current_call: ContextVar[CallState | None] = ContextVar('llm_call', default=None)
+
+
+def opaque_conversation_id(query, context):
+    if query is None:
+        return None
+    session = getattr(query, 'session', None)
+    session_id = getattr(session, 'session_id', None)
+    if not isinstance(session_id, (str, int)):
+        launcher = getattr(query, 'launcher_id', None)
+        if not isinstance(launcher, (str, int)):
+            return None
+        launcher_type = getattr(query, 'launcher_type', '')
+        session_id = [str(getattr(launcher_type, 'value', launcher_type)), str(launcher)]
+    scope = [context.workspace_uuid, query.bot_uuid, query.pipeline_uuid, session_id]
+    return 'lb-' + hashlib.sha256(json.dumps(scope, ensure_ascii=True).encode()).hexdigest()
+
+
+def error_metadata(exc):
+    status = getattr(exc, 'status_code', None)
+    status = status if isinstance(status, int) else None
+    response = getattr(exc, 'response', None)
+    headers = getattr(response, 'headers', {})
+    headers = headers if hasattr(headers, 'get') else {}
+    retry_after = None
+    raw = headers.get('retry-after')
+    if raw:
+        try:
+            retry_after = float(raw)
+        except (ValueError, TypeError):
+            try:
+                retry_after = parsedate_to_datetime(raw).timestamp() - time.time()
+            except (ValueError, TypeError, OverflowError):
+                pass
+    if retry_after is not None:
+        retry_after = max(0, retry_after) if math.isfinite(retry_after) else None
+    request_id = headers.get('x-request-id') or headers.get('request-id')
+    if not isinstance(request_id, str):
+        request_id = None
+    return {'status_code': status, 'retry_after': retry_after, 'request_id': request_id[:256] if request_id else None}
diff --git a/src/langbot/pkg/provider/modelmgr/requester.py b/src/langbot/pkg/provider/modelmgr/requester.py
index 331cb6fa7..1aa75e4d9 100644
--- a/src/langbot/pkg/provider/modelmgr/requester.py
+++ b/src/langbot/pkg/provider/modelmgr/requester.py
@@ -1,8 +1,9 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
+from contextlib import aclosing
 
 import abc
+import asyncio
 import typing
 import time
 
@@ -13,6 +14,7 @@ from ...workspace.errors import WorkspaceInvariantError
 import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
 from . import token
 from . import reasoning
+from . import request_policy
 import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
 import langbot_plugin.api.entities.builtin.provider.message as provider_message
 
@@ -100,7 +102,48 @@ class RuntimeProvider:
             raise WorkspaceInvariantError('LLM invocation requires an ExecutionContext when query is absent')
         return execution_context
 
-    @diagnostics.observe('api', 'model.invoke_llm', source='agent')
+    def _call_state(self, query, model, extra_args, context):
+        config = getattr(self.requester, 'requester_cfg', {})
+        policy = request_policy.RequestPolicy.resolve(config, model.model_entity.extra_args, extra_args)
+        if query is not None:
+            if query.variables is None:
+                query.variables = {}
+            query.variables.pop(LLM_USAGE_QUERY_VARIABLE, None)
+            query.variables.pop(STREAM_USAGE_QUERY_VARIABLE, None)
+        return request_policy.CallState(policy, request_policy.opaque_conversation_id(query, context))
+
+    async def _record_call(self, query, model, state, status, error):
+        state.mark('total_ms')
+        if query is None:
+            return
+        stream_usage = query.variables.pop(STREAM_USAGE_QUERY_VARIABLE, None)
+        usage = state.usage if state.usage is not None else stream_usage
+        if usage is not None:
+            _store_llm_usage(query, usage)
+        details = {'timing': dict(state.metrics), 'usage': usage, 'error': error}
+        query.variables['_llm_diagnostics'] = details
+        try:
+            from ...pipeline import monitoring_helper
+
+            await monitoring_helper.MonitoringHelper.record_llm_call(
+                ap=self.requester.ap,
+                query=query,
+                bot_id=query.bot_uuid or 'unknown',
+                bot_name=query.variables.get('_monitoring_bot_name', 'Unknown'),
+                pipeline_id=query.pipeline_uuid or 'unknown',
+                pipeline_name=query.variables.get('_monitoring_pipeline_name', 'Unknown'),
+                model_name=model.model_entity.name,
+                input_tokens=(usage or {}).get('prompt_tokens', 0) or 0,
+                output_tokens=(usage or {}).get('completion_tokens', 0) or 0,
+                duration_ms=int(state.metrics['total_ms']),
+                status=status,
+                error_message=error.get('message') if error else None,
+                message_id=query.variables.get('_monitoring_message_id'),
+                details=details,
+            )
+        except Exception as exc:
+            self.requester.ap.logger.error(f'[Monitoring] Failed to record LLM call: {exc}')
+
     async def invoke_llm(
         self,
         query: pipeline_query.Query | None,
@@ -111,81 +154,47 @@ class RuntimeProvider:
         remove_think: bool = False,
         execution_context: ExecutionContext | None = None,
     ) -> provider_message.Message:
-        """Bridge method for invoking LLM with monitoring"""
-        invocation_context = self._resolve_llm_execution_context(query, execution_context)
-        self._validate_invocation(model, invocation_context)
-        # Start timing for monitoring
-        start_time = time.time()
-        input_tokens = 0
-        output_tokens = 0
-        status = 'success'
-        error_message = None
-
+        """Preserve the legacy Message return while owning retries and deadlines."""
+        context = self._resolve_llm_execution_context(query, execution_context)
+        self._validate_invocation(model, context)
+        state = self._call_state(query, model, extra_args, context)
+        status, error = 'success', None
         try:
-            # Call the underlying requester
-            result = await self.requester.invoke_llm(
-                query=query,
-                model=model,
-                messages=messages,
-                funcs=funcs,
-                extra_args=extra_args,
-                remove_think=remove_think,
-            )
-
-            # Try to extract token usage if the requester returns it
-            # For requesters that return tuple (message, usage_info)
-            if isinstance(result, tuple):
-                msg, usage_info = result
-                if usage_info:
-                    _store_llm_usage(query, usage_info)
-                    input_tokens = usage_info.get('prompt_tokens', 0)
-                    output_tokens = usage_info.get('completion_tokens', 0)
-                return msg
-            else:
-                return result
-
-        except Exception as e:
+            while True:
+                state.metrics['attempts'] += 1
+                state.attempt_start = time.perf_counter()
+                try:
+                    result = await state.wait(
+                        lambda: self.requester.invoke_llm(
+                            query=query,
+                            model=model,
+                            messages=messages,
+                            funcs=funcs,
+                            extra_args=extra_args,
+                            remove_think=remove_think,
+                        ),
+                        'response',
+                        state.policy.total_timeout,
+                    )
+                    state.mark('response_completed_ms')
+                    if isinstance(result, tuple):
+                        message, state.usage = result
+                        return message
+                    return result
+                except Exception as exc:
+                    if not await state.retry(exc):
+                        raise
+        except (asyncio.CancelledError, GeneratorExit):
+            status, error = 'error', {'code': 'cancelled', 'message': 'LLM invocation cancelled'}
+            raise
+        except Exception as exc:
             status = 'error'
-            error_message = str(e)
+            error = exc.as_dict() if isinstance(exc, request_policy.STRUCTURED_ERRORS) else {'code': 'provider_error'}
+            error['message'] = str(exc)
             raise
         finally:
-            # Record LLM call monitoring data (only if query is provided)
-            if query is not None:
-                duration_ms = int((time.time() - start_time) * 1000)
+            await self._record_call(query, model, state, status, error)
 
-                # Import monitoring helper
-                try:
-                    from ...pipeline import monitoring_helper
-
-                    # Get monitoring metadata from query variables
-                    if query.variables:
-                        bot_name = query.variables.get('_monitoring_bot_name', 'Unknown')
-                        pipeline_name = query.variables.get('_monitoring_pipeline_name', 'Unknown')
-                        message_id = query.variables.get('_monitoring_message_id')
-                    else:
-                        bot_name = 'Unknown'
-                        pipeline_name = 'Unknown'
-                        message_id = None
-
-                    await monitoring_helper.MonitoringHelper.record_llm_call(
-                        ap=self.requester.ap,
-                        query=query,
-                        bot_id=query.bot_uuid or 'unknown',
-                        bot_name=bot_name,
-                        pipeline_id=query.pipeline_uuid or 'unknown',
-                        pipeline_name=pipeline_name,
-                        model_name=model.model_entity.name,
-                        input_tokens=input_tokens,
-                        output_tokens=output_tokens,
-                        duration_ms=duration_ms,
-                        status=status,
-                        error_message=error_message,
-                        message_id=message_id,
-                    )
-                except Exception as monitor_err:
-                    self.requester.ap.logger.error(f'[Monitoring] Failed to record LLM call: {monitor_err}')
-
-    @diagnostics.observe('api', 'model.invoke_llm_stream', source='agent')
     async def invoke_llm_stream(
         self,
         query: pipeline_query.Query | None,
@@ -195,80 +204,67 @@ class RuntimeProvider:
         extra_args: dict[str, typing.Any] = {},
         remove_think: bool = False,
         execution_context: ExecutionContext | None = None,
-    ) -> provider_message.MessageChunk:
-        """Bridge method for invoking LLM stream with monitoring"""
-        invocation_context = self._resolve_llm_execution_context(query, execution_context)
-        self._validate_invocation(model, invocation_context)
-        # Start timing for monitoring
-        start_time = time.time()
-        status = 'success'
-        error_message = None
-        input_tokens = 0
-        output_tokens = 0
-
+    ) -> typing.AsyncIterator[provider_message.MessageChunk]:
+        """Never retry after a raw upstream event or a legacy adapter chunk."""
+        context = self._resolve_llm_execution_context(query, execution_context)
+        self._validate_invocation(model, context)
+        state = self._call_state(query, model, extra_args, context)
+        status, error = 'success', None
         try:
-            # Stream the response
-            async for chunk in self.requester.invoke_llm_stream(
-                query=query,
-                model=model,
-                messages=messages,
-                funcs=funcs,
-                extra_args=extra_args,
-                remove_think=remove_think,
-            ):
-                yield chunk
-            # Extract usage from stream if available (stored by LiteLLM requester)
-            if query:
-                if query.variables is None:
-                    query.variables = {}
-                if STREAM_USAGE_QUERY_VARIABLE in query.variables:
-                    usage_info = query.variables[STREAM_USAGE_QUERY_VARIABLE]
-                    _store_llm_usage(query, usage_info)
-                    input_tokens = usage_info.get('prompt_tokens', 0)
-                    output_tokens = usage_info.get('completion_tokens', 0)
-                    del query.variables[STREAM_USAGE_QUERY_VARIABLE]
-        except Exception as e:
+            while True:
+                state.metrics['attempts'] += 1
+                state.attempt_start = time.perf_counter()
+                try:
+                    async with aclosing(
+                        self.requester.invoke_llm_stream(
+                            query=query,
+                            model=model,
+                            messages=messages,
+                            funcs=funcs,
+                            extra_args=extra_args,
+                            remove_think=remove_think,
+                        )
+                    ) as stream:
+                        while True:
+                            try:
+                                chunk = await state.wait(
+                                    lambda: anext(stream),
+                                    'total'
+                                    if getattr(self.requester, 'reports_raw_timing', False) is True
+                                    else ('stream_idle' if state.raw_started else 'first_event'),
+                                    state.policy.total_timeout
+                                    if getattr(self.requester, 'reports_raw_timing', False) is True
+                                    else (
+                                        state.policy.stream_idle_timeout
+                                        if state.raw_started
+                                        else state.policy.first_event_timeout
+                                    ),
+                                )
+                            except StopAsyncIteration:
+                                break
+                            if chunk is None:
+                                continue
+                            state.raw_started = True
+                            state.mark('first_event_ms')
+                            if chunk.content:
+                                state.mark('first_visible_chunk_ms')
+                            yield chunk
+                    state.mark('response_completed_ms')
+                    return
+                except Exception as exc:
+                    if not await state.retry(exc):
+                        raise
+        except (asyncio.CancelledError, GeneratorExit):
+            status, error = 'error', {'code': 'cancelled', 'message': 'LLM stream cancelled'}
+            raise
+        except Exception as exc:
             status = 'error'
-            error_message = str(e)
+            error = exc.as_dict() if isinstance(exc, request_policy.STRUCTURED_ERRORS) else {'code': 'provider_error'}
+            error['message'] = str(exc)
             raise
         finally:
-            # Record LLM call monitoring data (only if query is provided)
-            if query is not None:
-                duration_ms = int((time.time() - start_time) * 1000)
+            await self._record_call(query, model, state, status, error)
 
-                # Import monitoring helper
-                try:
-                    from ...pipeline import monitoring_helper
-
-                    # Get monitoring metadata from query variables
-                    if query.variables:
-                        bot_name = query.variables.get('_monitoring_bot_name', 'Unknown')
-                        pipeline_name = query.variables.get('_monitoring_pipeline_name', 'Unknown')
-                        message_id = query.variables.get('_monitoring_message_id')
-                    else:
-                        bot_name = 'Unknown'
-                        pipeline_name = 'Unknown'
-                        message_id = None
-
-                    await monitoring_helper.MonitoringHelper.record_llm_call(
-                        ap=self.requester.ap,
-                        query=query,
-                        bot_id=query.bot_uuid or 'unknown',
-                        bot_name=bot_name,
-                        pipeline_id=query.pipeline_uuid or 'unknown',
-                        pipeline_name=pipeline_name,
-                        model_name=model.model_entity.name,
-                        input_tokens=input_tokens,
-                        output_tokens=output_tokens,
-                        duration_ms=duration_ms,
-                        status=status,
-                        error_message=error_message,
-                        message_id=message_id,
-                    )
-                except Exception as monitor_err:
-                    self.requester.ap.logger.error(f'[Monitoring] Failed to record LLM stream call: {monitor_err}')
-
-    @diagnostics.observe('api', 'model.invoke_embedding', source='agent')
     async def invoke_embedding(
         self,
         model: RuntimeEmbeddingModel,
@@ -336,7 +332,6 @@ class RuntimeProvider:
             except Exception as monitor_err:
                 self.requester.ap.logger.error(f'[Monitoring] Failed to record embedding call: {monitor_err}')
 
-    @diagnostics.observe('api', 'model.invoke_rerank', source='agent')
     async def invoke_rerank(
         self,
         model: RuntimeRerankModel,
@@ -384,7 +379,7 @@ class RuntimeLLMModel:
     provider: RuntimeProvider
     """提供商实例"""
 
-    reasoning_config_override: dict[str, str] | None
+    reasoning_config_override: dict[str, typing.Any] | None
     """Request-scoped reasoning policy supplied by the active pipeline."""
 
     def __init__(
@@ -392,7 +387,7 @@ class RuntimeLLMModel:
         execution_context: ExecutionContext,
         model_entity: persistence_model.LLMModel,
         provider: RuntimeProvider,
-        reasoning_config_override: dict[str, str] | None = None,
+        reasoning_config_override: dict[str, typing.Any] | None = None,
     ):
         _ensure_same_execution_scope(provider.execution_context, execution_context, resource='LLM model')
         if model_entity.workspace_uuid != execution_context.workspace_uuid:
diff --git a/src/langbot/pkg/provider/modelmgr/requesters/codex.py b/src/langbot/pkg/provider/modelmgr/requesters/codex.py
index 5b793ef75..728382415 100644
--- a/src/langbot/pkg/provider/modelmgr/requesters/codex.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/codex.py
@@ -7,12 +7,16 @@ import json
 import secrets
 import time
 from collections import OrderedDict
+from contextlib import AsyncExitStack, aclosing
 
 import httpx
 import langbot
 import langbot_plugin.api.entities.builtin.provider.message as pm
 
 from .. import requester, reasoning
+from ..request_policy import current_call
+from ..http_pool import RequesterHTTPPool, requester_http_client
+from ..errors import RequesterError
 from ..codex_auth import BASE_URL, CodexAuth, LOGIN_REQUIRED
 from ..codex_errors import CodexProviderError
 
@@ -21,7 +25,8 @@ async def sse_events(response):
     """Decode SSE records, including CRLF, comments, and multiline data."""
     data = []
     size = 0
-    async for line in response.aiter_lines():
+    lines = response.aiter_lines()
+    async for line in lines:
         if not line:
             if data:
                 text = '\n'.join(data)
@@ -33,6 +38,18 @@ async def sse_events(response):
                         raise ValueError
                 except ValueError:
                     raise ValueError('Codex returned an invalid stream event') from None
+                if event.get('type') == 'response.completed':
+                    # Finish the HTTP body so a healthy connection can return
+                    # to the pool. Never wait indefinitely for a broken SSE
+                    # server after it has already sent the terminal result.
+                    try:
+                        async with asyncio.timeout(0.05):
+                            async for _ in lines:
+                                pass
+                    except (TimeoutError, httpx.HTTPError):
+                        pass
+                    yield event
+                    return
                 yield event
             data, size = [], 0
         elif line.startswith('data:'):
@@ -85,6 +102,7 @@ def _tool(item):
 def _usage(response):
     usage = response.get('usage') or {}
     return {
+        **usage,
         'prompt_tokens': usage.get('input_tokens', 0),
         'completion_tokens': usage.get('output_tokens', 0),
         'total_tokens': usage.get('total_tokens', usage.get('input_tokens', 0) + usage.get('output_tokens', 0)),
@@ -94,7 +112,10 @@ def _usage(response):
 
 
 class CodexRequester(requester.ProviderAPIRequester):
+    reports_raw_timing = True
+
     async def initialize(self):
+        self._http_pool = RequesterHTTPPool()
         self.auth = CodexAuth(self.ap)
         self.workspace = self.requester_cfg['workspace_uuid']
         self.provider = self.requester_cfg['provider_uuid']
@@ -104,6 +125,8 @@ class CodexRequester(requester.ProviderAPIRequester):
 
     async def aclose(self):
         self._replay.clear()
+        if getattr(self, '_http_pool', None) is not None:
+            await self._http_pool.aclose()
 
     def get_reasoning_capabilities(self, model):
         return {
@@ -184,6 +207,8 @@ class CodexRequester(requester.ProviderAPIRequester):
 
     def _body(self, query, model, messages, funcs, extra_args, tokens):
         args = {**(model.model_entity.extra_args or {}), **(extra_args or {})}
+        args.pop('request_policy', None)
+        args.pop('reasoning_config', None)
         # Never permit credentials, transport overrides, store/history or arbitrary
         # SDK kwargs to be smuggled through model advanced parameters.
         allowed = {'reasoning', 'text', 'parallel_tool_calls', 'tool_choice'}
@@ -235,13 +260,11 @@ class CodexRequester(requester.ProviderAPIRequester):
             'stream': True,
             'include': ['reasoning.encrypted_content'],
         }
-        reasoning_config = getattr(model, 'reasoning_config_override', None)
-        if reasoning_config is None:
-            reasoning_config = getattr(model.model_entity, 'reasoning_config', None)
-        level = reasoning.normalize_reasoning_config(reasoning_config)['level']
+        reasoning_config = reasoning.resolve_reasoning_config(model, extra_args)
+        level = reasoning_config['level']
         if level != 'provider_default':
             reasoning.validate_reasoning_capabilities(
-                {'level': level}, self.get_reasoning_capabilities(model), model.model_entity.name
+                reasoning_config, self.get_reasoning_capabilities(model), model.model_entity.name
             )
             body['reasoning'] = {'effort': level, 'summary': 'auto'}
         if funcs:
@@ -259,13 +282,39 @@ class CodexRequester(requester.ProviderAPIRequester):
 
     async def _events(self, query, model, messages, funcs, extra_args):
         tokens = await self.auth.access(self.workspace, self.provider)
+        state = current_call.get()
+        timeout = 120
+        if state:
+            p = state.policy
+            timeout = httpx.Timeout(
+                connect=p.connect_timeout, read=p.read_timeout, write=p.write_timeout, pool=p.pool_timeout
+            )
         try:
-            async with asyncio.timeout(300), httpx.AsyncClient(timeout=120, follow_redirects=False) as client:
+            async with (
+                asyncio.timeout(None if state else 300),
+                requester_http_client(self) as client,
+            ):
                 for attempt in range(2):
                     body = self._body(query, model, messages, funcs, extra_args, tokens)
-                    async with client.stream(
-                        'POST', BASE_URL + '/responses', json=body, headers=self._headers(tokens, stream=True)
-                    ) as response:
+                    if state:
+                        state.mark('request_prepared_ms')
+                        state.mark('request_started_ms')
+                    async with AsyncExitStack() as stack:
+                        response_context = client.stream(
+                            'POST',
+                            BASE_URL + '/responses',
+                            json=body,
+                            headers=self._headers(tokens, stream=True),
+                            timeout=timeout,
+                        )
+                        if state:
+                            response = await state.wait(
+                                lambda: stack.enter_async_context(response_context),
+                                'first_event',
+                                state.policy.first_event_timeout,
+                            )
+                        else:
+                            response = await stack.enter_async_context(response_context)
                         if response.status_code == 401 and attempt == 0:
                             tokens = await self.auth.access(
                                 self.workspace, self.provider, rejected_token=tokens['access_token']
@@ -273,85 +322,117 @@ class CodexRequester(requester.ProviderAPIRequester):
                             continue
                         if response.status_code != 200:
                             raise await self._response_error(response)
-                        async for event in sse_events(response):
+                        if state:
+                            state.mark('response_open_ms')
+                        events = sse_events(response)
+                        while True:
+                            try:
+                                if state:
+                                    event = await state.wait(
+                                        lambda: anext(events),
+                                        'stream_idle' if state.raw_started else 'first_event',
+                                        state.policy.stream_idle_timeout
+                                        if state.raw_started
+                                        else state.policy.first_event_timeout,
+                                    )
+                                else:
+                                    event = await anext(events)
+                            except StopAsyncIteration:
+                                break
+                            if state:
+                                state.raw_started = True
+                                state.mark('first_event_ms')
+                                kind = event.get('type', '')
+                                if kind == 'response.output_text.delta' and event.get('delta'):
+                                    state.mark('first_content_ms')
+                                elif 'reasoning' in kind and event.get('delta'):
+                                    state.mark('first_reasoning_ms')
                             yield event, tokens
                         return
-        except (httpx.HTTPError, TimeoutError):
-            raise ValueError('ChatGPT Codex network error or timeout. Please retry.') from None
+        except (httpx.HTTPError, TimeoutError) as exc:
+            raise RequesterError(
+                'ChatGPT Codex network error or timeout. Please retry.',
+                code='timeout' if isinstance(exc, (httpx.TimeoutException, TimeoutError)) else 'connection',
+                retryable=True,
+            ) from None
 
     async def _chunks(self, query, model, messages, funcs, extra_args, remove_think, usage_out):
         text = ''
         seen_calls = set()
         output_items = {}
         response_id = None
-        async for event, tokens in self._events(query, model, messages, funcs, extra_args):
-            kind = event.get('type')
-            response = event.get('response') or {}
-            response_id = response.get('id') or response_id
-            if kind in ('error', 'response.failed', 'response.incomplete'):
-                raise CodexProviderError('ChatGPT Codex response failed or was incomplete. Please retry.')
-            if kind == 'response.output_text.delta':
-                delta = event.get('delta', '')
-                text += delta
-                yield pm.MessageChunk(role='assistant', content=delta, resp_message_id=response_id)
-            elif kind in ('response.reasoning_summary_text.delta', 'response.reasoning_text.delta'):
-                if not remove_think:
-                    yield pm.MessageChunk(
-                        role='assistant',
-                        content='',
-                        provider_specific_fields={'reasoning_content': event.get('delta', '')},
-                    )
-            elif kind == 'response.output_item.done':
-                item = event.get('item') or {}
-                output_items[event.get('output_index', len(output_items))] = item
-                if item.get('type') == 'function_call' and item.get('call_id') not in seen_calls:
-                    seen_calls.add(item.get('call_id'))
-                    yield pm.MessageChunk(role='assistant', content='', tool_calls=[_tool(item)])
-            elif kind in ('response.completed', 'response.done'):
-                if response.get('status') not in (None, 'completed'):
-                    raise CodexProviderError('ChatGPT Codex response was not completed')
-                output = response.get('output') or [output_items[k] for k in sorted(output_items)]
-                for item in output:
+        async with aclosing(self._events(query, model, messages, funcs, extra_args)) as events:
+            async for event, tokens in events:
+                kind = event.get('type')
+                response = event.get('response') or {}
+                response_id = response.get('id') or response_id
+                if kind in ('error', 'response.failed', 'response.incomplete'):
+                    raise CodexProviderError('ChatGPT Codex response failed or was incomplete. Please retry.')
+                if kind == 'response.output_text.delta':
+                    delta = event.get('delta', '')
+                    text += delta
+                    yield pm.MessageChunk(role='assistant', content=delta, resp_message_id=response_id)
+                elif kind in ('response.reasoning_summary_text.delta', 'response.reasoning_text.delta'):
+                    if not remove_think:
+                        yield pm.MessageChunk(
+                            role='assistant',
+                            content='',
+                            provider_specific_fields={'reasoning_content': event.get('delta', '')},
+                        )
+                elif kind == 'response.output_item.done':
+                    item = event.get('item') or {}
+                    output_items[event.get('output_index', len(output_items))] = item
                     if item.get('type') == 'function_call' and item.get('call_id') not in seen_calls:
                         seen_calls.add(item.get('call_id'))
                         yield pm.MessageChunk(role='assistant', content='', tool_calls=[_tool(item)])
-                # Some servers send only the terminal output, without text deltas.
-                final_text = ''.join(
-                    p.get('text', '')
-                    for item in output
-                    if item.get('type') == 'message'
-                    for p in item.get('content', [])
-                    if p.get('type') == 'output_text'
-                )
-                if not text and final_text:
-                    text = final_text
-                    yield pm.MessageChunk(role='assistant', content=text, resp_message_id=response_id)
-                usage_out.update(_usage(response))
-                if query is not None:
-                    if query.variables is None:
-                        query.variables = {}
-                    query.variables[requester.STREAM_USAGE_QUERY_VARIABLE] = dict(usage_out)
-                fields = None
-                if query is not None and output:
-                    handle = secrets.token_urlsafe(24)
-                    self._replay[handle] = (self._scope(query, model, tokens), time.time() + 3600, output)
-                    while len(self._replay) > 64:
-                        self._replay.popitem(last=False)
-                    fields = {'codex_replay_id': handle}
-                yield pm.MessageChunk(
-                    role='assistant',
-                    content='',
-                    all_content=text,
-                    is_final=True,
-                    resp_message_id=response_id,
-                    provider_specific_fields=fields,
-                )
-                return
+                elif kind in ('response.completed', 'response.done'):
+                    if response.get('status') not in (None, 'completed'):
+                        raise CodexProviderError('ChatGPT Codex response was not completed')
+                    output = response.get('output') or [output_items[k] for k in sorted(output_items)]
+                    for item in output:
+                        if item.get('type') == 'function_call' and item.get('call_id') not in seen_calls:
+                            seen_calls.add(item.get('call_id'))
+                            yield pm.MessageChunk(role='assistant', content='', tool_calls=[_tool(item)])
+                    # Some servers send only the terminal output, without text deltas.
+                    final_text = ''.join(
+                        p.get('text', '')
+                        for item in output
+                        if item.get('type') == 'message'
+                        for p in item.get('content', [])
+                        if p.get('type') == 'output_text'
+                    )
+                    if not text and final_text:
+                        text = final_text
+                        yield pm.MessageChunk(role='assistant', content=text, resp_message_id=response_id)
+                    usage_out.update(_usage(response))
+                    if current_call.get():
+                        current_call.get().usage = dict(usage_out)
+                    if query is not None:
+                        if query.variables is None:
+                            query.variables = {}
+                        query.variables[requester.STREAM_USAGE_QUERY_VARIABLE] = dict(usage_out)
+                    fields = None
+                    if query is not None and output:
+                        handle = secrets.token_urlsafe(24)
+                        self._replay[handle] = (self._scope(query, model, tokens), time.time() + 3600, output)
+                        while len(self._replay) > 64:
+                            self._replay.popitem(last=False)
+                        fields = {'codex_replay_id': handle}
+                    yield pm.MessageChunk(
+                        role='assistant',
+                        content='',
+                        all_content=text,
+                        is_final=True,
+                        resp_message_id=response_id,
+                        provider_specific_fields=fields,
+                    )
+                    return
         raise CodexProviderError('ChatGPT Codex stream ended before completion. Please retry.')
 
     async def invoke_llm_stream(self, query, model, messages, funcs=None, extra_args=None, remove_think=False):
-        async for chunk in self._chunks(query, model, messages, funcs, extra_args, remove_think, {}):
-            yield chunk
+        async with aclosing(self._chunks(query, model, messages, funcs, extra_args, remove_think, {})) as chunks:
+            async for chunk in chunks:
+                yield chunk
 
     async def invoke_llm(self, query, model, messages, funcs=None, extra_args=None, remove_think=False):
         usage = {}
@@ -376,11 +457,12 @@ class CodexRequester(requester.ProviderAPIRequester):
     async def scan_models(self, api_key=None):
         tokens = await self.auth.access(self.workspace, self.provider)
         try:
-            async with asyncio.timeout(90), httpx.AsyncClient(timeout=30, follow_redirects=False) as client:
+            async with asyncio.timeout(90), requester_http_client(self) as client:
                 for attempt in range(2):
                     response = await client.get(
                         BASE_URL + '/models',
                         params={'client_version': langbot.__version__},
+                        timeout=30,
                         headers=self._headers(tokens),
                     )
                     if response.status_code == 401 and attempt == 0:
diff --git a/src/langbot/pkg/provider/modelmgr/requesters/litellmchat.py b/src/langbot/pkg/provider/modelmgr/requesters/litellmchat.py
index 0bbaa1db5..300fb7551 100644
--- a/src/langbot/pkg/provider/modelmgr/requesters/litellmchat.py
+++ b/src/langbot/pkg/provider/modelmgr/requesters/litellmchat.py
@@ -2,10 +2,26 @@
 
 from __future__ import annotations
 
+from ....i18n import t
+
 import typing
+import copy
+import time
+import os
+from collections import OrderedDict
+from contextlib import asynccontextmanager
+from openai import AsyncOpenAI
+from ..http_pool import RequesterHTTPPool
+from ..prompt_caching import apply_prompt_caching
+import asyncio
+import httpx
+from urllib.parse import urlsplit
+from ..request_policy import current_call, RequestPolicy, error_metadata
 
 import litellm
 from litellm import acompletion, aembedding, arerank
+from litellm.llms.custom_httpx.http_handler import AsyncHTTPHandler
+from litellm.secret_managers.main import get_secret, get_secret_bool
 
 from .. import errors, reasoning, requester
 from ....utils import httpclient
@@ -14,8 +30,43 @@ import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
 import langbot_plugin.api.entities.builtin.provider.message as provider_message
 
 
+class _LeasedHTTPHandler(AsyncHTTPHandler):
+    """Adapt a leased transport without LiteLLM's hidden connection-error retry."""
+
+    def __init__(self, client, timeout):
+        self.client = client
+        self.timeout = timeout
+        self.event_hooks = None
+        self.client_alias = 'langbot'
+
+    async def post(self, url, *, timeout=None, stream=False, logging_obj=None, **kwargs):
+        if isinstance(kwargs.get('data'), (str, bytes)):
+            kwargs['content'] = kwargs.pop('data')
+        request = self.client.build_request('POST', url, timeout=timeout or self.timeout, **kwargs)
+        response = await self.client.send(request, stream=stream)
+        try:
+            response.raise_for_status()
+        except BaseException:
+            await response.aclose()
+            raise
+        return response
+
+    async def close(self):
+        pass  # The requester's lease owns the transport.
+
+    async def __aexit__(self, *args):
+        pass
+
+    def __del__(self):
+        pass
+
+
 class _ThinkStripState:
-    """Stateful filter that drops think blocks across chunks."""
+    """Drop explicit think blocks, buffering only possible split tag prefixes.
+
+    Untagged text must stream immediately: an orphan closing tag cannot identify
+    earlier text as reasoning without delaying every ordinary answer until EOF.
+    """
 
     _THINK_OPEN = '<think>'
     _THINK_CLOSE = '</think>'
@@ -30,7 +81,6 @@ class _ThinkStripState:
         self._open_tags = tuple(open_tag for open_tag, _close_tag in self._pairs)
         self._buf = ''
         self._close_tag: str | None = None
-        self._pending_initial = True
 
     def feed(self, chunk: str) -> str:
         """Feed a streaming delta and return user-visible content."""
@@ -63,7 +113,6 @@ class _ThinkStripState:
         if close_idx != -1:
             self._close_tag = None
             self._buf = ''
-            self._pending_initial = False
             return self._process_visible_text(text[close_idx + len(close_tag) :])
 
         self._buf = self._close_prefix(text, close_tag)
@@ -74,37 +123,14 @@ class _ThinkStripState:
         index = 0
 
         while index < len(text):
-            if self._pending_initial:
-                open_idx, open_tag, close_tag = self._find_next_open(text, index)
-                orphan_close_idx, orphan_close_tag = self._find_next_close(text, index)
-
-                if orphan_close_idx != -1 and (open_idx == -1 or orphan_close_idx < open_idx):
-                    self._pending_initial = False
-                    index = orphan_close_idx + len(orphan_close_tag)
-                    continue
-
-                if open_idx == -1:
-                    self._buf = text[index:]
-                    return ''.join(out)
-
-                if open_idx > index:
-                    self._pending_initial = False
-                    out.append(text[index:open_idx])
-                    index = open_idx
-                    continue
-
             open_idx, open_tag, close_tag = self._find_next_open(text, index)
             if open_idx == -1:
                 emit_end = self._visible_emit_end(text, index)
                 out.append(text[index:emit_end])
-                if emit_end > index:
-                    self._pending_initial = False
                 self._buf = text[emit_end:]
                 return ''.join(out)
 
             out.append(text[index:open_idx])
-            if open_idx > index:
-                self._pending_initial = False
             body_start = open_idx + len(open_tag)
             close_idx = text.find(close_tag, body_start)
             if close_idx == -1:
@@ -112,7 +138,6 @@ class _ThinkStripState:
                 self._buf = self._close_prefix(text[body_start:], close_tag)
                 return ''.join(out)
 
-            self._pending_initial = False
             index = close_idx + len(close_tag)
 
         self._buf = ''
@@ -130,16 +155,6 @@ class _ThinkStripState:
                 best_close = close_tag
         return best_idx, best_open, best_close
 
-    def _find_next_close(self, text: str, start: int) -> tuple[int, str]:
-        best_idx = -1
-        best_close = ''
-        for _open_tag, close_tag in self._pairs:
-            idx = text.find(close_tag, start)
-            if idx != -1 and (best_idx == -1 or idx < best_idx):
-                best_idx = idx
-                best_close = close_tag
-        return best_idx, best_close
-
     def _visible_emit_end(self, text: str, start: int) -> int:
         visible = text[start:]
         limit = min(len(visible), max(len(open_tag) for open_tag in self._open_tags) - 1)
@@ -162,6 +177,8 @@ class _ThinkStripState:
 class LiteLLMRequester(requester.ProviderAPIRequester):
     """LiteLLM unified API requester supporting chat, embedding, and rerank."""
 
+    reports_raw_timing = True
+
     _EMBEDDING_MODEL_HINTS = ('embedding', 'embed', 'bge-', 'e5-', 'm3e', 'gte-', 'text-embedding')
     _RERANK_MODEL_HINTS = ('rerank', 're-rank', 're_rank')
     _QWEN_DEDICATED_THINKING_MODELS = frozenset(
@@ -210,9 +227,83 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
 
     async def initialize(self):
         """Initialize LiteLLM client settings."""
-        # LiteLLM doesn't require explicit client initialization
-        # Configuration is passed per-request via litellm params
-        pass
+        self._http_pool = RequesterHTTPPool(
+            follow_redirects=True, drain_sse=self._get_custom_llm_provider() == 'openai'
+        )
+
+    async def aclose(self):
+        if getattr(self, '_http_pool', None) is not None:
+            await self._http_pool.aclose()
+        if hasattr(self, '_metadata_cache'):
+            self._metadata_cache.clear()
+
+    @asynccontextmanager
+    async def _client_args(self, args):
+        pool = getattr(self, '_http_pool', None)
+        provider = args.get('custom_llm_provider') or self._get_custom_llm_provider()
+        if pool is None or provider not in ('openai', 'openrouter') or args.get('client') is not None:
+            yield args
+            return
+        # Preserve explicitly customized LiteLLM transports and TLS settings.
+        if (
+            any(args.get(key) is not None for key in ('shared_session', 'http_client', 'ssl_verify', 'ssl_certificate'))
+            or getattr(litellm, 'aclient_session', None) is not None
+            or getattr(litellm, 'ssl_verify', True) is not True
+            or getattr(litellm, 'ssl_certificate', None)
+            or getattr(litellm, 'force_ipv4', False)
+            or any(name in os.environ for name in ('SSL_VERIFY', 'SSL_CERTIFICATE'))
+        ):
+            yield args
+            return
+        if not args['model'].startswith(f'{provider}/'):
+            yield args  # A caller redirected the model; let LiteLLM resolve its transport.
+            return
+        model_name = args['model'].removeprefix(f'{provider}/')
+        if provider == 'openrouter' and model_name in litellm.open_ai_chat_completion_models:
+            yield args  # Preserve LiteLLM's alternate bare-model routing.
+            return
+        transport_started = time.perf_counter()
+        async with pool.lease() as client:
+            state = current_call.get()
+            if state:
+                state.metrics['transport_setup_ms'] = round(
+                    state.metrics.get('transport_setup_ms', 0) + (time.perf_counter() - transport_started) * 1000, 3
+                )
+            if provider == 'openrouter' or get_secret_bool('EXPERIMENTAL_OPENAI_BASE_LLM_HTTP_HANDLER'):
+                yield {**args, 'client': _LeasedHTTPHandler(client, args.get('timeout', 120))}
+                return
+            # A fresh lightweight SDK wrapper keeps rotating keys, organization
+            # and deadlines request-local; only its HTTP transport is reused.
+            sdk = AsyncOpenAI(
+                api_key=args.get('api_key') or litellm.api_key or litellm.openai_key or get_secret('OPENAI_API_KEY'),
+                base_url=(
+                    args.get('api_base')
+                    or litellm.api_base
+                    or get_secret('OPENAI_BASE_URL')
+                    or get_secret('OPENAI_API_BASE')
+                    or 'https://api.openai.com/v1'
+                ),
+                http_client=client,
+                organization=args.get('organization') or litellm.organization,
+                timeout=args.get('timeout', 120),
+                max_retries=0,
+            )
+            yield {**args, 'client': sdk}
+
+    def _cached_metadata(self, key, compute):
+        if not hasattr(self, '_metadata_cache'):
+            self._metadata_cache = OrderedDict()
+        cache = self._metadata_cache
+        cached = cache.get(key)
+        if cached is not None and time.monotonic() - cached[0] < 300:
+            cache.move_to_end(key)
+            return copy.deepcopy(cached[1])
+        value = compute()
+        cache[key] = (time.monotonic(), copy.deepcopy(value))
+        cache.move_to_end(key)
+        while len(cache) > 128:
+            cache.popitem(last=False)
+        return value
 
     def _build_litellm_model_name(self, model_name: str, custom_llm_provider: str | None = None) -> str:
         """Build LiteLLM model name with provider prefix if needed."""
@@ -228,8 +319,21 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
     def _get_custom_llm_provider(self) -> str | None:
         return self.requester_cfg.get('custom_llm_provider') or None
 
+    def _uses_openrouter_protocol(self) -> bool:
+        # OpenRouter's shipped metadata also serves OpenAI-compatible non-chat
+        # operations. Select gateway chat semantics independently of it.
+        return (
+            self._get_custom_llm_provider() == 'openrouter'
+            or self.requester_cfg.get('requester_name') == 'openrouter-chat-completions'
+        )
+
     def _safe_litellm_bool_helper(self, helper_name: str, model_name: str) -> bool:
         """Call a LiteLLM boolean capability helper without letting metadata gaps fail requests."""
+        return self._cached_metadata(
+            ('bool', helper_name, model_name), lambda: self._uncached_litellm_bool_helper(helper_name, model_name)
+        )
+
+    def _uncached_litellm_bool_helper(self, helper_name, model_name):
         helper = getattr(litellm, helper_name, None)
         if not callable(helper):
             return False
@@ -328,6 +432,9 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         return None
 
     def _safe_context_length(self, model_name: str) -> int | None:
+        return self._cached_metadata(('model', model_name), lambda: self._uncached_model_info(model_name))
+
+    def _uncached_model_info(self, model_name):
         helper = getattr(litellm, 'get_model_info', None)
         if not callable(helper):
             return self._known_context_length_fallback(model_name)
@@ -592,6 +699,20 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         }
         if family == 'qwen' and 'disabled' in capabilities['levels'] and 'enabled' not in capabilities['levels']:
             capabilities['legacy_levels'] = ['enabled']
+        if self._uses_openrouter_protocol():
+            capabilities['budget_tokens'] = {'min': 1, 'max': 32768}
+        elif family == 'anthropic' and not self._is_anthropic_adaptive_model(normalized_name):
+            if normalized_name.rsplit('/', 1)[-1].startswith(
+                ('claude-3-7-', 'claude-sonnet-4', 'claude-opus-4', 'claude-haiku-4')
+            ):
+                capabilities['budget_tokens'] = {'min': 1024, 'max': 32768}
+        elif family == 'qwen' and self._supports_qwen_thinking_budget(model_name):
+            capabilities['budget_tokens'] = {'min': 1, 'max': 32768}
+        elif family == 'gemini' and 'gemini-2.5' in normalized_name:
+            capabilities['budget_tokens'] = {
+                'min': 128 if 'pro' in normalized_name else (512 if 'flash-lite' in normalized_name else 1),
+                'max': 32768 if 'pro' in normalized_name else 24576,
+            }
         return capabilities
 
     def _build_reasoning_args(self, model: requester.RuntimeLLMModel) -> dict[str, typing.Any]:
@@ -599,14 +720,21 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         if level == 'provider_default':
             return {}
 
-        config = {'level': level}
+        config = reasoning.normalize_reasoning_config(self._reasoning_config_value(model))
         capabilities = self.get_reasoning_capabilities(model)
         try:
             reasoning.validate_reasoning_capabilities(config, capabilities, model.model_entity.name)
         except ValueError as exc:
-            raise errors.RequesterError(str(exc)) from exc
+            raise errors.RequesterError(str(exc), code='invalid_policy') from exc
 
         family = self._reasoning_family(model.model_entity.name, model)
+        budget = config.get('budget_tokens')
+        if budget is not None:
+            if self._uses_openrouter_protocol():
+                return {'extra_body': {'reasoning': {'max_tokens': budget}}}
+            if family == 'qwen':
+                return {'extra_body': {'enable_thinking': True, 'thinking_budget': budget}}
+            return {'thinking': {'type': 'enabled', 'budget_tokens': budget}}
         if level == 'disabled':
             if family in {'deepseek', 'kimi', 'mimo', 'doubao'}:
                 return {'extra_body': {'thinking': {'type': 'disabled'}}}
@@ -1058,7 +1186,7 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         if include_retry_params:
             if self.requester_cfg.get('drop_params'):
                 args['drop_params'] = self.requester_cfg['drop_params']
-            if self.requester_cfg.get('num_retries'):
+            if self.requester_cfg.get('num_retries') is not None:
                 args['num_retries'] = self.requester_cfg['num_retries']
             if self.requester_cfg.get('api_version'):
                 args['api_version'] = self.requester_cfg['api_version']
@@ -1066,24 +1194,102 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
 
     def _handle_litellm_error(self, e: Exception) -> None:
         """Convert LiteLLM exceptions to RequesterError. Never returns, always raises."""
-        # Check more specific exceptions first (they inherit from base exceptions)
-        if isinstance(e, litellm.ContextWindowExceededError):
-            raise errors.RequesterError(f'上下文长度超限: {str(e)}')
-        if isinstance(e, litellm.BadRequestError):
-            raise errors.RequesterError(f'请求参数错误: {str(e)}')
-        if isinstance(e, litellm.AuthenticationError):
-            raise errors.RequesterError(f'API key 无效: {str(e)}')
-        if isinstance(e, litellm.NotFoundError):
-            raise errors.RequesterError(f'模型或路径无效: {str(e)}')
-        if isinstance(e, litellm.RateLimitError):
-            raise errors.RequesterError(f'请求过于频繁或余额不足: {str(e)}')
-        if isinstance(e, litellm.Timeout):
-            raise errors.RequesterError(f'请求超时: {str(e)}')
-        if isinstance(e, litellm.APIConnectionError):
-            raise errors.RequesterError(f'连接错误: {str(e)}')
-        if isinstance(e, litellm.APIError):
-            raise errors.RequesterError(f'API 错误: {str(e)}')
-        raise errors.RequesterError(f'未知错误: {str(e)}')
+        if isinstance(e, errors.RequesterError):
+            raise e
+        metadata = error_metadata(e)
+        mappings = (
+            (litellm.ContextWindowExceededError, 'context_length', t('Context length exceeded'), False),
+            (litellm.AuthenticationError, 'authentication', t('Invalid API key'), False),
+            (litellm.NotFoundError, 'not_found', t('Invalid model or endpoint'), False),
+            (litellm.RateLimitError, 'rate_limit', t('Rate limit exceeded or insufficient balance'), True),
+            (litellm.Timeout, 'timeout', t('Request timed out'), True),
+            (litellm.APIConnectionError, 'connection', t('Connection failed'), True),
+            (litellm.BadRequestError, 'bad_request', t('Invalid request parameters'), False),
+        )
+        for exception_type, code, message, retryable in mappings:
+            if isinstance(e, exception_type):
+                body = getattr(e, 'body', None)
+                body = body if isinstance(body, dict) else {}
+                upstream_error = body.get('error', body)
+                upstream_code = upstream_error.get('code') if isinstance(upstream_error, dict) else None
+                if upstream_code in ('insufficient_quota', 'quota_exceeded', 'billing_hard_limit_reached'):
+                    code, retryable = 'quota_exceeded', False
+                raise errors.RequesterError(message, code=code, retryable=retryable, **metadata) from e
+        status = metadata['status_code']
+        raise errors.RequesterError(
+            t('API error') if isinstance(e, litellm.APIError) else t('Unknown error'),
+            code='upstream_error',
+            retryable=status in (408, 409, 429, 500, 502, 503, 504),
+            **metadata,
+        ) from e
+
+    def _apply_transport(self, args, model, extra_args):
+        state = current_call.get()
+        policy = (
+            state.policy
+            if state
+            else RequestPolicy.resolve(self.requester_cfg, model.model_entity.extra_args, extra_args)
+        )
+        args.pop('request_policy', None)
+        if state:
+            # The Host owns retries; do not multiply attempts in nested SDKs.
+            args['num_retries'] = 0
+            args['max_retries'] = 0
+            args['timeout'] = httpx.Timeout(
+                connect=policy.connect_timeout,
+                read=policy.read_timeout,
+                write=policy.write_timeout,
+                pool=policy.pool_timeout,
+            )
+            state.mark('request_prepared_ms')
+        base = args.get('api_base', '')
+        is_openrouter = urlsplit(base).hostname == 'openrouter.ai' or (
+            not base and self._get_custom_llm_provider() == 'openrouter'
+        )
+        headers = dict(args.get('extra_headers') or {})
+        existing = {key.lower() for key in headers}
+        if policy.attribution and is_openrouter:
+            if 'http-referer' not in existing:
+                headers['HTTP-Referer'] = 'https://langbot.app'
+            if not {'x-title', 'x-openrouter-title'} & existing:
+                headers['X-OpenRouter-Title'] = 'LangBot'
+        conversation_header = policy.conversation_header or ('x-session-id' if is_openrouter else None)
+        if state and state.conversation_id and conversation_header and conversation_header not in existing:
+            headers[conversation_header] = state.conversation_id
+        if headers:
+            args['extra_headers'] = headers
+
+    async def _completion(self, args):
+        state = current_call.get()
+        if state:
+            state.mark('request_started_ms')
+            result = await state.wait(
+                lambda: acompletion(**args),
+                'first_event' if args.get('stream') else 'response',
+                state.policy.first_event_timeout if args.get('stream') else state.policy.total_timeout,
+            )
+            state.mark('response_open_ms')
+            return result
+        return await acompletion(**args)
+
+    async def _raw_chunks(self, response):
+        state = current_call.get()
+        iterator = response.__aiter__()
+        while True:
+            try:
+                if state:
+                    chunk = await state.wait(
+                        lambda: anext(iterator),
+                        'stream_idle' if state.raw_started else 'first_event',
+                        state.policy.stream_idle_timeout if state.raw_started else state.policy.first_event_timeout,
+                    )
+                    state.raw_started = True
+                    state.mark('first_event_ms')
+                else:
+                    chunk = await anext(iterator)
+            except StopAsyncIteration:
+                return
+            yield chunk
 
     async def _build_completion_args(
         self,
@@ -1094,14 +1300,31 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         stream: bool = False,
     ) -> dict:
         """Build common completion arguments for invoke_llm and invoke_llm_stream."""
+        preparation_start = time.perf_counter()
+        state = current_call.get()
+        cache_preparation = state is not None and state.policy.retries > 0
+        scoped_model = copy.copy(model)
+        try:
+            scoped_model.reasoning_config_override = reasoning.resolve_reasoning_config(model, extra_args)
+        except ValueError as exc:
+            raise errors.RequesterError(str(exc), code='invalid_policy') from exc
+        model = scoped_model
         reasoning_family = self._reasoning_family(model.model_entity.name, model)
         reasoning_level = self._reasoning_level(model)
-        req_messages = self._convert_messages(
-            messages,
-            reasoning_family=reasoning_family,
-            include_reasoning_context=reasoning_level != 'disabled',
-        )
-        model_name = self._build_litellm_model_name(model.model_entity.name)
+        message_key = ('messages', id(messages), reasoning_family, reasoning_level != 'disabled')
+        if cache_preparation and message_key in state.prepared:
+            req_messages = copy.deepcopy(state.prepared[message_key])
+            state.metrics['preparation_cache_hits'] = state.metrics.get('preparation_cache_hits', 0) + 1
+        else:
+            req_messages = self._convert_messages(
+                messages, reasoning_family=reasoning_family, include_reasoning_context=reasoning_level != 'disabled'
+            )
+            if cache_preparation:
+                state.prepared[message_key] = copy.deepcopy(req_messages)
+        # OpenRouter shares requester metadata with OpenAI-compatible non-chat
+        # operations, but chat needs its own adapter to preserve cache markers.
+        chat_provider = 'openrouter' if self._uses_openrouter_protocol() else self._get_custom_llm_provider()
+        model_name = self._build_litellm_model_name(model.model_entity.name, chat_provider)
         api_key = model.provider.token_mgr.get_token()
 
         args = {
@@ -1114,10 +1337,14 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
             args['stream_options'] = {'include_usage': True}
         self._build_common_args(args)
 
+        if chat_provider == 'openrouter':
+            args['custom_llm_provider'] = chat_provider
+
         # Apply model-level extra_args first, then call-level extra_args
         if model.model_entity.extra_args:
             args.update(model.model_entity.extra_args)
         args.update(extra_args)
+        args.pop('reasoning_config', None)
 
         reasoning_args = self._build_reasoning_args(model)
         if reasoning_args:
@@ -1142,8 +1369,23 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                     raise errors.RequesterError('allowed_openai_params must be an array')
                 args['allowed_openai_params'] = list(dict.fromkeys([*allowed_openai_params, 'reasoning_effort']))
 
+        budget = model.reasoning_config_override.get('budget_tokens')
+        if budget is not None and reasoning_family == 'anthropic' and not self._uses_openrouter_protocol():
+            output_limit = args.get('max_completion_tokens', args.get('max_tokens'))
+            if type(output_limit) is not int or output_limit <= budget:
+                raise errors.RequesterError(
+                    'Anthropic budget_tokens requires explicit max_tokens greater than the budget',
+                    code='invalid_policy',
+                )
+
         if funcs:
-            tools = await self.ap.tool_mgr.generate_tools_for_openai(funcs)
+            tool_key = ('tools', id(funcs))
+            if cache_preparation and tool_key in state.prepared:
+                tools = copy.deepcopy(state.prepared[tool_key])
+            else:
+                tools = copy.deepcopy(await self.ap.tool_mgr.generate_tools_for_openai(funcs))
+                if cache_preparation:
+                    state.prepared[tool_key] = copy.deepcopy(tools)
             if tools:
                 args['tools'] = tools
                 args.setdefault('tool_choice', 'auto')
@@ -1152,6 +1394,15 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                 # and imports proxy-only dependencies such as FastAPI.
                 args['_skip_mcp_handler'] = True
 
+        cache_family = reasoning_family
+        if cache_family == 'openrouter':
+            cache_family = self._infer_reasoning_family_from_model_name(model.model_entity.name) or cache_family
+        apply_prompt_caching(args, state, provider=self._get_custom_llm_provider(), family=cache_family)
+        self._apply_transport(args, model, extra_args)
+        if state:
+            state.metrics['preparation_ms'] = round(
+                state.metrics.get('preparation_ms', 0) + (time.perf_counter() - preparation_start) * 1000, 3
+            )
         return args
 
     async def count_tokens(
@@ -1173,7 +1424,7 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
             count_args['tool_choice'] = args['tool_choice']
 
         try:
-            tokens = litellm.token_counter(**count_args)
+            tokens = await asyncio.to_thread(litellm.token_counter, **count_args)
         except Exception as e:
             self._handle_litellm_error(e)
 
@@ -1193,38 +1444,41 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         """Invoke LLM and return message with usage info."""
         args = await self._build_completion_args(model, messages, funcs, extra_args, stream=False)
 
-        try:
-            response = await acompletion(**args)
+        async with self._client_args(args) as args:
+            try:
+                response = await self._completion(args)
 
-            message_data = response.choices[0].message.model_dump()
-            if 'role' not in message_data or message_data['role'] is None:
-                message_data['role'] = 'assistant'
+                message_data = response.choices[0].message.model_dump()
+                if 'role' not in message_data or message_data['role'] is None:
+                    message_data['role'] = 'assistant'
 
-            content = message_data.get('content', '')
-            reasoning_content = message_data.get('reasoning_content', None)
-            thinking_blocks = message_data.get('thinking_blocks')
-            if reasoning_content or thinking_blocks:
-                provider_fields = dict(message_data.get('provider_specific_fields') or {})
-                if reasoning_content:
-                    provider_fields['reasoning_content'] = reasoning_content
-                if thinking_blocks:
-                    provider_fields['thinking_blocks'] = thinking_blocks
-                message_data['provider_specific_fields'] = provider_fields
-            display_reasoning = reasoning_content or self._thinking_blocks_text(thinking_blocks) or None
-            message_data['content'] = self._process_thinking_content(content, display_reasoning, remove_think)
+                content = message_data.get('content', '')
+                reasoning_content = message_data.get('reasoning_content', None)
+                thinking_blocks = message_data.get('thinking_blocks')
+                if reasoning_content or thinking_blocks:
+                    provider_fields = dict(message_data.get('provider_specific_fields') or {})
+                    if reasoning_content:
+                        provider_fields['reasoning_content'] = reasoning_content
+                    if thinking_blocks:
+                        provider_fields['thinking_blocks'] = thinking_blocks
+                    message_data['provider_specific_fields'] = provider_fields
+                display_reasoning = reasoning_content or self._thinking_blocks_text(thinking_blocks) or None
+                message_data['content'] = self._process_thinking_content(content, display_reasoning, remove_think)
 
-            if 'reasoning_content' in message_data:
-                del message_data['reasoning_content']
-            if 'thinking_blocks' in message_data:
-                del message_data['thinking_blocks']
+                if 'reasoning_content' in message_data:
+                    del message_data['reasoning_content']
+                if 'thinking_blocks' in message_data:
+                    del message_data['thinking_blocks']
 
-            message = provider_message.Message(**message_data)
-            usage_info = self._extract_usage(response)
+                message = provider_message.Message(**message_data)
+                usage_info = self._extract_usage(response)
+                if current_call.get():
+                    current_call.get().usage = usage_info
 
-            return message, usage_info
+                return message, usage_info
 
-        except Exception as e:
-            self._handle_litellm_error(e)
+            except Exception as e:
+                self._handle_litellm_error(e)
 
     async def invoke_llm_stream(
         self,
@@ -1245,132 +1499,151 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         reasoning_started = False
         reasoning_closed = False
         thinking_blocks_state: list[dict[str, typing.Any]] = []
+        response = None
 
-        try:
-            response = await acompletion(**args)
-            async for chunk in response:
-                # Capture usage whenever a chunk carries it.
-                #
-                # Important: many OpenAI-compatible gateways (e.g. new-api) and
-                # providers send the final usage payload in a chunk that STILL
-                # contains a (empty-delta) choice, not an empty `choices` list.
-                # The previous implementation only captured usage when `choices`
-                # was empty, so streamed calls always recorded 0 tokens.
-                # We therefore capture usage independently of `choices`, and then
-                # fall through to also process any content this chunk may carry.
-                if getattr(chunk, 'usage', None):
-                    usage_info = self._normalize_usage(chunk.usage)
-                    if query is not None:
-                        if query.variables is None:
-                            query.variables = {}
-                        query.variables[requester.STREAM_USAGE_QUERY_VARIABLE] = usage_info
+        async with self._client_args(args) as args:
+            try:
+                response = await self._completion(args)
+                async for chunk in self._raw_chunks(response):
+                    # Capture usage whenever a chunk carries it.
+                    #
+                    # Important: many OpenAI-compatible gateways (e.g. new-api) and
+                    # providers send the final usage payload in a chunk that STILL
+                    # contains a (empty-delta) choice, not an empty `choices` list.
+                    # The previous implementation only captured usage when `choices`
+                    # was empty, so streamed calls always recorded 0 tokens.
+                    # We therefore capture usage independently of `choices`, and then
+                    # fall through to also process any content this chunk may carry.
+                    if getattr(chunk, 'usage', None):
+                        if current_call.get():
+                            current_call.get().usage = self._extract_usage(chunk)
+                        usage_info = self._normalize_usage(chunk.usage)
+                        if query is not None:
+                            if query.variables is None:
+                                query.variables = {}
+                            query.variables[requester.STREAM_USAGE_QUERY_VARIABLE] = usage_info
 
-                if not hasattr(chunk, 'choices') or not chunk.choices:
-                    continue
+                    if not hasattr(chunk, 'choices') or not chunk.choices:
+                        continue
 
-                choice = chunk.choices[0]
-                delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
-                finish_reason = getattr(choice, 'finish_reason', None)
+                    choice = chunk.choices[0]
+                    delta = choice.delta.model_dump() if hasattr(choice, 'delta') else {}
+                    finish_reason = getattr(choice, 'finish_reason', None)
 
-                if 'role' in delta and delta['role']:
-                    role = delta['role']
+                    if 'role' in delta and delta['role']:
+                        role = delta['role']
 
-                delta_content = delta.get('content') or ''
-                reasoning_content = delta.get('reasoning_content') or ''
-                provider_fields = dict(delta.get('provider_specific_fields') or {})
-                if finish_reason:
-                    # Preserve an explicit provider stop even when the final delta is empty.
-                    provider_fields['finish_reason'] = finish_reason
-                raw_thinking_blocks = delta.get('thinking_blocks')
-                if raw_thinking_blocks:
-                    thinking_blocks_state = self._merge_thinking_blocks(thinking_blocks_state, raw_thinking_blocks)
-                    provider_fields['thinking_blocks'] = thinking_blocks_state
-                thinking_blocks_text = self._thinking_blocks_text(raw_thinking_blocks)
-                display_reasoning_content = reasoning_content or thinking_blocks_text
+                    state = current_call.get()
+                    if state and delta.get('content'):
+                        state.mark('first_content_ms')
+                    if state and (delta.get('reasoning_content') or delta.get('thinking_blocks')):
+                        state.mark('first_reasoning_ms')
+                    delta_content = delta.get('content') or ''
+                    reasoning_content = delta.get('reasoning_content') or ''
+                    provider_fields = dict(delta.get('provider_specific_fields') or {})
+                    if finish_reason:
+                        # Preserve an explicit provider stop even when the final delta is empty.
+                        provider_fields['finish_reason'] = finish_reason
+                    raw_thinking_blocks = delta.get('thinking_blocks')
+                    if raw_thinking_blocks:
+                        thinking_blocks_state = self._merge_thinking_blocks(thinking_blocks_state, raw_thinking_blocks)
+                        provider_fields['thinking_blocks'] = thinking_blocks_state
+                    thinking_blocks_text = self._thinking_blocks_text(raw_thinking_blocks)
+                    display_reasoning_content = reasoning_content or thinking_blocks_text
 
-                # Handle reasoning_content based on remove_think flag
-                if reasoning_content:
-                    provider_fields['reasoning_content'] = reasoning_content
-                    if remove_think:
-                        delta_content = delta_content or None
-                    else:
-                        # Stream explicit markers so downstream adapters and
-                        # the debug page see the same format as non-streaming
-                        # responses.
-                        if not reasoning_started:
-                            delta_content = '<think>\n'
-                            reasoning_started = True
+                    # Handle reasoning_content based on remove_think flag
+                    if reasoning_content:
+                        provider_fields['reasoning_content'] = reasoning_content
+                        if remove_think:
+                            delta_content = delta_content or None
                         else:
-                            delta_content = ''
-                        delta_content += display_reasoning_content
-                        if delta.get('content'):
-                            delta_content += f'\n</think>\n{delta.get("content")}'
-                            reasoning_closed = True
+                            # Stream explicit markers so downstream adapters and
+                            # the debug page see the same format as non-streaming
+                            # responses.
+                            if not reasoning_started:
+                                delta_content = '<think>\n'
+                                reasoning_started = True
+                            else:
+                                delta_content = ''
+                            delta_content += display_reasoning_content
+                            if delta.get('content'):
+                                delta_content += f'\n</think>\n{delta.get("content")}'
+                                reasoning_closed = True
 
-                elif display_reasoning_content:
-                    if remove_think:
-                        delta_content = delta_content or None
-                    else:
-                        if not reasoning_started:
-                            delta_content = '<think>\n'
-                            reasoning_started = True
+                    elif display_reasoning_content:
+                        if remove_think:
+                            delta_content = delta_content or None
                         else:
-                            delta_content = ''
-                        delta_content += display_reasoning_content
-                        if delta.get('content'):
-                            delta_content += f'\n</think>\n{delta.get("content")}'
-                            reasoning_closed = True
+                            if not reasoning_started:
+                                delta_content = '<think>\n'
+                                reasoning_started = True
+                            else:
+                                delta_content = ''
+                            delta_content += display_reasoning_content
+                            if delta.get('content'):
+                                delta_content += f'\n</think>\n{delta.get("content")}'
+                                reasoning_closed = True
 
-                elif delta_content and not remove_think and reasoning_started and not reasoning_closed:
-                    delta_content = f'\n</think>\n{delta_content}'
-                    reasoning_closed = True
+                    elif delta_content and not remove_think and reasoning_started and not reasoning_closed:
+                        delta_content = f'\n</think>\n{delta_content}'
+                        reasoning_closed = True
 
-                if finish_reason and not remove_think and reasoning_started and not reasoning_closed:
-                    delta_content = f'{delta_content}\n</think>\n'
-                    reasoning_closed = True
+                    if finish_reason and not remove_think and reasoning_started and not reasoning_closed:
+                        delta_content = f'{delta_content}\n</think>\n'
+                        reasoning_closed = True
 
-                if think_state is not None and delta_content:
-                    delta_content = think_state.feed(delta_content)
+                    if think_state is not None and delta_content:
+                        delta_content = think_state.feed(delta_content)
 
-                tool_calls = self._normalize_stream_tool_calls(delta.get('tool_calls'), tool_call_state)
+                    tool_calls = self._normalize_stream_tool_calls(delta.get('tool_calls'), tool_call_state)
 
-                if not delta_content and not tool_calls and not provider_fields and not finish_reason:
+                    if not delta_content and not tool_calls and not provider_fields and not finish_reason:
+                        chunk_idx += 1
+                        continue
+
+                    chunk_data: dict[str, typing.Any] = {
+                        'role': role,
+                        'content': delta_content if delta_content else None,
+                        'tool_calls': tool_calls,
+                        'is_final': bool(finish_reason),
+                    }
+
+                    # Preserve provider_specific_fields from delta (e.g., Gemini thought_signatures)
+                    if provider_fields:
+                        chunk_data['provider_specific_fields'] = provider_fields
+
+                    chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
+                    yield provider_message.MessageChunk(**chunk_data)
                     chunk_idx += 1
-                    continue
 
-                chunk_data: dict[str, typing.Any] = {
-                    'role': role,
-                    'content': delta_content if delta_content else None,
-                    'tool_calls': tool_calls,
-                    'is_final': bool(finish_reason),
-                }
-
-                # Preserve provider_specific_fields from delta (e.g., Gemini thought_signatures)
-                if provider_fields:
-                    chunk_data['provider_specific_fields'] = provider_fields
-
-                chunk_data = {k: v for k, v in chunk_data.items() if v is not None}
-                yield provider_message.MessageChunk(**chunk_data)
-                chunk_idx += 1
-
-            if reasoning_started and not reasoning_closed:
-                yield provider_message.MessageChunk(
-                    role=role,
-                    content='\n</think>\n',
-                    is_final=True,
-                )
-
-            if think_state is not None:
-                pending_content = think_state.flush()
-                if pending_content:
+                if reasoning_started and not reasoning_closed:
                     yield provider_message.MessageChunk(
                         role=role,
-                        content=pending_content,
+                        content='\n</think>\n',
                         is_final=True,
                     )
 
-        except Exception as e:
-            self._handle_litellm_error(e)
+                if think_state is not None:
+                    pending_content = think_state.flush()
+                    if pending_content:
+                        yield provider_message.MessageChunk(
+                            role=role,
+                            content=pending_content,
+                            is_final=True,
+                        )
+
+            except Exception as e:
+                self._handle_litellm_error(e)
+
+            finally:
+                # Closing a downstream async generator must release the provider's
+                # HTTP stream immediately, including cancellation while at yield.
+                close = getattr(response, 'aclose', None)
+                if callable(close):
+                    try:
+                        await asyncio.wait_for(close(), timeout=5.0)
+                    except Exception as exc:
+                        self.ap.logger.debug(f'Failed to close LLM stream: {type(exc).__name__}')
 
     async def invoke_embedding(
         self,
@@ -1543,9 +1816,9 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                 body = await httpclient.response_text(e.response)
             except Exception:
                 pass
-            raise errors.RequesterError(f'rerank 请求失败 (HTTP {e.response.status_code}): {body or str(e)}')
+            raise errors.RequesterError(t('API error') + f' (HTTP {e.response.status_code}): {body or str(e)}')
         except httpx.HTTPError as e:
-            raise errors.RequesterError(f'rerank 连接错误: {str(e)}')
+            raise errors.RequesterError(t('Connection failed') + f': {e}')
 
         raw_results = data.get('results', []) if isinstance(data, dict) else []
         results = []
diff --git a/src/langbot/pkg/provider/session/sessionmgr.py b/src/langbot/pkg/provider/session/sessionmgr.py
index 74b9a8c40..4e4f1076b 100644
--- a/src/langbot/pkg/provider/session/sessionmgr.py
+++ b/src/langbot/pkg/provider/session/sessionmgr.py
@@ -271,8 +271,22 @@ class SessionManager:
     async def get_session(self, query: pipeline_query.Query) -> provider_session.Session:
         """获取会话"""
         session_key, execution_context = _query_session_key(query)
+        from ...agent.runner.context_reset import get_reset_generation
+
+        generation = None
+        persistence = getattr(self.ap, 'persistence_mgr', None)
+        if persistence is not None:
+            generation = await get_reset_generation(
+                persistence.get_db_engine(),
+                execution_context.workspace_uuid,
+                query.bot_uuid,
+                f'{str(query.launcher_type.value).lower()}_{query.launcher_id}',
+            )
         now = time.monotonic()
         session = self._session_index.get(session_key)
+        if session is not None and getattr(session, '_context_generation', None) != generation:
+            self._remove_session(session)
+            session = None
         if session is not None:
             self._touch_session(session, session_key, now)
             return session
@@ -319,6 +333,7 @@ class SessionManager:
         object.__setattr__(session, 'bot_uuid', query.bot_uuid)
         object.__setattr__(session, '_execution_context', session_context)
         object.__setattr__(session, '_langbot_session_key', session_key)
+        object.__setattr__(session, '_context_generation', generation)
         object.__setattr__(session, '_langbot_session_concurrency', session_concurrency)
         session._semaphore = asyncio.Semaphore(session_concurrency)
         self._session_index[session_key] = session
diff --git a/src/langbot/pkg/provider/tools/loaders/mcp.py b/src/langbot/pkg/provider/tools/loaders/mcp.py
index 48501292b..3cb8b8cf5 100644
--- a/src/langbot/pkg/provider/tools/loaders/mcp.py
+++ b/src/langbot/pkg/provider/tools/loaders/mcp.py
@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+from ....i18n import t
+
 import dataclasses
 import enum
 import json
@@ -1464,9 +1466,6 @@ class RuntimeMCPSession:
         if query is None:
             return
         try:
-            from langbot.pkg.telemetry import features as telemetry_features
-
-            telemetry_features.increment(query, 'mcp_resource_reads', envelope.get('source') or 'unknown')
             query.variables.setdefault(MCP_RESOURCE_TRACE_QUERY_KEY, []).append(
                 {
                     'server_name': envelope.get('server_name'),
@@ -1493,7 +1492,7 @@ class RuntimeMCPSession:
             # Raw transport exceptions may echo command arguments, headers, or
             # environment values. Detailed diagnostics belong in AUDIT_VIEW
             # logs; resource-list responses expose only a stable status.
-            'error_message': 'MCP runtime failed' if self.error_message else None,
+            'error_message': t('MCP runtime failed') if self.error_message else None,
             'error_code': self._public_error_code if self.error_message else None,
             'error_phase': self.error_phase.value if self.error_phase else None,
             'retry_count': self.retry_count,
diff --git a/src/langbot/pkg/provider/tools/loaders/native.py b/src/langbot/pkg/provider/tools/loaders/native.py
index 74b067320..9d03deb26 100644
--- a/src/langbot/pkg/provider/tools/loaders/native.py
+++ b/src/langbot/pkg/provider/tools/loaders/native.py
@@ -313,7 +313,7 @@ class NativeToolLoader(loader.ToolLoader):
             include_visible=False,
             include_activated=True,
         )
-        referenced_skill_names = skill_loader.find_referenced_skill_names(command)
+        referenced_skill_names = skill_loader.find_referenced_skill_names(command, query)
 
         if selected_skill is None and referenced_skill_names:
             if len(referenced_skill_names) > 1:
@@ -334,8 +334,13 @@ class NativeToolLoader(loader.ToolLoader):
                 raise ValueError(f'Activated skill "{selected_skill_name}" has no package_root.')
 
             # Wrap command with Python venv bootstrap if the skill has a Python project.
-            # The venv is created inside the skill's mount path.
-            skill_mount = f'/workspace/.skills/{selected_skill_name}'
+            # Dependency state stays writable and separate from the published package.
+            skill_mount = skill_loader.get_virtual_skill_mount_path(selected_skill_name, selected_skill['revision'])
+            legacy_mount = skill_loader.get_virtual_skill_mount_path(selected_skill_name)
+            command = regex.sub(regex.escape(legacy_mount) + r'(?![A-Za-z0-9_-])', skill_mount, command)
+            if workdir == legacy_mount or workdir.startswith(legacy_mount + '/'):
+                workdir = skill_mount + workdir[len(legacy_mount) :]
+            parameters = {**parameters, 'command': command, 'workdir': workdir}
             python_project = selected_skill.get('python_project') is True
             if 'python_project' not in selected_skill:
                 python_project = skill_loader.should_prepare_skill_python_env(package_root)
@@ -350,9 +355,7 @@ class NativeToolLoader(loader.ToolLoader):
                     state_path=f'/workspace/.skill-envs/{selected_skill_name}/{revision_key}',
                 )
 
-        # All exec calls (with or without skills) go through the same container
-        # via execute_tool. Skills are mounted at /workspace/.skills/{name}/
-        # via read-only mounts composed by the Core skill loader.
+        # Package paths contain the pinned digest, so no shared alias changes underneath another run.
         result = await self.ap.box_service.execute_tool(
             parameters,
             query,
@@ -744,9 +747,13 @@ class NativeToolLoader(loader.ToolLoader):
         return f'{str(base).rstrip("/")}/{relative}'
 
     async def _run_workspace_file_script(self, script: str, query: pipeline_query.Query) -> dict:
+        # Sandbox images built from a host rootfs may only ship `python3`, and a
+        # read-only /usr prevents creating a `python` shim inside the jail, so
+        # resolve the interpreter instead of assuming its name.
+        command = f'PYTHON_BIN=$(command -v python3 || command -v python); "$PYTHON_BIN" - <<\'PY\'\n{script}\nPY'
         result = await self.ap.box_service.execute_tool(
             {
-                'command': f"python - <<'PY'\n{script}\nPY",
+                'command': command,
                 'timeout_sec': 30,
             },
             query,
diff --git a/src/langbot/pkg/provider/tools/loaders/skill.py b/src/langbot/pkg/provider/tools/loaders/skill.py
index 7ddaab86d..233268891 100644
--- a/src/langbot/pkg/provider/tools/loaders/skill.py
+++ b/src/langbot/pkg/provider/tools/loaders/skill.py
@@ -4,6 +4,8 @@ import os
 import re
 import typing
 
+from langbot_plugin.box.models import ARTIFACT_MOUNT_ROOT
+
 from ....api.http.context import ExecutionContext
 from ....utils.python_workspace import (
     should_prepare_python_env,
@@ -21,7 +23,9 @@ SKILL_MOUNT_PREFIX = '/workspace/.skills'
 _SKILL_MOUNT_PATTERN = re.compile(r'/workspace/\.skills/([A-Za-z0-9_-]+)')
 
 
-def get_virtual_skill_mount_path(skill_name: str) -> str:
+def get_virtual_skill_mount_path(skill_name: str, revision: str | None = None) -> str:
+    if revision is not None:
+        return f'{ARTIFACT_MOUNT_ROOT}/{revision.removeprefix("sha256:")}'
     return f'{SKILL_MOUNT_PREFIX}/{skill_name}'
 
 
@@ -96,7 +100,7 @@ def build_execution_mounts(ap: app.Application, query: pipeline_query.Query) ->
         mounts.append(
             {
                 'host_path': package_root,
-                'mount_path': get_virtual_skill_mount_path(skill_name),
+                'mount_path': get_virtual_skill_mount_path(skill_name, revision),
                 'mode': 'ro',
                 'content_digest': revision,
                 'manifest_path': manifest_path,
@@ -151,150 +155,37 @@ def get_activated_skill_names(query: pipeline_query.Query) -> list[str]:
     return normalize_skill_names(list(get_activated_skills(query).keys()))
 
 
-def get_activated_skill_bindings(query: pipeline_query.Query) -> list[dict[str, str]]:
-    """Return exact bindings suitable for persistence and restart recovery."""
+async def get_persisted_skill_revision(ap: app.Application, query: pipeline_query.Query, name: str) -> str | None:
+    """Read a pin only for the currently authorized Host run."""
+    session = getattr(query, '_agent_run_session', None)
+    if not isinstance(session, dict):
+        return None
+    from ....agent.runner.run_ledger_store import RunLedgerStore, TERMINAL_STATUSES
 
-    return [
-        {'name': name, 'revision': str(skill.get('revision', '') or '')}
-        for name, skill in get_activated_skills(query).items()
-    ]
-
-
-def normalize_skill_bindings(value: typing.Any) -> list[dict[str, str]]:
-    if not isinstance(value, list):
-        raise ValueError('Activated Skill recovery requires revision bindings.')
-    bindings: list[dict[str, str]] = []
-    seen: dict[str, str] = {}
-    for item in value:
-        if not isinstance(item, dict):
-            raise ValueError('Activated Skill recovery cannot use names without revisions.')
-        name = str(item.get('name', '') or '').strip()
-        revision = str(item.get('revision', '') or '').strip()
-        if not name or not revision:
-            raise ValueError('Each activated Skill binding requires name and revision.')
-        previous = seen.get(name)
-        if previous is not None and previous != revision:
-            raise ValueError(f'Conflicting pinned revisions supplied for Skill "{name}".')
-        if previous is None:
-            seen[name] = revision
-            bindings.append({'name': name, 'revision': revision})
-    return bindings
-
-
-async def restore_activated_skills(
-    ap: app.Application,
-    query: pipeline_query.Query,
-    skill_bindings: typing.Any,
-) -> list[str]:
-    """Restore exact revisions or fail explicitly; never resolve latest by name."""
-
-    repository = getattr(ap, 'skill_repository', None)
-    if repository is None:
-        raise ValueError('Skill repository is unavailable during run recovery.')
-    context = ExecutionContext(
-        instance_uuid=str(getattr(query, 'instance_uuid', '') or ''),
-        workspace_uuid=str(getattr(query, 'workspace_uuid', '') or ''),
-        placement_generation=getattr(query, 'placement_generation', 0) or 0,
-        bot_uuid=getattr(query, 'bot_uuid', None),
-        pipeline_uuid=getattr(query, 'pipeline_uuid', None),
-        query_uuid=getattr(query, 'query_uuid', None),
-    )
-    bound_names = get_bound_skill_names(query)
-    restored: list[str] = []
-    for binding in normalize_skill_bindings(skill_bindings):
-        skill_name = binding['name']
-        if bound_names is not None and skill_name not in bound_names:
-            raise ValueError(f'Skill "{skill_name}" is no longer authorized for this recoverable run.')
-        skill_data = await repository.get_skill(
-            context,
-            skill_name,
-            snapshot=True,
-            revision=binding['revision'],
-        )
-        if skill_data is None:
-            raise ValueError(f'Skill "{skill_name}" pinned revision {binding["revision"]} is unavailable.')
-        register_activated_skill(query, skill_data)
-        restored.append(skill_name)
-    return restored
-
-
-async def restore_activated_skills_from_state(
-    ap: app.Application,
-    query: pipeline_query.Query,
-    state: dict[str, dict[str, typing.Any]],
-) -> list[str]:
-    """Restore the exact activated Skill revisions stored by the Host."""
-    conversation_state = state.get('conversation', {}) if isinstance(state, dict) else {}
-    if not isinstance(conversation_state, dict):
-        raise ValueError('Activated Skill conversation state must be an object.')
-    if ACTIVATED_SKILL_NAMES_STATE_KEY not in conversation_state:
-        return []
-    bindings = conversation_state[ACTIVATED_SKILL_NAMES_STATE_KEY]
-    if isinstance(bindings, list) and bindings and all(isinstance(item, str) for item in bindings):
-        logger = getattr(ap, 'logger', None)
-        if logger is not None:
-            workspace_uuid = getattr(query, 'workspace_uuid', None) or getattr(
-                getattr(query, '_execution_context', None), 'workspace_uuid', 'unknown'
-            )
-            logger.warning(
-                f'Legacy name-only activated Skill state in Workspace {workspace_uuid}; '
-                'reactivate Skills to pin their revisions.'
-            )
-        return []
-    return await restore_activated_skills(ap, query, bindings)
+    context = getattr(query, '_execution_context', None)
+    workspace_id = str(getattr(context, 'workspace_uuid', None) or getattr(query, 'workspace_uuid', '') or '')
+    run = await RunLedgerStore(ap.persistence_mgr.get_db_engine()).get_run(str(session['run_id']))
+    if run is None or run['workspace_id'] != workspace_id or run['status'] in TERMINAL_STATUSES:
+        raise ValueError('Skill recovery requires an active run in the current Workspace')
+    return (run.get('metadata') or {}).get('activated_skills', {}).get(name)
 
 
 async def persist_activated_skill(
     ap: app.Application,
     query: pipeline_query.Query,
     skill_name: str,
-) -> None:
-    """Persist activated Skill revision bindings into Host-owned conversation state.
-
-    ``activate`` runs host-side. This writes the run's pinned Skill revisions
-    to the conversation-scope ``host.activated_skills`` snapshot so a later
-    run can restore them via ``restore_activated_skills_from_state``. Host writes
-    here and a runner ``state.updated`` to the same key follow last-write-wins.
-
-    Best-effort: a persistence failure must not fail the activation itself. No-op
-    when the call is not inside an authorized agent run, or when conversation
-    state is unavailable (state disabled / scope not enabled / no conversation).
-    """
+    revision: str,
+) -> str:
+    """Persist one pin in the run ledger, never in Runner-writable conversation state."""
     session = getattr(query, '_agent_run_session', None)
     if not isinstance(session, dict):
-        return
+        return revision
+    from ....agent.runner.run_ledger_store import RunLedgerStore
 
-    authorization = session.get('authorization')
-    if not isinstance(authorization, dict):
-        return
-
-    state_context = authorization.get('state_context')
-    if not isinstance(state_context, dict):
-        return
-
-    scope_keys = state_context.get('scope_keys')
-    conversation_scope_key = scope_keys.get('conversation') if isinstance(scope_keys, dict) else None
-    if not conversation_scope_key:
-        return
-
-    try:
-        from ....agent.runner.persistent_state_store import get_persistent_state_store
-
-        store = get_persistent_state_store(ap.persistence_mgr.get_db_engine())
-        await store.state_set(
-            scope_key=conversation_scope_key,
-            state_key=ACTIVATED_SKILL_NAMES_STATE_KEY,
-            value=get_activated_skill_bindings(query),
-            runner_id=str(session.get('runner_id', '') or ''),
-            binding_identity=str(state_context.get('binding_identity', 'unknown') or 'unknown'),
-            scope='conversation',
-            context=state_context,
-            logger=getattr(ap, 'logger', None),
-        )
-    except Exception as e:  # noqa: BLE001 - persistence is best-effort, must not break activation
-        logger = getattr(ap, 'logger', None)
-        if logger is not None:
-            logger.warning(f'Failed to persist activated skill "{skill_name}": {e}')
+    context = getattr(query, '_execution_context', None)
+    workspace_id = str(getattr(context, 'workspace_uuid', None) or getattr(query, 'workspace_uuid', '') or '')
+    store = RunLedgerStore(ap.persistence_mgr.get_db_engine())
+    return await store.pin_skill_revision(str(session['run_id']), workspace_id, skill_name, revision)
 
 
 def parse_skill_mount_path(sandbox_path: str) -> tuple[str | None, str]:
@@ -324,6 +215,12 @@ def resolve_virtual_skill_path(
     include_visible: bool,
     include_activated: bool,
 ) -> tuple[dict | None, str]:
+    if sandbox_path == ARTIFACT_MOUNT_ROOT or sandbox_path.startswith(ARTIFACT_MOUNT_ROOT + '/'):
+        for skill in get_activated_skills(query).values():
+            root = get_virtual_skill_mount_path(skill['name'], skill['revision'])
+            if sandbox_path == root or sandbox_path.startswith(root + '/'):
+                return skill, '/workspace' + sandbox_path[len(root) :]
+        raise ValueError('Artifact path does not belong to a Skill activated by this run')
     skill_name, rewritten_path = parse_skill_mount_path(sandbox_path)
     if skill_name is None:
         return None, rewritten_path
@@ -346,7 +243,7 @@ def resolve_virtual_skill_path(
     )
 
 
-def find_referenced_skill_names(text: str) -> list[str]:
+def find_referenced_skill_names(text: str, query=None) -> list[str]:
     if not text:
         return []
 
@@ -354,6 +251,20 @@ def find_referenced_skill_names(text: str) -> list[str]:
     for match in _SKILL_MOUNT_PATTERN.findall(text):
         if match not in seen:
             seen.append(match)
+    if query is not None:
+        for path in re.findall(re.escape(ARTIFACT_MOUNT_ROOT) + r'/[a-f0-9]{64}', text):
+            selected = next(
+                (
+                    name
+                    for name, skill in get_activated_skills(query).items()
+                    if get_virtual_skill_mount_path(name, skill['revision']) == path
+                ),
+                None,
+            )
+            if selected is None:
+                raise ValueError('Artifact command references a revision not activated by this run')
+            if selected not in seen:
+                seen.append(selected)
     return seen
 
 
diff --git a/src/langbot/pkg/provider/tools/loaders/skill_authoring.py b/src/langbot/pkg/provider/tools/loaders/skill_authoring.py
index bbe566fa7..3e8e0060f 100644
--- a/src/langbot/pkg/provider/tools/loaders/skill_authoring.py
+++ b/src/langbot/pkg/provider/tools/loaders/skill_authoring.py
@@ -63,7 +63,7 @@ class SkillToolLoader(loader.ToolLoader):
         if not self._read_only_tools:
             await self.initialize()
         tools = list(self._read_only_tools)
-        if sandbox_available:
+        if sandbox_available and self._skill_registration_available():
             tools.extend(self._sandbox_tools)
         return tools
 
@@ -73,10 +73,14 @@ class SkillToolLoader(loader.ToolLoader):
                 return tool
         return None
 
+    def _skill_registration_available(self) -> bool:
+        """Host skill scanning is disabled for Cloud-managed sandboxes."""
+        return not bool(getattr(getattr(self.ap, 'box_service', None), 'managed_admission_required', False))
+
     async def has_tool(self, name: str, *, sandbox_available: bool | None = None) -> bool:
         if not self._is_available() or name not in SKILL_TOOL_NAMES:
             return False
-        return name in READ_ONLY_SKILL_TOOL_NAMES or bool(sandbox_available)
+        return name in READ_ONLY_SKILL_TOOL_NAMES or (bool(sandbox_available) and self._skill_registration_available())
 
     @staticmethod
     def is_sandbox_tool(name: str) -> bool:
@@ -158,28 +162,34 @@ class SkillToolLoader(loader.ToolLoader):
 
         from . import skill as skill_loader
 
-        visible_skill = skill_loader.get_visible_skill(self.ap, query, skill_name)
-        if visible_skill is None:
-            visible_skills = skill_loader.get_visible_skills(self.ap, query)
-            available_names = ', '.join(sorted(visible_skills.keys())) or 'none'
-            raise ValueError(f'Skill "{skill_name}" not found. Available skills: {available_names}')
-
-        skill_data = await self.ap.skill_repository.get_skill(
-            self._execution_context(query),
-            skill_name,
-            snapshot=True,
-        )
+        bound_names = skill_loader.get_bound_skill_names(query)
+        if bound_names is not None and skill_name not in bound_names:
+            raise ValueError(f'Skill "{skill_name}" is not authorized for this run')
+        skill_data = skill_loader.get_activated_skill(query, skill_name)
         if skill_data is None:
-            raise ValueError(f'Skill "{skill_name}" is no longer available; reload the skill catalog.')
+            pinned_revision = await skill_loader.get_persisted_skill_revision(self.ap, query, skill_name)
+            if pinned_revision is None and skill_loader.get_visible_skill(self.ap, query, skill_name) is None:
+                raise ValueError(f'Skill "{skill_name}" not found in the authorized catalog')
+            skill_data = await self.ap.skill_repository.get_skill(
+                self._execution_context(query), skill_name, snapshot=True, revision=pinned_revision
+            )
+            if skill_data is None:
+                raise ValueError(f'Skill "{skill_name}" revision is unavailable')
 
-        # Keep the first revision pinned if the Skill is activated again in this run.
+        # Persist before exposing the activation; a failed write must not produce an unrecorded pin.
+        pinned_revision = await skill_loader.persist_activated_skill(self.ap, query, skill_name, skill_data['revision'])
+        if pinned_revision != skill_data['revision']:
+            skill_data = await self.ap.skill_repository.get_skill(
+                self._execution_context(query), skill_name, snapshot=True, revision=pinned_revision
+            )
+            if skill_data is None:
+                raise ValueError('The run-pinned Skill revision is unavailable')
         skill_data = skill_loader.register_activated_skill(query, skill_data)
-        await skill_loader.persist_activated_skill(self.ap, query, skill_name)
 
         instructions = skill_data.get('instructions', '')
         revision = str(skill_data.get('revision', '') or '')
         execution_available = await self._execution_available(query)
-        mount_path = skill_loader.get_virtual_skill_mount_path(skill_name) if execution_available else None
+        mount_path = skill_loader.get_virtual_skill_mount_path(skill_name, revision) if execution_available else None
 
         result_content = f'<command-message>The "{skill_name}" skill is activated</command-message>\n'
         result_content += '<skill-activation>\n'
@@ -191,9 +201,7 @@ class SkillToolLoader(loader.ToolLoader):
         result_content += '\n## Runtime Context\n'
         result_content += '- Use `list_skill_resources` and `read_skill_resource` for read-only package resources.\n'
         if execution_available:
-            result_content += (
-                f'- Box execution is available; executable package files will be mounted at {mount_path}.\n'
-            )
+            result_content += f'- Package files will be prepared read-only at {mount_path} before execution.\n'
         else:
             result_content += (
                 '- Box execution is unavailable. Do not attempt to run scripts or modify Workspace files.\n'
diff --git a/src/langbot/pkg/provider/tools/toolmgr.py b/src/langbot/pkg/provider/tools/toolmgr.py
index e3fb26519..b4116a9e0 100644
--- a/src/langbot/pkg/provider/tools/toolmgr.py
+++ b/src/langbot/pkg/provider/tools/toolmgr.py
@@ -1,9 +1,8 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import typing
 import time
+import asyncio
 import inspect
 from typing import TYPE_CHECKING
 
@@ -463,6 +462,19 @@ class ToolManager:
         start_time = time.perf_counter()
         try:
             result = await invoke()
+        except asyncio.CancelledError:
+            duration_ms = int((time.perf_counter() - start_time) * 1000)
+            await self._record_tool_call(
+                name=name,
+                source=source,
+                parameters=parameters,
+                query=query,
+                duration_ms=duration_ms,
+                status='error',
+                error_message='cancelled',
+            )
+            self._trace_tool_call(source=source, name=name, query=query, outcome='cancelled', error='cancelled')
+            raise
         except Exception as e:
             duration_ms = int((time.perf_counter() - start_time) * 1000)
             await self._record_tool_call(
@@ -474,6 +486,14 @@ class ToolManager:
                 status='error',
                 error_message=str(e),
             )
+            outcome = 'timeout' if isinstance(e, TimeoutError) else 'failed'
+            self._trace_tool_call(
+                source=source,
+                name=name,
+                query=query,
+                outcome=outcome,
+                error=str(e) or type(e).__name__,
+            )
             raise
 
         duration_ms = int((time.perf_counter() - start_time) * 1000)
@@ -486,9 +506,60 @@ class ToolManager:
             status='success',
             result=result,
         )
+        self._trace_tool_call(source=source, name=name, query=query, outcome='success')
         return result
 
-    @diagnostics.observe('api', 'tool.execute', source='agent', stage='execute')
+    def _trace_tool_call(
+        self,
+        *,
+        source: str,
+        name: str,
+        query: pipeline_query.Query,
+        outcome: str,
+        error: str = '',
+    ) -> None:
+        """Report one tool invocation as its own step of the execution chain.
+
+        Every tool call a Run makes - sandbox, skill, plugin or MCP - is the
+        observable action of that Run, so the chain shows it under the step that
+        issued it instead of collapsing the Run into a single opaque node.
+
+        An Agent's tool call is executed for the Runtime's RPC task, which never
+        inherited the chain context. The Run id the caller already carries is the
+        alias that task's record resolves the owning execution through.
+        """
+        # Imported here: telemetry pulls the application, and the tool manager is
+        # constructed while the application is still assembling.
+        from ...telemetry.execution import record as record_execution
+        from ...telemetry.platform import processing_mode
+
+        try:
+            record_execution(
+                self.ap,
+                get_query_execution_context(query),
+                execution_id=self._trace_execution_id(query),
+                family='tool',
+                operation=name,
+                adapter=source,
+                mode=processing_mode.get(),
+                outcome=outcome,
+                error=error,
+            )
+        except Exception as e:
+            self.ap.logger.warning(f'Failed to record tool call trace: {e}')
+
+    @staticmethod
+    def _trace_execution_id(query: pipeline_query.Query) -> str:
+        """The execution chain this call belongs to, when the caller knows it.
+
+        ``''`` lets the emitter fall back to the task's own trace context, which
+        is what a Pipeline-lane tool call has.
+        """
+        session = getattr(query, '_agent_run_session', None)
+        if not isinstance(session, dict):
+            return ''
+        return str(session.get('run_id') or '').strip()
+
     async def execute_func_call(
         self,
         name: str,
@@ -496,8 +567,6 @@ class ToolManager:
         query: pipeline_query.Query,
         source_ref: ToolSourceRef | None = None,
     ) -> typing.Any:
-        from langbot.pkg.telemetry import features as telemetry_features
-
         source_ref = source_ref or self.get_query_tool_source(query, name)
         if source_ref is not None:
             execution_context = get_query_execution_context(query)
@@ -553,7 +622,6 @@ class ToolManager:
                     return await loader.invoke_tool(name, parameters, query, source_id=source_id)
                 return await loader.invoke_tool(name, parameters, query)
 
-            telemetry_features.increment(query, 'tool_calls', telemetry_source)
             return await self._invoke_tool_with_monitoring(
                 source=telemetry_source,
                 name=name,
@@ -566,7 +634,6 @@ class ToolManager:
         await self._bind_plugin_workspace(execution_context)
         sandbox_available = await self._workspace_sandbox_available(execution_context)
         if sandbox_available and await self.native_tool_loader.has_tool(name):
-            telemetry_features.increment(query, 'tool_calls', 'native')
             return await self._invoke_tool_with_monitoring(
                 source='native',
                 name=name,
@@ -575,7 +642,6 @@ class ToolManager:
                 invoke=lambda: self.native_tool_loader.invoke_tool(name, parameters, query),
             )
         if await self.plugin_tool_loader.has_tool(name):
-            telemetry_features.increment(query, 'tool_calls', 'plugin')
             return await self._invoke_tool_with_monitoring(
                 source='plugin',
                 name=name,
@@ -584,7 +650,6 @@ class ToolManager:
                 invoke=lambda: self.plugin_tool_loader.invoke_tool(name, parameters, query),
             )
         if await self.mcp_tool_loader.has_tool(execution_context, name):
-            telemetry_features.increment(query, 'tool_calls', 'mcp')
             return await self._invoke_tool_with_monitoring(
                 source='mcp',
                 name=name,
@@ -596,7 +661,6 @@ class ToolManager:
             variables = getattr(query, 'variables', None)
             if isinstance(variables, dict):
                 variables['_skill_execution_available'] = sandbox_available
-            telemetry_features.increment(query, 'tool_calls', 'skill')
             return await self._invoke_tool_with_monitoring(
                 source='skill',
                 name=name,
diff --git a/src/langbot/pkg/rag/service/runtime.py b/src/langbot/pkg/rag/service/runtime.py
index b8b814f27..a685dc521 100644
--- a/src/langbot/pkg/rag/service/runtime.py
+++ b/src/langbot/pkg/rag/service/runtime.py
@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-from langbot.pkg.telemetry import diagnostics
-
 import posixpath
 import re
 from typing import TYPE_CHECKING, Any
@@ -73,7 +71,6 @@ class RAGRuntimeService:
             raise WorkspaceNotFoundError('Knowledge base not found')
         return kb_uuid
 
-    @diagnostics.observe('api', 'knowledge.vector_upsert', source='agent')
     async def vector_upsert(
         self,
         execution_context: ExecutionContext,
@@ -101,7 +98,6 @@ class RAGRuntimeService:
             documents=documents,
         )
 
-    @diagnostics.observe('api', 'knowledge.vector_search', source='agent')
     async def vector_search(
         self,
         execution_context: ExecutionContext,
@@ -126,7 +122,6 @@ class RAGRuntimeService:
             vector_weight=vector_weight,
         )
 
-    @diagnostics.observe('api', 'knowledge.vector_delete', source='agent')
     async def vector_delete(
         self,
         execution_context: ExecutionContext,
@@ -163,7 +158,6 @@ class RAGRuntimeService:
             )
         return count
 
-    @diagnostics.observe('api', 'knowledge.vector_list', source='agent')
     async def vector_list(
         self,
         execution_context: ExecutionContext,
@@ -192,7 +186,6 @@ class RAGRuntimeService:
             offset=offset,
         )
 
-    @diagnostics.observe('api', 'knowledge.get_file_stream', source='agent')
     async def get_file_stream(
         self,
         execution_context: ExecutionContext,
diff --git a/src/langbot/pkg/skill/manager.py b/src/langbot/pkg/skill/manager.py
index d0fef97c0..6b8f4e03e 100644
--- a/src/langbot/pkg/skill/manager.py
+++ b/src/langbot/pkg/skill/manager.py
@@ -59,12 +59,10 @@ class SkillManager:
         for existing_key in tuple(self._skills_by_scope):
             if existing_key[:2] == key[:2] and existing_key != key:
                 self._skills_by_scope.pop(existing_key, None)
-        self._skills_by_scope[key] = {}
 
         repository = getattr(self.ap, 'skill_repository', None)
         if repository is None:
-            self.ap.logger.info('Skill repository unavailable; skill cache will remain empty.')
-            return
+            raise RuntimeError('Skill repository unavailable')
 
         try:
             skills: dict[str, dict] = {}
@@ -77,6 +75,7 @@ class SkillManager:
             self.ap.logger.info(f'Loaded {len(skills)} skills for Workspace {execution_context.workspace_uuid}')
         except Exception as exc:
             self.ap.logger.warning(f'Failed to load skills for Workspace {execution_context.workspace_uuid}: {exc}')
+            raise
 
     async def ensure_loaded(self, context: TenantContext) -> None:
         key = self._scope_key(context)
diff --git a/src/langbot/pkg/telemetry/adapter_diagnostics.py b/src/langbot/pkg/telemetry/adapter_diagnostics.py
deleted file mode 100644
index 301dd9113..000000000
--- a/src/langbot/pkg/telemetry/adapter_diagnostics.py
+++ /dev/null
@@ -1,252 +0,0 @@
-"""Finite adapter acceptance evidence, without message contents or target IDs."""
-
-from __future__ import annotations
-
-from .diagnostic_catalog import catalog
-
-
-def boundary_fields(module, kind, operation, bound, parent):
-    """Only the actual adapter API / EBA conversion boundary counts as a test."""
-    entry = next((v for k, v in catalog().items() if k in module.split('.')), None)
-    if entry is None or '.platform.adapters.' not in module:
-        return {}
-    evidence = kind == 'event' and operation == 'platform.target2yiri'
-    resolved = operation
-    if kind == 'api':
-        evidence = True
-        if operation == 'call_platform_api':
-            action = bound.get('action')
-            if isinstance(action, str) and 'platform_api.' + action in entry['specific_apis']:
-                resolved = 'platform_api.' + action
-            elif action in entry['apis'] and action != 'call_platform_api':
-                resolved = action
-            else:
-                # Keep the generic boundary for investigation, not a false named test.
-                evidence = False
-        if (
-            entry['adapter'] == 'lark-omni'
-            and resolved
-            in {
-                'platform_api.check_tenant_access_token',
-                'platform_api.refresh_app_access_token',
-                'platform_api.refresh_tenant_access_token',
-            }
-            and getattr(bound.get('self'), 'config', {}).get('app_type', 'self') != 'isv'
-        ):
-            # Self-built apps return {'ok': True} without checking/refreshing a token.
-            evidence = False
-        # A forwarding API may call another decorated method. Count the outer call.
-        if parent and parent.adapter_api_active:
-            evidence = False
-    if not evidence:
-        return {'_adapter_api_active': kind == 'api'}
-    return {
-        '_adapter_api_active': kind == 'api',
-        '_adapter_void_ack': completed_void_api(module, resolved),
-        '_adapter_read_ack': completed_read_api(module, resolved),
-        'operation': resolved,
-        'attributes': {'adapter_evidence': True, **message_scenario(bound)},
-    }
-
-
-def message_scenario(bound):
-    """Read typed routing/media categories only; never serialize a payload."""
-    from langbot_plugin.api.entities.builtin.platform.events import MessageReceivedEvent
-    from langbot_plugin.api.entities.builtin.platform.message import MessageChain
-
-    event = bound.get('event') or bound.get('message_source')
-    target = bound.get('target_type') or bound.get('chat_type')
-    message = bound.get('message') or bound.get('new_content')
-    if isinstance(event, MessageReceivedEvent):
-        target = event.chat_type
-        message = event.message_chain
-    result = {}
-    if target is not None:
-        target = getattr(target, 'value', target)
-        if target == 'private':
-            target = 'person'
-        result['chat_type'] = target if target in ('person', 'group') else 'unknown'
-    if isinstance(message, MessageChain):
-        types = set()
-        for item in message:
-            name = type(item).__name__
-            types.add(
-                {
-                    'Plain': 'text',
-                    'Image': 'image',
-                    'Voice': 'audio',
-                    'Audio': 'audio',
-                    'Video': 'video',
-                    'File': 'file',
-                }.get(name, 'other')
-            )
-        result['content_type'] = next(iter(types)) if len(types) == 1 else 'mixed' if types else 'unknown'
-    return result
-
-
-# These exact API implementations always await an exception-raising SDK call,
-# including SDKs whose successful acknowledgement has no return payload.
-# Do not extend this to methods with conditional no-op or queued work.
-VOID_ACKNOWLEDGED_APIS = {
-    'aiocqhttp': frozenset(
-        {
-            'delete_message',
-            'set_group_name',
-            'mute_member',
-            'unmute_member',
-            'kick_member',
-            'leave_group',
-            'approve_friend_request',
-            'approve_group_invite',
-        }
-    ),
-    'discord': frozenset(
-        {
-            'edit_message',
-            'delete_message',
-            'mute_member',
-            'unmute_member',
-            'kick_member',
-            'leave_group',
-        }
-    ),
-    'kook': frozenset({'delete_message'}),
-}
-
-
-def completed_void_api(module, operation):
-    return module.endswith('.api_impl') and any(
-        directory in module.split('.') and operation in operations
-        for directory, operations in VOID_ACKNOWLEDGED_APIS.items()
-    )
-
-
-# These read methods return actual fetched/cached data or raise. Exclude the
-# known identity-only placeholders and file-ID passthroughs: a typed result alone
-# is not evidence that a lookup worked. Unsupported methods still raise normally.
-_READ_APIS = frozenset(
-    {
-        'get_message',
-        'get_group_info',
-        'get_group_list',
-        'get_group_member_list',
-        'get_group_member_info',
-        'get_user_info',
-        'get_friend_list',
-        'get_file_url',
-    }
-)
-READ_ACKNOWLEDGED_APIS = {
-    'aiocqhttp': _READ_APIS,
-    'telegram': _READ_APIS,
-    'discord': _READ_APIS - {'get_file_url'},
-    'dingtalk': _READ_APIS - {'get_group_info', 'get_user_info'},
-    'kook': _READ_APIS - {'get_file_url'},
-    'lark': _READ_APIS - {'get_user_info', 'get_group_member_info', 'get_file_url'},
-    'officialaccount': _READ_APIS,
-    'qqofficial': _READ_APIS,
-    'slack': _READ_APIS,
-    'wecom': _READ_APIS - {'get_user_info'},
-    'wecombot': _READ_APIS,
-    'wecomcs': _READ_APIS,
-}
-
-
-def completed_read_api(module, operation):
-    return module.endswith('.api_impl') and any(
-        directory in module.split('.') and operation in operations
-        for directory, operations in READ_ACKNOWLEDGED_APIS.items()
-    )
-
-
-def response_outcome(value, adapter=None):
-    """Finite acknowledgement contracts; never inspect or export message content.
-
-    A MessageResult may contain an *inbound* message ID even when nothing was
-    sent. Only its actual raw acknowledgement is eligible. Empty/queued/unknown
-    returns are unconfirmed, not failures. Unknown SDK shapes need an explicit
-    confirmation at their operation boundary, not a truthiness fallback.
-    """
-    from langbot_plugin.api.entities.builtin.platform.events import MessageResult
-
-    message_result = isinstance(value, MessageResult)
-    if message_result:
-        value = value.raw
-    if not isinstance(value, dict):
-        return 'skipped'
-    if (
-        value.get('ok') is False
-        or value.get('status') == 'failed'
-        or any(type(value.get(key)) is int and value[key] != 0 for key in ('retcode', 'errcode', 'code'))
-    ):
-        return 'failed'
-    if value.get('queued') is True or value.get('status') == 'async':
-        return 'skipped'
-    # The wrappers that batch sends retain the real responses in these fields.
-    if 'results' in value:
-        results = value['results']
-        if not isinstance(results, list) or not results:
-            return 'skipped'
-        outcomes = [response_outcome(item, adapter) for item in results]
-        return 'failed' if 'failed' in outcomes else 'skipped' if 'skipped' in outcomes else 'succeeded'
-    for key in ('result', 'raw'):
-        if key in value:
-            return response_outcome(value[key], adapter)
-    if (
-        value.get('ok') is True
-        or value.get('status') == 'ok'
-        or any(type(value.get(key)) is int and value[key] == 0 for key in ('retcode', 'errcode', 'code'))
-    ):
-        return 'succeeded'
-    if message_result:
-        # These wrappers put only SDK-returned IDs in raw (never the source ID).
-        if adapter in {'aiocqhttp-omni', 'discord-omni', 'telegram-omni'}:
-            message_id = value.get('message_id')
-            if type(message_id) in (str, int) and message_id:
-                return 'succeeded'
-        if adapter == 'lark-omni':
-            ids = value.get('message_ids')
-            if isinstance(ids, list) and ids and all(isinstance(item, str) and item for item in ids):
-                return 'succeeded'
-    return 'skipped'
-
-
-def record_api_result(value, *, edited_content=None):
-    """Record Telegram's Message/True acknowledgement and return it unchanged.
-
-    This runs after the SDK await, not after conversion or local stream setup.
-    No payload is retained; exceptions and the adapter's public return stay intact.
-    """
-    try:
-        from .diagnostics import current_span
-
-        span = current_span()
-        if span is not None and span.kind == 'api' and span.fields.get('attributes', {}).get('adapter_evidence'):
-            from telegram import Message
-
-            outcome = (
-                'succeeded'
-                if value is True or isinstance(value, Message)
-                else 'failed'
-                if value is False
-                else 'skipped'
-            )
-            if outcome == 'succeeded' and edited_content is not None:
-                from langbot_plugin.api.entities.builtin.platform.message import Plain
-
-                if any(not isinstance(item, Plain) for item in edited_content):
-                    outcome = 'partial'
-            # One unacknowledged component must not be hidden by a later success.
-            prior = span.adapter_api_result
-            span.adapter_api_result = (
-                'failed'
-                if 'failed' in (prior, outcome)
-                else 'skipped'
-                if 'skipped' in (prior, outcome)
-                else 'partial'
-                if 'partial' in (prior, outcome)
-                else outcome
-            )
-    except Exception:
-        pass
-    return value
diff --git a/src/langbot/pkg/telemetry/diagnostic_catalog.py b/src/langbot/pkg/telemetry/diagnostic_catalog.py
deleted file mode 100644
index 4eb92b045..000000000
--- a/src/langbot/pkg/telemetry/diagnostic_catalog.py
+++ /dev/null
@@ -1,93 +0,0 @@
-"""Packaged capability catalog and projections from existing runtime declarations."""
-
-from __future__ import annotations
-
-from functools import lru_cache
-from pathlib import Path
-
-import yaml
-
-from . import diagnostic_privacy as privacy
-
-
-@lru_cache(maxsize=1)
-def catalog():
-    result = {}
-    base = Path(__file__).resolve().parents[1] / 'platform'
-    for manifest in sorted((base / 'adapters').glob('*/manifest.yaml')):
-        data = yaml.safe_load(manifest.read_text(encoding='utf-8'))
-        spec = data.get('spec', {})
-        name = data['metadata']['name']
-        privacy.code_value('adapter', name)
-        events = spec.get('supported_events', [])
-        apis = spec.get('supported_apis', {})
-        apis = apis if isinstance(apis, list) else [v for items in apis.values() for v in items]
-        for event in events:
-            privacy.code_value('platform_event_type', event)
-        for operation in apis:
-            privacy.code_value('operation', operation)
-        specific_apis = []
-        for api in spec.get('platform_specific_apis', []):
-            privacy.code_value('operation', api['action'])
-            specific_apis.append(privacy.code_value('operation', 'platform_api.' + api['action']))
-        result[manifest.parent.name] = {
-            'adapter': name,
-            'events': events,
-            'apis': apis,
-            'specific_apis': specific_apis,
-        }
-    return result
-
-
-def adapter_fields(adapter):
-    module = type(adapter).__module__.split('.')
-    entry = next((entry for directory, entry in catalog().items() if directory in module), None)
-    return {'adapter': entry['adapter']} if entry else {}
-
-
-def _snapshot_bot(bot, *, listener_registered=False):
-    from . import diagnostics
-
-    manager = getattr(bot.ap, 'diagnostics', None)
-    if not isinstance(manager, diagnostics.DiagnosticsManager) or not manager.enabled:
-        return
-    fields = adapter_fields(bot.adapter)
-    if not fields:
-        return
-    fields['workspace_uuid'] = bot.execution_context.workspace_uuid
-    entry = next(e for e in catalog().values() if e['adapter'] == fields['adapter'])
-    for capability_type, method, declared in (
-        ('event', 'get_supported_events', entry['events']),
-        ('api', 'get_supported_apis', entry['apis'] + entry['specific_apis']),
-    ):
-        try:
-            supported = set(getattr(bot.adapter, method)() or [])
-        except Exception:
-            supported = set()
-        for name in declared:
-            if name == 'call_platform_api':
-                continue
-            manager.emit(
-                'capability',
-                name if capability_type == 'api' else 'platform.receive',
-                'succeeded',
-                source='platform',
-                stage='snapshot',
-                **fields,
-                platform_event_type=name if capability_type == 'event' else '',
-                attributes={
-                    'capability_type': capability_type,
-                    'capability_name': name,
-                    'supported': name in supported
-                    or (name in entry['specific_apis'] and 'call_platform_api' in supported),
-                    'configured': True,
-                    **({'listener_registered': listener_registered} if capability_type == 'event' else {}),
-                },
-            )
-
-
-def snapshot_bot(bot, *, listener_registered=False):
-    try:
-        _snapshot_bot(bot, listener_registered=listener_registered)
-    except Exception:
-        pass
diff --git a/src/langbot/pkg/telemetry/diagnostic_privacy.py b/src/langbot/pkg/telemetry/diagnostic_privacy.py
deleted file mode 100644
index 14dc21609..000000000
--- a/src/langbot/pkg/telemetry/diagnostic_privacy.py
+++ /dev/null
@@ -1,117 +0,0 @@
-"""Closed, content-free projection for Beta diagnostics; never serialize payloads."""
-
-from __future__ import annotations
-
-import builtins
-import hashlib
-import math
-import re
-from uuid import UUID
-
-KINDS = frozenset('lifecycle event route run api delivery interaction capability transport summary'.split())
-SOURCES = frozenset(
-    'platform pipeline agent event_processor webui_debug http websocket mcp plugin runtime startup internal synthetic'.split()
-)
-OUTCOMES = frozenset('started succeeded failed cancelled timeout skipped rejected waiting partial unknown'.split())
-PROCESSORS = frozenset(('', 'pipeline', 'agent', 'event_processor'))
-# These sets are populated ONLY by source-code decorators / packaged manifests,
-# never by requests, installed third-party plugin declarations or configuration.
-VOCABULARY: dict[str, set[str]] = {
-    'operation': {'diagnostics.transport', 'startup', 'route.primary', 'route.subscription', 'runner.run'},
-    'stage': {'execute', 'prepare', 'convert', 'dispatch', 'accepted', 'ack', 'resume', 'shutdown', 'snapshot'},
-    'adapter': set(),
-    'platform_event_type': set(),
-    'reason_code': set(
-        'runner_failed response_error route_not_found processor_incompatible processor_not_found discarded not_matched matched delivered waiting interaction_rejected generator_closed transport_loss'.split()
-    ),
-    'capability_type': {'event', 'api', 'processor'},
-    'tool_category': {'native', 'plugin', 'mcp', 'skill', 'platform', 'unknown'},
-    'transport': {'stdio', 'websocket', 'http', 'unknown'},
-    'runner_usage': {'agent', 'event'},
-    'os': {'linux', 'darwin', 'windows'},
-    'arch': {'x86_64', 'aarch64', 'arm64', 'amd64'},
-    'database': {'sqlite', 'postgresql'},
-    'edition': {'community', 'cloud', 'enterprise'},
-    'chat_type': {'person', 'group', 'unknown'},
-    'content_type': {'text', 'image', 'audio', 'video', 'file', 'mixed', 'other', 'unknown'},
-}
-BOOLS = frozenset(
-    'stream synthetic configured available previous_session_unclean recovered supported adapter_evidence listener_registered'.split()
-)
-NUMBERS = frozenset(
-    'attempts successes failures cancellations timeouts partial unknown generated queued acked dropped retried failed queue_size capacity result_count input_tokens output_tokens'.split()
-)
-VERSIONS = frozenset('sdk_version plugin_version runner_version python_version'.split())
-
-
-def code_value(field: str, value: str) -> str:
-    """Register a literal from trusted Core source, not a runtime string."""
-    VOCABULARY.setdefault(field, set()).add(value)
-    return value
-
-
-def category(field: str, value) -> str:
-    return value if type(value) is str and value in VOCABULARY.get(field, ()) else ''
-
-
-def opaque(value) -> str:
-    if not isinstance(value, str):
-        return ''
-    try:
-        return str(UUID(value))
-    except (ValueError, TypeError, AttributeError):
-        return ''
-
-
-def version_string(value) -> str:
-    # Public package versions only; do not permit arbitrary PEP440 local labels.
-    return value if isinstance(value, str) and re.fullmatch(r'[0-9][0-9.abrcdevpost-]{0,47}', value) else ''
-
-
-def attributes(values) -> dict:
-    if not isinstance(values, dict):
-        return {}
-    result = {}
-    for key, value in values.items():
-        if key in BOOLS and type(value) is bool:
-            result[key] = value
-        elif key in NUMBERS and type(value) is int and math.isfinite(value) and 0 <= value <= 1_000_000:
-            result[key] = value
-        elif key in VERSIONS and version_string(value):
-            result[key] = value
-        elif key in VOCABULARY and key not in {'operation', 'stage', 'adapter', 'platform_event_type', 'reason_code'}:
-            if category(key, value):
-                result[key] = value
-        elif key == 'capability_name' and any(category(f, value) for f in ('operation', 'platform_event_type')):
-            result[key] = value
-        elif key == 'capability_name' and value in PROCESSORS - {''}:
-            result[key] = value
-        elif key in ('event_types', 'api_operations', 'processor_types') and isinstance(value, (tuple, list)):
-            allowed = (
-                VOCABULARY['platform_event_type']
-                if key == 'event_types'
-                else VOCABULARY['operation']
-                if key == 'api_operations'
-                else PROCESSORS
-            )
-            result[key] = [v for v in value[:128] if type(v) is str and v in allowed]
-        elif (
-            key in ('source_revision', 'target_revision')
-            and isinstance(value, str)
-            and re.fullmatch('[a-f0-9]{40}', value)
-        ):
-            result[key] = value
-        elif key == 'latency_buckets' and isinstance(value, (list, tuple)):
-            result[key] = [v for v in value[:128] if type(v) is int and 0 <= v <= 1_000_000]
-    return result
-
-
-def error_fields(error: BaseException, operation: str, stage: str) -> dict:
-    cls = type(error)
-    # Third party exception names may be dynamically constructed from input.
-    name = cls.__name__ if getattr(builtins, cls.__name__, None) is cls else 'Exception'
-    if isinstance(error, TimeoutError):
-        name = 'TimeoutError'
-    # No message, traceback, filename, line number, or locals are inspected.
-    fingerprint = hashlib.sha256(f'{operation}|{stage}|{cls.__module__}|{cls.__qualname__}'.encode()).hexdigest()
-    return {'error_type': name, 'error_fingerprint': fingerprint}
diff --git a/src/langbot/pkg/telemetry/diagnostic_transport.py b/src/langbot/pkg/telemetry/diagnostic_transport.py
deleted file mode 100644
index 4220bd1a1..000000000
--- a/src/langbot/pkg/telemetry/diagnostic_transport.py
+++ /dev/null
@@ -1,399 +0,0 @@
-"""Bounded, memory-only Beta diagnostics transport, isolated from usage telemetry."""
-
-from __future__ import annotations
-
-import asyncio
-import contextlib
-import importlib.metadata
-import json
-import os
-import random
-import re
-import time
-import platform
-import sys
-from pathlib import Path
-from collections import Counter
-from datetime import datetime, timezone
-from uuid import uuid4
-
-import httpx
-from packaging.version import Version, InvalidVersion
-
-from . import diagnostic_privacy as privacy
-
-
-class DiagnosticsManager:
-    """All producers are synchronous; only this owned worker performs I/O.
-
-    Retention: 2048 events / 15 minutes, 5 attempts, 8 KiB per event. An in-flight
-    event stays in the same bounded queue until explicitly ACKed. No disk spool
-    means abrupt process loss cannot be counted after restart (documented).
-    """
-
-    def __init__(self, ap, *, version=None, instance_id=None, capacity=2048, marker_path=None):
-        from ..utils import constants
-        from .diagnostic_catalog import catalog
-
-        with contextlib.suppress(Exception):
-            catalog()
-        self.ap = ap
-        self.marker_path = Path(marker_path) if marker_path is not None else None
-        self._session_id = str(uuid4())
-        self._marker_active = False
-        raw_version = version or importlib.metadata.version('langbot')
-        try:
-            parsed = Version(raw_version)
-            self.version = str(parsed)
-            self.beta = bool(parsed.pre and parsed.pre[0] == 'b' and not parsed.local)
-        except InvalidVersion:
-            self.version, self.beta = '', False
-        self.instance_id = instance_id or constants.instance_id
-        try:
-            from langbot._build_info import CORE_REVISION
-        except ImportError:
-            CORE_REVISION = ''
-        candidate = os.getenv('LANGBOT_BUILD_REVISION', CORE_REVISION)
-        self.revision = candidate.lower() if re.fullmatch('[a-fA-F0-9]{40}', candidate or '') else ''
-        try:
-            self.sdk_version = privacy.version_string(importlib.metadata.version('langbot-plugin'))
-        except importlib.metadata.PackageNotFoundError:
-            self.sdk_version = ''
-        self.capacity = max(1, min(capacity, 2048))
-        self.pending: list[dict] = []
-        self.counters = Counter()
-        self._reported = Counter()
-        self._last_retry = False
-        self.max_attempts = 5
-        self.retention_seconds = 900
-        self.request_timeout = 10
-        self.client: httpx.AsyncClient | None = None
-        self._attempts: dict[str, int] = {}
-        self._born: dict[str, float] = {}
-        self._worker = None
-        self._request = None
-        self._closing = False
-        self._wake = asyncio.Event()
-        self._flush_lock = asyncio.Lock()
-
-    async def start_session(self):
-        """One bounded, content-free marker per process; never per-event disk I/O."""
-        if not self.enabled:
-            await self._clear_marker()
-            return
-        previous = False
-        if self.marker_path is not None:
-
-            def mark():
-                previous = self.marker_path.exists()
-                self.marker_path.parent.mkdir(parents=True, exist_ok=True)
-                self.marker_path.write_text(self._session_id)
-                return previous
-
-            with contextlib.suppress(Exception):
-                previous = await asyncio.to_thread(mark)
-                self._marker_active = True
-        privacy.code_value('operation', 'startup.session')
-        self.emit(
-            'lifecycle',
-            'startup.session',
-            'started',
-            source='startup',
-            stage='snapshot',
-            attributes={
-                'previous_session_unclean': previous,
-                'os': platform.system().lower(),
-                'arch': platform.machine().lower(),
-                'python_version': '.'.join(map(str, sys.version_info[:3])),
-                'database': self.ap.instance_config.data.get('database', {}).get('use', ''),
-                'edition': self.ap.instance_config.data.get('system', {}).get('edition', ''),
-            },
-        )
-
-    async def _clear_marker(self):
-        if self.marker_path is None:
-            return
-
-        def remove():
-            if self.marker_path.exists():
-                with self.marker_path.open() as handle:
-                    owner = handle.read(64)
-                if owner == self._session_id or not self._marker_active:
-                    self.marker_path.unlink(missing_ok=True)
-
-        with contextlib.suppress(Exception):
-            await asyncio.to_thread(remove)
-        self._marker_active = False
-
-    @property
-    def enabled(self):
-        config = getattr(getattr(self.ap, 'instance_config', None), 'data', {}).get('space', {})
-        return (
-            self.beta
-            and not self._closing
-            and not config.get('disable_telemetry', False)
-            and not config.get('disable_beta_diagnostics', False)
-            and bool(config.get('url'))
-        )
-
-    def clear(self):
-        self.pending.clear()
-        self._attempts.clear()
-        self._born.clear()
-        if self._request is not None and not self._request.done():
-            self._request.cancel()
-
-    def emit(
-        self,
-        kind,
-        operation,
-        outcome,
-        *,
-        source='internal',
-        workspace_uuid='',
-        attributes=None,
-        stage='execute',
-        adapter='',
-        processor_type='',
-        platform_event_type='',
-        reason_code='',
-        duration_ms=0,
-        error=None,
-        trace_id='',
-        span_id='',
-        parent_span_id='',
-        run_id='',
-        **ignored,
-    ):
-        """Project known scalar fields; failures in diagnostics never escape."""
-        try:
-            if not self.enabled:
-                self.clear()
-                return
-            if kind not in privacy.KINDS or outcome not in privacy.OUTCOMES:
-                return
-            self.counters['generated'] += 1
-            event = {
-                'schema_version': 1,
-                'event_id': str(uuid4()),
-                'kind': kind,
-                'instance_id': self.instance_id,
-                'workspace_uuid': privacy.opaque(workspace_uuid),
-                'core_version': self.version,
-                'core_revision': self.revision,
-                'sdk_version': self.sdk_version,
-                'release_channel': 'beta',
-                'source': source if source in privacy.SOURCES else 'internal',
-                'operation': privacy.category('operation', operation),
-                'stage': privacy.category('stage', stage),
-                'outcome': outcome,
-                'reason_code': privacy.category('reason_code', reason_code),
-                'adapter': privacy.category('adapter', adapter),
-                'processor_type': processor_type if processor_type in privacy.PROCESSORS else '',
-                'platform_event_type': privacy.category('platform_event_type', platform_event_type),
-                'occurred_at': datetime.now(timezone.utc).isoformat(),
-                'count': 1,
-                'sample_rate': 1,
-                'duration_ms': max(0, min(float(duration_ms), 86400000)),
-                'attributes': privacy.attributes(attributes),
-            }
-            for key, value in (
-                ('trace_id', trace_id),
-                ('span_id', span_id),
-                ('parent_span_id', parent_span_id),
-                ('run_id', run_id),
-            ):
-                if privacy.opaque(value):
-                    event[key] = privacy.opaque(value)
-            if error is not None:
-                event.update(privacy.error_fields(error, event['operation'], event['stage']))
-            encoded = json.dumps(event, allow_nan=False).encode()
-            if (
-                not isinstance(self.instance_id, str)
-                or not 0 < len(self.instance_id) <= 128
-                or len(encoded) > 8192
-                or len(self.pending) >= self.capacity
-            ):
-                self.counters['dropped'] += 1
-                return
-            self.pending.append(event)
-            self._born[event['event_id']] = time.monotonic()
-            self.counters['queued'] += 1
-            self._wake.set()
-        except Exception:
-            self.counters['dropped'] += 1
-
-    async def credentials(self, workspace_uuid):
-        # Resolve each Workspace independently in the background. No OSS secret.
-        token = os.getenv('LANGBOT_TELEMETRY_INGEST_TOKEN', '').strip()
-        if token:
-            return {'X-LangBot-Telemetry-Token': token}
-        if not workspace_uuid:
-            return {}
-        users = getattr(self.ap, 'user_service', None)
-        space = getattr(self.ap, 'space_service', None)
-        if users is None or space is None:
-            return {}
-        owner = await users.get_workspace_owner(workspace_uuid)
-        email = getattr(owner, 'user', None)
-        token = await space.get_valid_access_token(email) if email else None
-        return {'Authorization': f'Bearer {token}'} if token else {}
-
-    def _remove(self, ids):
-        self.pending[:] = [e for e in self.pending if e['event_id'] not in ids]
-        for event_id in ids:
-            self._attempts.pop(event_id, None)
-            self._born.pop(event_id, None)
-
-    async def flush_once(self):
-        async with self._flush_lock:
-            if not self.enabled:
-                self.clear()
-                return
-            expired = {
-                e['event_id']
-                for e in self.pending
-                if time.monotonic() - self._born.get(e['event_id'], 0) > self.retention_seconds
-            }
-            self.counters['dropped'] += len(expired)
-            self._remove(expired)
-            if not self.pending:
-                return
-            # Do not authenticate a mixed-Workspace batch with one owner's token.
-            workspace = self.pending[0]['workspace_uuid']
-            batch, size = [], 64
-            for event in self.pending:
-                n = len(json.dumps(event).encode()) + 2
-                if event['workspace_uuid'] != workspace:
-                    continue
-                if len(batch) == 50 or size + n > 256 * 1024:
-                    break
-                batch.append(event)
-                size += n
-            ids = {e['event_id'] for e in batch}
-            for event_id in ids:
-                self._attempts[event_id] = self._attempts.get(event_id, 0) + 1
-            acked, rejected = set(), set()
-            permanent = False
-            try:
-                async with asyncio.timeout(self.request_timeout):
-                    headers = {}
-                    try:
-                        async with asyncio.timeout(min(1.0, self.request_timeout / 3)):
-                            headers = await self.credentials(workspace)
-                    except Exception:
-                        pass
-                    if not self.enabled:
-                        self.clear()
-                        return
-                    if self.client is None:
-                        self.client = httpx.AsyncClient(timeout=self.request_timeout, follow_redirects=False)
-                    url = (
-                        self.ap.instance_config.data['space']['url'].rstrip('/') + '/api/v1/telemetry/diagnostics/batch'
-                    )
-                    # Stream the response to bound malicious/old server responses too.
-                    async with self.client.stream(
-                        'POST', url, json={'schema_version': 1, 'events': batch}, headers=headers
-                    ) as response:
-                        body = bytearray()
-                        async for chunk in response.aiter_bytes():
-                            body.extend(chunk)
-                            if len(body) > 65536:
-                                raise ValueError('diagnostics_response_limit')
-                        if response.status_code == 200:
-                            payload = json.loads(body)
-                            if payload.get('code') == 200 and isinstance(payload.get('data'), dict):
-                                data = payload['data']
-                                acked = {x for x in data.get('accepted_event_ids', []) if isinstance(x, str)} & ids
-                                rejected = {
-                                    x.get('event_id')
-                                    for x in data.get('rejected', [])
-                                    if isinstance(x, dict) and x.get('code') == 'invalid_event'
-                                } & ids
-                            elif isinstance(payload.get('code'), int):
-                                permanent = 400 <= payload['code'] < 500 and payload['code'] != 429
-                        else:
-                            permanent = 400 <= response.status_code < 500 and response.status_code != 429
-            except asyncio.CancelledError:
-                raise
-            except Exception:
-                pass
-            if not self.enabled:
-                self.clear()
-                return
-            rejected -= acked
-            remaining = ids - acked - rejected
-            if remaining:
-                self.counters['failed'] += 1
-            exhausted = {i for i in remaining if permanent or self._attempts.get(i, 0) >= self.max_attempts}
-            self.counters['acked'] += len(acked)
-            self.counters['dropped'] += len(rejected | exhausted)
-            self._last_retry = bool(remaining - exhausted)
-            self.counters['retried'] += len(remaining - exhausted)
-            self._remove(acked | rejected | exhausted)
-
-    def report_transport(self, stage='snapshot'):
-        # Snapshot interval deltas before enqueue; retries retain this event ID.
-        snapshot = self.counters.copy()
-        delta = snapshot - self._reported
-        before = self.counters['queued']
-        self.emit(
-            'transport',
-            'diagnostics.transport',
-            'partial' if delta['dropped'] else 'succeeded',
-            stage=stage,
-            attributes={**dict(delta), 'queue_size': len(self.pending), 'capacity': self.capacity},
-        )
-        if self.counters['queued'] > before:
-            self._reported = snapshot
-
-    def start(self):
-        if self.beta and self._worker is None:
-            self._worker = asyncio.create_task(self._loop(), name='beta-diagnostics')
-
-    async def _loop(self):
-        next_send, next_summary, backoff = 0.0, time.monotonic() + 60, 0.0
-        try:
-            while not self._closing:
-                if not self.enabled:
-                    self.clear()
-                    if self._marker_active:
-                        await self._clear_marker()
-                elif time.monotonic() >= next_summary:
-                    self.report_transport()
-                    next_summary = time.monotonic() + 60
-                if self._request is not None and self._request.done():
-                    with contextlib.suppress(asyncio.CancelledError, Exception):
-                        self._request.result()
-                    self._request = None
-                    backoff = min(30, max(1, backoff * 2)) if self._last_retry else 0
-                    next_send = time.monotonic() + random.uniform(0.5, 1.5) * backoff
-                if self.enabled and self.pending and self._request is None and time.monotonic() >= next_send:
-                    self._request = asyncio.create_task(self.flush_once(), name='beta-diagnostics-batch')
-                self._wake.clear()
-                with contextlib.suppress(TimeoutError):
-                    await asyncio.wait_for(self._wake.wait(), 0.1)
-        finally:
-            if self._request is not None:
-                self._request.cancel()
-                await asyncio.gather(self._request, return_exceptions=True)
-
-    async def shutdown(self, drain_timeout=2):
-        if self._worker is not None:
-            self._worker.cancel()
-            await asyncio.gather(self._worker, return_exceptions=True)
-            self._worker = None
-        if self.enabled and drain_timeout > 0:
-            self.report_transport(stage='shutdown')
-            with contextlib.suppress(Exception):
-                async with asyncio.timeout(drain_timeout):
-                    while self.pending:
-                        await self.flush_once()
-                        if self.pending:
-                            await asyncio.sleep(0.1)
-        self.counters['dropped'] += len(self.pending)
-        self._closing = True
-        self.clear()
-        await self._clear_marker()
-        if self.client is not None:
-            await self.client.aclose()
diff --git a/src/langbot/pkg/telemetry/diagnostics.py b/src/langbot/pkg/telemetry/diagnostics.py
deleted file mode 100644
index 871efc2f8..000000000
--- a/src/langbot/pkg/telemetry/diagnostics.py
+++ /dev/null
@@ -1,546 +0,0 @@
-"""Explicit diagnostic boundaries, preserving coroutine and generator semantics."""
-
-from __future__ import annotations
-
-import asyncio
-import contextlib
-import contextvars
-import functools
-import inspect
-import time
-from uuid import uuid4
-
-from .diagnostic_transport import DiagnosticsManager  # noqa: F401
-from . import diagnostic_privacy as privacy
-
-_CURRENT = contextvars.ContextVar('beta_diagnostic_span', default=None)
-
-
-def current_span():
-    return _CURRENT.get()
-
-
-def set_outcome(outcome, *, reason_code=''):
-    span = current_span()
-    if span is not None:
-        span.outcome = outcome if outcome in privacy.OUTCOMES else 'unknown'
-        span.fields['reason_code'] = privacy.category('reason_code', reason_code)
-
-
-def annotate(**fields):
-    """Trusted hook metadata is still projected at the transport boundary."""
-    span = current_span()
-    if span is not None:
-        span.fields.update(fields)
-
-
-def adapter_event_received(owner, event):
-    """Record one converted event before dispatch, including native callback paths."""
-    with contextlib.suppress(Exception):
-        manager = _manager(owner)
-        if manager is None:
-            return
-        from langbot_plugin.api.entities.builtin.platform.events import EBAEvent
-        from .adapter_diagnostics import message_scenario
-
-        if not isinstance(event, EBAEvent) or not _context_matches(manager, owner, _owner_context(owner)):
-            return
-        fields = _context_fields(owner, {'event': event})
-        fields['attributes'] = {
-            **fields.get('attributes', {}),
-            'adapter_evidence': True,
-            **message_scenario({'event': event}),
-        }
-        parent = current_span()
-        if (
-            parent
-            and parent.manager is manager
-            and not (
-                fields.get('workspace_uuid')
-                and parent.fields.get('workspace_uuid')
-                and fields['workspace_uuid'] != parent.fields['workspace_uuid']
-            )
-        ):
-            fields['trace_id'] = parent.fields['trace_id']
-            fields['parent_span_id'] = parent.fields['span_id']
-            if parent.fields.get('source') in ('webui_debug', 'synthetic'):
-                fields['source'] = parent.fields['source']
-                fields['attributes']['synthetic'] = True
-        fields.setdefault('source', 'platform')
-        privacy.code_value('operation', 'platform.adapter_event')
-        manager.emit('event', 'platform.adapter_event', 'succeeded', stage='accepted', **fields)
-
-
-def _owner_app(owner):
-    """An explicit owner (even absent/disabled) is an inheritance barrier."""
-    if owner is None:
-        return None, False
-    if hasattr(owner, 'ap'):
-        return owner.ap, True
-    if hasattr(owner, 'diagnostics') or hasattr(owner, 'instance_config'):
-        return owner, True
-    for name in ('requester', 'logger', 'adapter'):
-        nested = getattr(owner, name, None)
-        if nested is not None and nested is not owner:
-            app, explicit = _owner_app(nested)
-            if explicit:
-                return app, True
-    return None, False
-
-
-def _manager(owner):
-    app, _ = _owner_app(owner)
-    manager = getattr(app, 'diagnostics', None)
-    if manager is None or not callable(getattr(manager, 'emit', None)) or not getattr(manager, 'enabled', False):
-        return None
-    # Honor the shared Span/management producer interface without inheriting a
-    # different producer just because this application's producer is absent.
-    config = getattr(getattr(app, 'instance_config', None), 'data', {}).get('space', {})
-    if config.get('disable_telemetry', False) or config.get('disable_beta_diagnostics', False):
-        return None
-    # Direct manager holders (e.g. ReplyStreamSession) are intentional. An app
-    # must never borrow another app's manager, policy, or credential resolver.
-    if hasattr(app, 'instance_config') and getattr(manager, 'ap', app) is not app:
-        return None
-    return manager
-
-
-def _owner_context(owner):
-    context = getattr(owner, 'execution_context', None)
-    if context is not None:
-        return context
-    for name in ('requester', 'logger', 'adapter'):
-        nested = getattr(owner, name, None)
-        if nested is not None and nested is not owner:
-            context = _owner_context(nested)
-            if context is not None:
-                return context
-    return None
-
-
-def _execution_context(owner, bound):
-    context = bound.get('execution_context') or _owner_context(owner)
-    adapter_context = bound.get('adapter_context')
-    if isinstance(adapter_context, dict):
-        context = adapter_context.get('_execution_context') or context
-    query = bound.get('query')
-    if query is not None:
-        context = getattr(query, '_execution_context', None) or context
-    return context
-
-
-def _context_matches(manager, owner, context):
-    from ..api.http.context import ExecutionContext
-
-    if not isinstance(context, ExecutionContext):
-        return True
-    instance = getattr(getattr(manager.ap, 'workspace_service', None), 'instance_uuid', manager.instance_id)
-    if instance != context.instance_uuid:
-        return False
-    owned = _owner_context(owner)
-    return not isinstance(owned, ExecutionContext) or (
-        owned.instance_uuid,
-        owned.workspace_uuid,
-        owned.placement_generation,
-    ) == (context.instance_uuid, context.workspace_uuid, context.placement_generation)
-
-
-def _context_fields(owner, bound):
-    context = _execution_context(owner, bound)
-    query = bound.get('query')
-    from .diagnostic_catalog import adapter_fields
-
-    adapter = getattr(owner, 'adapter', None) or owner
-    fields = adapter_fields(adapter)
-    # ExecutionContext is constructed/validated by the existing auth boundary;
-    # do not infer Workspace identity from arbitrary payload dicts or event IDs.
-    if context is not None:
-        from ..api.http.context import ExecutionContext
-
-        if isinstance(context, ExecutionContext):
-            fields['workspace_uuid'] = context.workspace_uuid
-            saved = getattr(query, '_diagnostic_context', None) if query is not None else None
-            if isinstance(saved, dict) and saved.get('workspace_uuid') == context.workspace_uuid:
-                fields.update(saved)
-    binding = bound.get('binding')
-    if binding is not None:
-        fields['processor_type'] = getattr(binding, 'processor_type', '')
-    event = bound.get('event')
-    if event is not None:
-        fields['platform_event_type'] = getattr(event, 'event_type', None) or getattr(event, 'type', '')
-        delivery = getattr(event, 'delivery', None)
-        if getattr(delivery, 'surface', None) == 'webui':
-            fields['source'] = 'webui_debug'
-            fields['attributes'] = {'synthetic': True}
-    if getattr(owner, 'mock', False) is True:
-        fields['source'] = 'webui_debug'
-        fields['attributes'] = {'synthetic': True}
-    return fields
-
-
-class Span:
-    def __init__(self, manager, kind, operation, fields):
-        parent = current_span()
-        self.manager = manager
-        self.kind = kind
-        self.operation = operation
-        self.fields = dict(fields)
-        self.adapter_api_active = bool(self.fields.pop('_adapter_api_active', False))
-        self.adapter_void_ack = bool(self.fields.pop('_adapter_void_ack', False))
-        self.adapter_read_ack = bool(self.fields.pop('_adapter_read_ack', False))
-        self.adapter_api_result = None
-        if (
-            parent
-            and parent.fields.get('workspace_uuid')
-            and self.fields.get('workspace_uuid')
-            and parent.fields['workspace_uuid'] != self.fields['workspace_uuid']
-        ):
-            parent = None
-        if self.fields.get('operation'):
-            self.operation = privacy.category('operation', self.fields.pop('operation')) or operation
-        self.outcome = None
-        self.finished = False
-        self.started = time.monotonic()
-        self.fields['trace_id'] = (
-            parent.fields['trace_id']
-            if parent and parent.manager is manager
-            else (privacy.opaque(self.fields.get('trace_id')) or str(uuid4()))
-        )
-        self.fields['span_id'] = str(uuid4())
-        if parent and parent.manager is manager:
-            self.adapter_api_active = self.adapter_api_active or parent.adapter_api_active
-            self.fields['parent_span_id'] = parent.fields['span_id']
-            for key in ('workspace_uuid', 'adapter', 'processor_type', 'platform_event_type', 'run_id'):
-                if not self.fields.get(key) and parent.fields.get(key):
-                    self.fields[key] = parent.fields[key]
-            if parent.fields.get('source') in ('webui_debug', 'synthetic'):
-                self.fields['source'] = parent.fields['source']
-                self.fields['attributes'] = {**self.fields.get('attributes', {}), 'synthetic': True}
-        self.emit('started')
-
-    def emit(self, outcome, **extra):
-        if self.manager is not None:
-            self.manager.emit(self.kind, self.operation, outcome, **{**self.fields, **extra})
-
-    @contextlib.contextmanager
-    def activate(self):
-        token = _CURRENT.set(self)
-        try:
-            yield self
-        finally:
-            _CURRENT.reset(token)
-
-    def finish(self, error=None):
-        if self.finished:
-            return
-        self.finished = True
-        error = error if error is not None else self.fields.pop('error', None)
-        if isinstance(error, (asyncio.CancelledError, GeneratorExit)):
-            outcome = 'cancelled'
-        elif isinstance(error, TimeoutError):
-            outcome = 'timeout'
-        elif error is not None:
-            outcome = self.outcome if self.outcome in ('timeout', 'partial', 'rejected') else 'failed'
-        else:
-            outcome = self.outcome or 'succeeded'
-        self.emit(outcome, error=error, duration_ms=(time.monotonic() - self.started) * 1000)
-
-
-def result_outcome(value):
-    """Inspect only the SDK response's status, not arbitrary result contents."""
-    from langbot_plugin.api.entities.builtin.platform.events import EBAEvent
-    from langbot_plugin.runtime.io.handler import ActionResponse
-
-    span = current_span()
-    if span is not None and span.fields.get('stage') == 'convert':
-        if isinstance(value, EBAEvent):
-            annotate(platform_event_type=value.type)
-            if span.fields.get('attributes', {}).get('adapter_evidence'):
-                from .adapter_diagnostics import message_scenario
-
-                annotate(attributes={**span.fields['attributes'], **message_scenario({'event': value})})
-                # Successful conversion is counted once at dispatch, which also
-                # covers adapters constructing EBA events in native callbacks.
-                span.fields['attributes']['adapter_evidence'] = False
-        elif value is None:
-            set_outcome('skipped', reason_code='not_matched')
-        if not isinstance(value, EBAEvent) and span.fields.get('attributes', {}).get('adapter_evidence'):
-            span.fields['attributes']['adapter_evidence'] = False
-    if span is not None and span.kind == 'api' and span.fields.get('attributes', {}).get('adapter_evidence'):
-        from .adapter_diagnostics import response_outcome
-
-        # A normal return is not an acknowledgement. Fail closed for acceptance
-        # (but never for the business call) even if result inspection raises.
-        if span.outcome is None:
-            set_outcome('skipped')
-            outcome = span.adapter_api_result
-            if outcome is None:
-                if (value is None and span.adapter_void_ack) or (
-                    span.adapter_read_ack and value is not None and value != ''
-                ):
-                    outcome = 'succeeded'
-                else:
-                    outcome = response_outcome(value, span.fields.get('adapter'))
-            set_outcome(outcome, **({'reason_code': 'response_error'} if outcome == 'failed' else {}))
-    if isinstance(value, ActionResponse):
-        if value.code != 0:
-            set_outcome('failed', reason_code='response_error')
-
-
-def observe(kind, operation, *, source='internal', stage='execute', ap=None, fields=None):
-    """Explicit boundary with a stable/off fast path and transparent generators."""
-    privacy.code_value('operation', operation)
-    privacy.code_value('stage', stage)
-
-    def decorate(fn):
-        signature = inspect.signature(fn)
-
-        def span_for(args, kwargs):
-            try:
-                bound = signature.bind_partial(*args, **kwargs).arguments
-                owner = bound.get(next(iter(signature.parameters), ''))
-                manager_owner = (ap() if callable(ap) else ap) if ap is not None else owner
-                manager = _manager(manager_owner)
-                _, explicit = _owner_app(manager_owner)
-                context = _execution_context(owner, bound)
-                parent = current_span()
-                if manager is None and not explicit and ap is None and parent is not None:
-                    # Stateless converters may inherit, but a different explicit
-                    # Workspace must not select a parent's credentials/policy.
-                    if context is not None and getattr(context, 'workspace_uuid', None) != parent.fields.get(
-                        'workspace_uuid'
-                    ):
-                        return None
-                    manager = parent.manager
-                if manager is None or not manager.enabled:
-                    return None
-                if not _context_matches(manager, owner, context):
-                    return None
-                metadata = {'source': source, 'stage': stage}
-                metadata.update(_context_fields(owner, bound))
-                from .diagnostic_catalog import catalog
-
-                for directory, entry in catalog().items():
-                    if directory in fn.__module__.split('.') and '.platform.' in fn.__module__:
-                        metadata['adapter'] = entry['adapter']
-                        break
-                if fields:
-                    extra = fields(bound)
-                    extra['attributes'] = {**metadata.get('attributes', {}), **extra.get('attributes', {})}
-                    metadata.update(extra)
-                from .adapter_diagnostics import boundary_fields
-
-                if parent and (
-                    parent.manager is not manager
-                    or (
-                        metadata.get('workspace_uuid')
-                        and parent.fields.get('workspace_uuid')
-                        and metadata['workspace_uuid'] != parent.fields['workspace_uuid']
-                    )
-                ):
-                    parent = None
-                evidence = (
-                    boundary_fields(fn.__module__, kind, operation, bound, parent)
-                    if metadata.get('attributes', {}).get('adapter_evidence') is not False
-                    else {}
-                )
-                metadata['attributes'] = {**metadata.get('attributes', {}), **evidence.pop('attributes', {})}
-                metadata.update(evidence)
-                return Span(manager, kind, operation, metadata)
-            except Exception:
-                return None
-
-        def inspect_result(value):
-            with contextlib.suppress(Exception):
-                result_outcome(value)
-
-        def finish(span, error=None):
-            if span is not None:
-                with contextlib.suppress(Exception):
-                    span.finish(error)
-
-        if inspect.isasyncgenfunction(fn):
-            from collections.abc import AsyncGenerator
-
-            class ObservedGenerator(AsyncGenerator):
-                """Delegate each native protocol operation, without extra close.
-
-                A yield-based proxy cannot distinguish athrow(GeneratorExit)
-                (which may yield) from aclose() (which must reject a yield).
-                Let the native generator implement that distinction and retain
-                its own primary/cleanup exception and cancellation semantics.
-                """
-
-                def __init__(self, args, kwargs):
-                    self.gen = fn(*args, **kwargs)
-                    self.args, self.kwargs = args, kwargs
-                    self.span = None
-                    self.started = False
-
-                def __getattr__(self, name):
-                    return getattr(self.gen, name)
-
-                async def _advance(self, method, *values):
-                    if not self.started:
-                        self.started = True
-                        self.span = span_for(self.args, self.kwargs)
-                    token = _CURRENT.set(self.span)
-                    try:
-                        value = await method(*values)
-                        if self.span is not None and method != self.gen.aclose:
-                            inspect_result(value)
-                        if method == self.gen.aclose:
-                            finish(self.span, GeneratorExit())
-                        return value
-                    except StopAsyncIteration:
-                        finish(self.span)
-                        raise
-                    except BaseException as exc:
-                        # Rejected protocol calls (e.g. concurrent asend, a
-                        # yielded GeneratorExit) need not terminate the stream.
-                        if self.gen.ag_frame is None:
-                            finish(self.span, exc)
-                        raise
-                    finally:
-                        _CURRENT.reset(token)
-
-                def __anext__(self):
-                    return self._advance(self.gen.__anext__)
-
-                def asend(self, value):
-                    return self._advance(self.gen.asend, value)
-
-                def athrow(self, *values):
-                    return self._advance(self.gen.athrow, *values)
-
-                def aclose(self):
-                    return self._advance(self.gen.aclose)
-
-            @functools.wraps(fn)
-            def stream(*args, **kwargs):
-                return ObservedGenerator(args, kwargs)
-
-            return stream
-
-        @functools.wraps(fn)
-        async def call(*args, **kwargs):
-            span = span_for(args, kwargs)
-            if span is None:
-                token = _CURRENT.set(None)
-                try:
-                    return await fn(*args, **kwargs)
-                finally:
-                    _CURRENT.reset(token)
-            try:
-                with span.activate():
-                    value = await fn(*args, **kwargs)
-                    inspect_result(value)
-                finish(span)
-                return value
-            except BaseException as exc:
-                finish(span, exc)
-                raise
-
-        return call
-
-    return decorate
-
-
-def event(owner, kind, operation, outcome, **fields):
-    """Emit a point-in-time fact from an existing state transition."""
-    try:
-        manager = _manager(owner)
-        _, explicit = _owner_app(owner)
-        parent = current_span()
-        context = _owner_context(owner)
-        if manager is None and not explicit and parent:
-            if context is not None and getattr(context, 'workspace_uuid', None) != parent.fields.get('workspace_uuid'):
-                return
-            manager = parent.manager
-        if manager is not None and _context_matches(manager, owner, context):
-            inherited = dict(parent.fields) if parent and parent.manager is manager else {}
-            if context is not None and getattr(context, 'workspace_uuid', None) != inherited.get('workspace_uuid'):
-                inherited = _context_fields(owner, {})
-            inherited.update(fields)
-            manager.emit(kind, operation, outcome, **inherited)
-    except Exception:
-        pass
-
-
-def capture_context():
-    span = current_span()
-    if span is None or not span.manager.enabled:
-        return None
-    result = {
-        k: span.fields[k]
-        for k in ('trace_id', 'workspace_uuid', 'run_id', 'adapter', 'processor_type', 'platform_event_type', 'source')
-        if k in span.fields
-    }
-    result['parent_span_id'] = span.fields['span_id']
-    return result
-
-
-def link_context(saved):
-    """Link only after the existing Host run/installation validator accepted it."""
-    span = current_span()
-    if span is None or not isinstance(saved, dict):
-        return
-    if span.fields.get('workspace_uuid') and span.fields['workspace_uuid'] != saved.get('workspace_uuid'):
-        return
-    for key in ('trace_id', 'parent_span_id', 'workspace_uuid', 'run_id'):
-        if privacy.opaque(saved.get(key)):
-            span.fields[key] = saved[key]
-    for key in ('adapter', 'platform_event_type'):
-        if privacy.category(key, saved.get(key)):
-            span.fields[key] = saved[key]
-    if saved.get('processor_type') in privacy.PROCESSORS:
-        span.fields['processor_type'] = saved['processor_type']
-    if saved.get('source') in ('webui_debug', 'synthetic'):
-        span.fields['source'] = saved['source']
-        span.fields['attributes'] = {**span.fields.get('attributes', {}), 'synthetic': True}
-
-
-def declare_runner(descriptor):
-    """Allow only identifiers from a validated installed public Runner manifest."""
-    import re
-
-    pairs = {'plugin_id': descriptor.get_plugin_id(), 'runner_id': descriptor.id}
-    for field, value in pairs.items():
-        allowed = privacy.VOCABULARY.setdefault(field, set())
-        if (
-            isinstance(value, str)
-            and len(value) <= 128
-            and len(allowed) < 1024
-            and re.fullmatch(r'(?:plugin:)?[A-Za-z0-9_.-]+/[A-Za-z0-9_.-]+(?:/[A-Za-z0-9_.-]+)?', value)
-        ):
-            allowed.add(value)
-
-
-def runner_metadata(owner, descriptor, processor_type):
-    manager = _manager(owner)
-    if manager is None or not manager.enabled:
-        return
-    metadata = {
-        'plugin_id': descriptor.get_plugin_id(),
-        'runner_id': descriptor.id,
-        'plugin_version': getattr(descriptor, 'plugin_version', ''),
-        'runner_usage': 'event' if processor_type == 'event_processor' else 'agent',
-    }
-    annotate(attributes=metadata)
-    event(
-        owner,
-        'capability',
-        'runner.run',
-        'succeeded',
-        source='runtime',
-        stage='snapshot',
-        processor_type=processor_type,
-        attributes={
-            **metadata,
-            'capability_type': 'processor',
-            'capability_name': processor_type,
-            'supported': True,
-            'configured': True,
-            'available': True,
-        },
-    )
diff --git a/src/langbot/pkg/telemetry/execution.py b/src/langbot/pkg/telemetry/execution.py
new file mode 100644
index 000000000..80123d4b7
--- /dev/null
+++ b/src/langbot/pkg/telemetry/execution.py
@@ -0,0 +1,625 @@
+"""Point-wise execution records for the existing telemetry sender.
+
+One execution is reported as a chain of independent records that all carry the
+same ``event_id``: one ``execution_node`` per workflow node, emitted the moment
+that node completes, plus exactly one ``execution_chain`` emitted when the
+execution closes. The sender never assembles a trace locally, so there is no
+per-trace buffer, no TTL and no per-process trace cap; memory is bounded by the
+outbound record buffer, the flush batch size and the small cross-task registry.
+
+Platform events are reportable only after a business route, matching plugin
+subscription or validated interaction callback admits the execution. Pending
+events enqueue neither nodes nor closures, even on failure or in Cloud mode.
+Independent API/WebUI executions remain eligible without a Bot route.
+
+Sampling is decided exactly once per event, at ingress, from a stable hash of
+the execution identity, so every record of one execution carries the same
+decision. A chain that broke is always emitted even when sampling excluded it,
+which is why a chain may arrive with zero or partial nodes.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import contextlib
+import contextvars
+import json
+import re
+import time
+import zlib
+from datetime import datetime, timezone
+
+from . import trace as trace_mod
+from .identity import workspace_identity
+from .resources import snapshot as resources_snapshot
+from .resources import begin as resources_begin
+from .resources import end as resources_end
+from .trace import TraceState
+
+MODES = frozenset({'pipeline', 'agent', 'event_processor', 'none'})
+OUTCOMES = frozenset({'success', 'failed', 'cancelled', 'timeout', 'skipped', 'unknown'})
+FAMILIES = frozenset({'platform_event', 'event_route', 'pipeline', 'runner', 'platform_api', 'tool'})
+ORIGINS = frozenset({'platform', 'webui', 'api'})
+CLOSED_BY = frozenset({'event_done', 'pipeline_done', 'runner_done', 'timeout'})
+CHAIN_FAILURES = frozenset({'failed', 'cancelled', 'timeout'})
+
+SCHEMA = 2
+TRACE_MODES = frozenset({'off', 'failures', 'sampled', 'all'})
+DEFAULT_TRACE_MODE = 'all'
+DEFAULT_TRACE_SAMPLE = 20
+
+# Sampling bounds: outbound record buffer, one flush batch, flush cadence.
+MAX_BUFFERED_RECORDS = 512
+MAX_RECORDS_PER_FLUSH = 128
+MAX_FLUSH_BYTES = 200 * 1024
+DEFAULT_FLUSH_SECONDS = 180
+MAX_FLUSH_ATTEMPTS = 3
+FLUSH_RETRY_BACKOFF_SECONDS = 0.2
+
+# Cross-task registry: execution id -> owning execution, bounded and time-limited.
+MAX_REGISTRY_KEYS = 256
+REGISTRY_TTL_SECONDS = 300
+
+# Failure details are operator-facing: keep the failure class and a short
+# machine-ish prefix, never the free text a user or a model produced.
+ERROR_LIMIT = 200
+ERROR_PREFIX_CHARS = 120
+ERROR_PREFIX_TOKENS = 12
+_ERROR_CLASS = re.compile(r'[A-Za-z_][\w.]*')
+
+# Set by cross-task call sites (plugin/RPC actions) that run without the
+# ingress context but know which execution they belong to.
+_execution_id: contextvars.ContextVar[str] = contextvars.ContextVar('telemetry_execution_id', default='')
+
+
+def set_execution_id(execution_id: str) -> contextvars.Token:
+    return _execution_id.set(str(execution_id or '').strip())
+
+
+def reset_execution_id(token: contextvars.Token | None) -> None:
+    if token is None:
+        return
+    try:
+        _execution_id.reset(token)
+    except (ValueError, RuntimeError):
+        # A token from another context must never break execution.
+        pass
+
+
+def current_execution_id() -> str:
+    return _execution_id.get()
+
+
+def _short_prefix(text: str) -> str:
+    """Keep at most a short machine-ish prefix of one failure message."""
+    if len(text) > ERROR_PREFIX_CHARS:
+        text = text[:ERROR_PREFIX_CHARS]
+    parts = text.split(' ')
+    if len(parts) > ERROR_PREFIX_TOKENS:
+        text = ' '.join(parts[:ERROR_PREFIX_TOKENS])
+    return text
+
+
+def _bounded_error(value, limit: int = ERROR_LIMIT) -> str:
+    """Bound and redact one failure detail before it leaves the process.
+
+    The failure class and a short machine-ish prefix survive; control characters
+    are stripped, whitespace is collapsed and the result is capped, so a noisy
+    failure cannot smuggle user or model content into the payload.
+    """
+    if value is None:
+        return ''
+    text = ''.join(ch if ch.isprintable() else ' ' for ch in str(value))
+    text = ' '.join(text.split())
+    if not text:
+        return ''
+    head, separator, rest = text.partition(':')
+    head = head.strip()
+    if separator and len(head) <= 80 and _ERROR_CLASS.fullmatch(head):
+        prefix = _short_prefix(rest.strip())
+        text = f'{head}: {prefix}' if prefix else head
+    else:
+        text = _short_prefix(text)
+    return text[:limit]
+
+
+def _duration_ms(started_at: str, ended_at: str) -> int:
+    """Wall-clock duration of one execution; 0 when the interval is unusable."""
+    try:
+        delta = datetime.fromisoformat(ended_at) - datetime.fromisoformat(started_at)
+        return max(int(delta.total_seconds() * 1000), 0)
+    except Exception:
+        return 0
+
+
+class ExecutionCounters:
+    """Point-wise execution reporting: one record per node, one per closure."""
+
+    def __init__(self, manager):
+        self.manager = manager
+        self.records: list[dict] = []
+        self.task: asyncio.Task | None = None
+        # Exposed health: records dropped or accepted by the sender.
+        self.dropped = 0
+        self.sent = 0
+        # Execution identity -> owning execution, for nodes recorded by tasks
+        # that no longer share the ingress context (plugin/RPC platform calls).
+        # It resolves identity and the sampling decision only; node payloads are
+        # never accumulated here.
+        self.registry: dict[str, TraceState] = {}
+        self.registry_deadlines: dict[str, float] = {}
+
+    # ------------------------------------------------------------------ config
+
+    def trace_mode(self) -> str:
+        # Cloud activity counts must not be computed from sampled successes.
+        # Telemetry opt-out is still enforced before recording observations.
+        from ..utils import constants
+
+        if constants.edition == 'cloud':
+            return 'all'
+        mode = str(self.manager.telemetry_config.get('execution_trace', DEFAULT_TRACE_MODE) or '').strip().lower()
+        return mode if mode in TRACE_MODES else DEFAULT_TRACE_MODE
+
+    def trace_sample(self) -> int:
+        try:
+            sample = int(self.manager.telemetry_config.get('execution_trace_sample', DEFAULT_TRACE_SAMPLE))
+        except (TypeError, ValueError):
+            sample = DEFAULT_TRACE_SAMPLE
+        return sample if 1 <= sample <= 100000 else DEFAULT_TRACE_SAMPLE
+
+    def flush_seconds(self) -> int:
+        try:
+            seconds = int(self.manager.telemetry_config.get('telemetry_flush_seconds', DEFAULT_FLUSH_SECONDS))
+        except (TypeError, ValueError):
+            seconds = DEFAULT_FLUSH_SECONDS
+        return seconds if 1 <= seconds <= 86400 else DEFAULT_FLUSH_SECONDS
+
+    def health(self) -> dict:
+        """Real telemetry health for the instance liveness snapshot."""
+        return {
+            'records_buffered': len(self.records),
+            'records_sent': self.sent,
+            'records_dropped': self.dropped,
+            'executions_in_flight': len({id(state) for state in self.registry.values()}),
+        }
+
+    # ----------------------------------------------------------------- sampling
+
+    def _sampling_decision(self, event_id: str) -> tuple[str, int, bool]:
+        """Decide once, deterministically, whether this event is reported.
+
+        The identity is hashed instead of being parsed, so every id shape
+        (``platform:...``, uuids, debug ids) samples the same way.
+        """
+        mode = self.trace_mode()
+        if mode == 'sampled':
+            denominator = self.trace_sample()
+            try:
+                sampled_in = zlib.crc32(str(event_id).encode('utf-8')) % denominator == 0
+            except Exception:
+                sampled_in = False
+            return mode, denominator, sampled_in
+        # 'off' reports nothing, 'failures' keeps nodes out but still reports a
+        # broken chain, 'all' reports everything; denominator is 1 for all three.
+        return mode, 1, mode == 'all'
+
+    def configure(
+        self,
+        state: TraceState,
+        context=None,
+        *,
+        debug: bool = False,
+        origin: str = 'platform',
+        identity=None,
+    ) -> None:
+        """Bind the execution identity, the sampling decision and the flags."""
+        if state.configured:
+            return
+        state.configured = True
+        state.debug = bool(debug)
+        state.origin = origin if origin in ORIGINS else 'platform'
+        mode, denominator, sampled_in = self._sampling_decision(state.event_id)
+        state.mode = mode
+        state.denominator = denominator
+        # Debug chains are always emitted (unless reporting is off entirely).
+        state.emit_nodes = mode != 'off' and (state.debug or sampled_in)
+        if identity is not None:
+            state.identity = dict(identity)
+        elif context is not None:
+            try:
+                state.identity = workspace_identity(context)
+            except Exception:
+                pass
+
+    # --------------------------------------------------------------- recording
+
+    def record(
+        self,
+        context,
+        *,
+        family: str,
+        operation: str,
+        mode: str = 'none',
+        adapter: str = '',
+        runner: str = '',
+        outcome: str = 'unknown',
+        synthetic: bool = False,
+        error: str = '',
+        execution_id: str | None = None,
+        node: str = '',
+    ):
+        try:
+            if not self._enabled() or trace_mod.reporting_suppressed():
+                return
+            if family not in FAMILIES or mode not in MODES or outcome not in OUTCOMES:
+                return
+            # Only code-defined identifiers are accepted; never pass user values.
+            if any(not isinstance(v, str) or len(v.encode('utf-8')) > 512 for v in (operation, adapter, runner, node)):
+                return
+            state = trace_mod.current()
+            if state is None:
+                # Cross-task work has no inherited context; resolve the owning
+                # execution through the registry. It owns no step here, so the
+                # record attaches to the innermost step the chain has open -
+                # the step that caused this work.
+                lookup = str(execution_id or '').strip() or current_execution_id()
+                if lookup:
+                    state = self._resolve(lookup)
+            if state is None or state.closed or not state.reportable:
+                # Every observation belongs to exactly one execution chain.
+                return
+            self.configure(state)
+            try:
+                identity = workspace_identity(context)
+            except Exception:
+                # A cross-task caller may only carry the Workspace id; reuse the
+                # identity already bound to the owning execution.
+                identity = state.identity or {}
+            if not identity.get('instance_id') or not identity.get('workspace_uuid'):
+                return
+            state.identity = identity
+
+            ended_at = datetime.now(timezone.utc).isoformat()
+            node_id, seq, started_at, parent = trace_mod.resolve_node(state, str(node or ''), ended_at)
+            root = parent == '' and not state.root_seen
+            if root:
+                state.root_seen = True
+            bounded_error = _bounded_error(error)
+            state.note_node(outcome=outcome, adapter=adapter, runner=runner, family=family, error=bounded_error)
+            if not state.emit_nodes:
+                # Sampling excluded this execution: the node is never buffered,
+                # only counted so the terminal chain can explain the gap.
+                state.dropped_nodes += 1
+                return
+            self._enqueue_record(
+                {
+                    'event_type': 'execution_node',
+                    'schema': SCHEMA,
+                    **self._common_fields(state, ended_at),
+                    'node_id': node_id,
+                    'parent_node_id': parent,
+                    'root': root,
+                    'seq': seq,
+                    'family': family,
+                    'operation': operation,
+                    'mode': mode,
+                    'adapter': adapter,
+                    'runner': runner,
+                    'outcome': outcome,
+                    'error': bounded_error,
+                    'route_ref': trace_mod.current_route(),
+                    'run_id': trace_mod.current_run(),
+                    'synthetic': bool(synthetic),
+                    'started_at': started_at,
+                    'ended_at': ended_at,
+                }
+            )
+        except Exception:
+            # Observability must never change execution behavior.
+            return
+
+    def _common_fields(self, state: TraceState, timestamp: str) -> dict:
+        identity = state.identity
+        from ..utils import constants
+
+        return {
+            'event_id': state.event_id,
+            'instance_id': identity['instance_id'],
+            'workspace_uuid': identity['workspace_uuid'],
+            'version': constants.semantic_version,
+            'edition': constants.edition,
+            'timestamp': timestamp,
+            'debug': bool(state.debug),
+            'sample': {'mode': state.mode, 'denominator': state.denominator},
+        }
+
+    def _enabled(self) -> bool:
+        try:
+            cfg = self.manager.telemetry_config
+        except Exception:
+            return False
+        return bool(cfg) and not cfg.get('disable_telemetry', False) and bool(cfg.get('url'))
+
+    def _ensure_loop(self) -> None:
+        if self.task is None or self.task.done():
+            self.task = asyncio.create_task(self._loop())
+
+    # --------------------------------------------------------------- registry
+
+    def register_trace(self, state: TraceState, execution_id) -> None:
+        """Alias one more execution identity onto an in-flight execution."""
+        try:
+            exec_id = str(execution_id or '').strip()
+            if not exec_id:
+                return
+            self.configure(state)
+            now = time.monotonic()
+            self._sweep_registry(now)
+            if exec_id not in self.registry and len(self.registry) >= MAX_REGISTRY_KEYS:
+                oldest = next(iter(self.registry))
+                self.registry.pop(oldest, None)
+                self.registry_deadlines.pop(oldest, None)
+            self.registry[exec_id] = state
+            self.registry_deadlines[exec_id] = now + REGISTRY_TTL_SECONDS
+        except Exception:
+            return
+
+    def _resolve(self, execution_id: str) -> TraceState | None:
+        self._sweep_registry(time.monotonic())
+        return self.registry.get(execution_id)
+
+    def _unregister_state(self, state: TraceState) -> None:
+        for exec_id in [key for key, value in self.registry.items() if value is state]:
+            self.registry.pop(exec_id, None)
+            self.registry_deadlines.pop(exec_id, None)
+
+    def _sweep_registry(self, now: float) -> None:
+        """Drop aliases whose owning execution never closed (leaked bindings)."""
+        for exec_id in [key for key, deadline in self.registry_deadlines.items() if deadline <= now]:
+            self.registry.pop(exec_id, None)
+            self.registry_deadlines.pop(exec_id, None)
+
+    # ------------------------------------------------------------------ chains
+
+    def close_trace(self, state: TraceState, reason: str = 'event_done', error: str = '') -> None:
+        """Emit the one terminal ``execution_chain`` record of this execution."""
+        try:
+            if state.closed:
+                return
+            state.closed = True
+            self._unregister_state(state)
+            if not self._enabled():
+                # Nothing can be uploaded; never grow the buffer for a record
+                # the sender would refuse to send.
+                return
+            self.configure(state)
+            if not self._chain_emitted(state):
+                return
+            self._enqueue_record(self._build_chain_record(state, reason, error))
+        except Exception:
+            return
+        finally:
+            # The snapshot was read while building the chain record; stop
+            # collecting so a late sibling task cannot leak into the next chain.
+            resources_end()
+
+    def _chain_emitted(self, state: TraceState) -> bool:
+        if not state.reportable or state.mode == 'off':
+            return False
+        if state.emit_nodes or state.debug:
+            return True
+        # A chain that broke is always uploaded, even when sampling excluded it.
+        return bool(state.failure_reason) or state.outcome() in CHAIN_FAILURES
+
+    def _build_chain_record(self, state: TraceState, reason: str, error: str = '') -> dict:
+        ended_at = datetime.now(timezone.utc).isoformat()
+        record = {
+            'event_type': 'execution_chain',
+            'schema': SCHEMA,
+            **self._common_fields(state, ended_at),
+            'closed_by': reason if reason in CLOSED_BY else 'event_done',
+            'outcome': state.outcome(),
+            # First failure detail in this chain, so Space can show why it broke.
+            'error': _bounded_error(error)
+            or _bounded_error(state.failure_reason)
+            or _bounded_error(state.first_error)
+            or '',
+            'duration_ms': _duration_ms(state.started_at, ended_at),
+            'stage_count': state.stage_count,
+            'dropped_nodes': state.dropped_nodes,
+            'origin': state.origin,
+            'model_name': state.model_name or '',
+            'adapter': state.adapter or state.node_adapter,
+            'runner': state.runner or state.node_runner,
+            'runner_category': state.runner_category or '',
+            'pipeline_plugins': state.pipeline_plugins,
+        }
+        # Bound resources of this execution, when any lane reported them. The
+        # key is omitted entirely when nothing was noted, so the payload stays
+        # byte-identical for executions with no bindings.
+        resources = resources_snapshot()
+        if resources:
+            record['resources'] = resources
+        return record
+
+    def _enqueue_record(self, record: dict) -> None:
+        if len(self.records) >= MAX_BUFFERED_RECORDS:
+            self.records.pop(0)
+            self.dropped += 1
+        self.records.append(record)
+        self._ensure_loop()
+
+    # ----------------------------------------------------------------- flushing
+
+    async def _loop(self):
+        interval = self.flush_seconds()
+        try:
+            # The outbound loop exists only to flush the record buffer.
+            while self.records:
+                await asyncio.sleep(interval)
+                await self.flush()
+        finally:
+            self.task = None
+
+    async def flush(self) -> bool:
+        """Upload the buffered prefix; records stay buffered until accepted."""
+        if not self.records:
+            return True
+        batch: list[dict] = []
+        size = 0
+        for record in self.records:
+            if len(batch) >= MAX_RECORDS_PER_FLUSH:
+                break
+            encoded = len(json.dumps(record, ensure_ascii=False, default=str))
+            if batch and size + encoded > MAX_FLUSH_BYTES:
+                break
+            batch.append(record)
+            size += encoded
+        if not batch:
+            return True
+        payload = {'records': batch}
+        for attempt in range(MAX_FLUSH_ATTEMPTS):
+            try:
+                if await self.manager.send(payload):
+                    # Peek-then-send: nothing is removed until the sender
+                    # accepted the exact batch it was given.
+                    del self.records[: len(batch)]
+                    self.sent += len(batch)
+                    return True
+            except asyncio.CancelledError:
+                raise
+            except Exception:
+                pass
+            if attempt + 1 < MAX_FLUSH_ATTEMPTS:
+                await asyncio.sleep(FLUSH_RETRY_BACKOFF_SECONDS * (2**attempt))
+        # The batch stays buffered for the next cycle; a permanently unaccepted
+        # record is dropped by the buffer bound above and counted there.
+        return False
+
+    async def shutdown(self):
+        if self.task is not None:
+            self.task.cancel()
+            await asyncio.gather(self.task, return_exceptions=True)
+            self.task = None
+        try:
+            deadline = time.monotonic() + 5
+            while self.records and time.monotonic() < deadline:
+                remaining = len(self.records)
+                await asyncio.wait_for(self.flush(), timeout=2)
+                if len(self.records) >= remaining:
+                    break
+        except (Exception, asyncio.CancelledError):
+            pass
+        self.dropped += len(self.records)
+        self.records.clear()
+        self.registry.clear()
+        self.registry_deadlines.clear()
+
+
+def _execution_counters(ap):
+    counters = getattr(getattr(ap, 'telemetry', None), 'execution', None)
+    return counters if isinstance(counters, ExecutionCounters) else None
+
+
+def record(ap, context, **observation):
+    """Best-effort bridge usable with optional telemetry and test doubles."""
+    try:
+        counters = _execution_counters(ap)
+        if counters is not None:
+            counters.record(context, **observation)
+    except Exception:
+        pass
+
+
+def close_trace(ap, state, reason: str = 'event_done', error: str = '') -> None:
+    """Best-effort chain close usable with optional telemetry and test doubles."""
+    try:
+        counters = _execution_counters(ap)
+        if counters is not None:
+            counters.close_trace(state, reason, error)
+    except Exception:
+        pass
+
+
+def bind_trace(ap, execution_id: str | None = None) -> trace_mod.TraceBinding:
+    """Bind an execution and alias its identity for cross-task nodes."""
+    binding = trace_mod.bind(execution_id)
+    if binding.created:
+        resources_begin()
+    if execution_id:
+        try:
+            counters = _execution_counters(ap)
+            if counters is not None:
+                counters.register_trace(binding.state, execution_id)
+        except Exception:
+            pass
+    return binding
+
+
+@contextlib.contextmanager
+def ingress(
+    ap,
+    reason: str = 'event_done',
+    context=None,
+    execution_id: str | None = None,
+    *,
+    debug: bool = False,
+    origin: str = 'platform',
+    synthetic_event: str = '',
+    require_route: bool = False,
+):
+    """Open one execution boundary; only the owner of the chain closes it.
+
+    ``require_route`` starts a platform event without reporting permission;
+    dispatch calls ``trace.admit`` when a delivery matches. Nested entrypoints
+    must reuse that decision instead of implicitly admitting the parent.
+
+    Nested boundaries (an event routed into a Pipeline, a Runner invoked from a
+    dispatch) reuse the in-flight chain instead of starting a second one, and
+    only alias their own execution id onto it. ``debug``/``origin`` apply when
+    this boundary starts the chain: a WebUI/API run without a real platform
+    event synthesizes its virtual inbound event here, so the chain has a real
+    identity and a node for the event it stands for.
+    """
+    binding = trace_mod.bind(execution_id)
+    state = binding.state
+    if binding.created:
+        if require_route:
+            state.reportable = False
+        resources_begin()
+    counters = _execution_counters(ap)
+    if counters is not None:
+        if binding.created:
+            counters.configure(state, context, debug=debug, origin=origin)
+        if execution_id:
+            counters.register_trace(state, execution_id)
+    scope = None
+    if binding.created and synthetic_event:
+        scope = trace_mod.stage_scope()
+        node = scope.__enter__()
+        record(
+            ap,
+            context,
+            family='platform_event',
+            operation=str(synthetic_event)[:512],
+            outcome='success',
+            synthetic=True,
+            node=node,
+        )
+    failure = ''
+    try:
+        yield binding
+    except BaseException as exc:
+        # The boundary itself failed: keep the reason on the chain so the
+        # uploaded payload explains the failure instead of reporting a bare
+        # outcome, even when sampling had excluded this execution.
+        if binding.created:
+            failure = f'{type(exc).__name__}: {exc}'
+            state.mark_failure(failure)
+        raise
+    finally:
+        if scope is not None:
+            scope.__exit__(None, None, None)
+        if trace_mod.unbind_root(binding):
+            close_trace(ap, state, reason, failure)
diff --git a/src/langbot/pkg/telemetry/features.py b/src/langbot/pkg/telemetry/features.py
deleted file mode 100644
index ab66f9884..000000000
--- a/src/langbot/pkg/telemetry/features.py
+++ /dev/null
@@ -1,102 +0,0 @@
-"""Per-query telemetry feature counters.
-
-Collects anonymous, content-free usage signals (tool call counts, knowledge
-base usage, sandbox executions, ...) into ``query.variables`` during query
-processing. The chat handler reads the accumulated dict when building the
-telemetry payload and ships it as the ``features`` JSON object.
-
-Every helper here is defensive: telemetry must NEVER break the pipeline, so
-all mutations are wrapped and failures are silently ignored.
-"""
-
-from __future__ import annotations
-
-import typing
-
-if typing.TYPE_CHECKING:
-    import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-
-
-FEATURES_KEY = '_telemetry_features'
-
-
-def get_features(query: pipeline_query.Query) -> dict:
-    """Return the mutable features dict for this query, creating it if needed."""
-    try:
-        return query.variables.setdefault(FEATURES_KEY, {})
-    except Exception:
-        return {}
-
-
-def increment(query: pipeline_query.Query, group: str, key: str | None = None, amount: int = 1) -> None:
-    """Increment a counter.
-
-    ``increment(q, 'sandbox', 'execs')`` -> features['sandbox']['execs'] += 1
-    ``increment(q, 'tool_call_rounds')`` -> features['tool_call_rounds'] += 1
-    """
-    try:
-        features = get_features(query)
-        if key is None:
-            features[group] = int(features.get(group, 0)) + amount
-        else:
-            nested = features.setdefault(group, {})
-            if isinstance(nested, dict):
-                nested[key] = int(nested.get(key, 0)) + amount
-    except Exception:
-        pass
-
-
-def set_value(query: pipeline_query.Query, group: str, value: typing.Any) -> None:
-    """Set a feature value (overwrites)."""
-    try:
-        get_features(query)[group] = value
-    except Exception:
-        pass
-
-
-def collect_features(query: pipeline_query.Query) -> dict:
-    """Build the final ``features`` object for the telemetry payload.
-
-    Combines the counters accumulated during processing with end-of-query
-    snapshots (activated skills, bound MCP servers). Returns a plain dict
-    that must be JSON-serializable; non-serializable values are dropped.
-    """
-    features: dict = {}
-    try:
-        accumulated = query.variables.get(FEATURES_KEY)
-        if isinstance(accumulated, dict):
-            features.update(accumulated)
-    except Exception:
-        pass
-
-    # Activated skills (names only, registered by the activate tool)
-    try:
-        activated = query.variables.get('_activated_skills', {})
-        if isinstance(activated, dict) and activated:
-            features['activated_skills'] = sorted(activated.keys())
-    except Exception:
-        pass
-
-    # MCP servers bound to the pipeline (names only; None means "all enabled")
-    try:
-        bound_mcp = query.variables.get('_pipeline_bound_mcp_servers', None)
-        if bound_mcp is not None:
-            features['mcp_servers'] = list(bound_mcp)
-    except Exception:
-        pass
-
-    # Drop anything that is not JSON-serializable
-    import json
-
-    try:
-        json.dumps(features)
-        return features
-    except Exception:
-        safe: dict = {}
-        for k, v in features.items():
-            try:
-                json.dumps({k: v})
-                safe[k] = v
-            except Exception:
-                continue
-        return safe
diff --git a/src/langbot/pkg/telemetry/heartbeat.py b/src/langbot/pkg/telemetry/heartbeat.py
index 69b73a97b..e30b4fad8 100644
--- a/src/langbot/pkg/telemetry/heartbeat.py
+++ b/src/langbot/pkg/telemetry/heartbeat.py
@@ -13,6 +13,7 @@ credentials.
 from __future__ import annotations
 
 import asyncio
+import time
 import typing
 from datetime import datetime, timezone
 
@@ -36,10 +37,66 @@ class WorkspaceResourceSnapshot(typing.TypedDict):
     mcp_server_count: int
     extension_count: int
     skill_count: int
+    agent_count: int
+    event_processor_count: int
+    plugin_processor_count: int
+    event_binding_count: int
     adapters: list[str]
     execution_generation: int
 
 
+# Agent rows are configuration, not runtime registries, so a Workspace count
+# needs one scoped query. The heartbeat runs every five minutes on Cloud; a
+# one-hour cache keeps the cost at one bounded query per Workspace per hour
+# instead of one per cycle. Unavailable counts stay -1, never a fake zero.
+_AGENT_COUNT_CACHE_TTL_SECONDS = 3600
+_agent_count_cache: dict[str, tuple[float, dict[str, int]]] = {}
+
+
+async def _workspace_agent_kind_counts(ap: core_app.Application, workspace_uuid: str) -> dict[str, int]:
+    """Count one Workspace's Agents per ``kind``; empty when unavailable."""
+    if not workspace_uuid:
+        return {}
+    now = time.monotonic()
+    cached = _agent_count_cache.get(workspace_uuid)
+    if cached is not None and cached[0] > now:
+        return dict(cached[1])
+    counts: dict[str, int] = {}
+    try:
+        from ..entity.persistence import agent as persistence_agent
+
+        statement = sqlalchemy.select(persistence_agent.Agent.kind, sqlalchemy.func.count()).group_by(
+            persistence_agent.Agent.kind
+        )
+        tenant_uow = getattr(ap.persistence_mgr, 'tenant_uow', None)
+        if callable(tenant_uow):
+            async with tenant_uow(workspace_uuid):
+                result = await ap.persistence_mgr.execute_async(statement)
+        else:
+            result = await ap.persistence_mgr.execute_async(statement)
+        for kind, count in result.all():
+            counts[str(kind or 'agent')] = int(count or 0)
+    except Exception:
+        counts = {}
+    _agent_count_cache[workspace_uuid] = (now + _AGENT_COUNT_CACHE_TTL_SECONDS, dict(counts))
+    return counts
+
+
+def _bot_processor_counts(bots) -> tuple[int, int]:
+    """Count plugin-processor subscriptions and event bindings across bots."""
+    processors = 0
+    bindings = 0
+    for bot in bots:
+        entity = getattr(bot, 'bot_entity', bot)
+        declared_processors = getattr(entity, 'plugin_processors', None)
+        if isinstance(declared_processors, list):
+            processors += len(declared_processors)
+        declared_bindings = getattr(entity, 'event_bindings', None)
+        if isinstance(declared_bindings, list):
+            bindings += len(declared_bindings)
+    return processors, bindings
+
+
 async def _count(
     ap: core_app.Application,
     table,
@@ -81,6 +138,10 @@ async def _cloud_workspace_resource_counts(ap: core_app.Application, bindings) -
             'mcp_server_count': 0,
             'extension_count': 0,
             'skill_count': 0,
+            'agent_count': -1,
+            'event_processor_count': -1,
+            'plugin_processor_count': 0,
+            'event_binding_count': 0,
             'adapters': [],
             'execution_generation': binding.placement_generation,
         }
@@ -91,6 +152,9 @@ async def _cloud_workspace_resource_counts(ap: core_app.Application, bindings) -
     for key, bot in getattr(ap.platform_mgr, '_bots_by_key', {}).items():
         if len(key) >= 2 and key[1] in counts:
             counts[key[1]]['bot_count'] += 1
+            processors, bindings = _bot_processor_counts([bot])
+            counts[key[1]]['plugin_processor_count'] += processors
+            counts[key[1]]['event_binding_count'] += bindings
             adapter = getattr(bot, 'adapter', None)
             if adapter is not None and getattr(bot, 'enable', False):
                 adapter_sets[key[1]].add(adapter.__class__.__name__)
@@ -113,6 +177,10 @@ async def _cloud_workspace_resource_counts(ap: core_app.Application, bindings) -
     for workspace_uuid, resource in counts.items():
         resource['extension_count'] = resource['plugin_count'] + resource['mcp_server_count']
         resource['adapters'] = sorted(adapter_sets[workspace_uuid])
+        agent_counts = await _workspace_agent_kind_counts(ap, workspace_uuid)
+        if agent_counts:
+            resource['agent_count'] = int(agent_counts.get('agent', 0))
+            resource['event_processor_count'] = int(agent_counts.get('event_processor', 0))
     return list(counts.values())
 
 
@@ -163,6 +231,9 @@ async def build_heartbeat_payload(
             features['bot_count'] = len(platform_mgr.bots)
             adapters = sorted({bot.adapter.__class__.__name__ for bot in enabled_bots if getattr(bot, 'adapter', None)})
             features['adapters'] = adapters
+            processors, bindings = _bot_processor_counts(platform_mgr.bots)
+            features['plugin_processor_count'] = processors
+            features['event_binding_count'] = bindings
     except Exception:
         pass
 
@@ -189,10 +260,10 @@ async def build_heartbeat_payload(
             cloud_counter=lambda: len(ap.platform_mgr._bots_by_key),
         )
 
-    # Plugin count (from plugin runtime)
+    # Scoped Cloud counts are already available; never issue tenantless RPCs.
     try:
         plugin_connector = getattr(ap, 'plugin_connector', None)
-        if plugin_connector is not None:
+        if workspace_resource is None and plugin_connector is not None:
             plugins = await plugin_connector.list_plugins()
             features['plugin_count'] = len(plugins)
     except Exception:
@@ -206,6 +277,13 @@ async def build_heartbeat_payload(
     except Exception:
         pass
 
+    # Agent processors: configuration rows, so one scoped count per Workspace.
+    # The Cloud snapshot supplies these already and overrides below.
+    if workspace_resource is None:
+        agent_counts = await _workspace_agent_kind_counts(ap, workspace_uuid)
+        features['agent_count'] = int(agent_counts.get('agent', 0)) if agent_counts else -1
+        features['event_processor_count'] = int(agent_counts.get('event_processor', 0)) if agent_counts else -1
+
     if workspace_resource is not None:
         features.update({key: value for key, value in workspace_resource.items() if key != 'workspace_uuid'})
 
@@ -271,4 +349,8 @@ async def heartbeat_loop(ap: core_app.Application) -> None:
                 ap.logger.debug(f'Telemetry heartbeat failed: {e}')
             except Exception:
                 pass
-        await asyncio.sleep(HEARTBEAT_INTERVAL_SECONDS)
+        # Cloud dashboards need fresh scoped snapshots, including Workspaces
+        # discovered after startup. A transient startup/projection failure must
+        # not suppress synchronization for an entire day.
+        cloud_runtime = getattr(getattr(ap.persistence_mgr, 'mode', None), 'value', None) == 'cloud_runtime'
+        await asyncio.sleep(300 if cloud_runtime else HEARTBEAT_INTERVAL_SECONDS)
diff --git a/src/langbot/pkg/telemetry/platform.py b/src/langbot/pkg/telemetry/platform.py
new file mode 100644
index 000000000..0842d466b
--- /dev/null
+++ b/src/langbot/pkg/telemetry/platform.py
@@ -0,0 +1,123 @@
+"""Observe adapter calls without collecting their arguments or response content."""
+
+from __future__ import annotations
+
+import asyncio
+import functools
+from contextvars import ContextVar
+
+from .execution import record
+
+processing_mode: ContextVar[str] = ContextVar('telemetry_processing_mode', default='none')
+
+
+def result_failure_detail(result) -> str:
+    """Pull a bounded human-readable reason out of a failed platform API result."""
+    raw = getattr(result, 'raw', result)
+    if isinstance(raw, dict):
+        for key in ('error', 'message', 'msg', 'detail', 'reason'):
+            value = raw.get(key)
+            if isinstance(value, str) and value.strip():
+                return value
+            if isinstance(value, dict):
+                for inner in ('message', 'msg', 'detail'):
+                    if isinstance(value.get(inner), str) and value[inner].strip():
+                        return value[inner]
+    return ''
+
+
+def result_outcome(result):
+    # Empty returns do not prove a remote operation succeeded.
+    if result is None:
+        return 'unknown'
+    if result is False:
+        return 'failed'
+    raw = getattr(result, 'raw', result)
+    if isinstance(raw, dict):
+        if raw.get('ok') is False or raw.get('success') is False or raw.get('status') == 'failed':
+            return 'failed'
+        for key in ('errcode', 'retcode'):
+            if key in raw and raw[key] not in (0, '0', None):
+                return 'failed'
+        if raw.get('error'):
+            return 'failed'
+        if isinstance(raw.get('results'), list):
+            outcomes = [result_outcome(item) for item in raw['results']]
+            if 'failed' in outcomes:
+                return 'failed'
+            if not outcomes or 'unknown' in outcomes:
+                return 'unknown'
+            return 'success'
+    # Arbitrary error envelopes or empty mappings are not affirmative evidence.
+    if isinstance(raw, dict):
+        if raw.get('ok') is True or raw.get('success') is True or raw.get('status') == 'ok':
+            return 'success'
+        if any(raw.get(key) not in (None, '') for key in ('message_id', 'id')):
+            return 'success'
+        if any(key in raw and raw[key] in (0, '0') for key in ('errcode', 'retcode')):
+            return 'success'
+        return 'unknown'
+    return 'success' if result is True or getattr(result, 'message_id', None) is not None else 'unknown'
+
+
+def observe_adapter(ap, context, adapter):
+    """Install once on a concrete bot; identity never comes from task-local tenants."""
+    if getattr(adapter, '_execution_observed', False):
+        return
+    try:
+        declared = frozenset(adapter.get_supported_apis())
+    except Exception:
+        return
+    for name in declared:
+        if '.' in name:
+            continue
+        original = getattr(adapter, name, None)
+        if not asyncio.iscoroutinefunction(original):
+            continue
+
+        def make_wrapper(method, operation):
+            @functools.wraps(method)
+            async def wrapped(*args, **kwargs):
+                observed_operation = operation
+                if operation == 'call_platform_api':
+                    action = args[0] if args else kwargs.get('action')
+                    if action in declared:
+                        observed_operation = action
+                outcome = 'unknown'
+                error_detail = ''
+                try:
+                    result = await method(*args, **kwargs)
+                    outcome = result_outcome(result)
+                    if outcome != 'success':
+                        error_detail = result_failure_detail(result)
+                    return result
+                except asyncio.CancelledError:
+                    outcome = 'cancelled'
+                    error_detail = 'cancelled'
+                    raise
+                except TimeoutError as exc:
+                    outcome = 'timeout'
+                    error_detail = str(exc) or 'timeout'
+                    raise
+                except Exception as exc:
+                    outcome = 'failed'
+                    error_detail = str(exc)
+                    raise
+                finally:
+                    record(
+                        ap,
+                        context,
+                        family='platform_api',
+                        operation=observed_operation,
+                        adapter=adapter.__class__.__name__,
+                        mode=processing_mode.get(),
+                        outcome=outcome,
+                        error=error_detail,
+                    )
+
+            return wrapped
+
+        # Adapters are Pydantic models: method instrumentation is not a model
+        # field assignment. Keep wrappers instance-local, never on the class.
+        object.__setattr__(adapter, name, make_wrapper(original, name))
+    object.__setattr__(adapter, '_execution_observed', True)
diff --git a/src/langbot/pkg/telemetry/resources.py b/src/langbot/pkg/telemetry/resources.py
new file mode 100644
index 000000000..ff959fd7c
--- /dev/null
+++ b/src/langbot/pkg/telemetry/resources.py
@@ -0,0 +1,244 @@
+"""Bound-resource reporting for one execution.
+
+Answers "what was this execution given", never "what did it use": the ids and
+names of the Plugins, MCP servers, knowledge bases, tools, skills and models
+that were bound or granted to a Pipeline lane or an Agent run. Nothing here is
+user content - only code-defined identifiers (uuids, plugin names, tool names)
+and booleans.
+
+Two sources feed one execution-scoped snapshot:
+
+* the Host policy actually granted to the run (``ResourcePolicy`` /
+  materialized ``AgentResources``), which is authoritative for tools, skills and
+  platform APIs;
+* the Runner's own configuration form (``DynamicFormItemSchema``), which is
+  where knowledge bases and models are selected. Selector field types are read
+  through :mod:`langbot.pkg.agent.runner.config_schema`, so a Runner that
+  declares a new selector type is picked up without a change here.
+
+The snapshot is attached to the terminal ``execution_chain`` record. It is
+bounded twice - per list and in total - and it is best effort: telemetry must
+never change execution behaviour, so every entry point swallows failures.
+"""
+
+from __future__ import annotations
+
+import contextlib
+import contextvars
+import json
+import typing
+
+MAX_ITEMS_PER_LIST = 64
+MAX_NAME_CHARS = 512
+MAX_RESOURCES_BYTES = 32 * 1024
+
+_resources: contextvars.ContextVar[dict[str, typing.Any] | None] = contextvars.ContextVar(
+    'telemetry_resources', default=None
+)
+
+
+def begin() -> None:
+    """Start collecting for one execution (called when a chain is opened)."""
+    _resources.set({})
+
+
+def end() -> None:
+    """Stop collecting; the snapshot has already been read."""
+    _resources.set(None)
+
+
+def _bucket(section: str) -> dict[str, typing.Any] | None:
+    root = _resources.get()
+    if root is None:
+        return None
+    bucket = root.get(section)
+    if not isinstance(bucket, dict):
+        bucket = {}
+        root[section] = bucket
+    return bucket
+
+
+def _names(values: typing.Any) -> list[str]:
+    """Bound one list of code-defined identifiers, preserving source order."""
+    if not isinstance(values, (list, tuple)):
+        return []
+    names: list[str] = []
+    seen: set[str] = set()
+    for value in values:
+        name = ''
+        if isinstance(value, str):
+            name = value
+        elif isinstance(value, dict):
+            name = str(value.get('name') or value.get('kb_id') or value.get('tool_name') or '')
+        if not name:
+            continue
+        name = name[:MAX_NAME_CHARS]
+        if name in seen:
+            continue
+        seen.add(name)
+        names.append(name)
+        if len(names) >= MAX_ITEMS_PER_LIST:
+            break
+    return names
+
+
+def _tool_refs(values: typing.Any) -> list[dict[str, str]]:
+    """Bound one list of tool grants, keeping their Host source."""
+    if not isinstance(values, (list, tuple)):
+        return []
+    refs: list[dict[str, str]] = []
+    seen: set[str] = set()
+    for value in values:
+        if isinstance(value, str):
+            name, source, source_id = value, '', ''
+        elif isinstance(value, dict):
+            name = str(value.get('tool_name') or value.get('name') or '')
+            source = str(value.get('tool_type') or value.get('source') or '')
+            raw_id = value.get('source_id')
+            source_id = raw_id if isinstance(raw_id, str) else ''
+        else:
+            continue
+        name = name[:MAX_NAME_CHARS]
+        if not name or name in seen:
+            continue
+        seen.add(name)
+        ref = {'name': name}
+        if source:
+            ref['source'] = source[:64]
+        if source_id:
+            ref['source_id'] = source_id[:MAX_NAME_CHARS]
+        refs.append(ref)
+        if len(refs) >= MAX_ITEMS_PER_LIST:
+            break
+    return refs
+
+
+def _set(bucket: dict[str, typing.Any] | None, key: str, value: typing.Any) -> None:
+    if bucket is None or value in (None, [], {}, ''):
+        return
+    bucket[key] = value
+
+
+def note_pipeline(
+    *,
+    plugins: typing.Any = None,
+    plugins_all: typing.Any = None,
+    mcp_servers: typing.Any = None,
+    mcp_all: typing.Any = None,
+    knowledge_bases: typing.Any = None,
+    skills: typing.Any = None,
+    tools: typing.Any = None,
+    tools_all: typing.Any = None,
+    runner_id: str = '',
+    selectors: typing.Any = None,
+) -> None:
+    """Record the resources one Pipeline lane is bound to.
+
+    ``plugins``/``mcp_servers`` accept ``None`` to mean "all enabled", matching
+    the binding model; the ``*_all`` flags make that explicit in the payload.
+    """
+    try:
+        bucket = _bucket('pipeline')
+        if bucket is None:
+            return
+        _set(bucket, 'plugins', _names(plugins))
+        _set(bucket, 'mcp_servers', _names(mcp_servers))
+        _set(bucket, 'knowledge_bases', _names(knowledge_bases))
+        _set(bucket, 'skills', _names(skills))
+        _set(bucket, 'tools', _tool_refs(tools))
+        _set(bucket, 'runner_id', str(runner_id or '')[:MAX_NAME_CHARS])
+        if isinstance(selectors, dict):
+            cleaned = {
+                str(kind)[:64]: _names(fields)
+                for kind, fields in selectors.items()
+                if isinstance(fields, (list, tuple)) and fields
+            }
+            _set(bucket, 'selectors', cleaned)
+        if plugins_all is not None:
+            bucket['plugins_all'] = bool(plugins_all)
+        if mcp_all is not None:
+            bucket['mcp_all'] = bool(mcp_all)
+        if tools_all is not None:
+            bucket['tools_all'] = bool(tools_all)
+    except Exception:
+        return
+
+
+def note_agent(
+    *,
+    runner_id: str = '',
+    resources: typing.Any = None,
+    selectors: typing.Any = None,
+    capabilities: typing.Any = None,
+    tools_all: typing.Any = None,
+) -> None:
+    """Record the resources one Agent run was granted.
+
+    ``resources`` is the materialized ``AgentResources`` mapping the Host
+    actually handed to the Runner (tools carry their source), ``selectors``
+    names the Runner config-form fields that selected a resource, and
+    ``capabilities`` records which resource categories the Runner declared.
+    """
+    try:
+        bucket = _bucket('agent')
+        if bucket is None:
+            return
+        _set(bucket, 'runner_id', str(runner_id or '')[:MAX_NAME_CHARS])
+        if isinstance(resources, dict):
+            _set(bucket, 'knowledge_bases', _names(resources.get('knowledge_bases')))
+            _set(bucket, 'skills', _names(resources.get('skills')))
+            _set(bucket, 'models', _names(resources.get('models')))
+            _set(bucket, 'tools', _tool_refs(resources.get('tools')))
+            _set(bucket, 'platform_apis', _names(resources.get('platform_tools')))
+        if isinstance(selectors, dict):
+            cleaned = {
+                str(kind)[:64]: _names(fields)
+                for kind, fields in selectors.items()
+                if isinstance(fields, (list, tuple)) and fields
+            }
+            _set(bucket, 'selectors', cleaned)
+        if isinstance(capabilities, dict):
+            declared = {key: bool(value) for key, value in capabilities.items() if value}
+            _set(bucket, 'capabilities', declared)
+        if tools_all is not None:
+            bucket['tools_all'] = bool(tools_all)
+    except Exception:
+        return
+
+
+def snapshot() -> dict[str, typing.Any] | None:
+    """Return the bounded snapshot of this execution; None when nothing was noted."""
+    try:
+        root = _resources.get()
+        if not isinstance(root, dict) or not root:
+            return None
+
+        def encoded_size(value: typing.Any) -> int:
+            return len(json.dumps(value, ensure_ascii=False, default=str).encode('utf-8'))
+
+        if encoded_size(root) <= MAX_RESOURCES_BYTES:
+            return root
+        # Over budget: shed the widest detail (per-tool grants) first, then fall
+        # back to a marker, so an oversized binding can never block the chain
+        # record or bloat the upload.
+        reduced = json.loads(json.dumps(root, ensure_ascii=False, default=str))
+        for section in ('pipeline', 'agent'):
+            bucket = reduced.get(section)
+            if isinstance(bucket, dict):
+                bucket.pop('tools', None)
+        reduced['truncated'] = True
+        if encoded_size(reduced) <= MAX_RESOURCES_BYTES:
+            return reduced
+        return {'truncated': True}
+    except Exception:
+        return None
+
+
+@contextlib.contextmanager
+def collected() -> typing.Iterator[None]:
+    """Bind a collection scope, for callers outside an execution boundary."""
+    token = _resources.set({})
+    try:
+        yield
+    finally:
+        _resources.reset(token)
diff --git a/src/langbot/pkg/telemetry/telemetry.py b/src/langbot/pkg/telemetry/telemetry.py
index 3d3e29898..22ac71ae4 100644
--- a/src/langbot/pkg/telemetry/telemetry.py
+++ b/src/langbot/pkg/telemetry/telemetry.py
@@ -9,9 +9,7 @@ import httpx
 
 from ..core import app as core_app
 from ..utils import httpclient
-
-
-_MAX_INFLIGHT_TELEMETRY_TASKS = 8
+from .execution import ExecutionCounters
 
 
 class TelemetryManager:
@@ -26,34 +24,18 @@ class TelemetryManager:
         self.ap = ap
 
         self.telemetry_config: dict[str, typing.Any] = {}
-        self.send_tasks: list[asyncio.Task] = []
         self._client: httpx.AsyncClient | None = None
+        self.execution = ExecutionCounters(self)
 
     async def initialize(self):
         self.telemetry_config = self.ap.instance_config.data.get('space', {})
 
-    async def start_send_task(self, payload: dict):
-        self.send_tasks = [task for task in self.send_tasks if not task.done()]
-        if len(self.send_tasks) >= _MAX_INFLIGHT_TELEMETRY_TASKS:
-            self.ap.logger.debug('Telemetry queue is full; dropping best-effort event')
-            return
-        task = asyncio.create_task(self.send(payload))
-        self.send_tasks.append(task)
-        task.add_done_callback(self._send_task_done)
-
-    def _send_task_done(self, task: asyncio.Task) -> None:
-        try:
-            self.send_tasks.remove(task)
-        except ValueError:
-            pass
+    def health(self) -> dict:
+        """Telemetry health for the instance liveness snapshot."""
+        return self.execution.health()
 
     async def shutdown(self) -> None:
-        tasks = list(self.send_tasks)
-        for task in tasks:
-            task.cancel()
-        if tasks:
-            await asyncio.gather(*tasks, return_exceptions=True)
-        self.send_tasks.clear()
+        await self.execution.shutdown()
         if self._client is not None:
             await self._client.aclose()
             self._client = None
@@ -92,45 +74,51 @@ class TelemetryManager:
             url = server.rstrip('/') + '/api/v1/telemetry'
 
             try:
-                # Sanitize payload so string fields are strings and not nulls
+                # Sanitize payload so string fields are strings and not nulls.
+                # A record batch is an envelope of whole execution records; it
+                # carries no top-level record fields, so only a single-record
+                # payload gets the legacy defaults.
                 sanitized = dict(payload)
-                if 'query_id' in sanitized:
-                    try:
-                        sanitized['query_id'] = '' if sanitized['query_id'] is None else str(sanitized['query_id'])
-                    except Exception:
-                        sanitized['query_id'] = str(sanitized.get('query_id', ''))
+                if isinstance(sanitized.get('records'), list):
+                    sanitized['records'] = [item for item in sanitized['records'] if isinstance(item, dict)]
+                else:
+                    if 'query_id' in sanitized:
+                        try:
+                            sanitized['query_id'] = '' if sanitized['query_id'] is None else str(sanitized['query_id'])
+                        except Exception:
+                            sanitized['query_id'] = str(sanitized.get('query_id', ''))
 
-                for sfield in (
-                    'adapter',
-                    'runner',
-                    'runner_category',
-                    'model_name',
-                    'version',
-                    'edition',
-                    'error',
-                    'timestamp',
-                    'event_type',
-                ):
-                    if sfield not in sanitized:
-                        continue
-                    v = sanitized.get(sfield)
-                    sanitized[sfield] = '' if v is None else str(v)
+                    for sfield in (
+                        'adapter',
+                        'runner',
+                        'runner_category',
+                        'model_name',
+                        'version',
+                        'edition',
+                        'error',
+                        'timestamp',
+                        'event_type',
+                    ):
+                        if sfield not in sanitized:
+                            continue
+                        v = sanitized.get(sfield)
+                        sanitized[sfield] = '' if v is None else str(v)
 
-                # event_type defaults to 'query' for backward compatibility
-                if not sanitized.get('event_type'):
-                    sanitized['event_type'] = 'query'
+                    # event_type defaults to 'query' for backward compatibility
+                    if not sanitized.get('event_type'):
+                        sanitized['event_type'] = 'query'
 
-                # features must be a JSON object
-                if 'features' in sanitized and not isinstance(sanitized['features'], dict):
-                    sanitized['features'] = {}
+                    # features must be a JSON object
+                    if 'features' in sanitized and not isinstance(sanitized['features'], dict):
+                        sanitized['features'] = {}
 
-                if 'duration_ms' in sanitized:
-                    try:
-                        sanitized['duration_ms'] = (
-                            int(sanitized['duration_ms']) if sanitized['duration_ms'] is not None else 0
-                        )
-                    except Exception:
-                        sanitized['duration_ms'] = 0
+                    if 'duration_ms' in sanitized:
+                        try:
+                            sanitized['duration_ms'] = (
+                                int(sanitized['duration_ms']) if sanitized['duration_ms'] is not None else 0
+                            )
+                        except Exception:
+                            sanitized['duration_ms'] = 0
 
                 async with self._client_context() as client:
                     try:
@@ -167,18 +155,17 @@ class TelemetryManager:
 
                         if resp.status_code >= 400:
                             body = await httpclient.response_text(resp, max_chars=200)
-                            self.ap.logger.warning(
-                                f'Telemetry post to {url} returned status {resp.status_code} - {body}'
-                            )
+                            self.ap.logger.debug(f'Telemetry post to {url} returned status {resp.status_code} - {body}')
                         else:
                             # Detect application-level errors inside HTTP 200 responses
                             app_err = False
+                            j = None
                             try:
                                 j = await httpclient.parse_json_response(resp)
                                 app_code = j.get('code') if isinstance(j, dict) else None
                                 if app_code is not None and int(app_code) >= 400:
                                     app_err = True
-                                    self.ap.logger.warning(
+                                    self.ap.logger.debug(
                                         f'Telemetry post to {url} returned application error code {j.get("code")} - {j.get("msg")}'
                                     )
                             except Exception:
@@ -186,7 +173,7 @@ class TelemetryManager:
 
                             if app_err:
                                 body = await httpclient.response_text(resp, max_chars=200)
-                                self.ap.logger.warning(
+                                self.ap.logger.debug(
                                     f'Telemetry post to {url} returned app-level error - response: {body}'
                                 )
                             else:
@@ -194,20 +181,35 @@ class TelemetryManager:
                                 self.ap.logger.debug(
                                     f'Telemetry posted to {url}, status {resp.status_code} - response: {body}'
                                 )
+                            if not app_err:
+                                if isinstance(sanitized.get('records'), list):
+                                    ack = j.get('data') if isinstance(j, dict) else None
+                                    if (
+                                        not isinstance(ack, dict)
+                                        or type(ack.get('stored')) is not int
+                                        or type(ack.get('rejected')) is not int
+                                        or ack['rejected'] != 0
+                                        or ack['stored'] != len(sanitized['records'])
+                                    ):
+                                        self.ap.logger.warning(
+                                            'Telemetry batch was not fully acknowledged; retaining records'
+                                        )
+                                        return False
+                                return True
                     except asyncio.TimeoutError:
-                        self.ap.logger.warning(f'Telemetry post to {url} timed out')
+                        self.ap.logger.debug(f'Telemetry post to {url} timed out')
                     except Exception as e:
-                        self.ap.logger.warning(f'Failed to post telemetry to {url}: {e}', exc_info=True)
+                        self.ap.logger.debug(f'Failed to post telemetry to {url}: {e}', exc_info=True)
             except Exception as e:
                 try:
-                    self.ap.logger.warning(
+                    self.ap.logger.debug(
                         f'Failed to create HTTP client for telemetry or sanitize payload: {e}', exc_info=True
                     )
                 except Exception:
                     pass
         except Exception as e:
-            # Never raise from telemetry; surface as warning for visibility
+            # Never raise from telemetry; diagnostics are debug-only.
             try:
-                self.ap.logger.warning(f'Unexpected telemetry error: {e}', exc_info=True)
+                self.ap.logger.debug(f'Unexpected telemetry error: {e}', exc_info=True)
             except Exception:
                 pass
diff --git a/src/langbot/pkg/telemetry/trace.py b/src/langbot/pkg/telemetry/trace.py
new file mode 100644
index 000000000..ddc5f5d05
--- /dev/null
+++ b/src/langbot/pkg/telemetry/trace.py
@@ -0,0 +1,351 @@
+"""Per-execution telemetry context for the v2 execution chain.
+
+One execution identity (a Pipeline query, inbound platform event or Agent run)
+is one chain. The chain is not assembled here: every workflow node is emitted
+by ``telemetry.execution`` the moment it completes, carrying its own identity
+and the identity of the execution it belongs to. This module only holds the
+small execution-scoped context the emitter needs:
+
+* the execution state (identity, the once-per-event sampling decision, the
+  execution-scoped display fields);
+* the routing/run identity pinned around the block that produced a node;
+* the open-step stack, which lives in a ContextVar holding *one task's own
+  immutable tuple*. Child tasks inherit a snapshot at creation; no shared
+  mutable stack exists, so sibling tasks can never corrupt each other's parent
+  links.
+
+Cross-task callers (plugin/RPC work with no inherited context) resolve their
+execution through the owner's execution-id registry instead, which only
+carries the execution state - never node payloads. Such a caller owns no step
+of its own, yet its work is still caused by a step of the chain: the sandbox
+tool a Runner invoked, the platform API a plugin called back over RPC. The
+execution state therefore also keeps the ordered set of steps any task
+currently has open, and a record emitted with no open step of its own attaches
+to the innermost one of those instead of becoming a root of the chain.
+"""
+
+from __future__ import annotations
+
+import contextlib
+import contextvars
+import typing
+from datetime import datetime, timezone
+from uuid import uuid4
+
+_current: contextvars.ContextVar['TraceState | None'] = contextvars.ContextVar('telemetry_trace', default=None)
+_route: contextvars.ContextVar[str] = contextvars.ContextVar('telemetry_trace_route', default='')
+_run: contextvars.ContextVar[str] = contextvars.ContextVar('telemetry_trace_run', default='')
+_reporting_suppressed: contextvars.ContextVar[bool] = contextvars.ContextVar(
+    'telemetry_reporting_suppressed', default=False
+)
+# Open workflow step frames of the current task, innermost last. The tuple is
+# never mutated in place: opening a step publishes a new tuple on this task's
+# context only, and closing it restores the previous one.
+_node: contextvars.ContextVar[tuple['NodeFrame', ...]] = contextvars.ContextVar('telemetry_trace_node', default=())
+
+
+def _now() -> str:
+    return datetime.now(timezone.utc).isoformat()
+
+
+class NodeFrame(typing.NamedTuple):
+    """One workflow step open in the task that opened it."""
+
+    node_id: str
+    seq: int
+    started_at: str
+
+
+class TraceState:
+    """The small per-execution context; node payloads never accumulate here."""
+
+    __slots__ = (
+        'event_id',
+        'identity',
+        'configured',
+        'reportable',
+        'mode',
+        'denominator',
+        'emit_nodes',
+        'debug',
+        'origin',
+        'started_at',
+        'node_seq',
+        'root_seen',
+        'stage_count',
+        'dropped_nodes',
+        'last_outcome',
+        'first_error',
+        'failure_reason',
+        'runner_failed',
+        'closed',
+        'adapter',
+        'runner',
+        'runner_category',
+        'model_name',
+        'pipeline_plugins',
+        'node_adapter',
+        'node_runner',
+        'open_frames',
+    )
+
+    def __init__(self, execution_id: str | None = None) -> None:
+        # The execution identity every record of this chain carries.
+        self.event_id = str(execution_id).strip() if execution_id else str(uuid4())
+        self.identity: dict[str, str] = {}
+        self.configured = False
+        # Platform ingress starts pending; an actual route admits the shared
+        # execution before work starts. Independent API/debug runs are eligible.
+        self.reportable = not reporting_suppressed()
+        # Sampling, decided exactly once per event (at ingress or first touch).
+        self.mode = 'all'
+        self.denominator = 1
+        self.emit_nodes = False
+        self.debug = False
+        self.origin = 'platform'
+        self.started_at = _now()
+        self.node_seq = 0
+        self.root_seen = False
+        self.stage_count = 0
+        self.dropped_nodes = 0
+        self.last_outcome = ''
+        self.first_error = ''
+        self.failure_reason = ''
+        # Set when a runner node already reported a non-success outcome.
+        self.runner_failed = False
+        self.closed = False
+        # Execution-scoped record fields attached by the owning lane when known.
+        self.adapter = ''
+        self.runner = ''
+        self.runner_category = ''
+        self.model_name = ''
+        self.pipeline_plugins: typing.Any = None
+        # First adapter/runner seen on a node, used for the chain record.
+        self.node_adapter = ''
+        self.node_runner = ''
+        # Steps currently open anywhere in this chain, in the order they were
+        # opened. A record emitted by a task that owns no step of its own
+        # attaches to the innermost entry here.
+        self.open_frames: dict[str, NodeFrame] = {}
+
+    def allocate_seq(self) -> int:
+        """Mint the next chain-local node order."""
+        seq = self.node_seq
+        self.node_seq += 1
+        return seq
+
+    def open_lane(self) -> str:
+        """The innermost step open in this chain, whichever task opened it."""
+        frames = list(self.open_frames)
+        return frames[-1] if frames else ''
+
+    def open_step(self, frame: NodeFrame) -> None:
+        self.open_frames[frame.node_id] = frame
+
+    def close_step(self, frame: NodeFrame) -> None:
+        self.open_frames.pop(frame.node_id, None)
+
+    def note_node(
+        self,
+        *,
+        outcome: str,
+        adapter: str = '',
+        runner: str = '',
+        family: str = '',
+        error: str = '',
+    ) -> None:
+        """Remember only what the terminal chain record needs from a node."""
+        self.stage_count += 1
+        self.last_outcome = outcome or 'unknown'
+        if adapter and not self.node_adapter:
+            self.node_adapter = adapter
+        if runner and not self.node_runner:
+            self.node_runner = runner
+        if family == 'runner' and outcome not in ('success', 'skipped'):
+            # The lane records its own failed runner step only when the runner
+            # orchestrator has not already landed one.
+            self.runner_failed = True
+        if outcome in ('failed', 'cancelled', 'timeout') and error and not self.first_error:
+            # The chain must explain itself even when the break was reported as
+            # a failed node rather than as an exception at the boundary.
+            self.first_error = error
+
+    def mark_failure(self, reason: str) -> None:
+        """Remember why this chain broke so the payload can explain itself."""
+        if not self.failure_reason:
+            self.failure_reason = reason
+
+    def outcome(self) -> str:
+        """Terminal outcome of this chain; a broken chain never reports success."""
+        if self.failure_reason and self.last_outcome in ('', 'success', 'skipped', 'unknown'):
+            return 'failed'
+        return self.last_outcome or 'unknown'
+
+
+class TraceBinding(typing.NamedTuple):
+    state: TraceState
+    created: bool
+    token: typing.Any
+
+
+def bind(execution_id: str | None = None) -> TraceBinding:
+    """Start a chain unless one is already in flight in this context.
+
+    The execution id becomes the chain identity only when this call starts it;
+    a nested boundary keeps the enclosing execution's identity.
+    """
+    existing = _current.get()
+    if existing is not None:
+        return TraceBinding(existing, False, None)
+    state = TraceState(execution_id)
+    return TraceBinding(state, True, _current.set(state))
+
+
+def unbind(binding: TraceBinding) -> bool:
+    """Detach this binding. Returns True when this caller owns the chain."""
+    if binding.token is not None:
+        try:
+            _current.reset(binding.token)
+        except (ValueError, RuntimeError):
+            # A token from another context must never break execution.
+            pass
+    return binding.created
+
+
+def unbind_root(binding: TraceBinding) -> bool:
+    """Detach only when this caller started the chain, else leave it in place."""
+    if not binding.created:
+        return False
+    return unbind(binding)
+
+
+def current() -> TraceState | None:
+    return _current.get()
+
+
+def reporting_suppressed() -> bool:
+    return _reporting_suppressed.get()
+
+
+def admit() -> None:
+    """Admit a matched delivery without revoking any concurrent sibling's choice."""
+    state = current()
+    if state is not None and not state.closed and not reporting_suppressed():
+        state.reportable = True
+
+
+@contextlib.contextmanager
+def suppress_reporting():
+    """Keep discarded branch side effects local even if a sibling is admitted."""
+    token = _reporting_suppressed.set(True)
+    try:
+        yield
+    finally:
+        _reporting_suppressed.reset(token)
+
+
+def node_stack() -> tuple[NodeFrame, ...]:
+    """This task's own open-step frames, innermost last."""
+    return _node.get()
+
+
+def current_parent() -> str:
+    """The step node that records emitted right now attach to ('' when none)."""
+    stack = _node.get()
+    return stack[-1].node_id if stack else ''
+
+
+def enclosing_parent() -> str:
+    """The step enclosing the innermost open one; '' when it is a root step."""
+    stack = _node.get()
+    return stack[-2].node_id if len(stack) > 1 else ''
+
+
+def resolve_node(
+    state: TraceState,
+    node: str,
+    fallback_started_at: str,
+) -> tuple[str, int, str, str]:
+    """Resolve one record to ``(node_id, seq, started_at, parent_node_id)``.
+
+    A record stamped with a ``node`` the calling task has open reuses that
+    step's identity, order and start time; anything else (a plain observation,
+    or a node opened by another task) is a fresh node under the innermost step
+    open in this task. A task that inherited no step at all - plugin/RPC work
+    that only knows the execution - attaches to the innermost step the chain
+    has open, which is the step that caused the work.
+    """
+    stack = _node.get()
+    if node:
+        for index in range(len(stack) - 1, -1, -1):
+            frame = stack[index]
+            if frame.node_id == node:
+                parent = stack[index - 1].node_id if index > 0 else ''
+                return frame.node_id, frame.seq, frame.started_at, parent
+    parent = stack[-1].node_id if stack else state.open_lane()
+    return node or uuid4().hex, state.allocate_seq(), fallback_started_at, parent
+
+
+@contextlib.contextmanager
+def stage_scope() -> typing.Iterator[str]:
+    """Open one workflow step node for the duration of the block.
+
+    Allocates a globally unique node id and the chain-local order, publishes the
+    frame on this task's own stack so nested observations attach to it as their
+    parent, and yields the node id so the step's own record can be stamped with
+    ``node=``. Outside any execution nothing is allocated and '' is yielded.
+    """
+    state = _current.get()
+    if state is None:
+        yield ''
+        return
+    frame = NodeFrame(uuid4().hex, state.allocate_seq(), _now())
+    token = _node.set(_node.get() + (frame,))
+    state.open_step(frame)
+    try:
+        yield frame.node_id
+    finally:
+        state.close_step(frame)
+        try:
+            _node.reset(token)
+        except (ValueError, RuntimeError):
+            # A token from another context must never break execution.
+            pass
+
+
+def set_run(run_id: str) -> typing.Any:
+    return _run.set(run_id)
+
+
+def reset_run(token: typing.Any) -> None:
+    if token is None:
+        return
+    try:
+        _run.reset(token)
+    except (ValueError, RuntimeError):
+        # A token from another context must never break execution.
+        pass
+
+
+@contextlib.contextmanager
+def scope(*, route_ref: str | None = None) -> typing.Iterator[None]:
+    """Pin routing identity for nodes recorded inside this block."""
+    if not route_ref:
+        yield
+        return
+    token = _route.set(route_ref)
+    try:
+        yield
+    finally:
+        try:
+            _route.reset(token)
+        except (ValueError, RuntimeError):
+            pass
+
+
+def current_route() -> str:
+    return _route.get()
+
+
+def current_run() -> str:
+    return _run.get()
diff --git a/src/langbot/pkg/utils/bounded_executor.py b/src/langbot/pkg/utils/bounded_executor.py
index 006714181..cc53beb6a 100644
--- a/src/langbot/pkg/utils/bounded_executor.py
+++ b/src/langbot/pkg/utils/bounded_executor.py
@@ -173,6 +173,13 @@ class BoundedThreadPoolExecutor(concurrent.futures.ThreadPoolExecutor):
             max_workers=max_workers,
             thread_name_prefix=thread_name_prefix,
         )
+        # asyncio DNS uses the default executor too. Keep a separately bounded
+        # resolver pool so long-lived SDK calls cannot starve control-plane DNS.
+        self._dns_executor = (
+            BoundedThreadPoolExecutor(max_workers=2, max_pending=32, thread_name_prefix='langbot-dns')
+            if thread_name_prefix != 'langbot-dns'
+            else None
+        )
         self.max_workers = max_workers
         self.max_pending = max_pending
         self.max_inflight_per_scope = max_inflight_per_scope
@@ -194,6 +201,10 @@ class BoundedThreadPoolExecutor(concurrent.futures.ThreadPoolExecutor):
         *args: Any,
         **kwargs: Any,
     ) -> concurrent.futures.Future:
+        import socket
+
+        if self._dns_executor is not None and fn in (socket.getaddrinfo, socket.getnameinfo):
+            return self._dns_executor.submit(fn, *args, **kwargs)
         scope = current_blocking_work_scope()
         if not self._capacity.acquire(blocking=False):
             with self._stats_lock:
@@ -246,6 +257,11 @@ class BoundedThreadPoolExecutor(concurrent.futures.ThreadPoolExecutor):
         future.add_done_callback(complete)
         return future
 
+    def shutdown(self, wait: bool = True, *, cancel_futures: bool = False) -> None:
+        if self._dns_executor is not None:
+            self._dns_executor.shutdown(wait=wait, cancel_futures=cancel_futures)
+        super().shutdown(wait=wait, cancel_futures=cancel_futures)
+
     def _release_scope_locked(self, scope: str | None) -> None:
         if scope is None:
             return
diff --git a/src/langbot/pkg/utils/execution_lease.py b/src/langbot/pkg/utils/execution_lease.py
new file mode 100644
index 000000000..037441de7
--- /dev/null
+++ b/src/langbot/pkg/utils/execution_lease.py
@@ -0,0 +1,145 @@
+"""Host liveness leases; no age-based timeout for live executions."""
+
+import asyncio
+import datetime
+import uuid
+
+import sqlalchemy as sa
+
+from ..entity.persistence.agent_run import AgentRun
+from ..entity.persistence.monitoring import MonitoringMessage, MonitoringError
+from .inflight import inflight_hub
+
+OWNER_ID = str(uuid.uuid4())
+LEASE_SECONDS = 120
+INTERRUPTED_REASON = 'Execution interrupted: the LangBot host stopped renewing its lease.'
+
+
+def host_execution_lease():
+    return {
+        'execution_owner_id': OWNER_ID,
+        'execution_lease_expires_at': datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None)
+        + datetime.timedelta(seconds=LEASE_SECONDS),
+    }
+
+
+async def reconcile_execution_leases(engine, workspace_uuid, *, now=None, owner_id=OWNER_ID):
+    """Renew our live leases, then conditionally finish bounded expired batches.
+
+    Never adopt a lease that already expired: another host may be reaping it.
+    Remote queued/claimed runs and pre-upgrade rows without ownership are excluded.
+    """
+    now = now or datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None)
+    changed = 0
+    workspaces = set()
+    for model, statuses, scope, identity in (
+        (AgentRun, ['running'], AgentRun.workspace_id, AgentRun.run_id),
+        (MonitoringMessage, ['pending'], MonitoringMessage.workspace_uuid, MonitoringMessage.id),
+    ):
+        conditions = [model.status.in_(statuses), scope == workspace_uuid]
+        if model is AgentRun:
+            conditions.extend([AgentRun.queue_name.is_(None), AgentRun.claimed_by_runtime_id.is_(None)])
+        async with engine.begin() as connection:
+            if engine.dialect.name == 'postgresql':
+                await connection.execute(
+                    sa.text("SELECT set_config('langbot.workspace_uuid', :workspace, true)"),
+                    {'workspace': workspace_uuid},
+                )
+            await connection.execute(
+                sa.update(model)
+                .where(
+                    *conditions,
+                    model.execution_owner_id == owner_id,
+                    model.execution_lease_expires_at > now,
+                )
+                .values(execution_lease_expires_at=now + datetime.timedelta(seconds=LEASE_SECONDS))
+            )
+            expired = [*conditions, model.execution_lease_expires_at <= now]
+            rows = (
+                await connection.execute(
+                    sa.select(identity, scope).where(*expired).order_by(model.execution_lease_expires_at).limit(500)
+                )
+            ).all()
+            if not rows:
+                continue
+            values = (
+                {'status': 'failed', 'status_reason': INTERRUPTED_REASON, 'finished_at': now, 'updated_at': now}
+                if model is AgentRun
+                else {
+                    'status': 'error',
+                    'level': 'error',
+                }
+            )
+            if model is MonitoringMessage:
+                messages = (
+                    (
+                        await connection.execute(
+                            sa.select(model).where(
+                                model.id.in_([row[0] for row in rows]),
+                            )
+                        )
+                    )
+                    .mappings()
+                    .all()
+                )
+                for message in messages:
+                    result = await connection.execute(
+                        sa.update(model)
+                        .where(
+                            *expired,
+                            model.id == message['id'],
+                        )
+                        .values(**values)
+                    )
+                    changed += result.rowcount
+                    if result.rowcount:
+                        await connection.execute(
+                            sa.insert(MonitoringError).values(
+                                id=str(uuid.uuid4()),
+                                timestamp=now,
+                                error_type='ExecutionInterrupted',
+                                error_message=INTERRUPTED_REASON,
+                                message_id=message['id'],
+                                **{
+                                    key: message[key]
+                                    for key in (
+                                        'workspace_uuid',
+                                        'bot_id',
+                                        'bot_name',
+                                        'pipeline_id',
+                                        'pipeline_name',
+                                        'session_id',
+                                    )
+                                },
+                            )
+                        )
+            else:
+                result = await connection.execute(
+                    sa.update(model)
+                    .where(
+                        *expired,
+                        identity.in_([row[0] for row in rows]),
+                    )
+                    .values(**values)
+                )
+                changed += result.rowcount
+            workspaces.update(row[1] for row in rows)
+    for workspace in workspaces:
+        inflight_hub.notify(workspace)
+    return changed
+
+
+async def maintain_execution_leases(ap):
+    while True:
+        try:
+            count = 0
+            for binding in await ap.workspace_service.list_active_execution_bindings():
+                count += await reconcile_execution_leases(
+                    ap.persistence_mgr.get_db_engine(),
+                    binding.workspace_uuid,
+                )
+            if count:
+                ap.logger.info(f'Marked {count} interrupted executions as failed after host lease expiry')
+        except Exception:
+            ap.logger.exception('Execution lease maintenance failed')
+        await asyncio.sleep(20)
diff --git a/src/langbot/pkg/utils/inflight.py b/src/langbot/pkg/utils/inflight.py
new file mode 100644
index 000000000..8d6b8c0b4
--- /dev/null
+++ b/src/langbot/pkg/utils/inflight.py
@@ -0,0 +1,100 @@
+"""Bounded, process-local fan-out. One snapshot producer per watched Workspace.
+
+Writes wake the producer; quiet workspaces reconcile every 15 seconds to cover
+external workers. A shared broker can replace notify without changing clients.
+No payloads or subscriptions survive the final viewer disconnecting.
+"""
+
+import asyncio
+import contextlib
+import logging
+import time
+
+
+logger = logging.getLogger(__name__)
+
+
+class InflightHub:
+    def __init__(self):
+        self.channels = {}
+        self.run_workspaces = {}
+
+    def notify(self, workspace):
+        channel = self.channels.get(workspace)
+        if channel:
+            channel['dirty'].set()
+
+    def notify_run(self, run_id, event_type):
+        workspace = self.run_workspaces.get(run_id)
+        channel = self.channels.get(workspace)
+        if channel and channel['phases'].get(run_id) != event_type:
+            channel['phases'][run_id] = event_type
+            channel['dirty'].set()
+
+    def track(self, workspace, run_ids):
+        self.run_workspaces = {key: value for key, value in self.run_workspaces.items() if value != workspace}
+        if workspace not in self.channels:
+            return
+        self.run_workspaces.update({run_id: workspace for run_id in run_ids})
+        if workspace in self.channels:
+            phases = self.channels[workspace]['phases']
+            self.channels[workspace]['phases'] = {key: value for key, value in phases.items() if key in run_ids}
+
+    async def watch(self, workspace, loader):
+        if sum(len(c['queues']) for c in self.channels.values()) >= 1024:
+            raise RuntimeError('Too many monitoring connections')
+        channel = self.channels.get(workspace)
+        if channel is None:
+            channel = dict(queues=set(), dirty=asyncio.Event(), phases={}, snapshot=None)
+            self.channels[workspace] = channel
+            channel['task'] = asyncio.create_task(self._produce(channel, loader))
+        if len(channel['queues']) >= 64:
+            raise RuntimeError('Too many Workspace monitoring connections')
+        queue = asyncio.Queue(maxsize=1)
+        channel['queues'].add(queue)
+        if channel['snapshot'] is not None:
+            queue.put_nowait(channel['snapshot'])
+        try:
+            # Reconnect periodically so HTTP authentication/authorization is renewed.
+            deadline = time.monotonic() + 60
+            while time.monotonic() < deadline:
+                try:
+                    yield await asyncio.wait_for(queue.get(), timeout=min(10, deadline - time.monotonic()))
+                except asyncio.TimeoutError:
+                    yield {'kind': 'heartbeat'}
+        finally:
+            channel['queues'].discard(queue)
+            if not channel['queues']:
+                self.channels.pop(workspace, None)
+                self.track(workspace, [])
+                channel['task'].cancel()
+                with contextlib.suppress(asyncio.CancelledError):
+                    await channel['task']
+
+    async def _produce(self, channel, loader):
+        previous = None
+        while True:
+            channel['dirty'].clear()
+            try:
+                snapshot = await loader()
+                frame = {'kind': 'snapshot', 'data': snapshot}
+            except Exception:
+                if previous is None or previous['kind'] != 'unavailable':
+                    logger.exception('Failed to load in-flight monitoring snapshot')
+                frame = {'kind': 'unavailable'}
+            # Heartbeats are sent separately; unchanged snapshots generate no traffic.
+            if frame != previous:
+                channel['snapshot'] = frame
+                for queue in tuple(channel['queues']):
+                    if queue.full():
+                        queue.get_nowait()
+                    queue.put_nowait(frame)
+                previous = frame
+            await asyncio.sleep(0.75)
+            try:
+                await asyncio.wait_for(channel['dirty'].wait(), timeout=15)
+            except asyncio.TimeoutError:
+                pass
+
+
+inflight_hub = InflightHub()
diff --git a/src/langbot/pkg/vector/mgr.py b/src/langbot/pkg/vector/mgr.py
index 21e868223..03424af10 100644
--- a/src/langbot/pkg/vector/mgr.py
+++ b/src/langbot/pkg/vector/mgr.py
@@ -356,7 +356,10 @@ class VectorDBManager:
             scope = await self._resolve_pgvector_scope(
                 execution_context,
                 knowledge_base_uuid,
-                expected_dimension=len(query_vector),
+                # A text-only search sends no query vector; there is no dimension
+                # to bind, and validating an empty vector as dimension 0 would
+                # mask the backend capability error.
+                expected_dimension=len(query_vector) or None,
                 initialize_dimension=False,
             )
             results = await pgvector.search(
diff --git a/src/langbot/templates/config.yaml b/src/langbot/templates/config.yaml
index 81af5c004..de5092370 100644
--- a/src/langbot/templates/config.yaml
+++ b/src/langbot/templates/config.yaml
@@ -293,6 +293,20 @@ plugin:
     binary_storage:
         # Max bytes for a single plugin binary storage value
         max_value_bytes: 10485760
+    runtime_ops:
+        # Cloud-only ops telemetry: aggregates this instance's plugin Runtime
+        # health, desired installations and reconcile failures and pushes them
+        # to the Space control plane for the internal ops dashboard. This is
+        # separate from product telemetry (space.disable_telemetry) and has no
+        # effect outside the cloud edition. Defaults to on for the cloud
+        # edition; set false to stop pushing samples.
+        enabled: true
+        # Seconds between samples pushed to Space.
+        interval_seconds: 30
+        # Per-request timeout for the Runtime /healthz probe and the upload.
+        timeout_seconds: 10
+        # Maximum number of failed installations included per sample.
+        max_failures: 5
 runner:
     # Host-level admin permissions for trusted control plugins. These plugins
     # can use existing plugin action handlers to inspect or manage Runner
@@ -401,7 +415,7 @@ box:
         workspace_quota_mb: 0
         readiness_cache_sec: 15
     local:
-        profile: 'default'
+        profile: 'default'  # Network enabled by default. Use 'offline' to disable network by default.
         image: ''  # Custom local sandbox image. Leave empty to use the profile default.
         host_root: './data/box'  # Base host directory for local workspace mounts. Docker deployments should override this with an absolute host path.
         default_workspace: ''  # Defaults to '<host_root>/default'. Relative paths are resolved under host_root.
@@ -432,8 +446,38 @@ space:
     # OAuth authorization page URL (user will be redirected here)
     oauth_authorize_url: 'https://space.langbot.app/auth/authorize'
     disable_models_service: false
-    # Master OFF for usage telemetry and Beta diagnostics.
+    # Optional telemetry is enabled by default. Set true to disable usage,
+    # heartbeat and execution reporting; restart after changing this setting.
     disable_telemetry: false
-    # Beta builds only: opt out of content-free quality diagnostics separately.
-    # Stable, alpha and RC builds never enable this diagnostics producer.
-    disable_beta_diagnostics: false
+    # Per-event execution traces: off | failures | sampled | all.
+    # Every eligible execution is reported by default: one record per workflow
+    # node plus one terminal chain record. 'failures' keeps only broken chains;
+    # 'sampled' keeps broken chains plus about 1/N of successful ones, selected
+    # by event hash (N is execution_trace_sample); 'off' reports no executions.
+    # Unmatched/discarded platform events are never reported unless a plugin
+    # subscription matches. Matched delivery failures and valid interaction
+    # resumptions remain eligible, as do independent API/WebUI executions.
+    # Cloud uses 'all' but still applies this route eligibility filter.
+    execution_trace: all
+    execution_trace_sample: 20
+    # How often buffered execution records are uploaded, in seconds.
+    # Complete execution records are batched as {"records": [...]}; records are
+    # removed from the buffer only once the sender accepted them.
+    telemetry_flush_seconds: 180
+
+# Host-owned LLM policy. Models/calls may override through extra_args.request_policy.
+llm_request_policy:
+    retries: 0
+    total_timeout: 300
+    first_event_timeout: 120
+    stream_idle_timeout: 120
+    connect_timeout: 10
+    read_timeout: 120
+    write_timeout: 30
+    pool_timeout: 10
+    backoff_base: 0.5
+    backoff_max: 8
+    attribution: true
+    # Optional: x-session-id, x-litellm-session-id, x-langbot-conversation-id.
+    # OpenRouter uses x-session-id automatically; other providers default to none.
+    conversation_header: null
diff --git a/src/langbot/templates/default-pipeline-config.json b/src/langbot/templates/default-pipeline-config.json
index 23666ab03..b692c3e0e 100644
--- a/src/langbot/templates/default-pipeline-config.json
+++ b/src/langbot/templates/default-pipeline-config.json
@@ -55,7 +55,7 @@
         },
         "misc": {
             "exception-handling": "show-hint",
-            "failure-hint": "Request failed.",
+            "failure-hint": "",
             "at-sender": true,
             "quote-origin": true,
             "track-function-calls": false,
diff --git a/src/langbot/templates/dingtalk_stream_card.json b/src/langbot/templates/dingtalk_stream_card.json
new file mode 100644
index 000000000..e11cb3154
--- /dev/null
+++ b/src/langbot/templates/dingtalk_stream_card.json
@@ -0,0 +1,6 @@
+{
+  "editorData": "{\"schemaVersion\":\"3.0.0\",\"schema\":{\"config\":null,\"componentsMap\":[{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"AIPending\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"AIPending\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"AICardStatusContainer\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"AICardStatusContainer\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"BaseText\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"BaseText\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"AICardContent\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"AICardContent\",\"children\":[]},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"AICardContainer\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"AICardContainer\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"MarkdownBlock\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"MarkdownBlock\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"CollapsePanel\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"CollapsePanel\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"AIGenerationProcessing\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"AIGenerationProcessing\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"Feedback\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"Feedback\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"Grid\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"Grid\"},{\"package\":\"@ali/dxComponent\",\"version\":\"1.0.0\",\"exportName\":\"Image\",\"main\":\"./src/index.tsx\",\"destructuring\":false,\"subName\":\"\",\"componentName\":\"Image\"}],\"componentsTree\":[{\"componentName\":\"AICardContainer\",\"id\":\"node_root\",\"props\":{\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enablePending\":true,\"enableWriting\":true,\"enableDoing\":true,\"enableFailed\":true,\"summaryContent\":{\"type\":\"variableValue\",\"variableType\":\"global\",\"variable\":\"\"},\"enableTitle\":false,\"flowStatusVar\":{\"type\":\"variableValue\",\"variableType\":\"global\",\"variable\":\"flowStatus\"},\"operationPenalType\":\"prompt\",\"enableFlowAbort\":false,\"innerOffset\":0,\"enableGradientBorder\":true,\"cardSizeMode\":\"adaptive\",\"cardSizeHeightMode\":\"adaptive\",\"cardSizeWidthMode\":\"adaptive\",\"cardSizeHeight\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":226,\"variable\":\"\",\"variableType\":\"global\"},\"hasBackground\":false,\"backgroundType\":\"Standard\",\"standardBackgroundColor\":\"gray\",\"backgroundColor\":\"#F6F6F6\",\"darkModeBackgroundColor\":\"#3C3C3C\",\"enableEngineUpgrade\":false,\"enableExposeStatPoint\":false,\"enableDebugTool\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AICardStatusContainer\",\"id\":\"node_status_pending\",\"props\":{\"status\":1,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enableExtend\":false,\"autoFoldConfig\":{\"needFold\":true,\"heightLimit\":480,\"foldStatusLocalDataKey\":\"_cardFoldStatusLocalDataKey\"},\"innerOffset\":0,\"enableCollapse\":false,\"margin\":-2},\"title\":\"处理中\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AIPending\",\"id\":\"node_pending_inner\",\"props\":{\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"pendingTip\":{\"type\":\"dynamicString\",\"content\":\"正在处理…\",\"i18n\":false},\"style\":\"embed\",\"hideIcon\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"AICardStatusContainer\",\"id\":\"node_status_writing\",\"props\":{\"status\":2,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enableExtend\":false,\"autoFoldConfig\":{\"needFold\":false,\"heightLimit\":480,\"foldStatusLocalDataKey\":\"_cardFoldStatusLocalDataKey\"},\"innerOffset\":0,\"enableCollapse\":false,\"margin\":-2},\"title\":\"回复中\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AICardContent\",\"id\":\"node_status_writing_content\",\"props\":{\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"innerOffset\":0,\"disabledWhileForward\":false,\"statPoint\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"statPointParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"margin\":-2,\"transformToEventChain\":false,\"enableStatPoint\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"CollapsePanel\",\"id\":\"langbot_reasoning_2\",\"props\":{\"id\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"reasoning_2\"},\"title\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"思考过程 · 输出中\"},\"contentVisible\":true,\"marginLeft\":12,\"marginRight\":12,\"marginBottom\":8,\"marginTop\":4,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"variable\",\"variable\":\"hasReasoning\",\"variableType\":\"global\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"margin\":-2,\"innerOffset\":0},\"title\":\"思考过程\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"MarkdownBlock\",\"id\":\"node_2_reasoning\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"reasoning\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"reasoning\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"MarkdownBlock\",\"id\":\"node_2_answer\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"answer\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"answer\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]}]},{\"componentName\":\"AICardStatusContainer\",\"id\":\"node_status_doing\",\"props\":{\"status\":4,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enableExtend\":false,\"autoFoldConfig\":{\"needFold\":false,\"heightLimit\":480,\"foldStatusLocalDataKey\":\"_cardFoldStatusLocalDataKey\"},\"innerOffset\":0,\"enableCollapse\":false,\"margin\":-2},\"title\":\"执行中\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AICardContent\",\"id\":\"node_status_doing_content\",\"props\":{\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"innerOffset\":0,\"disabledWhileForward\":false,\"statPoint\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"statPointParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"margin\":-2,\"transformToEventChain\":false,\"enableStatPoint\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"CollapsePanel\",\"id\":\"langbot_reasoning_4\",\"props\":{\"id\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"reasoning_4\"},\"title\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"思考过程 · 输出中\"},\"contentVisible\":true,\"marginLeft\":12,\"marginRight\":12,\"marginBottom\":8,\"marginTop\":4,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"variable\",\"variable\":\"hasReasoning\",\"variableType\":\"global\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"margin\":-2,\"innerOffset\":0},\"title\":\"思考过程\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"MarkdownBlock\",\"id\":\"node_4_reasoning\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"reasoning\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"reasoning\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"MarkdownBlock\",\"id\":\"node_4_answer\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"answer\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"answer\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]}]},{\"componentName\":\"AICardStatusContainer\",\"id\":\"node_status_done\",\"props\":{\"status\":3,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enableExtend\":false,\"autoFoldConfig\":{\"needFold\":false,\"heightLimit\":480,\"foldStatusLocalDataKey\":\"_cardFoldStatusLocalDataKey\"},\"innerOffset\":0,\"enableCollapse\":false,\"margin\":-2},\"title\":\"已完成\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AICardContent\",\"id\":\"node_done_content\",\"props\":{\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"innerOffset\":0,\"disabledWhileForward\":false,\"statPoint\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"statPointParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"margin\":-2,\"transformToEventChain\":false,\"enableStatPoint\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"CollapsePanel\",\"id\":\"langbot_reasoning_3\",\"props\":{\"id\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"reasoning_3\"},\"title\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"思考过程\"},\"contentVisible\":false,\"marginLeft\":12,\"marginRight\":12,\"marginBottom\":8,\"marginTop\":4,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"variable\",\"variable\":\"hasReasoning\",\"variableType\":\"global\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"margin\":-2,\"innerOffset\":0},\"title\":\"思考过程\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"MarkdownBlock\",\"id\":\"node_3_reasoning\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"reasoning\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"reasoning\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"MarkdownBlock\",\"id\":\"node_3_answer\",\"props\":{\"mdVer\":0,\"icon\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"emoji\"},\"content\":{\"variable\":\"answer\",\"variableType\":\"global\",\"type\":\"variableValue\",\"varType\":\"markdown\"},\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"isStreaming\":false,\"enableLinkStatPoint\":false,\"linkStatPoint\":{\"type\":\"dynamicString\",\"content\":\"Page_InteractiveCard__Click_markdownOpenlink\",\"i18n\":false},\"linkStatPointParams\":[],\"marginTop\":8,\"marginBottom\":8,\"marginLeft\":12,\"marginRight\":12},\"title\":\"answer\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"Grid\",\"id\":\"langbot_operation_row\",\"props\":{\"direction\":\"horizontal\",\"isAutoHeight\":true,\"isAutoWidth\":false,\"autoWidth\":false,\"isFixedWidth\":false,\"fixWidth\":false,\"width\":100,\"childGravity\":\"leftTop\",\"hasBackground\":false,\"hasGradientBackground\":false,\"hasBorder\":false,\"marginLeft\":12,\"marginRight\":12,\"marginTop\":0,\"marginBottom\":0,\"actionType\":\"none\"},\"children\":[{\"componentName\":\"Grid\",\"id\":\"langbot_link\",\"props\":{\"direction\":\"horizontal\",\"isAutoHeight\":false,\"isAutoWidth\":true,\"autoWidth\":true,\"isFixedWidth\":false,\"fixWidth\":false,\"childGravity\":\"leftCenter\",\"hasBackground\":false,\"hasGradientBackground\":false,\"hasBorder\":false,\"marginLeft\":0,\"marginRight\":8,\"marginTop\":0,\"marginBottom\":0,\"actionType\":\"url\",\"urlType\":\"all\",\"url\":{\"type\":\"dynamicLink\",\"valueType\":\"fixed\",\"value\":\"https://langbot.app?utm_source=dingtalk&utm_medium=bot_card&utm_campaign=langbot\",\"variable\":\"\",\"variableType\":\"global\"},\"height\":28},\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"Image\",\"id\":\"langbot_logo\",\"props\":{\"images\":{\"defaultLang\":\"zh_Hans\",\"content\":{\"zh_Hans\":{\"value\":\"https://langbot.app/logo-blue.png\",\"valueType\":\"fixed\",\"type\":\"dynamicImage\"}},\"i18n\":false},\"darkModeImages\":{\"defaultLang\":\"zh_Hans\",\"content\":{\"zh_Hans\":{\"value\":\"\",\"valueType\":\"fixed\",\"type\":\"dynamicImage\"}},\"i18n\":false},\"single\":true,\"height\":{\"type\":\"dynamicWidth\",\"valueType\":\"fixed\",\"value\":12,\"variableType\":\"global\",\"variable\":\"\",\"full\":false,\"adaptive\":false},\"width\":{\"type\":\"dynamicWidth\",\"valueType\":\"fixed\",\"value\":12,\"variableType\":\"global\",\"variable\":\"\",\"full\":false,\"adaptive\":false},\"cornerRadius\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":0,\"variable\":\"\",\"variableType\":\"global\"},\"scaleType\":\"fitCenter\",\"enablePreview\":false,\"marginRight\":4,\"marginTop\":0,\"marginBottom\":0,\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"adaptiveSize\":false,\"fixedRatio\":false,\"hoverText\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"enableBorder\":false,\"borderColor\":\"#297f8790\",\"statPoint\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"statPointParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"aspectRatio\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":2,\"variable\":\"\",\"variableType\":\"global\"},\"offsetLeft\":0,\"offsetRight\":0,\"marginLeft\":0,\"innerOffset\":0,\"enableStatPoint\":false,\"margin\":-2},\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"},{\"componentName\":\"BaseText\",\"id\":\"langbot_name\",\"props\":{\"text\":{\"type\":\"dynamicString\",\"content\":\"LangBot\",\"i18n\":false},\"hoverText\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"iconType\":\"iconCode\",\"iconFont\":{\"type\":\"icon\",\"iconType\":\"ddIcon\",\"icon\":\"icon_question_fill\"},\"icon\":{\"type\":\"dynamicLink\",\"value\":\"\",\"valueType\":\"fixed\",\"variable\":\"\",\"variableType\":\"global\"},\"darkIcon\":{\"type\":\"dynamicLink\",\"value\":\"\",\"valueType\":\"fixed\",\"variable\":\"\",\"variableType\":\"global\"},\"autoWidth\":true,\"maxWidth\":0,\"fixedWidth\":0,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"fontColorType\":\"Standard\",\"enableHighlight\":false,\"maxLine\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":1},\"color\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"common_level3_base_color\"},\"customLightColor\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"#35404b\",\"variable\":\"\",\"variableType\":\"global\"},\"customDarkColor\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"#f6f6f6\",\"variable\":\"\",\"variableType\":\"global\"},\"gravity\":\"left\",\"fontSizeType\":\"Standard\",\"styleType\":\"custom\",\"styleToken\":\"common_body_text_style\",\"size\":\"small\",\"customFontSize\":12,\"customFontLineHeight\":16,\"bold\":false,\"italic\":false,\"strikeThrough\":false,\"lineHeight\":\"normal\",\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"autoMaxWidth\":false,\"innerOffset\":0,\"enableIcon\":false,\"widthMode\":\"wrap_content\"},\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"Grid\",\"id\":\"langbot_native_operations\",\"props\":{\"direction\":\"horizontal\",\"isAutoHeight\":true,\"isAutoWidth\":false,\"autoWidth\":false,\"isFixedWidth\":false,\"fixWidth\":false,\"width\":100,\"childGravity\":\"leftTop\",\"hasBackground\":false,\"hasGradientBackground\":false,\"hasBorder\":false,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"actionType\":\"none\"},\"children\":[{\"componentName\":\"Feedback\",\"id\":\"node_ocmv2bczc8a\",\"props\":{\"params\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"marginConfigurable\":false,\"enableLikeDislike\":true,\"enableCopy\":true,\"copyType\":\"common\",\"copyVariable\":{\"type\":\"variableValue\",\"variable\":\"answer\",\"variableType\":\"global\",\"varType\":\"markdown\"},\"enableDivider\":false,\"margin\":-2,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"innerOffset\":0,\"enableRefresh\":false,\"actionType\":\"none\",\"actionId\":{\"type\":\"dynamicString\",\"content\":\"sys_action_like_component\",\"i18n\":false},\"refreshParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"refreshSuccessCondition\":{\"op\":\"and\",\"conditions\":[]},\"refreshSuccessToast\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"refreshFailureToast\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"refreshVisible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}}},\"title\":\"操作区\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[]}],\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}],\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]},{\"componentName\":\"AICardStatusContainer\",\"id\":\"node_status_failed\",\"props\":{\"status\":5,\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"enableExtend\":false,\"autoFoldConfig\":{\"needFold\":true,\"heightLimit\":480,\"foldStatusLocalDataKey\":\"_cardFoldStatusLocalDataKey\"},\"innerOffset\":0,\"enableCollapse\":false,\"margin\":-2},\"title\":\"失败\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"AICardContent\",\"id\":\"node_failed_content\",\"props\":{\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"marginLeft\":0,\"marginRight\":0,\"marginTop\":0,\"marginBottom\":0,\"innerOffset\":0,\"disabledWhileForward\":false,\"statPoint\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"statPointParams\":[{\"type\":\"fixed\",\"variable\":\"\",\"value\":\"\",\"name\":\"\",\"variableType\":\"global\",\"id\":\"1\"}],\"margin\":-2,\"transformToEventChain\":false,\"enableStatPoint\":false},\"hidden\":false,\"title\":\"\",\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\",\"children\":[{\"componentName\":\"BaseText\",\"id\":\"node_failed_text\",\"props\":{\"text\":{\"i18n\":false,\"type\":\"dynamicString\",\"content\":\"操作失败，请稍后重试。\"},\"hoverText\":{\"type\":\"dynamicString\",\"content\":\"\",\"i18n\":false},\"iconType\":\"iconCode\",\"iconFont\":{\"type\":\"icon\",\"icon\":\"\",\"iconType\":\"ddIcon\"},\"icon\":{\"type\":\"dynamicLink\",\"value\":\"\",\"valueType\":\"fixed\",\"variable\":\"\",\"variableType\":\"global\"},\"darkIcon\":{\"type\":\"dynamicLink\",\"value\":\"\",\"valueType\":\"fixed\",\"variable\":\"\",\"variableType\":\"global\"},\"autoWidth\":false,\"maxWidth\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":0,\"variable\":\"\",\"variableType\":\"global\"},\"fixedWidth\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":0,\"variable\":\"\",\"variableType\":\"global\"},\"marginLeft\":10,\"marginRight\":10,\"marginTop\":10,\"marginBottom\":10,\"fontColorType\":\"Standard\",\"enableHighlight\":false,\"maxLine\":{\"type\":\"dynamicNumber\",\"valueType\":\"fixed\",\"value\":2,\"variable\":\"\",\"variableType\":\"global\"},\"color\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"common_level1_base_color\",\"variable\":\"\",\"variableType\":\"global\"},\"customLightColor\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"#35404b\",\"variable\":\"\",\"variableType\":\"global\"},\"customDarkColor\":{\"type\":\"dynamicColor\",\"valueType\":\"fixed\",\"value\":\"#f6f6f6\",\"variable\":\"\",\"variableType\":\"global\"},\"gravity\":\"center\",\"fontSizeType\":\"Standard\",\"styleType\":\"custom\",\"styleToken\":\"common_body_text_style\",\"size\":\"middle\",\"customFontSize\":15,\"customFontLineHeight\":22,\"bold\":false,\"italic\":false,\"strikeThrough\":false,\"lineHeight\":\"normal\",\"visible\":{\"type\":\"dynamicVisible\",\"value\":true,\"valueType\":\"fixed\",\"condition\":{\"op\":\"and\",\"conditions\":[]}},\"autoMaxWidth\":false,\"innerOffset\":0,\"enableIcon\":false,\"widthMode\":\"match_parent\",\"margin\":-2},\"title\":\"基础文本\",\"hidden\":false,\"isLocked\":false,\"condition\":true,\"conditionGroup\":\"\"}]}]}]}],\"i18n\":{},\"version\":\"1.0.0\"},\"mockData\":{\"cardData\":{\"flowStatus\":\"3\",\"content\":\"\",\"reasoning\":\"> 这里展示模型返回的思考内容。\",\"answer\":\"这里展示正式回答。\",\"query\":\"\",\"config\":{\"autoLayout\":true},\"brand\":\"\",\"hasReasoning\":\"true\",\"autoLayout\":\"true\"},\"cardPrivateData\":{},\"localData\":{\"flowStatus\":\"\",\"_cardFoldStatusLocalDataKey\":\"\"},\"richTextData\":{}},\"renderContext\":{\"regenerateEnabled\":\"1\",\"regenerateIndex\":\"2\",\"regenerateTotal\":\"5\"},\"editVersion\":0,\"customWidgetInfo\":\"\",\"useCustomWidgetInfo\":false,\"variableList\":[{\"id\":\"content\",\"name\":\"content\",\"type\":\"markdown\",\"description\":\"content\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"reasoning\",\"name\":\"reasoning\",\"type\":\"markdown\",\"description\":\"reasoning\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"answer\",\"name\":\"answer\",\"type\":\"markdown\",\"description\":\"answer\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"flowStatus\",\"name\":\"flowStatus\",\"type\":\"string\",\"description\":\"flowStatus\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"query\",\"name\":\"query\",\"type\":\"string\",\"description\":\"query\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"config\",\"name\":\"config\",\"type\":\"object\",\"description\":\"config\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false,\"schema\":[{\"id\":\"config.autoLayout\",\"name\":\"autoLayout\",\"type\":\"boolean\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false,\"description\":\"Wide card layout\"}]},{\"id\":\"brand\",\"name\":\"brand\",\"type\":\"markdown\",\"description\":\"brand\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"hasReasoning\",\"name\":\"hasReasoning\",\"type\":\"string\",\"description\":\"hasReasoning\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false},{\"id\":\"autoLayout\",\"name\":\"autoLayout\",\"type\":\"string\",\"description\":\"Card auto layout\",\"private\":false,\"editorVarType\":\"variables\",\"disabled\":false}],\"formList\":[],\"customContextList\":[],\"expList\":[],\"localList\":[],\"hsfList\":[],\"lwpList\":[],\"pageData\":{},\"extension\":{\"extendType\":\"AI\",\"aiStatusList\":[3,1,5,2,4],\"fileTypeList\":[]}}",
+  "widgetInfo": "",
+  "type": "im",
+  "mode": "card"
+}
diff --git a/src/langbot/templates/embed/widget.js b/src/langbot/templates/embed/widget.js
index 72e3f5355..6cd9643a4 100644
--- a/src/langbot/templates/embed/widget.js
+++ b/src/langbot/templates/embed/widget.js
@@ -174,6 +174,17 @@
     bubbleIcon: "__LANGBOT_BUBBLE_ICON__",
   };
 
+  // Wizard previews must use the browser-visible backend URL. Behind a TLS
+  // proxy the server's request URL can be HTTP even when widget.js uses HTTPS.
+  // External embeds retain their configured public webhook prefix.
+  if (scriptEl && scriptEl.src) {
+    var scriptUrl = new URL(scriptEl.src, document.baseURI);
+    if (scriptUrl.searchParams.get("preview") === "wizard") {
+      CONFIG.baseUrl = scriptUrl.origin;
+      CONFIG.logoUrl = CONFIG.baseUrl + "/api/v1/embed/logo";
+    }
+  }
+
   // ========== Styles ==========
   var STYLES =
     '\
@@ -524,18 +535,15 @@
       return;
     }
 
-    var protocol = CONFIG.baseUrl.indexOf("https") === 0 ? "wss:" : "ws:";
-    var host = CONFIG.baseUrl.replace(/^https?:\/\//, "");
-    var url =
-      protocol +
-      "//" +
-      host +
-      "/api/v1/embed/" +
-      CONFIG.botUuid +
-      "/ws/connect?session_type=" +
-      CONFIG.sessionType +
-      "&session_id=" +
-      encodeURIComponent(state.sessionId);
+    var endpoint = new URL(CONFIG.baseUrl || "/", window.location.href);
+    endpoint.protocol = endpoint.protocol === "https:" || endpoint.protocol === "wss:" ? "wss:" : "ws:";
+    endpoint.pathname = endpoint.pathname.replace(/\/$/, "") +
+      "/api/v1/embed/" + encodeURIComponent(CONFIG.botUuid) + "/ws/connect";
+    endpoint.search = "";
+    endpoint.hash = "";
+    endpoint.searchParams.set("session_type", CONFIG.sessionType);
+    endpoint.searchParams.set("session_id", state.sessionId);
+    var url = endpoint.toString();
 
     try {
       state.ws = new WebSocket(url);
diff --git a/src/langbot/templates/metadata/pipeline/output.yaml b/src/langbot/templates/metadata/pipeline/output.yaml
index d5e0fae07..d3acfc693 100644
--- a/src/langbot/templates/metadata/pipeline/output.yaml
+++ b/src/langbot/templates/metadata/pipeline/output.yaml
@@ -106,11 +106,81 @@ stages:
           en_US: Failure Hint Text
           zh_Hans: 失败提示文本
         description:
-          en_US: The text to display when a request fails. Only effective when Exception Handling Strategy is set to "Show Failure Hint"
-          zh_Hans: 请求失败时显示的提示文本，仅在异常处理策略设置为"仅文字提示"时生效
+          en_US: Leave empty to use the Workspace language. Used for request and runner execution failures; custom text is not translated.
+          zh_Hans: 留空时使用工作区语言的默认提示。用于请求和运行器执行失败；自定义文本不会自动翻译。
         type: string
         required: false
-        default: 'Request failed.'
+        default: ''
+      - name: rate-limit-hint
+        label:
+          en_US: Rate Limit Hint
+          zh_Hans: 限速提示
+        description:
+          en_US: "Leave empty to use the Workspace language."
+          zh_Hans: "留空时使用工作区语言的默认提示。"
+        type: string
+        required: false
+        default: ''
+      - name: content-filter-hint
+        label:
+          en_US: Content Filter Hint
+          zh_Hans: 内容过滤提示
+        description:
+          en_US: "Leave empty to use the Workspace language."
+          zh_Hans: "留空时使用工作区语言的默认提示。"
+        type: string
+        required: false
+        default: ''
+      - name: content-filter-error-hint
+        label:
+          en_US: Content Check Failure Hint
+          zh_Hans: 内容检查失败提示
+        description:
+          en_US: "Leave empty to use the Workspace language."
+          zh_Hans: "留空时使用工作区语言的默认提示。"
+        type: string
+        required: false
+        default: ''
+      - name: tool-call-hint
+        label:
+          en_US: Tool Call Hint
+          zh_Hans: 工具调用提示
+        description:
+          en_US: "Leave empty to use the Workspace language. Use {tools} for tool names. Requires Track Function Calls."
+          zh_Hans: "留空时使用工作区语言的默认提示。 可用 {tools} 插入工具名称，需开启跟踪函数调用。"
+        type: string
+        required: false
+        default: ''
+      - name: runner-not-found-hint
+        label:
+          en_US: Runner Missing Hint
+          zh_Hans: 运行器不存在提示
+        description:
+          en_US: "Leave empty to use the Workspace language. Use {runner} for the runner ID."
+          zh_Hans: "留空时使用工作区语言的默认提示。 可用 {runner} 插入运行器 ID。"
+        type: string
+        required: false
+        default: ''
+      - name: runner-not-authorized-hint
+        label:
+          en_US: Runner Authorization Hint
+          zh_Hans: 运行器未授权提示
+        description:
+          en_US: "Leave empty to use the Workspace language."
+          zh_Hans: "留空时使用工作区语言的默认提示。"
+        type: string
+        required: false
+        default: ''
+      - name: runner-unavailable-hint
+        label:
+          en_US: Runner Unavailable Hint
+          zh_Hans: 运行器不可用提示
+        description:
+          en_US: "Leave empty to use the Workspace language."
+          zh_Hans: "留空时使用工作区语言的默认提示。"
+        type: string
+        required: false
+        default: ''
       - name: at-sender
         label:
           en_US: At Sender
diff --git a/tests/e2e/live_aiocqhttp_eba_probe.py b/tests/e2e/live_aiocqhttp_eba_probe.py
index 493d208e3..dc193c9bd 100644
--- a/tests/e2e/live_aiocqhttp_eba_probe.py
+++ b/tests/e2e/live_aiocqhttp_eba_probe.py
@@ -7,6 +7,7 @@ import json
 import time
 from collections import Counter
 from pathlib import Path
+from types import SimpleNamespace
 
 from langbot.pkg.platform.adapters.aiocqhttp.adapter import AiocqhttpAdapter
 from langbot_plugin.api.definition.abstract.platform.event_logger import AbstractEventLogger
@@ -36,9 +37,8 @@ def dump_event(event: platform_events.Event) -> dict:
 
 async def main():
     parser = argparse.ArgumentParser(description='Live OneBot v11 / aiocqhttp EBA probe for Matcha or a real endpoint.')
-    parser.add_argument('--host', default='127.0.0.1')
-    parser.add_argument('--port', type=int, default=2280)
-    parser.add_argument('--access-token', default='')
+    parser.add_argument('--ws-url', required=True, help='Protocol endpoint Universal WebSocket URL')
+    parser.add_argument('--access-token', required=True)
     parser.add_argument('--timeout', type=int, default=120)
     parser.add_argument('--target-type', choices=['private', 'group'], default=None)
     parser.add_argument('--target-id', default=None)
@@ -54,10 +54,18 @@ async def main():
     out_fp = out_path.open('a', encoding='utf-8')
 
     adapter = AiocqhttpAdapter(
-        {'host': args.host, 'port': args.port, 'access-token': args.access_token},
+        {'connection_mode': 'forward_ws', 'ws_url': args.ws_url, 'access-token': args.access_token},
         ProbeLogger(),
     )
 
+    # This standalone converter probe uses a single explicit local runtime.
+    runtime = SimpleNamespace(enable=True, workspace_uuid='probe', bot_entity=SimpleNamespace(uuid='probe'))
+
+    async def resolve_probe(_):
+        return runtime
+
+    runtime.ap = SimpleNamespace(platform_mgr=SimpleNamespace(resolve_public_bot=resolve_probe))
+    adapter.bind_runtime(runtime)
     observed: list[platform_events.Event] = []
     first_message = asyncio.Event()
 
@@ -79,7 +87,7 @@ async def main():
             return None
 
     task = asyncio.create_task(adapter.run_async())
-    print(f'Listening on ws://{args.host}:{args.port}/ws/ . Trigger events from Matcha now.')
+    print('Connecting to the protocol endpoint. Trigger events now.')
 
     api_results: dict[str, str] = {}
     try:
diff --git a/tests/e2e/test_llm_request_chain.py b/tests/e2e/test_llm_request_chain.py
new file mode 100644
index 000000000..fdfabd9be
--- /dev/null
+++ b/tests/e2e/test_llm_request_chain.py
@@ -0,0 +1,442 @@
+"""Real HTTP -> Core -> Plugin Runtime -> Runner -> Core -> HTTP provider tests.
+
+Only the remote provider is deterministic. No requester, SDK action transport,
+model manager, persistence or monitoring implementation is mocked.
+"""
+
+import json
+import importlib.metadata
+import os
+from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+from pathlib import Path
+import shutil
+import select
+import socket
+import sqlite3
+import threading
+import time
+
+import pytest
+import yaml
+import httpx
+from packaging.version import Version
+
+from tests.e2e import test_runner_plugin_runtime as runtime
+
+# Reuse the real process lifecycle; override the package/config and port below.
+runner_runtime_ports = runtime.runner_runtime_ports
+runner_runtime_process = runtime.runner_runtime_process
+runner_langbot_process = runtime.runner_langbot_process
+runner_client = runtime.runner_client
+
+pytestmark = pytest.mark.e2e
+RUNNER = runtime.QA_RUNNER_ID
+USAGE = {
+    'prompt_tokens': 24,
+    'completion_tokens': 4,
+    'total_tokens': 28,
+    'prompt_tokens_details': {'cached_tokens': 16},
+    'completion_tokens_details': {'reasoning_tokens': 2},
+}
+
+
+@pytest.fixture(scope='session')
+def runner_e2e_tmpdir(tmp_path_factory):
+    # Retain logs/SQLite in pytest's bounded temp history for failure diagnosis.
+    return tmp_path_factory.mktemp('llm-chain')
+
+
+@pytest.fixture(scope='session')
+def runner_e2e_port():
+    return runtime._free_port()
+
+
+@pytest.fixture(scope='session')
+def runner_e2e_config_path(runner_e2e_tmpdir, runner_e2e_port, runner_runtime_ports):
+    path = runtime.runner_e2e_config_path.__wrapped__(runner_e2e_tmpdir, runner_e2e_port, runner_runtime_ports)
+    root = runner_e2e_tmpdir / 'agent-runner-qa-package'
+    manifest = root / 'components/runner/default.yaml'
+    data = yaml.safe_load(manifest.read_text(encoding='utf-8'))
+    data['spec']['permissions'] = {'models': ['invoke', 'stream', 'count_tokens']}
+    data['spec']['config'] = [
+        {'name': 'model', 'label': {'en_US': 'Model'}, 'type': 'llm-model-selector', 'required': True},
+    ]
+    manifest.write_text(yaml.safe_dump(data), encoding='utf-8')
+    shutil.copyfile(Path(__file__).parent / 'utils/llm_probe_runner.py', root / 'components/runner/default.py')
+    if legacy_root := os.environ.get('LLM_E2E_LEGACY_API_DIR'):
+        for filename in ('legacy_api.py', 'legacy_message.py'):
+            shutil.copyfile(Path(legacy_root) / filename, root / filename)
+    shutil.make_archive(str(runner_e2e_tmpdir / 'agent-runner-qa'), 'zip', root_dir=root)
+    return path
+
+
+@pytest.fixture(scope='module')
+def provider_wire():
+    records = []
+
+    class Handler(BaseHTTPRequestHandler):
+        protocol_version = 'HTTP/1.1'
+
+        def log_message(self, *args):
+            pass
+
+        def do_POST(self):
+            body = json.loads(self.rfile.read(int(self.headers['Content-Length'])))
+            scenario = self.headers.get('x-e2e-case', 'normal')
+            row = {'body': body, 'headers': dict(self.headers), 'case': scenario, 'closed': False}
+            records.append(row)
+            attempt = sum(x['case'] == scenario for x in records)
+            status = 503 if scenario.startswith('retry') and attempt == 1 else 200
+            if scenario == 'auth':
+                status = 401
+            if status != 200:
+                data = json.dumps({'error': {'message': 'injected failure', 'type': 'api_error'}}).encode()
+                self.send_response(status)
+                self.send_header('Content-Type', 'application/json')
+                self.send_header('Content-Length', str(len(data)))
+                self.send_header('Retry-After', '0')
+                self.end_headers()
+                self.wfile.write(data)
+                return
+            if scenario in ('first-timeout', 'total-timeout'):
+                time.sleep(1)
+            try:
+                if not body.get('stream'):
+                    data = json.dumps(
+                        {
+                            'id': 'e2e',
+                            'object': 'chat.completion',
+                            'created': 1,
+                            'model': body['model'],
+                            'choices': [
+                                {
+                                    'index': 0,
+                                    'message': {'role': 'assistant', 'content': 'hello world'},
+                                    'finish_reason': 'stop',
+                                }
+                            ],
+                            'usage': USAGE,
+                        }
+                    ).encode()
+                    self.send_response(200)
+                    self.send_header('Content-Type', 'application/json')
+                    self.send_header('Content-Length', str(len(data)))
+                    self.end_headers()
+                    self.wfile.write(data)
+                    return
+                self.send_response(200)
+                self.send_header('Content-Type', 'text/event-stream')
+                self.send_header('Transfer-Encoding', 'chunked')
+                self.end_headers()
+
+                def send(event):
+                    data = ('data: ' + (event if isinstance(event, str) else json.dumps(event)) + '\n\n').encode()
+                    self.wfile.write(f'{len(data):X}\r\n'.encode() + data + b'\r\n')
+                    self.wfile.flush()
+
+                base = {'id': 'e2e', 'object': 'chat.completion.chunk', 'created': 1, 'model': body['model']}
+                if scenario.startswith('reasoning-'):
+                    send(
+                        {
+                            **base,
+                            'choices': [
+                                {'index': 0, 'delta': {'reasoning_content': 'private thought'}, 'finish_reason': None}
+                            ],
+                        }
+                    )
+                for index, content in enumerate(['hello ', 'world']):
+                    send({**base, 'choices': [{'index': 0, 'delta': {'content': content}, 'finish_reason': None}]})
+                    if index == 0 and (scenario in ('idle-timeout', 'disconnect') or scenario.startswith('cancel-')):
+                        if scenario == 'disconnect':
+                            self.close_connection = True
+                            return
+                        deadline = time.monotonic() + 1
+                        while time.monotonic() < deadline:
+                            if select.select([self.connection], [], [], 0.02)[0]:
+                                if not self.connection.recv(1, socket.MSG_PEEK):
+                                    row['closed'] = True
+                                    return
+                send({**base, 'choices': [{'index': 0, 'delta': {}, 'finish_reason': 'stop'}]})
+                send({**base, 'choices': [], 'usage': USAGE})
+                send('[DONE]')
+                self.wfile.write(b'0\r\n\r\n')
+                self.wfile.flush()
+            except (BrokenPipeError, ConnectionResetError, ConnectionAbortedError):
+                row['closed'] = True
+
+    server = ThreadingHTTPServer(('127.0.0.1', 0), Handler)
+    thread = threading.Thread(target=server.serve_forever, daemon=True)
+    thread.start()
+    yield f'http://127.0.0.1:{server.server_port}/v1', records
+    server.shutdown()
+    server.server_close()
+    thread.join(timeout=5)
+
+
+@pytest.fixture(scope='module')
+def chain(request, provider_wire):
+    request.getfixturevalue('runner_langbot_process')
+    port = request.getfixturevalue('runner_e2e_port')
+    client = httpx.Client(base_url=f'http://127.0.0.1:{port}', timeout=90, trust_env=False)
+    request.addfinalizer(client.close)
+    tmpdir = request.getfixturevalue('runner_e2e_tmpdir')
+    token = runtime._init_and_auth(client)
+    runtime._ensure_qa_plugin(client, token, tmpdir / 'agent-runner-qa.zip')
+    client.headers['Authorization'] = f'Bearer {token}'
+
+    def post(path, body):
+        response = client.post('/api/v1/' + path, json=body)
+        assert response.status_code == 200, response.text
+        payload = response.json()
+        assert payload['code'] == 0, payload
+        return payload['data']
+
+    provider = post(
+        'provider/providers',
+        {
+            'name': 'LLM chain E2E',
+            'requester': 'openai-chat-completions',
+            'base_url': provider_wire[0],
+            'api_keys': ['e2e-not-a-secret'],
+        },
+    )['uuid']
+    model = post(
+        'provider/models/llm',
+        {
+            'name': 'gpt-4o',
+            'provider_uuid': provider,
+            'abilities': ['func_call', 'vision'],
+            'extra_args': {
+                'temperature': 0.1,
+                'request_policy': {'retries': 1, 'backoff_base': 0.01, 'backoff_max': 0.01},
+            },
+        },
+    )['uuid']
+    agent = post(
+        'agents',
+        {
+            'kind': 'agent',
+            'name': 'LLM chain E2E',
+            'config': {
+                'runner': {'id': RUNNER},
+                'runner_config': {RUNNER: {'model': model}},
+            },
+        },
+    )['uuid']
+
+    def invoke(mode, case='normal', extra=None, *, configuration=None, options=None):
+        selected_model, selected_agent = model, agent
+        if configuration:
+            configured_provider = post(
+                'provider/providers',
+                {
+                    'name': case,
+                    'requester': configuration.get('requester', 'openai-chat-completions'),
+                    'base_url': provider_wire[0],
+                    'api_keys': ['e2e-not-a-secret'],
+                },
+            )['uuid']
+            selected_model = post(
+                'provider/models/llm',
+                {
+                    'name': configuration.get('name', 'gpt-4o'),
+                    'provider_uuid': configured_provider,
+                    'abilities': ['func_call', 'vision'],
+                    'extra_args': configuration.get('extra_args', {}),
+                },
+            )['uuid']
+            selected_agent = post(
+                'agents',
+                {
+                    'kind': 'agent',
+                    'name': case,
+                    'config': {
+                        'runner': {'id': RUNNER},
+                        'runner_config': {RUNNER: {'model': selected_model}},
+                    },
+                },
+            )['uuid']
+        result = post(
+            f'agents/{selected_agent}/debug',
+            {
+                'text': json.dumps(
+                    {'mode': mode, 'model': selected_model, 'case': case, 'extra': extra or {}, **(options or {})}
+                ),
+                'conversation_id': 'e2e-' + case,
+                'event_type': 'message.received',
+            },
+        )
+        assert result.get('final_text'), result
+        return json.loads(result['final_text'])
+
+    return invoke, provider_wire[1], tmpdir
+
+
+def test_multimodal_and_tool_definitions_survive_plugin_transport(chain):
+    invoke, records, _ = chain
+    image = 'data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVR42mP8/x8AAwMCAO+jRZkAAAAASUVORK5CYII='
+    messages = [
+        {
+            'role': 'user',
+            'content': [
+                {'type': 'text', 'text': 'Describe this image'},
+                {'type': 'image_url', 'image_url': {'url': image}},
+            ],
+        }
+    ]
+    result = invoke('runner-events', 'multimodal', options={'messages': messages, 'tool': True})
+    assert result['text'] == 'hello world', result
+    wire = records[-1]['body']
+    assert wire['messages'][0]['content'][1]['image_url']['url'] == image
+    assert wire['tools'][0]['function']['name'] == 'e2e_lookup'
+
+
+def test_openrouter_budget_cache_and_explicit_attribution(chain):
+    invoke, records, _ = chain
+    result = invoke(
+        'runner-events',
+        'openrouter-budget-cache',
+        configuration={
+            'requester': 'openrouter-chat-completions',
+            'name': 'anthropic/claude-sonnet-4',
+            'extra_args': {
+                'reasoning_config': {'budget_tokens': 2048},
+                'prompt_caching': {'mode': 'anthropic'},
+                'max_tokens': 4096,
+            },
+        },
+        extra={
+            'extra_headers': {'HTTP-Referer': 'https://langbot.app', 'X-OpenRouter-Title': 'LangBot'},
+            'request_policy': {'conversation_header': 'x-langbot-conversation-id'},
+        },
+        options={'messages': [{'role': 'system', 'content': 'Stable prefix'}, {'role': 'user', 'content': 'Hello'}]},
+    )
+    assert result['text'] == 'hello world', result
+    wire = records[-1]
+    assert wire['body']['reasoning']['max_tokens'] == 2048
+    assert wire['body']['messages'][0]['content'][0]['cache_control']['type'] == 'ephemeral'
+    headers = {k.lower(): v for k, v in wire['headers'].items()}
+    assert headers['http-referer'] == 'https://langbot.app'
+    assert headers['x-openrouter-title'] == 'LangBot'
+    assert headers['x-langbot-conversation-id'].startswith('lb-')
+    assert 'prompt_caching' not in wire['body'] and 'reasoning_config' not in wire['body']
+
+
+def test_openai_cache_options_and_call_override(chain):
+    invoke, records, _ = chain
+    result = invoke(
+        'runner-usage',
+        'openai-cache',
+        configuration={
+            'extra_args': {'temperature': 0.1, 'prompt_caching': {'mode': 'openai', 'retention': 'in_memory'}},
+        },
+        extra={'temperature': 0.7, 'prompt_cache_key': 'explicit-test-cache'},
+    )
+    assert result['text'] == 'hello world', result
+    assert records[-1]['body']['temperature'] == 0.7
+    assert records[-1]['body']['prompt_cache_key'] == 'explicit-test-cache'
+    assert records[-1]['body']['prompt_cache_retention'] == 'in_memory'
+
+
+@pytest.mark.parametrize(
+    'mode',
+    [
+        'runner-message',
+        'runner-usage',
+        'runner-stream',
+        'runner-events',
+        'plugin-message',
+        'plugin-usage',
+        'plugin-stream',
+        'plugin-events',
+    ],
+)
+def test_public_api_shapes_and_usage(chain, mode):
+    invoke, records, _ = chain
+    result = invoke(mode, mode)
+    assert result['text'] == 'hello world', result
+    assert not result.get('error'), result
+    if mode.endswith(('usage', 'events')):
+        assert result['usage']['prompt_tokens_details']['cached_tokens'] == 16
+        assert result['usage']['completion_tokens_details']['reasoning_tokens'] == 2
+    wire = records[-1]['body']
+    assert wire['temperature'] == 0.1
+    assert 'request_policy' not in wire
+
+
+@pytest.mark.parametrize(
+    'case,extra,attempts,error',
+    [
+        ('retry-once', {}, 2, False),
+        ('retry-disabled', {'num_retries': 0}, 1, True),
+        ('retry-max-disabled', {'max_retries': 0}, 1, True),
+        ('auth', {}, 1, True),
+        ('first-timeout', {'request_policy': {'first_event_timeout': 0.2, 'retries': 0}}, 1, True),
+        ('total-timeout', {'request_policy': {'total_timeout': 0.2, 'retries': 0}}, 1, True),
+        ('idle-timeout', {'request_policy': {'stream_idle_timeout': 0.2}}, 1, True),
+        ('disconnect', {}, 1, True),
+    ],
+)
+def test_failures_retry_boundaries_and_recovery(chain, case, extra, attempts, error):
+    invoke, records, _ = chain
+    result = invoke('runner-events', case, extra)
+    assert bool(result.get('error')) == error, result
+    assert sum(row['case'] == case for row in records) == attempts
+    assert invoke('runner-events', 'recover-' + case)['text'] == 'hello world'
+
+
+@pytest.mark.parametrize('surface', ['runner-events', 'runner-stream', 'plugin-events', 'plugin-stream'])
+def test_cancellation_releases_upstream_and_next_request_works(chain, surface):
+    invoke, records, tmpdir = chain
+    case = 'cancel-' + surface
+    result = invoke(case, case)
+    assert result['text'] == 'hello ', result
+    time.sleep(1.5)
+    cancelled = [row for row in records if row['case'] == case]
+    with sqlite3.connect(tmpdir / 'data/langbot.db') as db:
+        diagnostics = db.execute('SELECT session_id, details FROM monitoring_llm_calls').fetchall()
+    assert len(cancelled) == 1
+    assert invoke('runner-events', 'after-' + case)['text'] == 'hello world'
+    if (
+        not cancelled[0]['closed']
+        and Version(importlib.metadata.version('langbot-plugin')) <= Version('0.7.11')
+        and not os.environ.get('LLM_E2E_REQUIRE_STREAM_CLOSE')
+    ):
+        pytest.xfail('Published SDK <=0.7.11 omits aclosing in public proxy wrappers; requires companion SDK fix')
+    assert cancelled[0]['closed'], diagnostics[-2:]
+
+
+@pytest.mark.parametrize('remove_think', [True, False])
+def test_reasoning_visibility_through_runner(chain, remove_think):
+    result = chain[0]('runner-events', f'reasoning-{remove_think}', options={'remove_think': remove_think})
+    assert result['text'].endswith('hello world'), result
+    assert ('private thought' in result['text']) == (not remove_think)
+
+
+def test_invalid_budget_fails_before_network_and_recovers(chain):
+    invoke, records, _ = chain
+    result = invoke('runner-events', 'invalid-budget', {'reasoning_config': {'budget_tokens': 1}})
+    assert result.get('error'), result
+    assert not any(row['case'] == 'invalid-budget' for row in records)
+    assert invoke('runner-events', 'after-invalid-budget')['text'] == 'hello world'
+
+
+def test_monitoring_persists_complete_usage_and_structured_errors(chain):
+    invoke, _, tmpdir = chain
+    invoke('runner-events', 'monitoring')
+    with sqlite3.connect(tmpdir / 'data/langbot.db') as db:
+        rows = db.execute('SELECT details FROM monitoring_llm_calls').fetchall()
+    details = [json.loads(row[0]) for row in rows if row[0]]
+    assert any(x.get('usage', {}).get('prompt_tokens_details', {}).get('cached_tokens') == 16 for x in details)
+    assert any(x.get('timing', {}).get('transport_setup_ms') is not None for x in details)
+    assert any(x.get('error', {}).get('code') for x in details if x.get('error'))
+
+
+@pytest.mark.skipif(not os.environ.get('LLM_E2E_LEGACY_API_DIR'), reason='Historical SDK sources not supplied')
+@pytest.mark.parametrize('mode', ['legacy-message', 'legacy-usage', 'legacy-stream', 'legacy-events'])
+def test_historical_sdk_proxy_and_message_entities(chain, mode):
+    result = chain[0](mode, mode)
+    assert result['text'] == 'hello world', result
+    assert not result.get('error'), result
+    if mode.endswith(('usage', 'events')):
+        assert result['usage']['prompt_tokens_details']['cached_tokens'] == 16
diff --git a/tests/e2e/utils/llm_probe_runner.py b/tests/e2e/utils/llm_probe_runner.py
new file mode 100644
index 000000000..9b1115408
--- /dev/null
+++ b/tests/e2e/utils/llm_probe_runner.py
@@ -0,0 +1,93 @@
+"""Packaged into the isolated E2E plugin; calls only public SDK LLM APIs."""
+
+from contextlib import aclosing
+import json
+
+from langbot_plugin.api.definition.components.runner.runner import Runner
+from langbot_plugin.api.entities.builtin.provider.message import Message
+from langbot_plugin.api.entities.builtin.runner.result import RunnerResult
+from langbot_plugin.api.proxies.langbot_api import LangBotAPIProxy
+
+
+class DefaultRunner(Runner):
+    async def run(self, ctx):
+        options = json.loads(ctx.input.to_text())
+        mode = options['mode']
+        cancel = mode.startswith('cancel-')
+        if cancel:
+            mode = mode.removeprefix('cancel-')
+        api = LangBotAPIProxy(self._plugin_runtime_handler) if mode.startswith('plugin-') else ctx.api
+        message_type = Message
+        if mode.startswith('legacy-'):
+            # Historical files are supplied verbatim from a released SDK tag.
+            # Its proxy and serializers use the real current runtime transport.
+            import legacy_api
+            import legacy_message
+
+            legacy_api.provider_message = legacy_message
+            api = legacy_api.LangBotAPIProxy(self._plugin_runtime_handler)
+            message_type = legacy_message.Message
+        extra = {
+            **options['extra'],
+            'extra_headers': {
+                **options['extra'].get('extra_headers', {}),
+                'x-e2e-case': options['case'],
+            },
+        }
+        kwargs = {
+            'llm_model_uuid': options['model'],
+            'messages': [
+                message_type.model_validate(m)
+                for m in options.get(
+                    'messages',
+                    [
+                        {'role': 'user', 'content': 'Hello'},
+                    ],
+                )
+            ],
+            'extra_args': extra,
+        }
+        if options.get('tool'):
+            from langbot_plugin.api.entities.builtin.resource.tool import LLMTool
+
+            async def lookup(**kwargs):
+                return 'result'
+
+            kwargs['funcs'] = [
+                LLMTool(
+                    name='e2e_lookup',
+                    human_desc='Lookup',
+                    description='Lookup',
+                    parameters={'type': 'object', 'properties': {}},
+                    func=lookup,
+                )
+            ]
+        result = {'text': '', 'usage': None}
+        if 'remove_think' in options:
+            kwargs['remove_think'] = options['remove_think']
+        try:
+            if mode.endswith('message'):
+                message = await api.invoke_llm(**kwargs)
+                assert isinstance(message, message_type)
+                result['text'] = message.content
+            elif mode.endswith('usage'):
+                response = await api.invoke_llm_with_usage(**kwargs)
+                result.update(
+                    text=response.message.content, usage=response.usage.model_dump() if response.usage else None
+                )
+            else:
+                events = mode.endswith('events')
+                method = api.invoke_llm_stream_events if events else api.invoke_llm_stream
+                async with aclosing(method(**kwargs)) as stream:
+                    async for item in stream:
+                        chunk = item.chunk if events else item
+                        if chunk:
+                            result['text'] += chunk.content or ''
+                        if events and item.usage:
+                            result['usage'] = item.usage.model_dump()
+                        if cancel and result['text']:
+                            break
+        except Exception as exc:
+            result['error'] = type(exc).__name__ + ': ' + str(exc)
+        yield RunnerResult.message_completed(ctx.run_id, Message(role='assistant', content=json.dumps(result)))
+        yield RunnerResult.run_completed(ctx.run_id, finish_reason='stop')
diff --git a/tests/integration/api/test_box_security.py b/tests/integration/api/test_box_security.py
index 0a40f230a..a0409d908 100644
--- a/tests/integration/api/test_box_security.py
+++ b/tests/integration/api/test_box_security.py
@@ -35,6 +35,7 @@ async def box_security_api():
         'owner-token': SimpleNamespace(uuid='owner-account', user='owner@example.com'),
     }
     application = Mock()
+    application._message_language_cache = None
     application.deployment = SimpleNamespace(multi_workspace_enabled=False)
     application.persistence_mgr = SimpleNamespace(tenant_uow=None)
     application.user_service.get_authenticated_account = AsyncMock(side_effect=lambda token: accounts[token])
diff --git a/tests/integration/api/test_passkey_account_scope.py b/tests/integration/api/test_passkey_account_scope.py
new file mode 100644
index 000000000..8f6ce38bc
--- /dev/null
+++ b/tests/integration/api/test_passkey_account_scope.py
@@ -0,0 +1,139 @@
+"""Account scoping for the passkey and password routes.
+
+A passkey belongs to the Account, so these routes must answer for the Account the
+caller authenticated as. They previously required a Workspace (which the WebUI
+never sends for them) and then resolved the Account by looking the login name up
+as an email address, so every Account whose login name is not its email answered
+``404`` and the passkey feature was unreachable.
+"""
+
+from __future__ import annotations
+
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, Mock
+
+import pytest
+import quart
+
+from langbot.pkg.api.http.controller.groups.user import UserRouterGroup
+
+pytestmark = pytest.mark.integration
+
+ACCOUNT_UUID = '1905ad7c-168f-49f4-893e-e43614e4c4bf'
+LOGIN_NAME = 'dev'
+ACCOUNT_EMAIL = 'dev@local.test'
+
+
+@pytest.fixture
+async def account_api():
+    # The login name deliberately differs from the email: that is what made the
+    # email lookup miss the authenticated Account.
+    account = SimpleNamespace(
+        uuid=ACCOUNT_UUID,
+        user=LOGIN_NAME,
+        normalized_email=ACCOUNT_EMAIL,
+        password='',
+        account_type='space',
+    )
+    application = Mock()
+    application.deployment = SimpleNamespace(multi_workspace_enabled=False)
+    application.instance_config.data = {'system': {'allow_modify_login_info': True}}
+    application.persistence_mgr = SimpleNamespace(tenant_uow=None)
+    application.user_service.get_authenticated_account = AsyncMock(return_value=account)
+    application.user_service.get_user_by_email = AsyncMock(return_value=None)
+    application.user_service.get_user_passkeys = AsyncMock(return_value=[])
+    application.user_service.generate_passkey_registration_options = AsyncMock(
+        return_value=({'challenge': 'probe'}, 'challenge-token')
+    )
+    application.user_service.verify_and_save_passkey_registration = AsyncMock(
+        return_value=SimpleNamespace(uuid='credential-uuid', name='probe', created_at=None)
+    )
+    application.user_service.rename_user_passkey = AsyncMock(
+        return_value=SimpleNamespace(uuid='credential-uuid', name='renamed')
+    )
+    application.user_service.delete_user_passkey = AsyncMock(return_value=True)
+    application.user_service.set_password = AsyncMock()
+    application.user_service.change_password = AsyncMock()
+
+    quart_app = quart.Quart(__name__)
+    router = UserRouterGroup(application, quart_app)
+    await router.initialize()
+    return application, quart_app.test_client()
+
+
+def _headers() -> dict[str, str]:
+    # The WebUI sends an account session and no Workspace for these routes.
+    return {'Authorization': 'Bearer account-token'}
+
+
+async def test_passkey_list_answers_without_a_workspace(account_api):
+    application, client = account_api
+
+    response = await client.get('/api/v1/user/passkeys', headers=_headers())
+
+    assert response.status_code == 200, await response.get_data(as_text=True)
+    application.user_service.get_user_passkeys.assert_awaited_once_with(ACCOUNT_UUID)
+    application.user_service.get_user_by_email.assert_not_awaited()
+
+
+async def test_passkey_registration_options_bind_the_authenticated_account(account_api):
+    application, client = account_api
+
+    response = await client.post(
+        '/api/v1/user/passkey/register/options',
+        json={'origin': 'http://localhost:3000'},
+        headers=_headers(),
+    )
+
+    assert response.status_code == 200, await response.get_data(as_text=True)
+    assert (await response.get_json())['data']['challenge_token'] == 'challenge-token'
+    options_call = application.user_service.generate_passkey_registration_options.await_args
+    assert options_call.kwargs['account_uuid'] == ACCOUNT_UUID
+    assert options_call.kwargs['rp_id'] == 'localhost'
+    application.user_service.get_user_by_email.assert_not_awaited()
+
+
+async def test_passkey_verify_and_revoke_address_the_authenticated_account(account_api):
+    application, client = account_api
+
+    verified = await client.post(
+        '/api/v1/user/passkey/register/verify',
+        json={'challenge_token': 'challenge-token', 'credential': {'id': 'cred'}},
+        headers=_headers(),
+    )
+    renamed = await client.patch(
+        '/api/v1/user/passkey/credential-uuid',
+        json={'name': 'renamed'},
+        headers=_headers(),
+    )
+    deleted = await client.delete('/api/v1/user/passkey/credential-uuid', headers=_headers())
+
+    assert verified.status_code == 200, await verified.get_data(as_text=True)
+    assert renamed.status_code == 200, await renamed.get_data(as_text=True)
+    assert deleted.status_code == 200, await deleted.get_data(as_text=True)
+    assert application.user_service.rename_user_passkey.await_args.kwargs['account_uuid'] == ACCOUNT_UUID
+    assert application.user_service.delete_user_passkey.await_args.kwargs['account_uuid'] == ACCOUNT_UUID
+    application.user_service.get_user_by_email.assert_not_awaited()
+
+
+async def test_password_routes_target_the_authenticated_account(account_api):
+    application, client = account_api
+
+    set_password = await client.post(
+        '/api/v1/user/set-password',
+        json={'new_password': 'new-secret'},
+        headers=_headers(),
+    )
+    change_password = await client.post(
+        '/api/v1/user/change-password',
+        json={'current_password': 'old-secret', 'new_password': 'new-secret'},
+        headers=_headers(),
+    )
+
+    assert set_password.status_code == 200, await set_password.get_data(as_text=True)
+    assert change_password.status_code == 200, await change_password.get_data(as_text=True)
+    # The service resolves the row by email, so the route hands it the email of the
+    # account the token identified -- never the login name it was given.
+    assert application.user_service.set_password.await_args.args[0] == ACCOUNT_EMAIL
+    assert application.user_service.change_password.await_args.args[0] == ACCOUNT_EMAIL
+    application.user_service.get_user_by_email.assert_not_awaited()
diff --git a/tests/integration/api/test_plugins_security.py b/tests/integration/api/test_plugins_security.py
index 52442e371..2f773595a 100644
--- a/tests/integration/api/test_plugins_security.py
+++ b/tests/integration/api/test_plugins_security.py
@@ -88,6 +88,7 @@ async def plugin_security_api(plugin_module):
     }
 
     application = Mock()
+    application._message_language_cache = None
     application.deployment = SimpleNamespace(multi_workspace_enabled=False)
     application.user_service.get_authenticated_account = AsyncMock(side_effect=lambda token: accounts[token])
     application.workspace_collaboration_service.resolve_account_workspace = AsyncMock(
diff --git a/tests/integration/api/test_providers.py b/tests/integration/api/test_providers.py
index df72d9084..bedd1d871 100644
--- a/tests/integration/api/test_providers.py
+++ b/tests/integration/api/test_providers.py
@@ -329,7 +329,7 @@ class TestModelEndpoints:
         assert response.status_code == 400
         data = await response.get_json()
         assert data['code'] == 'model_test_failed'
-        assert data['msg'] == '模型请求失败: API key invalid'
+        assert data['msg'] == 'Model request failed: API key invalid'
 
 
 @pytest.mark.usefixtures('mock_circular_import_chain')
diff --git a/tests/integration/api/test_skills_entitlements.py b/tests/integration/api/test_skills_entitlements.py
index a936597ab..0df480e27 100644
--- a/tests/integration/api/test_skills_entitlements.py
+++ b/tests/integration/api/test_skills_entitlements.py
@@ -24,6 +24,7 @@ async def skills_api():
         execution=SimpleNamespace(instance_uuid='instance-a', placement_generation=1),
     )
     application = Mock()
+    application._message_language_cache = None
     application.deployment = SimpleNamespace(multi_workspace_enabled=False)
     application.persistence_mgr = SimpleNamespace(tenant_uow=None)
     application.user_service.get_authenticated_account = AsyncMock(return_value=account)
@@ -70,4 +71,6 @@ async def test_list_skills_does_not_hide_other_entitlement_failures(skills_api):
         },
     )
 
-    assert response.status_code == 500
+    assert response.status_code == 403
+    payload = await response.get_json()
+    assert payload['code'] == 'entitlement_unavailable'
diff --git a/tests/integration/api/test_support_admin_launch.py b/tests/integration/api/test_support_admin_launch.py
index e71cb4374..dd578794f 100644
--- a/tests/integration/api/test_support_admin_launch.py
+++ b/tests/integration/api/test_support_admin_launch.py
@@ -24,6 +24,7 @@ from langbot.pkg.api.http.controller.groups.user import UserRouterGroup
 from langbot.pkg.cloud.launch import SpaceLaunchError, SpaceLaunchService
 from langbot.pkg.cloud.support_admin import SupportAdminSessionService
 from langbot.pkg.entity.persistence.base import Base
+from langbot.pkg.entity.persistence.metadata import WorkspaceMetadata
 from langbot.pkg.entity.persistence.support_admin import SupportAdminTemporarySession
 from langbot.pkg.entity.persistence.user import User
 from langbot.pkg.entity.persistence.workspace import (
@@ -124,6 +125,9 @@ class _TenantUow:
         self._transaction = await self.session.begin()
         return self
 
+    async def execute(self, statement):
+        return await self.session.execute(statement)
+
     async def __aexit__(self, exc_type, exc, traceback):
         try:
             if exc_type is None:
@@ -173,6 +177,7 @@ async def support_admin_api(tmp_path):
             tables=[
                 User.__table__,
                 Workspace.__table__,
+                WorkspaceMetadata.__table__,
                 WorkspaceExecutionState.__table__,
                 WorkspaceMembership.__table__,
                 SupportAdminTemporarySession.__table__,
diff --git a/tests/integration/api/test_totp_workspace_scope.py b/tests/integration/api/test_totp_workspace_scope.py
new file mode 100644
index 000000000..4f822062b
--- /dev/null
+++ b/tests/integration/api/test_totp_workspace_scope.py
@@ -0,0 +1,124 @@
+"""Workspace scoping for the owner/admin second-factor oversight endpoints.
+
+An owner or admin answers for the Accounts of its own Workspace. The oversight
+endpoints must never list, revoke, or re-bind another tenant's Account.
+"""
+
+from __future__ import annotations
+
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, Mock
+
+import pytest
+import quart
+
+from langbot.pkg.api.http.controller.groups.user import UserRouterGroup
+from langbot.pkg.workspace.errors import WorkspaceNotFoundError
+
+pytestmark = pytest.mark.integration
+
+WORKSPACE_UUID = '11111111-1111-4111-8111-111111111111'
+MEMBER_UUID = 'member-account'
+FOREIGN_UUID = 'foreign-account'
+
+
+def _access(account_uuid: str) -> SimpleNamespace:
+    return SimpleNamespace(
+        workspace=SimpleNamespace(uuid=WORKSPACE_UUID),
+        membership=SimpleNamespace(
+            uuid=f'membership-{account_uuid}',
+            account_uuid=account_uuid,
+            role='owner',
+            projection_revision=1,
+        ),
+        execution=SimpleNamespace(instance_uuid='instance-a', placement_generation=1),
+    )
+
+
+async def _resolve(account_uuid: str, _workspace_uuid: str | None) -> SimpleNamespace:
+    # The collaboration service hides Accounts that are not members here.
+    if account_uuid == FOREIGN_UUID:
+        raise WorkspaceNotFoundError('Workspace not found')
+    return _access(account_uuid)
+
+
+@pytest.fixture
+async def totp_admin_api():
+    accounts = {'member-token': SimpleNamespace(uuid=MEMBER_UUID, user='member@example.com')}
+    application = Mock()
+    application._message_language_cache = None
+    application.deployment = SimpleNamespace(multi_workspace_enabled=False)
+    application.instance_config.data = {'system': {'allow_modify_login_info': True}}
+    application.persistence_mgr = SimpleNamespace(tenant_uow=None)
+    application.user_service.get_authenticated_account = AsyncMock(
+        side_effect=lambda token: accounts[token]
+    )
+    application.workspace_collaboration_service.resolve_account_workspace = AsyncMock(
+        side_effect=_resolve
+    )
+    application.workspace_collaboration_service.list_members = AsyncMock(
+        return_value=[SimpleNamespace(membership=SimpleNamespace(account_uuid=MEMBER_UUID))]
+    )
+    application.totp_service.list_account_states = AsyncMock(
+        return_value=[{'account_uuid': MEMBER_UUID, 'user': 'member', 'enabled': False}]
+    )
+    application.totp_service.revoke_for_account = AsyncMock(return_value=True)
+    application.totp_service.get_account = AsyncMock(
+        return_value=SimpleNamespace(uuid=FOREIGN_UUID, user='foreign@example.com')
+    )
+    application.totp_service.begin_enrollment = AsyncMock()
+    application.totp_service.confirm_enrollment = AsyncMock()
+
+    quart_app = quart.Quart(__name__)
+    router = UserRouterGroup(application, quart_app)
+    await router.initialize()
+    return application, quart_app.test_client()
+
+
+def _headers() -> dict[str, str]:
+    return {'Authorization': 'Bearer member-token', 'X-Workspace-Id': WORKSPACE_UUID}
+
+
+async def test_account_list_covers_only_the_callers_workspace(totp_admin_api):
+    application, client = totp_admin_api
+
+    response = await client.get('/api/v1/user/totp/accounts', headers=_headers())
+
+    assert response.status_code == 200
+    payload = await response.get_json()
+    assert [item['account_uuid'] for item in payload['data']['accounts']] == [MEMBER_UUID]
+    application.totp_service.list_account_states.assert_awaited_once_with(
+        account_uuids=[MEMBER_UUID]
+    )
+
+
+async def test_foreign_account_cannot_be_revoked(totp_admin_api):
+    application, client = totp_admin_api
+
+    response = await client.delete(
+        f'/api/v1/user/totp/accounts/{FOREIGN_UUID}',
+        headers=_headers(),
+    )
+
+    assert response.status_code == 404
+    assert (await response.get_json())['code'] == 'account_not_found'
+    application.totp_service.revoke_for_account.assert_not_awaited()
+
+
+async def test_foreign_account_cannot_be_re_bound(totp_admin_api):
+    application, client = totp_admin_api
+
+    enroll = await client.post(
+        f'/api/v1/user/totp/accounts/{FOREIGN_UUID}/enroll',
+        headers=_headers(),
+    )
+    confirm = await client.post(
+        f'/api/v1/user/totp/accounts/{FOREIGN_UUID}/enroll/confirm',
+        json={'code': '123456'},
+        headers=_headers(),
+    )
+
+    assert enroll.status_code == 404
+    assert confirm.status_code == 404
+    application.totp_service.begin_enrollment.assert_not_awaited()
+    application.totp_service.confirm_enrollment.assert_not_awaited()
diff --git a/tests/integration/api/test_user_space_oauth.py b/tests/integration/api/test_user_space_oauth.py
index af2bab140..ad49c83a6 100644
--- a/tests/integration/api/test_user_space_oauth.py
+++ b/tests/integration/api/test_user_space_oauth.py
@@ -27,6 +27,7 @@ async def space_oauth_api():
         execution=SimpleNamespace(instance_uuid='instance-a', placement_generation=1),
     )
     application = Mock()
+    application._message_language_cache = None
     application.deployment = SimpleNamespace(multi_workspace_enabled=False, mode='oss')
     application.directory_projection_service = None
     application.persistence_mgr = None
diff --git a/tests/integration/api/test_workspaces.py b/tests/integration/api/test_workspaces.py
index a561cd44c..dd2baf961 100644
--- a/tests/integration/api/test_workspaces.py
+++ b/tests/integration/api/test_workspaces.py
@@ -824,6 +824,10 @@ async def test_cloud_projection_is_selected_explicitly_and_collaboration_runs_in
                 not_before=1,
                 expires_at=4102444800,
                 plan_name='free',
+                plan_revision=3,
+                monthly_bonus_credits=12,
+                subscription_started_at=1720000000,
+                limits={'bots.max': 2, 'pipelines.max': 3, 'plugins.max': 4, 'knowledge_bases.max': 5},
             )
 
     application.entitlement_resolver = PlanResolver()
@@ -843,6 +847,11 @@ async def test_cloud_projection_is_selected_explicitly_and_collaboration_runs_in
     assert current['workspace']['source'] == 'cloud_projection'
     assert current['placement_generation'] == 12
     assert current['plan_name'] == 'free'
+    assert current['limitation']['max_bots'] == 2
+    assert current['current_plan']['subscription_started_at'] == 1720000000
+    assert current['current_plan']['revision'] == 3
+    assert current['current_plan']['monthly_bonus_credits'] == 12
+    assert current['current_plan']['limits']['bots.max'] == 2
 
     create_workspace = await client.post(
         '/api/v1/workspaces',
@@ -926,3 +935,64 @@ async def test_account_bootstrap_does_not_disclose_non_member_workspaces(workspa
     )
     assert current.status_code == 404
     assert (await current.get_json())['code'] == 'resource_not_found'
+
+
+async def test_workspace_language_persists_and_controls_api_errors(workspace_api):
+    from langbot.pkg.i18n.workspace import get_workspace_language, set_workspace_language
+
+    application, client, engine, owner_token = workspace_api
+    current = await client.get('/api/v1/workspaces/current', headers=_auth(owner_token))
+    current = (await current.get_json())['data']
+    workspace_uuid = current['workspace']['uuid']
+    owner_uuid = current['membership']['account_uuid']
+    headers = {**_auth(owner_token, workspace_uuid), 'X-LangBot-Language': 'ja-JP'}
+    initial = await client.get('/api/v1/workspaces/current/language', headers=headers)
+    assert (await initial.get_json())['data']['language'] == 'en_US'
+
+    updated = await client.put('/api/v1/workspaces/current/language', headers=headers, json={'language': 'zh_Hans'})
+    assert updated.status_code == 200
+    assert await get_workspace_language(application, workspace_uuid) == 'zh_Hans'
+    restarted = SimpleNamespace(persistence_mgr=application.persistence_mgr)
+    assert await get_workspace_language(restarted, workspace_uuid) == 'zh_Hans'
+
+    invalid = await client.put('/api/v1/workspaces/current/language', headers=headers, json={'language': '../bad'})
+    assert invalid.status_code == 400
+    assert await get_workspace_language(application, workspace_uuid, refresh=True) == 'zh_Hans'
+
+    # A second Workspace has independent persisted metadata and cache entries.
+    async with engine.begin() as connection:
+        await connection.execute(
+            sqlalchemy.insert(Workspace).values(
+                uuid='language-other',
+                instance_uuid='instance-workspace-api',
+                name='Other',
+                slug='language-other',
+                source='cloud_projection',
+                type='team',
+            )
+        )
+    await set_workspace_language(application, 'language-other', 'es_ES')
+    assert await get_workspace_language(application, 'language-other') == 'es_ES'
+    assert await get_workspace_language(application, workspace_uuid) == 'zh_Hans'
+
+    # Read-only members can inspect the setting but cannot modify it. The
+    # verified Workspace setting wins over the browser language hint.
+    async with engine.begin() as connection:
+        await connection.execute(
+            sqlalchemy.update(WorkspaceMembership)
+            .where(
+                WorkspaceMembership.workspace_uuid == workspace_uuid,
+                WorkspaceMembership.account_uuid == owner_uuid,
+            )
+            .values(role='viewer')
+        )
+    denied = await client.put('/api/v1/workspaces/current/language', headers=headers, json={'language': 'ja_JP'})
+    assert denied.status_code == 403
+    assert (await denied.get_json())['msg'] == '没有执行此操作的权限'
+    visible = await client.get('/api/v1/workspaces/current/language', headers=headers)
+    assert visible.status_code == 200
+    assert (await visible.get_json())['data']['language'] == 'zh_Hans'
+
+    # An untrusted selector never causes us to read another tenant's language.
+    unauthorized = await client.get('/api/v1/workspaces/current/language', headers=_auth(owner_token, 'language-other'))
+    assert unauthorized.status_code in (403, 404)
diff --git a/tests/integration/persistence/fixtures/release_4_10_11_audit.md b/tests/integration/persistence/fixtures/release_4_10_11_audit.md
new file mode 100644
index 000000000..768c64c6d
--- /dev/null
+++ b/tests/integration/persistence/fixtures/release_4_10_11_audit.md
@@ -0,0 +1,39 @@
+# v4.10.11 database upgrade audit
+
+Baseline: `v4.10.11`, commit `273b8839b91cb74dfd083644efea18849a3fcee1`.
+Audited target: `48c6a1f536bab421cbaab98a1e880fe767034463`.
+
+## Changes covered
+
+- Agent definitions, run ledger, interactions/delivery results, event logs and transcripts.
+- Bot legacy pipeline routing to event bindings, followed by legacy-column removal; plugin processors.
+- Workspace tenancy/ownership repairs and migration branch merges.
+- Unified runner state, assistant conversations and pipeline migration snapshots.
+- Passkey and TOTP credentials.
+- Knowledge-base initialization/drafts and engine document identity.
+- Operation traceability, monitoring links/tool calls, execution host leases and metadata text storage.
+
+The existing migration chain covers the audited ORM changes. No additional runtime migration was required by the exercised upgrade paths.
+
+## Added regression
+
+`release_4_10_11_schema.json` contains immutable dialect-specific DDL exported from the tagged release ORM, rather than current metadata. The test applies historical migrations through `0023_bot_scoped_sessions` to include migration-owned objects such as pgvector, seeds representative legacy records, then runs the current persistence schema initialization twice.
+
+Assertions cover legacy bot/pipeline preservation, converted bot event bindings, default plugin processors, initialized knowledge bases, SQLite foreign-key consistency, and missing tables/columns/indexes/constraints or column type/nullability differences against current metadata. Equivalent existing index/constraint column sets are accepted; extra migration-owned objects and removed-object diffs are not treated as missing schema. This is representative upgrade coverage, not an exhaustive test of every possible historical data value.
+
+## Executed verification
+
+Disposable SQLite files and an isolated PostgreSQL/pgvector container were used; no business database was changed.
+
+- New tagged-release upgrade test: **2 passed** (SQLite and PostgreSQL), including repeated initialization.
+- Existing PostgreSQL migration / release-migration / RAG identity suite: **85 passed, 4 failed** on the first completed run. The four failures were release configuration validation because the test URL omitted the required port.
+- Release-migration suite rerun with explicit `:5432`: **5 passed**.
+- Ruff check and format: passed for the new regression test.
+
+Reproduce the new regression with `TEST_POSTGRES_URL` pointing to a disposable PostgreSQL/pgvector instance whose test role can create/drop databases:
+
+```sh
+pytest tests/integration/persistence/test_release_41011_upgrade.py -q
+```
+
+Without `TEST_POSTGRES_URL`, PostgreSQL is explicitly skipped rather than reported as tested.
diff --git a/tests/integration/persistence/fixtures/release_4_10_11_schema.json b/tests/integration/persistence/fixtures/release_4_10_11_schema.json
new file mode 100644
index 000000000..cf8a2beda
--- /dev/null
+++ b/tests/integration/persistence/fixtures/release_4_10_11_schema.json
@@ -0,0 +1,250 @@
+{
+  "source_commit": "273b8839b91cb74dfd083644efea18849a3fcee1",
+  "revision": "0023_bot_scoped_sessions",
+  "dialects": {
+    "sqlite": [
+      "\nCREATE TABLE directory_projection_inbox (\n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tevent_uuid VARCHAR(36) NOT NULL, \n\tcursor BIGINT NOT NULL, \n\tevent_type VARCHAR(128) NOT NULL, \n\trevision BIGINT NOT NULL, \n\tfingerprint TEXT NOT NULL, \n\treceived_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tapplied_at DATETIME, \n\tPRIMARY KEY (instance_uuid, event_uuid), \n\tCONSTRAINT uq_directory_projection_inbox_cursor UNIQUE (instance_uuid, cursor), \n\tCONSTRAINT ck_directory_projection_inbox_cursor CHECK (cursor > 0), \n\tCONSTRAINT ck_directory_projection_inbox_revision CHECK (revision > 0), \n\tCONSTRAINT ck_directory_projection_inbox_fingerprint CHECK (length(fingerprint) = 64)\n)\n\n",
+      "CREATE INDEX ix_directory_projection_inbox_pending ON directory_projection_inbox (instance_uuid, applied_at, cursor)",
+      "\nCREATE TABLE directory_projection_states (\n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tcursor BIGINT DEFAULT '0' NOT NULL, \n\tsnapshot_coverage_cursor BIGINT DEFAULT '0' NOT NULL, \n\tsnapshot_fingerprint TEXT NOT NULL, \n\tlast_applied_at DATETIME NOT NULL, \n\tlease_expires_at DATETIME, \n\tPRIMARY KEY (instance_uuid), \n\tCONSTRAINT ck_directory_projection_state_cursor CHECK (cursor >= 0), \n\tCONSTRAINT ck_directory_projection_state_snapshot_coverage CHECK (snapshot_coverage_cursor >= 0 AND snapshot_coverage_cursor <= cursor), \n\tCONSTRAINT ck_directory_projection_state_fingerprint CHECK (length(snapshot_fingerprint) = 64)\n)\n\n",
+      "\nCREATE TABLE metadata (\n\t\"key\" VARCHAR(255) NOT NULL, \n\tvalue VARCHAR(255), \n\tPRIMARY KEY (\"key\")\n)\n\n",
+      "\nCREATE TABLE users (\n\tid INTEGER NOT NULL, \n\tuuid VARCHAR(36) NOT NULL, \n\tuser VARCHAR(255) NOT NULL, \n\tnormalized_email VARCHAR(320) NOT NULL, \n\tpassword VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\taccount_type VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tspace_account_uuid VARCHAR(255), \n\tspace_access_token TEXT, \n\tspace_refresh_token TEXT, \n\tspace_access_token_expires_at DATETIME, \n\tspace_api_key VARCHAR(255), \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT ck_users_normalized_email CHECK (normalized_email = trim(normalized_email) AND length(normalized_email) > 0 AND length(normalized_email) <= 320), \n\tCONSTRAINT ck_users_status CHECK (status IN ('active', 'disabled', 'deleted')), \n\tCONSTRAINT ck_users_source CHECK (source IN ('local', 'cloud_projection'))\n)\n\n",
+      "CREATE UNIQUE INDEX uq_users_normalized_email ON users (normalized_email)",
+      "CREATE UNIQUE INDEX uq_users_uuid ON users (uuid)",
+      "\nCREATE TABLE workspaces (\n\tuuid VARCHAR(36) NOT NULL, \n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tslug VARCHAR(255) NOT NULL, \n\ttype VARCHAR(32) DEFAULT 'team' NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tcreated_by_account_uuid VARCHAR(36), \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_workspaces_instance_slug UNIQUE (instance_uuid, slug), \n\tCONSTRAINT ck_workspaces_type CHECK (type IN ('personal', 'team')), \n\tCONSTRAINT ck_workspaces_status CHECK (status IN ('provisioning', 'active', 'suspended', 'archived', 'deleted')), \n\tCONSTRAINT ck_workspaces_source CHECK (source IN ('local', 'cloud_projection')), \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_workspaces_instance_status ON workspaces (instance_uuid, status)",
+      "CREATE UNIQUE INDEX uq_workspaces_local_instance ON workspaces (instance_uuid) WHERE source = 'local'",
+      "\nCREATE TABLE api_keys (\n\tid INTEGER NOT NULL, \n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tcreated_by_account_uuid VARCHAR(36), \n\tname VARCHAR(255) NOT NULL, \n\tkey_hash VARCHAR(64) NOT NULL, \n\tscopes JSON DEFAULT '[]' NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\texpires_at DATETIME, \n\tlast_used_at DATETIME, \n\tdescription VARCHAR(512), \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT ck_api_keys_status CHECK (status IN ('active', 'revoked')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_api_keys_workspace_name ON api_keys (workspace_uuid, name)",
+      "CREATE INDEX ix_api_keys_workspace_status ON api_keys (workspace_uuid, status)",
+      "CREATE UNIQUE INDEX uq_api_keys_key_hash ON api_keys (key_hash)",
+      "CREATE UNIQUE INDEX uq_api_keys_uuid ON api_keys (uuid)",
+      "\nCREATE TABLE binary_storages (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tunique_key VARCHAR(255) NOT NULL, \n\t\"key\" VARCHAR(255) NOT NULL, \n\towner_type VARCHAR(255) NOT NULL, \n\towner VARCHAR(255) NOT NULL, \n\tvalue BLOB NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (workspace_uuid, unique_key), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_binary_storages_workspace_owner ON binary_storages (workspace_uuid, owner_type, owner)",
+      "\nCREATE TABLE bots (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tdescription VARCHAR(255) NOT NULL, \n\tadapter VARCHAR(255) NOT NULL, \n\tadapter_config JSON NOT NULL, \n\tenable BOOLEAN NOT NULL, \n\tuse_pipeline_name VARCHAR(255), \n\tuse_pipeline_uuid VARCHAR(255), \n\tpipeline_routing_rules JSON DEFAULT '[]' NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_bots_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_bots_workspace_name ON bots (workspace_uuid, name)",
+      "\nCREATE TABLE knowledge_bases (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR, \n\tdescription TEXT, \n\temoji VARCHAR(10), \n\tcreated_at DATETIME, \n\tupdated_at DATETIME, \n\tknowledge_engine_plugin_id VARCHAR, \n\tcollection_id VARCHAR, \n\tlegacy_vector_collection BOOLEAN DEFAULT 0 NOT NULL, \n\tcreation_settings JSON, \n\tretrieval_settings JSON, \n\tembedding_dimension INTEGER, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_knowledge_bases_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tCONSTRAINT ck_knowledge_bases_embedding_dimension_positive CHECK (embedding_dimension IS NULL OR embedding_dimension > 0), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_bases_name ON knowledge_bases (name)",
+      "CREATE INDEX ix_knowledge_bases_workspace_name ON knowledge_bases (workspace_uuid, name)",
+      "CREATE UNIQUE INDEX uq_knowledge_bases_workspace_collection ON knowledge_bases (workspace_uuid, collection_id) WHERE collection_id IS NOT NULL",
+      "\nCREATE TABLE legacy_pipelines (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tdescription VARCHAR(255) NOT NULL, \n\temoji VARCHAR(10), \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tfor_version VARCHAR(255) NOT NULL, \n\tis_default BOOLEAN NOT NULL, \n\tstages JSON NOT NULL, \n\tconfig JSON NOT NULL, \n\textensions_preferences JSON NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_legacy_pipelines_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_legacy_pipelines_workspace_default ON legacy_pipelines (workspace_uuid, is_default)",
+      "CREATE INDEX ix_legacy_pipelines_workspace_name ON legacy_pipelines (workspace_uuid, name)",
+      "\nCREATE TABLE mcp_servers (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tenable BOOLEAN NOT NULL, \n\tmode VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\treadme TEXT DEFAULT '' NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_mcp_servers_workspace_name UNIQUE (workspace_uuid, name), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_mcp_servers_workspace_enable ON mcp_servers (workspace_uuid, enable)",
+      "\nCREATE TABLE model_providers (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\trequester VARCHAR(255) NOT NULL, \n\tbase_url VARCHAR(512) NOT NULL, \n\tapi_keys JSON NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_model_providers_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_model_providers_workspace_name ON model_providers (workspace_uuid, name)",
+      "CREATE INDEX ix_model_providers_workspace_requester ON model_providers (workspace_uuid, requester)",
+      "\nCREATE TABLE monitoring_embedding_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\tmodel_name VARCHAR(255) NOT NULL, \n\tprompt_tokens INTEGER NOT NULL, \n\ttotal_tokens INTEGER NOT NULL, \n\tduration INTEGER NOT NULL, \n\tinput_count INTEGER NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\terror_message TEXT, \n\tknowledge_base_id VARCHAR(255), \n\tquery_text TEXT, \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\tcall_type VARCHAR(50), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_embedding_calls_knowledge_base_id ON monitoring_embedding_calls (knowledge_base_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_message_id ON monitoring_embedding_calls (message_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_session_id ON monitoring_embedding_calls (session_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_timestamp ON monitoring_embedding_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_embedding_calls_workspace_kb ON monitoring_embedding_calls (workspace_uuid, knowledge_base_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_workspace_timestamp ON monitoring_embedding_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_errors (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\terror_type VARCHAR(255) NOT NULL, \n\terror_message TEXT NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255), \n\tstack_trace TEXT, \n\tmessage_id VARCHAR(255), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_errors_bot_id ON monitoring_errors (bot_id)",
+      "CREATE INDEX ix_monitoring_errors_message_id ON monitoring_errors (message_id)",
+      "CREATE INDEX ix_monitoring_errors_pipeline_id ON monitoring_errors (pipeline_id)",
+      "CREATE INDEX ix_monitoring_errors_timestamp ON monitoring_errors (timestamp)",
+      "CREATE INDEX ix_monitoring_errors_workspace_session ON monitoring_errors (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_errors_workspace_timestamp ON monitoring_errors (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_feedback (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\tfeedback_id VARCHAR(255) NOT NULL, \n\tfeedback_type INTEGER NOT NULL, \n\tfeedback_content TEXT, \n\tinaccurate_reasons TEXT, \n\tbot_id VARCHAR(255), \n\tbot_name VARCHAR(255), \n\tpipeline_id VARCHAR(255), \n\tpipeline_name VARCHAR(255), \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\tstream_id VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tplatform VARCHAR(255), \n\tPRIMARY KEY (id), \n\tCONSTRAINT uq_monitoring_feedback_workspace_feedback_id UNIQUE (workspace_uuid, feedback_id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_feedback_bot_id ON monitoring_feedback (bot_id)",
+      "CREATE INDEX ix_monitoring_feedback_feedback_id ON monitoring_feedback (feedback_id)",
+      "CREATE INDEX ix_monitoring_feedback_message_id ON monitoring_feedback (message_id)",
+      "CREATE INDEX ix_monitoring_feedback_pipeline_id ON monitoring_feedback (pipeline_id)",
+      "CREATE INDEX ix_monitoring_feedback_session_id ON monitoring_feedback (session_id)",
+      "CREATE INDEX ix_monitoring_feedback_stream_id ON monitoring_feedback (stream_id)",
+      "CREATE INDEX ix_monitoring_feedback_timestamp ON monitoring_feedback (timestamp)",
+      "CREATE INDEX ix_monitoring_feedback_workspace_session ON monitoring_feedback (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_feedback_workspace_timestamp ON monitoring_feedback (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_llm_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\tmodel_name VARCHAR(255) NOT NULL, \n\tinput_tokens INTEGER NOT NULL, \n\toutput_tokens INTEGER NOT NULL, \n\ttotal_tokens INTEGER NOT NULL, \n\tduration INTEGER NOT NULL, \n\tcost FLOAT, \n\tstatus VARCHAR(50) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\terror_message TEXT, \n\tmessage_id VARCHAR(255), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_llm_calls_bot_id ON monitoring_llm_calls (bot_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_message_id ON monitoring_llm_calls (message_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_pipeline_id ON monitoring_llm_calls (pipeline_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_timestamp ON monitoring_llm_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_llm_calls_workspace_session ON monitoring_llm_calls (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_workspace_timestamp ON monitoring_llm_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_messages (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tmessage_content TEXT NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\tlevel VARCHAR(50) NOT NULL, \n\tplatform VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tuser_name VARCHAR(255), \n\trunner_name VARCHAR(255), \n\tvariables TEXT, \n\trole VARCHAR(50), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_messages_bot_id ON monitoring_messages (bot_id)",
+      "CREATE INDEX ix_monitoring_messages_pipeline_id ON monitoring_messages (pipeline_id)",
+      "CREATE INDEX ix_monitoring_messages_session_id ON monitoring_messages (session_id)",
+      "CREATE INDEX ix_monitoring_messages_timestamp ON monitoring_messages (timestamp)",
+      "CREATE INDEX ix_monitoring_messages_workspace_session ON monitoring_messages (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_messages_workspace_timestamp ON monitoring_messages (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_sessions (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tmessage_count INTEGER NOT NULL, \n\tstart_time DATETIME NOT NULL, \n\tlast_activity DATETIME NOT NULL, \n\tis_active BOOLEAN NOT NULL, \n\tplatform VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tuser_name VARCHAR(255), \n\tPRIMARY KEY (workspace_uuid, bot_id, session_id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_sessions_bot_id ON monitoring_sessions (bot_id)",
+      "CREATE INDEX ix_monitoring_sessions_is_active ON monitoring_sessions (is_active)",
+      "CREATE INDEX ix_monitoring_sessions_last_activity ON monitoring_sessions (last_activity)",
+      "CREATE INDEX ix_monitoring_sessions_pipeline_id ON monitoring_sessions (pipeline_id)",
+      "CREATE INDEX ix_monitoring_sessions_start_time ON monitoring_sessions (start_time)",
+      "CREATE INDEX ix_monitoring_sessions_workspace_active ON monitoring_sessions (workspace_uuid, is_active)",
+      "CREATE INDEX ix_monitoring_sessions_workspace_activity ON monitoring_sessions (workspace_uuid, last_activity)",
+      "\nCREATE TABLE monitoring_tool_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp DATETIME NOT NULL, \n\ttool_name VARCHAR(255) NOT NULL, \n\ttool_source VARCHAR(50) NOT NULL, \n\tduration INTEGER NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\targuments TEXT, \n\tresult TEXT, \n\terror_message TEXT, \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_tool_calls_bot_id ON monitoring_tool_calls (bot_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_message_id ON monitoring_tool_calls (message_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_pipeline_id ON monitoring_tool_calls (pipeline_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_session_id ON monitoring_tool_calls (session_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_timestamp ON monitoring_tool_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_tool_calls_workspace_session ON monitoring_tool_calls (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_workspace_timestamp ON monitoring_tool_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE plugin_settings (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tplugin_author VARCHAR(255) NOT NULL, \n\tplugin_name VARCHAR(255) NOT NULL, \n\tinstallation_uuid VARCHAR(36) NOT NULL, \n\tartifact_digest VARCHAR(64) NOT NULL, \n\truntime_revision INTEGER NOT NULL, \n\tenabled BOOLEAN NOT NULL, \n\tpriority INTEGER NOT NULL, \n\tconfig JSON NOT NULL, \n\tinstall_source VARCHAR(255) NOT NULL, \n\tinstall_info JSON NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (workspace_uuid, plugin_author, plugin_name), \n\tCONSTRAINT uq_plugin_settings_installation_uuid UNIQUE (installation_uuid), \n\tCONSTRAINT ck_plugin_settings_runtime_revision_positive CHECK (runtime_revision >= 1), \n\tCONSTRAINT ck_plugin_settings_artifact_digest_length CHECK (length(artifact_digest) = 64), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_plugin_settings_workspace_enabled ON plugin_settings (workspace_uuid, enabled)",
+      "CREATE UNIQUE INDEX ix_plugin_settings_workspace_installation ON plugin_settings (workspace_uuid, installation_uuid)",
+      "\nCREATE TABLE support_admin_temporary_sessions (\n\tgrant_jti_hash VARCHAR(64) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tactor_account_uuid VARCHAR(36) NOT NULL, \n\tissued_at DATETIME NOT NULL, \n\texpires_at DATETIME NOT NULL, \n\trevoked_at DATETIME, \n\tlast_used_at DATETIME, \n\tPRIMARY KEY (grant_jti_hash), \n\tCONSTRAINT ck_support_admin_sessions_grant_jti_hash CHECK (length(grant_jti_hash) = 64), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_support_admin_sessions_workspace_expiry ON support_admin_temporary_sessions (workspace_uuid, expires_at)",
+      "\nCREATE TABLE webhooks (\n\tid INTEGER NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\turl VARCHAR(1024) NOT NULL, \n\tdescription VARCHAR(512), \n\tenabled BOOLEAN NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_webhooks_workspace_name ON webhooks (workspace_uuid, name)",
+      "\nCREATE TABLE workspace_execution_states (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tactive_generation BIGINT DEFAULT '1' NOT NULL, \n\tstate VARCHAR(32) DEFAULT 'active' NOT NULL, \n\twrite_fenced BOOLEAN DEFAULT 0 NOT NULL, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tdesired_state_revision BIGINT DEFAULT '0' NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (workspace_uuid), \n\tCONSTRAINT ck_workspace_execution_generation CHECK (active_generation > 0), \n\tCONSTRAINT ck_workspace_execution_state CHECK (state IN ('provisioning', 'active', 'migrating', 'draining', 'inactive')), \n\tCONSTRAINT ck_workspace_execution_source CHECK (source IN ('local', 'cloud')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_workspace_execution_states_instance_state ON workspace_execution_states (instance_uuid, state)",
+      "\nCREATE TABLE workspace_invitations (\n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tnormalized_email VARCHAR(320) NOT NULL, \n\trole VARCHAR(32) NOT NULL, \n\ttoken_hash VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'pending' NOT NULL, \n\texpires_at DATETIME NOT NULL, \n\taccepted_at DATETIME, \n\trevoked_at DATETIME, \n\tcreated_by_account_uuid VARCHAR(36) NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT ck_workspace_invitations_role CHECK (role IN ('admin', 'developer', 'operator', 'viewer')), \n\tCONSTRAINT ck_workspace_invitations_status CHECK (status IN ('pending', 'accepted', 'revoked', 'expired')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE UNIQUE INDEX uq_workspace_invitations_pending_email ON workspace_invitations (workspace_uuid, normalized_email) WHERE status = 'pending'",
+      "CREATE UNIQUE INDEX uq_workspace_invitations_token_hash ON workspace_invitations (token_hash)",
+      "\nCREATE TABLE workspace_memberships (\n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\taccount_uuid VARCHAR(36) NOT NULL, \n\trole VARCHAR(32) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tinvited_by_account_uuid VARCHAR(36), \n\tjoined_at DATETIME, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_workspace_membership_account UNIQUE (workspace_uuid, account_uuid), \n\tCONSTRAINT ck_workspace_memberships_role CHECK (role IN ('owner', 'admin', 'developer', 'operator', 'viewer')), \n\tCONSTRAINT ck_workspace_memberships_status CHECK (status IN ('active', 'disabled', 'removed')), \n\tCONSTRAINT ck_workspace_memberships_source CHECK (source IN ('local', 'cloud_projection')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(account_uuid) REFERENCES users (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(invited_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_workspace_memberships_account_status ON workspace_memberships (account_uuid, status)",
+      "CREATE UNIQUE INDEX uq_workspace_memberships_one_active_owner ON workspace_memberships (workspace_uuid) WHERE role = 'owner' AND status = 'active'",
+      "\nCREATE TABLE workspace_metadata (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\t\"key\" VARCHAR(255) NOT NULL, \n\tvalue VARCHAR(255), \n\tPRIMARY KEY (workspace_uuid, \"key\"), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "\nCREATE TABLE bot_admins (\n\tid INTEGER NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tbot_uuid VARCHAR(255) NOT NULL, \n\tlauncher_type VARCHAR(64) NOT NULL, \n\tlauncher_id VARCHAR(255) NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT uq_bot_admin UNIQUE (workspace_uuid, bot_uuid, launcher_type, launcher_id), \n\tCONSTRAINT fk_bot_admins_workspace_bot FOREIGN KEY(workspace_uuid, bot_uuid) REFERENCES bots (workspace_uuid, uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "\nCREATE TABLE codex_credentials (\n\tprovider_uuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tpayload JSON NOT NULL, \n\tversion INTEGER NOT NULL, \n\tlease_owner VARCHAR(64), \n\tlease_until FLOAT NOT NULL, \n\tPRIMARY KEY (provider_uuid), \n\tCONSTRAINT fk_codex_credentials_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_codex_credentials_workspace ON codex_credentials (workspace_uuid)",
+      "\nCREATE TABLE embedding_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_embedding_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_embedding_models_workspace_name ON embedding_models (workspace_uuid, name)",
+      "CREATE INDEX ix_embedding_models_workspace_provider ON embedding_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE knowledge_base_files (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tkb_id VARCHAR(255), \n\tfile_name VARCHAR, \n\textension VARCHAR, \n\tcreated_at DATETIME, \n\tstatus VARCHAR, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_knowledge_base_files_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tCONSTRAINT fk_knowledge_base_files_workspace_kb FOREIGN KEY(workspace_uuid, kb_id) REFERENCES knowledge_bases (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_base_files_workspace_kb ON knowledge_base_files (workspace_uuid, kb_id)",
+      "\nCREATE TABLE llm_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\tabilities JSON NOT NULL, \n\tcontext_length INTEGER, \n\treasoning_config JSON DEFAULT '{\"level\":\"provider_default\"}' NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_llm_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_llm_models_workspace_name ON llm_models (workspace_uuid, name)",
+      "CREATE INDEX ix_llm_models_workspace_provider ON llm_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE pipeline_run_records (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tpipeline_uuid VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(255) NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tstarted_at DATETIME NOT NULL, \n\tfinished_at DATETIME NOT NULL, \n\tresult JSON NOT NULL, \n\tknowledge_base_uuid VARCHAR(255), \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_pipeline_run_records_workspace_pipeline FOREIGN KEY(workspace_uuid, pipeline_uuid) REFERENCES legacy_pipelines (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_pipeline_run_records_workspace_created ON pipeline_run_records (workspace_uuid, created_at)",
+      "CREATE INDEX ix_pipeline_run_records_workspace_pipeline ON pipeline_run_records (workspace_uuid, pipeline_uuid)",
+      "\nCREATE TABLE rerank_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tupdated_at DATETIME DEFAULT CURRENT_TIMESTAMP NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_rerank_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_rerank_models_workspace_name ON rerank_models (workspace_uuid, name)",
+      "CREATE INDEX ix_rerank_models_workspace_provider ON rerank_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE knowledge_base_chunks (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tfile_id VARCHAR(255), \n\ttext TEXT, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_knowledge_base_chunks_workspace_file FOREIGN KEY(workspace_uuid, file_id) REFERENCES knowledge_base_files (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_base_chunks_workspace_file ON knowledge_base_chunks (workspace_uuid, file_id)"
+    ],
+    "postgresql": [
+      "\nCREATE TABLE directory_projection_inbox (\n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tevent_uuid VARCHAR(36) NOT NULL, \n\tcursor BIGINT NOT NULL, \n\tevent_type VARCHAR(128) NOT NULL, \n\trevision BIGINT NOT NULL, \n\tfingerprint TEXT NOT NULL, \n\treceived_at TIMESTAMP WITH TIME ZONE DEFAULT now() NOT NULL, \n\tapplied_at TIMESTAMP WITH TIME ZONE, \n\tPRIMARY KEY (instance_uuid, event_uuid), \n\tCONSTRAINT uq_directory_projection_inbox_cursor UNIQUE (instance_uuid, cursor), \n\tCONSTRAINT ck_directory_projection_inbox_cursor CHECK (cursor > 0), \n\tCONSTRAINT ck_directory_projection_inbox_revision CHECK (revision > 0), \n\tCONSTRAINT ck_directory_projection_inbox_fingerprint CHECK (length(fingerprint) = 64)\n)\n\n",
+      "CREATE INDEX ix_directory_projection_inbox_pending ON directory_projection_inbox (instance_uuid, applied_at, cursor)",
+      "\nCREATE TABLE directory_projection_states (\n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tcursor BIGINT DEFAULT '0' NOT NULL, \n\tsnapshot_coverage_cursor BIGINT DEFAULT '0' NOT NULL, \n\tsnapshot_fingerprint TEXT NOT NULL, \n\tlast_applied_at TIMESTAMP WITH TIME ZONE NOT NULL, \n\tlease_expires_at TIMESTAMP WITH TIME ZONE, \n\tPRIMARY KEY (instance_uuid), \n\tCONSTRAINT ck_directory_projection_state_cursor CHECK (cursor >= 0), \n\tCONSTRAINT ck_directory_projection_state_snapshot_coverage CHECK (snapshot_coverage_cursor >= 0 AND snapshot_coverage_cursor <= cursor), \n\tCONSTRAINT ck_directory_projection_state_fingerprint CHECK (length(snapshot_fingerprint) = 64)\n)\n\n",
+      "\nCREATE TABLE metadata (\n\tkey VARCHAR(255) NOT NULL, \n\tvalue VARCHAR(255), \n\tPRIMARY KEY (key)\n)\n\n",
+      "\nCREATE TABLE users (\n\tid SERIAL NOT NULL, \n\tuuid VARCHAR(36) NOT NULL, \n\t\"user\" VARCHAR(255) NOT NULL, \n\tnormalized_email VARCHAR(320) NOT NULL, \n\tpassword VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\taccount_type VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tspace_account_uuid VARCHAR(255), \n\tspace_access_token TEXT, \n\tspace_refresh_token TEXT, \n\tspace_access_token_expires_at TIMESTAMP WITHOUT TIME ZONE, \n\tspace_api_key VARCHAR(255), \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT ck_users_normalized_email CHECK (normalized_email = trim(normalized_email) AND length(normalized_email) > 0 AND length(normalized_email) <= 320), \n\tCONSTRAINT ck_users_status CHECK (status IN ('active', 'disabled', 'deleted')), \n\tCONSTRAINT ck_users_source CHECK (source IN ('local', 'cloud_projection'))\n)\n\n",
+      "CREATE UNIQUE INDEX uq_users_normalized_email ON users (normalized_email)",
+      "CREATE UNIQUE INDEX uq_users_uuid ON users (uuid)",
+      "\nCREATE TABLE workspaces (\n\tuuid VARCHAR(36) NOT NULL, \n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tslug VARCHAR(255) NOT NULL, \n\ttype VARCHAR(32) DEFAULT 'team' NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tcreated_by_account_uuid VARCHAR(36), \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_workspaces_instance_slug UNIQUE (instance_uuid, slug), \n\tCONSTRAINT ck_workspaces_type CHECK (type IN ('personal', 'team')), \n\tCONSTRAINT ck_workspaces_status CHECK (status IN ('provisioning', 'active', 'suspended', 'archived', 'deleted')), \n\tCONSTRAINT ck_workspaces_source CHECK (source IN ('local', 'cloud_projection')), \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_workspaces_instance_status ON workspaces (instance_uuid, status)",
+      "CREATE UNIQUE INDEX uq_workspaces_local_instance ON workspaces (instance_uuid) WHERE source = 'local'",
+      "\nCREATE TABLE api_keys (\n\tid SERIAL NOT NULL, \n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tcreated_by_account_uuid VARCHAR(36), \n\tname VARCHAR(255) NOT NULL, \n\tkey_hash VARCHAR(64) NOT NULL, \n\tscopes JSON DEFAULT '[]' NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\texpires_at TIMESTAMP WITHOUT TIME ZONE, \n\tlast_used_at TIMESTAMP WITHOUT TIME ZONE, \n\tdescription VARCHAR(512), \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT ck_api_keys_status CHECK (status IN ('active', 'revoked')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_api_keys_workspace_name ON api_keys (workspace_uuid, name)",
+      "CREATE INDEX ix_api_keys_workspace_status ON api_keys (workspace_uuid, status)",
+      "CREATE UNIQUE INDEX uq_api_keys_key_hash ON api_keys (key_hash)",
+      "CREATE UNIQUE INDEX uq_api_keys_uuid ON api_keys (uuid)",
+      "\nCREATE TABLE binary_storages (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tunique_key VARCHAR(255) NOT NULL, \n\tkey VARCHAR(255) NOT NULL, \n\towner_type VARCHAR(255) NOT NULL, \n\towner VARCHAR(255) NOT NULL, \n\tvalue BYTEA NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (workspace_uuid, unique_key), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_binary_storages_workspace_owner ON binary_storages (workspace_uuid, owner_type, owner)",
+      "\nCREATE TABLE bots (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tdescription VARCHAR(255) NOT NULL, \n\tadapter VARCHAR(255) NOT NULL, \n\tadapter_config JSON NOT NULL, \n\tenable BOOLEAN NOT NULL, \n\tuse_pipeline_name VARCHAR(255), \n\tuse_pipeline_uuid VARCHAR(255), \n\tpipeline_routing_rules JSON DEFAULT '[]' NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_bots_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_bots_workspace_name ON bots (workspace_uuid, name)",
+      "\nCREATE TABLE knowledge_bases (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR, \n\tdescription TEXT, \n\temoji VARCHAR(10), \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE, \n\tknowledge_engine_plugin_id VARCHAR, \n\tcollection_id VARCHAR, \n\tlegacy_vector_collection BOOLEAN DEFAULT false NOT NULL, \n\tcreation_settings JSON, \n\tretrieval_settings JSON, \n\tembedding_dimension INTEGER, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_knowledge_bases_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tCONSTRAINT ck_knowledge_bases_embedding_dimension_positive CHECK (embedding_dimension IS NULL OR embedding_dimension > 0), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_bases_name ON knowledge_bases (name)",
+      "CREATE INDEX ix_knowledge_bases_workspace_name ON knowledge_bases (workspace_uuid, name)",
+      "CREATE UNIQUE INDEX uq_knowledge_bases_workspace_collection ON knowledge_bases (workspace_uuid, collection_id) WHERE collection_id IS NOT NULL",
+      "\nCREATE TABLE legacy_pipelines (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tdescription VARCHAR(255) NOT NULL, \n\temoji VARCHAR(10), \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tfor_version VARCHAR(255) NOT NULL, \n\tis_default BOOLEAN NOT NULL, \n\tstages JSON NOT NULL, \n\tconfig JSON NOT NULL, \n\textensions_preferences JSON NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_legacy_pipelines_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_legacy_pipelines_workspace_default ON legacy_pipelines (workspace_uuid, is_default)",
+      "CREATE INDEX ix_legacy_pipelines_workspace_name ON legacy_pipelines (workspace_uuid, name)",
+      "\nCREATE TABLE mcp_servers (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tenable BOOLEAN NOT NULL, \n\tmode VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\treadme TEXT DEFAULT '' NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_mcp_servers_workspace_name UNIQUE (workspace_uuid, name), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_mcp_servers_workspace_enable ON mcp_servers (workspace_uuid, enable)",
+      "\nCREATE TABLE model_providers (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\trequester VARCHAR(255) NOT NULL, \n\tbase_url VARCHAR(512) NOT NULL, \n\tapi_keys JSON NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_model_providers_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_model_providers_workspace_name ON model_providers (workspace_uuid, name)",
+      "CREATE INDEX ix_model_providers_workspace_requester ON model_providers (workspace_uuid, requester)",
+      "\nCREATE TABLE monitoring_embedding_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tmodel_name VARCHAR(255) NOT NULL, \n\tprompt_tokens INTEGER NOT NULL, \n\ttotal_tokens INTEGER NOT NULL, \n\tduration INTEGER NOT NULL, \n\tinput_count INTEGER NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\terror_message TEXT, \n\tknowledge_base_id VARCHAR(255), \n\tquery_text TEXT, \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\tcall_type VARCHAR(50), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_embedding_calls_knowledge_base_id ON monitoring_embedding_calls (knowledge_base_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_message_id ON monitoring_embedding_calls (message_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_session_id ON monitoring_embedding_calls (session_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_timestamp ON monitoring_embedding_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_embedding_calls_workspace_kb ON monitoring_embedding_calls (workspace_uuid, knowledge_base_id)",
+      "CREATE INDEX ix_monitoring_embedding_calls_workspace_timestamp ON monitoring_embedding_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_errors (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\terror_type VARCHAR(255) NOT NULL, \n\terror_message TEXT NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255), \n\tstack_trace TEXT, \n\tmessage_id VARCHAR(255), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_errors_bot_id ON monitoring_errors (bot_id)",
+      "CREATE INDEX ix_monitoring_errors_message_id ON monitoring_errors (message_id)",
+      "CREATE INDEX ix_monitoring_errors_pipeline_id ON monitoring_errors (pipeline_id)",
+      "CREATE INDEX ix_monitoring_errors_timestamp ON monitoring_errors (timestamp)",
+      "CREATE INDEX ix_monitoring_errors_workspace_session ON monitoring_errors (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_errors_workspace_timestamp ON monitoring_errors (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_feedback (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tfeedback_id VARCHAR(255) NOT NULL, \n\tfeedback_type INTEGER NOT NULL, \n\tfeedback_content TEXT, \n\tinaccurate_reasons TEXT, \n\tbot_id VARCHAR(255), \n\tbot_name VARCHAR(255), \n\tpipeline_id VARCHAR(255), \n\tpipeline_name VARCHAR(255), \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\tstream_id VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tplatform VARCHAR(255), \n\tPRIMARY KEY (id), \n\tCONSTRAINT uq_monitoring_feedback_workspace_feedback_id UNIQUE (workspace_uuid, feedback_id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_feedback_bot_id ON monitoring_feedback (bot_id)",
+      "CREATE INDEX ix_monitoring_feedback_feedback_id ON monitoring_feedback (feedback_id)",
+      "CREATE INDEX ix_monitoring_feedback_message_id ON monitoring_feedback (message_id)",
+      "CREATE INDEX ix_monitoring_feedback_pipeline_id ON monitoring_feedback (pipeline_id)",
+      "CREATE INDEX ix_monitoring_feedback_session_id ON monitoring_feedback (session_id)",
+      "CREATE INDEX ix_monitoring_feedback_stream_id ON monitoring_feedback (stream_id)",
+      "CREATE INDEX ix_monitoring_feedback_timestamp ON monitoring_feedback (timestamp)",
+      "CREATE INDEX ix_monitoring_feedback_workspace_session ON monitoring_feedback (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_feedback_workspace_timestamp ON monitoring_feedback (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_llm_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tmodel_name VARCHAR(255) NOT NULL, \n\tinput_tokens INTEGER NOT NULL, \n\toutput_tokens INTEGER NOT NULL, \n\ttotal_tokens INTEGER NOT NULL, \n\tduration INTEGER NOT NULL, \n\tcost FLOAT, \n\tstatus VARCHAR(50) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\terror_message TEXT, \n\tmessage_id VARCHAR(255), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_llm_calls_bot_id ON monitoring_llm_calls (bot_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_message_id ON monitoring_llm_calls (message_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_pipeline_id ON monitoring_llm_calls (pipeline_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_timestamp ON monitoring_llm_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_llm_calls_workspace_session ON monitoring_llm_calls (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_llm_calls_workspace_timestamp ON monitoring_llm_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_messages (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tmessage_content TEXT NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\tlevel VARCHAR(50) NOT NULL, \n\tplatform VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tuser_name VARCHAR(255), \n\trunner_name VARCHAR(255), \n\tvariables TEXT, \n\trole VARCHAR(50), \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_messages_bot_id ON monitoring_messages (bot_id)",
+      "CREATE INDEX ix_monitoring_messages_pipeline_id ON monitoring_messages (pipeline_id)",
+      "CREATE INDEX ix_monitoring_messages_session_id ON monitoring_messages (session_id)",
+      "CREATE INDEX ix_monitoring_messages_timestamp ON monitoring_messages (timestamp)",
+      "CREATE INDEX ix_monitoring_messages_workspace_session ON monitoring_messages (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_messages_workspace_timestamp ON monitoring_messages (workspace_uuid, timestamp)",
+      "\nCREATE TABLE monitoring_sessions (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tmessage_count INTEGER NOT NULL, \n\tstart_time TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tlast_activity TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tis_active BOOLEAN NOT NULL, \n\tplatform VARCHAR(255), \n\tuser_id VARCHAR(255), \n\tuser_name VARCHAR(255), \n\tPRIMARY KEY (workspace_uuid, bot_id, session_id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_sessions_bot_id ON monitoring_sessions (bot_id)",
+      "CREATE INDEX ix_monitoring_sessions_is_active ON monitoring_sessions (is_active)",
+      "CREATE INDEX ix_monitoring_sessions_last_activity ON monitoring_sessions (last_activity)",
+      "CREATE INDEX ix_monitoring_sessions_pipeline_id ON monitoring_sessions (pipeline_id)",
+      "CREATE INDEX ix_monitoring_sessions_start_time ON monitoring_sessions (start_time)",
+      "CREATE INDEX ix_monitoring_sessions_workspace_active ON monitoring_sessions (workspace_uuid, is_active)",
+      "CREATE INDEX ix_monitoring_sessions_workspace_activity ON monitoring_sessions (workspace_uuid, last_activity)",
+      "\nCREATE TABLE monitoring_tool_calls (\n\tid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\ttimestamp TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\ttool_name VARCHAR(255) NOT NULL, \n\ttool_source VARCHAR(50) NOT NULL, \n\tduration INTEGER NOT NULL, \n\tstatus VARCHAR(50) NOT NULL, \n\tbot_id VARCHAR(255) NOT NULL, \n\tbot_name VARCHAR(255) NOT NULL, \n\tpipeline_id VARCHAR(255) NOT NULL, \n\tpipeline_name VARCHAR(255) NOT NULL, \n\tsession_id VARCHAR(255), \n\tmessage_id VARCHAR(255), \n\targuments TEXT, \n\tresult TEXT, \n\terror_message TEXT, \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_monitoring_tool_calls_bot_id ON monitoring_tool_calls (bot_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_message_id ON monitoring_tool_calls (message_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_pipeline_id ON monitoring_tool_calls (pipeline_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_session_id ON monitoring_tool_calls (session_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_timestamp ON monitoring_tool_calls (timestamp)",
+      "CREATE INDEX ix_monitoring_tool_calls_workspace_session ON monitoring_tool_calls (workspace_uuid, session_id)",
+      "CREATE INDEX ix_monitoring_tool_calls_workspace_timestamp ON monitoring_tool_calls (workspace_uuid, timestamp)",
+      "\nCREATE TABLE plugin_settings (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tplugin_author VARCHAR(255) NOT NULL, \n\tplugin_name VARCHAR(255) NOT NULL, \n\tinstallation_uuid VARCHAR(36) NOT NULL, \n\tartifact_digest VARCHAR(64) NOT NULL, \n\truntime_revision INTEGER NOT NULL, \n\tenabled BOOLEAN NOT NULL, \n\tpriority INTEGER NOT NULL, \n\tconfig JSON NOT NULL, \n\tinstall_source VARCHAR(255) NOT NULL, \n\tinstall_info JSON NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (workspace_uuid, plugin_author, plugin_name), \n\tCONSTRAINT uq_plugin_settings_installation_uuid UNIQUE (installation_uuid), \n\tCONSTRAINT ck_plugin_settings_runtime_revision_positive CHECK (runtime_revision >= 1), \n\tCONSTRAINT ck_plugin_settings_artifact_digest_length CHECK (length(artifact_digest) = 64), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_plugin_settings_workspace_enabled ON plugin_settings (workspace_uuid, enabled)",
+      "CREATE UNIQUE INDEX ix_plugin_settings_workspace_installation ON plugin_settings (workspace_uuid, installation_uuid)",
+      "\nCREATE TABLE support_admin_temporary_sessions (\n\tgrant_jti_hash VARCHAR(64) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tactor_account_uuid VARCHAR(36) NOT NULL, \n\tissued_at TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\texpires_at TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\trevoked_at TIMESTAMP WITHOUT TIME ZONE, \n\tlast_used_at TIMESTAMP WITHOUT TIME ZONE, \n\tPRIMARY KEY (grant_jti_hash), \n\tCONSTRAINT ck_support_admin_sessions_grant_jti_hash CHECK (length(grant_jti_hash) = 64), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_support_admin_sessions_workspace_expiry ON support_admin_temporary_sessions (workspace_uuid, expires_at)",
+      "\nCREATE TABLE webhooks (\n\tid SERIAL NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\turl VARCHAR(1024) NOT NULL, \n\tdescription VARCHAR(512), \n\tenabled BOOLEAN NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (id), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_webhooks_workspace_name ON webhooks (workspace_uuid, name)",
+      "\nCREATE TABLE workspace_execution_states (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tinstance_uuid VARCHAR(255) NOT NULL, \n\tactive_generation BIGINT DEFAULT '1' NOT NULL, \n\tstate VARCHAR(32) DEFAULT 'active' NOT NULL, \n\twrite_fenced BOOLEAN DEFAULT false NOT NULL, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tdesired_state_revision BIGINT DEFAULT '0' NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (workspace_uuid), \n\tCONSTRAINT ck_workspace_execution_generation CHECK (active_generation > 0), \n\tCONSTRAINT ck_workspace_execution_state CHECK (state IN ('provisioning', 'active', 'migrating', 'draining', 'inactive')), \n\tCONSTRAINT ck_workspace_execution_source CHECK (source IN ('local', 'cloud')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_workspace_execution_states_instance_state ON workspace_execution_states (instance_uuid, state)",
+      "\nCREATE TABLE workspace_invitations (\n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tnormalized_email VARCHAR(320) NOT NULL, \n\trole VARCHAR(32) NOT NULL, \n\ttoken_hash VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'pending' NOT NULL, \n\texpires_at TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\taccepted_at TIMESTAMP WITHOUT TIME ZONE, \n\trevoked_at TIMESTAMP WITHOUT TIME ZONE, \n\tcreated_by_account_uuid VARCHAR(36) NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT ck_workspace_invitations_role CHECK (role IN ('admin', 'developer', 'operator', 'viewer')), \n\tCONSTRAINT ck_workspace_invitations_status CHECK (status IN ('pending', 'accepted', 'revoked', 'expired')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(created_by_account_uuid) REFERENCES users (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE UNIQUE INDEX uq_workspace_invitations_pending_email ON workspace_invitations (workspace_uuid, normalized_email) WHERE status = 'pending'",
+      "CREATE UNIQUE INDEX uq_workspace_invitations_token_hash ON workspace_invitations (token_hash)",
+      "\nCREATE TABLE workspace_memberships (\n\tuuid VARCHAR(36) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\taccount_uuid VARCHAR(36) NOT NULL, \n\trole VARCHAR(32) NOT NULL, \n\tstatus VARCHAR(32) DEFAULT 'active' NOT NULL, \n\tinvited_by_account_uuid VARCHAR(36), \n\tjoined_at TIMESTAMP WITHOUT TIME ZONE, \n\tsource VARCHAR(32) DEFAULT 'local' NOT NULL, \n\tprojection_revision BIGINT DEFAULT '0' NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_workspace_membership_account UNIQUE (workspace_uuid, account_uuid), \n\tCONSTRAINT ck_workspace_memberships_role CHECK (role IN ('owner', 'admin', 'developer', 'operator', 'viewer')), \n\tCONSTRAINT ck_workspace_memberships_status CHECK (status IN ('active', 'disabled', 'removed')), \n\tCONSTRAINT ck_workspace_memberships_source CHECK (source IN ('local', 'cloud_projection')), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(account_uuid) REFERENCES users (uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(invited_by_account_uuid) REFERENCES users (uuid) ON DELETE SET NULL\n)\n\n",
+      "CREATE INDEX ix_workspace_memberships_account_status ON workspace_memberships (account_uuid, status)",
+      "CREATE UNIQUE INDEX uq_workspace_memberships_one_active_owner ON workspace_memberships (workspace_uuid) WHERE role = 'owner' AND status = 'active'",
+      "\nCREATE TABLE workspace_metadata (\n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tkey VARCHAR(255) NOT NULL, \n\tvalue VARCHAR(255), \n\tPRIMARY KEY (workspace_uuid, key), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "\nCREATE TABLE bot_admins (\n\tid SERIAL NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tbot_uuid VARCHAR(255) NOT NULL, \n\tlauncher_type VARCHAR(64) NOT NULL, \n\tlauncher_id VARCHAR(255) NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (id), \n\tCONSTRAINT uq_bot_admin UNIQUE (workspace_uuid, bot_uuid, launcher_type, launcher_id), \n\tCONSTRAINT fk_bot_admins_workspace_bot FOREIGN KEY(workspace_uuid, bot_uuid) REFERENCES bots (workspace_uuid, uuid) ON DELETE CASCADE, \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "\nCREATE TABLE codex_credentials (\n\tprovider_uuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tpayload JSON NOT NULL, \n\tversion INTEGER NOT NULL, \n\tlease_owner VARCHAR(64), \n\tlease_until FLOAT NOT NULL, \n\tPRIMARY KEY (provider_uuid), \n\tCONSTRAINT fk_codex_credentials_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_codex_credentials_workspace ON codex_credentials (workspace_uuid)",
+      "\nCREATE TABLE embedding_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_embedding_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_embedding_models_workspace_name ON embedding_models (workspace_uuid, name)",
+      "CREATE INDEX ix_embedding_models_workspace_provider ON embedding_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE knowledge_base_files (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tkb_id VARCHAR(255), \n\tfile_name VARCHAR, \n\textension VARCHAR, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE, \n\tstatus VARCHAR, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT uq_knowledge_base_files_workspace_uuid UNIQUE (workspace_uuid, uuid), \n\tCONSTRAINT fk_knowledge_base_files_workspace_kb FOREIGN KEY(workspace_uuid, kb_id) REFERENCES knowledge_bases (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_base_files_workspace_kb ON knowledge_base_files (workspace_uuid, kb_id)",
+      "\nCREATE TABLE llm_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\tabilities JSON NOT NULL, \n\tcontext_length INTEGER, \n\treasoning_config JSON DEFAULT '{\"level\":\"provider_default\"}' NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_llm_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_llm_models_workspace_name ON llm_models (workspace_uuid, name)",
+      "CREATE INDEX ix_llm_models_workspace_provider ON llm_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE pipeline_run_records (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tpipeline_uuid VARCHAR(255) NOT NULL, \n\tstatus VARCHAR(255) NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tstarted_at TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tfinished_at TIMESTAMP WITHOUT TIME ZONE NOT NULL, \n\tresult JSON NOT NULL, \n\tknowledge_base_uuid VARCHAR(255), \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_pipeline_run_records_workspace_pipeline FOREIGN KEY(workspace_uuid, pipeline_uuid) REFERENCES legacy_pipelines (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_pipeline_run_records_workspace_created ON pipeline_run_records (workspace_uuid, created_at)",
+      "CREATE INDEX ix_pipeline_run_records_workspace_pipeline ON pipeline_run_records (workspace_uuid, pipeline_uuid)",
+      "\nCREATE TABLE rerank_models (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tname VARCHAR(255) NOT NULL, \n\tprovider_uuid VARCHAR(255) NOT NULL, \n\textra_args JSON NOT NULL, \n\tprefered_ranking INTEGER NOT NULL, \n\tcreated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tupdated_at TIMESTAMP WITHOUT TIME ZONE DEFAULT now() NOT NULL, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_rerank_models_workspace_provider FOREIGN KEY(workspace_uuid, provider_uuid) REFERENCES model_providers (workspace_uuid, uuid), \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_rerank_models_workspace_name ON rerank_models (workspace_uuid, name)",
+      "CREATE INDEX ix_rerank_models_workspace_provider ON rerank_models (workspace_uuid, provider_uuid)",
+      "\nCREATE TABLE knowledge_base_chunks (\n\tuuid VARCHAR(255) NOT NULL, \n\tworkspace_uuid VARCHAR(36) NOT NULL, \n\tfile_id VARCHAR(255), \n\ttext TEXT, \n\tPRIMARY KEY (uuid), \n\tCONSTRAINT fk_knowledge_base_chunks_workspace_file FOREIGN KEY(workspace_uuid, file_id) REFERENCES knowledge_base_files (workspace_uuid, uuid) ON DELETE CASCADE, \n\tUNIQUE (uuid), \n\tFOREIGN KEY(workspace_uuid) REFERENCES workspaces (uuid) ON DELETE CASCADE\n)\n\n",
+      "CREATE INDEX ix_knowledge_base_chunks_workspace_file ON knowledge_base_chunks (workspace_uuid, file_id)"
+    ]
+  }
+}
diff --git a/tests/integration/persistence/test_migration_branch_convergence.py b/tests/integration/persistence/test_migration_branch_convergence.py
index 759b46b03..6020c9b37 100644
--- a/tests/integration/persistence/test_migration_branch_convergence.py
+++ b/tests/integration/persistence/test_migration_branch_convergence.py
@@ -370,7 +370,7 @@ def _legacy_shared_certification(**overrides):
 
 
 @pytest.mark.asyncio
-async def test_certification_artifact_digest_backfill_is_safe_and_enables_shared_placement(convergence_engine):
+async def test_certification_artifact_digest_backfill_is_safe_and_keeps_legacy_dedicated(convergence_engine):
     engine = convergence_engine
     async with engine.begin() as conn:
         await conn.run_sync(Base.metadata.create_all)
@@ -461,7 +461,7 @@ async def test_certification_artifact_digest_backfill_is_safe_and_enables_shared
                 artifact_digest=eligible_digest,
                 install_info=eligible_info,
             )
-            is PluginExecutionMode.SHARED_CERTIFIED
+            is PluginExecutionMode.DEDICATED
         )
 
     for name, (original_certification, artifact_digest) in rows.items():
diff --git a/tests/integration/persistence/test_migrations_postgres.py b/tests/integration/persistence/test_migrations_postgres.py
index babe2c207..1c0eb95c7 100644
--- a/tests/integration/persistence/test_migrations_postgres.py
+++ b/tests/integration/persistence/test_migrations_postgres.py
@@ -929,7 +929,7 @@ class TestPostgreSQLTenantRuntime:
                                 description='',
                                 adapter='capacity-probe',
                                 adapter_config={},
-                                enable=False,
+                                enable=True,
                                 event_bindings=[],
                             ),
                             persistence_pipeline.LegacyPipeline(
diff --git a/tests/integration/persistence/test_monitoring_postgres.py b/tests/integration/persistence/test_monitoring_postgres.py
index ebfee7a7e..d46ff2107 100644
--- a/tests/integration/persistence/test_monitoring_postgres.py
+++ b/tests/integration/persistence/test_monitoring_postgres.py
@@ -7,6 +7,7 @@ no existing tables are dropped. Without that URL these tests are skipped.
 
 from __future__ import annotations
 
+import datetime
 import logging
 import os
 import uuid
@@ -19,6 +20,7 @@ from sqlalchemy.ext.asyncio import create_async_engine
 
 from langbot.pkg.api.http.context import ExecutionContext
 from langbot.pkg.api.http.service.monitoring import MonitoringService
+from langbot.pkg.entity.persistence.agent_run import AgentRun
 from langbot.pkg.entity.persistence import monitoring as models
 from langbot.pkg.entity.persistence.workspace import Workspace
 from langbot.pkg.persistence.mgr import PersistenceManager, PersistenceMode
@@ -158,6 +160,56 @@ async def _read(service, method, context, *args, **kwargs):
         return await getattr(service, method)(context, *args, **kwargs)
 
 
+async def test_inflight_snapshot_reads_recent_utc_rows_and_terminal_transitions(service, cloud_database):
+    _, admin = cloud_database
+    now = datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None)
+    old = now - datetime.timedelta(hours=1)
+    async with admin.begin() as conn:
+        for workspace, run_id, status, created_at in [
+            (WORKSPACE_A, 'inflight-running', 'running', old),
+            (WORKSPACE_A, 'inflight-recent', 'completed', now),
+            (WORKSPACE_B, 'inflight-private', 'running', now),
+        ]:
+            await conn.execute(
+                sa.insert(AgentRun).values(
+                    run_id=run_id, workspace_id=workspace, runner_id='runner', status=status, created_at=created_at
+                )
+            )
+        for message_id, timestamp in [('inflight-message', now), ('inflight-old-message', old)]:
+            await conn.execute(
+                sa.insert(models.MonitoringMessage).values(
+                    id=message_id,
+                    workspace_uuid=WORKSPACE_A,
+                    timestamp=timestamp,
+                    **RESOURCE,
+                    message_content='hello',
+                    session_id='session',
+                    status='success',
+                    level='info',
+                )
+            )
+
+    try:
+        # The SSE producer owns its tenant scope after the HTTP handler has returned.
+        snapshot = await service.get_inflight_snapshot(_context(WORKSPACE_A))
+        assert snapshot['active_total'] == 1
+        assert {row['id'] for row in snapshot['items']} == {'inflight-running', 'inflight-recent', 'inflight-message'}
+        async with admin.begin() as conn:
+            await conn.execute(
+                sa.update(AgentRun).where(AgentRun.run_id == 'inflight-running').values(status='completed')
+            )
+        snapshot = await service.get_inflight_snapshot(_context(WORKSPACE_A), snapshot['items'])
+        assert snapshot['active_total'] == 0
+        assert next(row for row in snapshot['items'] if row['id'] == 'inflight-running')['status_group'] == 'completed'
+    finally:
+        async with admin.begin() as conn:
+            await conn.execute(
+                sa.delete(AgentRun).where(
+                    AgentRun.run_id.in_(['inflight-running', 'inflight-recent', 'inflight-private'])
+                )
+            )
+
+
 def _query(context, sender_id):
     return SimpleNamespace(
         _execution_context=context,
diff --git a/tests/integration/persistence/test_release_41011_upgrade.py b/tests/integration/persistence/test_release_41011_upgrade.py
new file mode 100644
index 000000000..ad73be4e6
--- /dev/null
+++ b/tests/integration/persistence/test_release_41011_upgrade.py
@@ -0,0 +1,130 @@
+"""Upgrade the immutable v4.10.11 ORM schema, not current metadata."""
+
+import json
+
+import pytest
+import sqlalchemy as sa
+from alembic.autogenerate import compare_metadata
+from alembic.migration import MigrationContext
+
+from langbot.pkg.entity.persistence.base import Base
+from langbot.pkg.persistence.alembic_runner import run_alembic_upgrade
+from tests.integration.persistence.test_migration_branch_convergence import (
+    FIXTURES,
+    _seed_common,
+)
+
+
+@pytest.fixture(params=['sqlite', 'postgresql'])
+async def convergence_engine(request, tmp_path):
+    import os
+    import uuid
+    from sqlalchemy.ext.asyncio import create_async_engine
+
+    if request.param == 'sqlite':
+        engine = create_async_engine(f'sqlite+aiosqlite:///{tmp_path / "release.db"}')
+        try:
+            yield engine
+        finally:
+            await engine.dispose()
+        return
+    url = os.environ.get('TEST_POSTGRES_URL')
+    if not url:
+        pytest.skip('TEST_POSTGRES_URL required for disposable PostgreSQL')
+    name = 'release_audit_' + uuid.uuid4().hex
+    admin = create_async_engine(url, isolation_level='AUTOCOMMIT')
+    async with admin.connect() as conn:
+        await conn.exec_driver_sql(f'CREATE DATABASE {name}')
+    engine = create_async_engine(sa.engine.make_url(url).set(database=name))
+    try:
+        yield engine
+    finally:
+        await engine.dispose()
+        async with admin.connect() as conn:
+            await conn.exec_driver_sql(f'DROP DATABASE {name}')
+        await admin.dispose()
+
+
+@pytest.mark.asyncio
+async def test_release_41011_upgrade(convergence_engine, monkeypatch):
+    from langbot.pkg.utils import constants
+
+    monkeypatch.setattr(constants, 'instance_id', 'migration-convergence')
+    engine = convergence_engine
+    fixture = json.loads((FIXTURES / 'release_4_10_11_schema.json').read_text())
+    async with engine.begin() as conn:
+        for statement in fixture['dialects'][engine.dialect.name]:
+            await conn.exec_driver_sql(statement)
+    # Apply the historical migrations too: pgvector is migration-owned,
+    # not represented in ORM metadata.
+    await run_alembic_upgrade(engine, fixture['revision'])
+    await _seed_common(engine)
+    async with engine.begin() as conn:
+        await conn.execute(sa.text("DELETE FROM metadata WHERE key='oss_workspace_uuid'"))
+        await conn.execute(
+            sa.text(
+                "INSERT INTO metadata (key,value) VALUES ('oss_workspace_uuid', '41100000-0000-4000-8000-000000000001')"
+            )
+        )
+    from types import SimpleNamespace
+    from unittest.mock import MagicMock
+    from langbot.pkg.persistence.mgr import PersistenceManager
+
+    manager = PersistenceManager(SimpleNamespace(logger=MagicMock(), instance_uuid='migration-convergence'))
+    manager.db = SimpleNamespace(engine=engine, get_engine=lambda: engine)
+    await manager._initialize_managed_schema()
+    async with engine.connect() as conn:
+        assert await conn.scalar(sa.text('SELECT name FROM bots')) == 'Fixture bot'
+        assert await conn.scalar(sa.text('SELECT name FROM legacy_pipelines')) == 'Fixture pipeline'
+        bot = await conn.run_sync(lambda sync: sa.Table('bots', sa.MetaData(), autoload_with=sync))
+        bindings = await conn.scalar(sa.select(bot.c.event_bindings))
+        assert any(binding['target_uuid'] == 'pipeline-1' for binding in bindings)
+        assert await conn.scalar(sa.select(bot.c.plugin_processors)) == []
+        kb = await conn.run_sync(lambda sync: sa.Table('knowledge_bases', sa.MetaData(), autoload_with=sync))
+        assert all(await conn.scalars(sa.select(kb.c.initialized)))
+        if engine.dialect.name == 'sqlite':
+            assert (await conn.exec_driver_sql('PRAGMA foreign_key_check')).all() == []
+    async with engine.connect() as conn:
+
+        def differences(sync):
+            context = MigrationContext.configure(sync, opts={'compare_type': True})
+            return compare_metadata(context, Base.metadata)
+
+        diffs = await conn.run_sync(differences)
+
+        def substantive(sync, diff):
+            if isinstance(diff, list):
+                return True
+            if diff[0] not in ('add_table', 'add_column', 'add_index', 'add_constraint'):
+                return False
+            if diff[0] in ('add_index', 'add_constraint'):
+                obj = diff[1]
+                inspector = sa.inspect(sync)
+                columns = [c.name for c in obj.columns]
+                existing = inspector.get_indexes(obj.table.name) + inspector.get_unique_constraints(obj.table.name)
+                if any(
+                    item['column_names'] == columns and (diff[0] != 'add_constraint' or item.get('unique', True))
+                    for item in existing
+                ):
+                    return False
+                if (
+                    diff[0] == 'add_constraint'
+                    and inspector.get_pk_constraint(obj.table.name)['constrained_columns'] == columns
+                ):
+                    return False
+            return True
+
+        actionable = await conn.run_sync(lambda sync: [d for d in diffs if substantive(sync, d)])
+        assert not actionable, repr(actionable)
+    await manager._initialize_managed_schema()
+    async with engine.connect() as conn:
+        assert await conn.scalar(sa.text('SELECT name FROM bots')) == 'Fixture bot'
+        assert await conn.scalar(sa.text('SELECT name FROM legacy_pipelines')) == 'Fixture pipeline'
+        bot = await conn.run_sync(lambda sync: sa.Table('bots', sa.MetaData(), autoload_with=sync))
+        bindings = await conn.scalar(sa.select(bot.c.event_bindings))
+        assert any(binding['target_uuid'] == 'pipeline-1' for binding in bindings)
+        assert await conn.scalar(sa.select(bot.c.plugin_processors)) == []
+        kb = await conn.run_sync(lambda sync: sa.Table('knowledge_bases', sa.MetaData(), autoload_with=sync))
+        assert all(await conn.scalars(sa.select(kb.c.initialized)))
+        if engine.dialect.name == 'sqlite':
+            assert (await conn.exec_driver_sql('PRAGMA foreign_key_check')).all() == []
diff --git a/tests/integration/persistence/test_sqlite_concurrency.py b/tests/integration/persistence/test_sqlite_concurrency.py
new file mode 100644
index 000000000..fbf3355e5
--- /dev/null
+++ b/tests/integration/persistence/test_sqlite_concurrency.py
@@ -0,0 +1,34 @@
+from types import SimpleNamespace
+
+import pytest
+import sqlalchemy as sa
+
+from langbot.pkg.core import app  # noqa: F401 - initialize manager registration before database imports
+from langbot.pkg.persistence.databases.sqlite import SQLiteDatabaseManager
+
+pytestmark = [pytest.mark.integration, pytest.mark.asyncio]
+
+
+async def test_writer_commits_while_independent_reader_is_open(tmp_path):
+    application = SimpleNamespace(
+        instance_config=SimpleNamespace(data={'database': {'sqlite': {'path': str(tmp_path / 'concurrency.db')}}})
+    )
+    manager = SQLiteDatabaseManager(application)
+    await manager.initialize()
+    engine = manager.get_engine()
+    try:
+        async with engine.begin() as connection:
+            await connection.execute(sa.text('CREATE TABLE probe (id INTEGER PRIMARY KEY, value TEXT)'))
+            await connection.execute(sa.text("INSERT INTO probe VALUES (1, 'old')"))
+        async with engine.connect() as reader:
+            await reader.execute(sa.text('BEGIN'))
+            assert (await reader.execute(sa.text('SELECT value FROM probe'))).scalar_one() == 'old'
+            async with engine.connect() as writer:
+                await writer.execute(sa.text('PRAGMA busy_timeout=250'))
+                await writer.execute(sa.text("UPDATE probe SET value='new' WHERE id=1"))
+                await writer.commit()
+            await reader.rollback()
+        async with engine.connect() as connection:
+            assert (await connection.execute(sa.text('SELECT value FROM probe'))).scalar_one() == 'new'
+    finally:
+        await engine.dispose()
diff --git a/tests/integration/plugin/test_certified_plugin_admission.py b/tests/integration/plugin/test_certified_plugin_admission.py
index 0a5126ad9..74cd4a2bd 100644
--- a/tests/integration/plugin/test_certified_plugin_admission.py
+++ b/tests/integration/plugin/test_certified_plugin_admission.py
@@ -29,7 +29,8 @@ pytestmark = pytest.mark.integration
     ('deployment', 'archive_kind', 'administrator_force', 'expected_profile'),
     [
         ('cloud', 'signed_shared', False, 'shared-runtime-v1'),
-        ('oss', 'signed_shared', False, 'shared-runtime-v1'),
+        ('cloud', 'legacy', False, 'dedicated'),
+        ('oss', 'signed_shared', False, 'dedicated'),
         ('oss', 'legacy', False, 'dedicated'),
         ('oss', 'forged_shared', True, 'dedicated'),
     ],
@@ -75,7 +76,7 @@ async def test_install_plugin_admits_archive_before_persistence_and_applies_sele
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize('archive_kind', ['legacy', 'invalid_shared'])
+@pytest.mark.parametrize('archive_kind', ['invalid_shared'])
 async def test_cloud_rejects_untrusted_archive_before_storage_persistence_or_runtime_apply(archive_kind: str) -> None:
     package, trusted_public_keys = _archive(archive_kind)
     connector, _execution_context, _binding = _connector('cloud', trusted_public_keys)
@@ -180,26 +181,20 @@ async def test_marketplace_version_selection_keeps_certificate_gate_and_single_a
     if requested_version is not None:
         info['plugin_version'] = requested_version
     task_context = TaskContext.new()
-    if archive_kind == 'legacy':
-        with pytest.raises(ValueError, match='CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_REQUIRED'):
-            await connector.install_plugin(PluginInstallSource.MARKETPLACE, info, task_context)
-        connector._store_artifact_package.assert_not_awaited()
-        connector._persist_installation_package.assert_not_awaited()
-        connector.handler.apply_plugin_installation.assert_not_awaited()
-        connector._refresh_runner_registry.assert_not_awaited()
-    else:
-        await connector.install_plugin(PluginInstallSource.MARKETPLACE, info, task_context)
-        persisted_info = connector._persist_installation_package.await_args.kwargs['install_info']
-        assert persisted_info['plugin_version'] == '1.0.0'
-        assert persisted_info['_certification']['runtime_profile'] == 'shared-runtime-v1'
-        connector.handler.apply_plugin_installation.assert_awaited_once_with(
-            binding,
-            artifact_package=package,
-            enabled=True,
-            execution_mode=PluginExecutionMode.SHARED_CERTIFIED,
-        )
-        connector._refresh_runner_registry.assert_awaited_once()
-        assert task_context.metadata['progress_percent'] == 100
+    await connector.install_plugin(PluginInstallSource.MARKETPLACE, info, task_context)
+    persisted_info = connector._persist_installation_package.await_args.kwargs['install_info']
+    assert persisted_info['plugin_version'] == '1.0.0'
+    assert persisted_info['_certification']['runtime_profile'] == (
+        'shared-runtime-v1' if archive_kind == 'signed_shared' else 'dedicated'
+    )
+    connector.handler.apply_plugin_installation.assert_awaited_once_with(
+        binding,
+        artifact_package=package,
+        enabled=True,
+        execution_mode=(PluginExecutionMode.SHARED_CERTIFIED if archive_kind == 'signed_shared' else PluginExecutionMode.DEDICATED),
+    )
+    connector._refresh_runner_registry.assert_awaited_once()
+    assert task_context.metadata['progress_percent'] == 100
     if requested_version is not None:
         # Confirmed migrations must fetch the reviewed release, never latest/MCP/skill.
         assert len(requests) == 1
@@ -249,7 +244,10 @@ def _connector(deployment: str, trusted_public_keys: dict[str, str]):
 def _archive(kind: str) -> tuple[bytes, dict[str, str]]:
     manifest = {
         'metadata': {'author': 'certified', 'name': 'example', 'version': '1.0.0'},
-        'execution': {'sharedRuntime': 'shared-runtime-v1'},
+        'execution': {
+            'sharedRuntime': 'shared-runtime-v1',
+            'componentModel': 'stateless-v1',
+        },
     }
     if kind == 'legacy':
         manifest.pop('execution')
diff --git a/tests/integration_tests/box/test_cloud_box_admission_integration.py b/tests/integration_tests/box/test_cloud_box_admission_integration.py
index 4be9ccfc6..c0bb8b961 100644
--- a/tests/integration_tests/box/test_cloud_box_admission_integration.py
+++ b/tests/integration_tests/box/test_cloud_box_admission_integration.py
@@ -40,6 +40,7 @@ _UTC = dt.timezone.utc
 
 
 class _AdmissionBackend(BaseSandboxBackend):
+    execution_scoped_mounts = True
     name = 'nsjail'
 
     def __init__(self, logger):
@@ -397,10 +398,10 @@ async def test_cloud_core_skills_mount_generically_and_do_not_require_box_entitl
             read_only_mounts=skill_loader.build_execution_mounts(service.ap, query),
         )
 
-        mounted_spec = backend.started_specs[-1]
+        mounted_spec = backend.executed_specs[-1]
         assert len(mounted_spec.extra_mounts) == 1
         assert mounted_spec.extra_mounts[0].host_path == refreshed_skill['package_root']
-        assert mounted_spec.extra_mounts[0].mount_path == '/workspace/.skills/runner'
+        assert mounted_spec.extra_mounts[0].mount_path == '/opt/langbot-artifacts/' + refreshed_skill['revision'][7:]
         assert mounted_spec.extra_mounts[0].mode.value == 'ro'
         assert mounted_spec.extra_mounts[0].content_digest == refreshed_skill['revision']
 
@@ -419,7 +420,7 @@ async def test_cloud_core_skills_mount_generically_and_do_not_require_box_entitl
 
 
 @pytest.mark.asyncio
-async def test_activated_skill_mount_changes_recreate_shared_box_session(tmp_path):
+async def test_activated_skill_mount_changes_preserve_shared_box_session(tmp_path):
     service, runtime, backend, entitlements, server_task, client_task = await _stack(tmp_path)
     context = _context('workspace-a')
     entitlements.snapshots[context.workspace_uuid] = _snapshot(context.workspace_uuid)
@@ -438,12 +439,12 @@ async def test_activated_skill_mount_changes_recreate_shared_box_session(tmp_pat
             second_query,
             read_only_mounts=skill_loader.build_execution_mounts(service.ap, second_query),
         )
-        assert len(backend.started_specs) == 2
-        assert backend.stopped_sessions == [backend.started_specs[0].session_id]
+        assert len(backend.started_specs) == 1
+        assert backend.stopped_sessions == []
 
         await service.execute_tool({'command': 'echo first query again'}, first_query)
-        assert len(backend.started_specs) == 3
-        assert backend.stopped_sessions == [backend.started_specs[0].session_id] * 2
+        assert len(backend.started_specs) == 1
+        assert backend.stopped_sessions == []
     finally:
         server_task.cancel()
         client_task.cancel()
diff --git a/tests/manual/dump_execution_trace_payload.py b/tests/manual/dump_execution_trace_payload.py
new file mode 100644
index 000000000..0cd852f09
--- /dev/null
+++ b/tests/manual/dump_execution_trace_payload.py
@@ -0,0 +1,97 @@
+"""Dump one real execution-trace payload for the Space wire-format contract test.
+
+Run from the repository root with the project interpreter, e.g.:
+
+    PYTHONPATH=src python tests/manual/dump_execution_trace_payload.py
+
+`langbot-space` keeps the output as
+`internal/service/testdata/execution_trace_payload.json` and asserts that its
+ingest validator accepts it. Regenerate that file whenever the payload shape
+changes; the Space test documents the same command.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import json
+import sys
+import types
+from importlib import import_module
+
+
+class CaptureManager:
+    """Captures what TelemetryManager would have posted."""
+
+    def __init__(self):
+        self.telemetry_config = {'url': 'https://space.langbot.test', 'execution_trace': 'all'}
+        self.sent: list[dict] = []
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+    def start_send_task(self, payload: dict) -> None:
+        self.sent.append(payload)
+
+
+async def main() -> int:
+    execution = import_module('langbot.pkg.telemetry.execution')
+    trace = import_module('langbot.pkg.telemetry.trace')
+
+    manager = CaptureManager()
+    counters = execution.ExecutionCounters(manager)
+    context = types.SimpleNamespace(instance_uuid='instance-1', workspace_uuid='workspace-1')
+    route_ref = 'agent:11111111-1111-4111-8111-111111111111'
+    run_id = '22222222-2222-4222-8222-222222222222'
+
+    with execution.ingress(types.SimpleNamespace(telemetry=types.SimpleNamespace(execution=counters))):
+        # Inbound platform event.
+        counters.record(
+            context,
+            family='platform_event',
+            operation='message.received',
+            adapter='AiocqhttpAdapter',
+            outcome='success',
+        )
+        # Route decision for that event.
+        with trace.scope(route_ref=route_ref):
+            counters.record(
+                context,
+                family='event_route',
+                operation='message.received',
+                mode='agent',
+                adapter='AiocqhttpAdapter',
+                outcome='success',
+            )
+            # Runner execution of the routed processor.
+            run_token = trace.set_run(run_id)
+            try:
+                counters.record(
+                    context,
+                    family='runner',
+                    operation='execute',
+                    mode='agent',
+                    runner='langbot/runner-demo',
+                    outcome='success',
+                )
+                # Outbound platform API call made by the runner.
+                counters.record(
+                    context,
+                    family='platform_api',
+                    operation='send_message',
+                    mode='agent',
+                    adapter='AiocqhttpAdapter',
+                    outcome='success',
+                )
+            finally:
+                trace.reset_run(run_token)
+
+    if len(manager.sent) != 1:
+        print(f'expected exactly one payload, captured {len(manager.sent)}', file=sys.stderr)
+        return 1
+    print(json.dumps(manager.sent[0], indent=2, ensure_ascii=False, sort_keys=True))
+    return 0
+
+
+if __name__ == '__main__':
+    raise SystemExit(asyncio.run(main()))
diff --git a/tests/unit_tests/agent/test_debug_prompt.py b/tests/unit_tests/agent/test_debug_prompt.py
index 1ba35abf1..1c29c4915 100644
--- a/tests/unit_tests/agent/test_debug_prompt.py
+++ b/tests/unit_tests/agent/test_debug_prompt.py
@@ -27,6 +27,7 @@ from langbot.pkg.agent.runner.host_models import AgentBinding, AgentEventEnvelop
 from langbot.pkg.agent.runner.orchestrator import AgentRunOrchestrator
 from langbot.pkg.agent.runner.persistent_state_store import reset_persistent_state_store
 from langbot.pkg.agent.runner.session_registry import get_session_registry
+from langbot.pkg.api.http.service.model import LLMModelsService
 from langbot.pkg.api.http.context import (
     ExecutionContext,
     PrincipalContext,
@@ -151,6 +152,7 @@ async def harness():
         runner_registry=ScopedRunnerRegistry(),
     )
     ap.plugin_connector = PromptPullConnector(ap)
+    ap.llm_model_service = LLMModelsService(ap)
     ap.agent_run_orchestrator = AgentRunOrchestrator(ap, ap.runner_registry)
     ap.agent_service = AgentService(ap)
     try:
@@ -315,4 +317,4 @@ async def test_non_debug_query_keeps_effective_prompt_instead_of_static_config(h
         assert query.prompt is original
         assert pulled == [message.model_dump(mode='json') for message in original.messages]
     else:
-        assert pulled == []
+        assert pulled == [Message(**message).model_dump(mode='json') for message in DEFAULT_PROMPT]
diff --git a/tests/unit_tests/agent/test_event_first_protocol.py b/tests/unit_tests/agent/test_event_first_protocol.py
index c052fabbb..29429d9ea 100644
--- a/tests/unit_tests/agent/test_event_first_protocol.py
+++ b/tests/unit_tests/agent/test_event_first_protocol.py
@@ -60,6 +60,18 @@ class TestQueryToEventEnvelope:
 
         assert event.conversation_id == 'conv-uuid-123'
 
+    @pytest.mark.parametrize('launcher_id', [123456789, -100123456789, 'launcher-123'])
+    def test_query_to_event_accepts_platform_launcher_ids(self, mock_query, launcher_id):
+        """Telegram private and group IDs are numeric at the Query boundary."""
+        mock_query.launcher_id = launcher_id
+        mock_query.session.using_conversation.uuid = None
+
+        event = QueryEntryAdapter.query_to_event(mock_query)
+        context = QueryEntryAdapter._build_conversation_context(mock_query)
+
+        assert context.launcher_id == str(launcher_id)
+        assert event.conversation_id == f'person_{launcher_id}'
+
     def test_query_to_event_prefers_variable_conversation_id_when_conversation_uuid_missing(self, mock_query):
         """Pipeline variables can provide the conversation identity for state scope."""
         mock_query.session.using_conversation.uuid = None
diff --git a/tests/unit_tests/agent/test_orchestrator_integration.py b/tests/unit_tests/agent/test_orchestrator_integration.py
index 955e41ccd..61802e15e 100644
--- a/tests/unit_tests/agent/test_orchestrator_integration.py
+++ b/tests/unit_tests/agent/test_orchestrator_integration.py
@@ -80,6 +80,7 @@ class FakePluginConnector:
         self.results = results or []
         self.error = error
         self.delay = delay
+        self.delay_entered = asyncio.Event()
         self.calls: list[dict] = []
         self.contexts: list[dict] = []
         self.sessions_during_run: list[dict | None] = []
@@ -100,6 +101,7 @@ class FakePluginConnector:
 
         for result in self.results:
             if self.delay:
+                self.delay_entered.set()
                 await asyncio.sleep(self.delay)
             yield result
 
@@ -142,6 +144,7 @@ class FakeApplication:
             get_knowledge_base_by_uuid=AsyncMock(return_value=FakeKnowledgeBase('kb_001'))
         )
         self.skill_mgr = types.SimpleNamespace(
+            reload_skills=AsyncMock(),
             get_skills=lambda context: {
                 'demo': {
                     'name': 'demo',
@@ -153,7 +156,7 @@ class FakeApplication:
                     'display_name': 'Hidden Skill',
                     'description': 'Not bound to this pipeline.',
                 },
-            }
+            },
         )
 
 
@@ -600,6 +603,35 @@ async def test_orchestrator_stops_after_cancel_request(clean_agent_state):
     assert run['status_reason'] == 'user stopped'
 
 
+@pytest.mark.asyncio
+@pytest.mark.parametrize('cancel_while_waiting', [False, True])
+async def test_consumer_exit_finalizes_cancelled_ledger(clean_agent_state, cancel_while_waiting):
+    connector = FakePluginConnector(
+        results=[
+            {'type': 'message.delta', 'data': {'chunk': {'role': 'assistant', 'content': 'first'}}},
+            {'type': 'run.completed', 'data': {'finish_reason': 'stop'}},
+        ]
+    )
+    orchestrator = AgentRunOrchestrator(FakeApplication(connector, clean_agent_state), FakeRegistry(make_descriptor()))
+    plan = orchestrator.query_bridge.build_plan(make_query())
+    stream = orchestrator.run(plan.event, plan.binding, adapter_context={'_execution_context': TEST_CONTEXT})
+    await anext(stream)
+    if cancel_while_waiting:
+        connector.delay = 10
+        pending = asyncio.create_task(anext(stream))
+        # Cancel the provider wait, not an arbitrary journal DB operation.
+        await asyncio.wait_for(connector.delay_entered.wait(), timeout=2)
+        pending.cancel()
+        with pytest.raises(asyncio.CancelledError):
+            await pending
+    else:
+        await stream.aclose()
+    run = await RunLedgerStore(clean_agent_state).get_run(connector.contexts[0]['run_id'])
+    assert run['status'] == 'cancelled'
+    assert run['status_reason'] == 'consumer_cancelled'
+    assert await get_session_registry().list_active_runs() == []
+
+
 @pytest.mark.asyncio
 async def test_orchestrator_does_not_package_query_messages_into_context(clean_agent_state):
     """Host should not build an agent working-context window from query.messages."""
@@ -1007,7 +1039,7 @@ class TestQueryEntrySessionQueryId:
 
     @pytest.mark.asyncio
     async def test_no_query_id_for_pure_event_first_flow(self, clean_agent_state):
-        """Pure event-first flow restores authorized Skill revisions without a Query."""
+        """Pure event-first flow ignores the previous run's conversation pins."""
         from langbot.pkg.agent.runner.host_models import (
             AgentEventEnvelope,
             AgentBinding,
@@ -1131,7 +1163,8 @@ class TestQueryEntrySessionQueryId:
         assert execution_query.message_event.type == event.event_type
         assert '_host_box_scope' not in execution_query.variables
         assert execution_query.variables['_pipeline_bound_skills'] == ['demo', 'hidden']
-        assert execution_query.variables[ACTIVATED_SKILLS_KEY]['demo']['revision'] == revision
+        assert ACTIVATED_SKILLS_KEY not in execution_query.variables
+        ap.skill_repository.get_skill.assert_not_awaited()
         assert execution_query.variables['_pipeline_mcp_resource_attachments'][0]['server_uuid'] == 'srv-1'
         assert execution_query.variables['_pipeline_mcp_resource_agent_read_enabled'] is True
         assert 'MCP resource context selected by LangBot host:' in plugin_connector.contexts[0]['input']['text']
@@ -1363,8 +1396,8 @@ class TestQueryEntryAdapterHostCapabilities:
 
     @pytest.mark.asyncio
     @pytest.mark.parametrize('legacy_state', [False, True])
-    async def test_run_from_query_restores_activated_skills_from_state(self, clean_agent_state, legacy_state):
-        """Pinned revisions restore; old name-only state requires reactivation."""
+    async def test_new_query_does_not_restore_conversation_skill_pins(self, clean_agent_state, legacy_state):
+        """New runs ignore both revision and legacy name-only conversation pins."""
         from langbot.pkg.agent.runner.persistent_state_store import get_persistent_state_store
         from langbot.pkg.provider.tools.loaders.skill import (
             ACTIVATED_SKILL_NAMES_STATE_KEY,
@@ -1408,16 +1441,8 @@ class TestQueryEntryAdapterHostCapabilities:
         messages = [message async for message in orchestrator.run_from_query(query)]
 
         assert len(messages) == 1
-        if legacy_state:
-            assert ACTIVATED_SKILLS_KEY not in query.variables
-            assert ap.skill_repository.get_skill.await_count == 0
-            assert (
-                len([warning for warning in ap.logger.warnings if 'Legacy name-only activated Skill state' in warning])
-                == 1
-            )
-        else:
-            assert query.variables[ACTIVATED_SKILLS_KEY]['demo']['name'] == 'demo'
-            assert query.variables[ACTIVATED_SKILLS_KEY]['demo']['revision'] == revision
+        assert ACTIVATED_SKILLS_KEY not in query.variables
+        ap.skill_repository.get_skill.assert_not_awaited()
 
     @pytest.mark.asyncio
     async def test_event_log_and_transcript_written(self, clean_agent_state):
@@ -1524,59 +1549,6 @@ async def test_synthetic_event_query_exposes_trusted_workspace_to_tools(clean_ag
     assert received[0].workspace_uuid == context.workspace_uuid
 
 
-@pytest.mark.asyncio
-async def test_beta_diagnostics_close_releases_real_orchestrator_session(clean_agent_state):
-    from langbot.pkg.telemetry.diagnostics import DiagnosticsManager
-
-    plugin_connector = FakePluginConnector(
-        results=[{'type': 'message.completed', 'data': {'message': {'role': 'assistant', 'content': 'CANARY'}}}]
-    )
-    ap = FakeApplication(plugin_connector, clean_agent_state)
-    ap.instance_config = types.SimpleNamespace(data={'space': {'url': 'https://example.invalid'}})
-    ap.diagnostics = DiagnosticsManager(ap, version='4.11.0b2', instance_id='instance-test')
-    orchestrator = AgentRunOrchestrator(ap, FakeRegistry(make_descriptor()))
-    gen = orchestrator.run_from_query(make_query())
-    assert (await anext(gen)).content == 'CANARY'
-    run_id = plugin_connector.contexts[0]['run_id']
-    assert await get_session_registry().get(run_id) is not None
-    await gen.aclose()
-    assert await get_session_registry().get(run_id) is None
-    records = [e for e in ap.diagnostics.pending if e['operation'] == 'runner.run']
-    assert records[-1]['outcome'] == 'cancelled'
-    import json
-
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('terminal', ['run.completed', 'run.failed'])
-async def test_beta_diagnostics_real_runner_terminal(clean_agent_state, terminal):
-    from langbot.pkg.telemetry.diagnostics import DiagnosticsManager
-
-    plugin_connector = FakePluginConnector(
-        results=[
-            {
-                'type': terminal,
-                'data': {'finish_reason': 'stop'} if terminal == 'run.completed' else {'error': 'CANARY'},
-            }
-        ]
-    )
-    ap = FakeApplication(plugin_connector, clean_agent_state)
-    ap.instance_config = types.SimpleNamespace(data={'space': {'url': 'https://example.invalid'}})
-    ap.diagnostics = DiagnosticsManager(ap, version='4.11.0b2', instance_id='instance-test')
-    orchestrator = AgentRunOrchestrator(ap, FakeRegistry(make_descriptor()))
-    try:
-        _ = [v async for v in orchestrator.run_from_query(make_query())]
-    except RunnerExecutionError:
-        assert terminal == 'run.failed'
-    records = [e for e in ap.diagnostics.pending if e['operation'] == 'runner.run']
-    assert records[-1]['outcome'] == ('succeeded' if terminal == 'run.completed' else 'failed')
-    assert records[-1]['run_id'] == plugin_connector.contexts[0]['run_id']
-    import json
-
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
 @pytest.mark.asyncio
 async def test_agent_cannot_deliver_exported_files_without_reply_api(clean_agent_state):
     """Returning file handles must not bypass the Agent's reply authorization."""
diff --git a/tests/unit_tests/agent/test_run_journal.py b/tests/unit_tests/agent/test_run_journal.py
index 8908681ba..1ae134a01 100644
--- a/tests/unit_tests/agent/test_run_journal.py
+++ b/tests/unit_tests/agent/test_run_journal.py
@@ -52,7 +52,19 @@ async def test_agent_run_snapshot_records_identity_and_redacts_inline_attachment
     assert raw_input['contents'][0]['image_base64'] == 'large-image-data'
     assert raw_input['attachments'][0]['content'] == 'large-file-data'
 
-    if event_type != 'message.received':
-        assert saved['metadata']['input_event'] == event.data
-    else:
-        assert 'input_event' not in saved['metadata']
+    assert saved['metadata']['input_event'] == event.data
+
+
+def test_event_snapshot_preserves_custom_values_and_redacts_nested_binary():
+    data = {
+        'member': {'id': 'u1', 'active': False},
+        'count': 0,
+        'custom': {'contents': [{'type': 'image_base64', 'image_base64': 'secret'}]},
+        'attachments': [{'name': 'file.txt', 'content': 'secret'}],
+    }
+    saved = AgentRunJournal._sanitize_event_data(data)
+    assert saved['member'] == data['member']
+    assert saved['count'] == 0
+    assert saved['custom']['contents'][0]['image_base64'] is None
+    assert saved['attachments'][0]['content'] is None
+    assert data['custom']['contents'][0]['image_base64'] == 'secret'
diff --git a/tests/unit_tests/agent/test_run_ledger_store.py b/tests/unit_tests/agent/test_run_ledger_store.py
index 4359d87c5..22c97baea 100644
--- a/tests/unit_tests/agent/test_run_ledger_store.py
+++ b/tests/unit_tests/agent/test_run_ledger_store.py
@@ -499,3 +499,91 @@ async def test_agent_run_pages_include_old_bindings_without_leaking_other_agents
     page, cursor, more, total = await store.list_runs(agent_id='one', workspace_id='w', before_id=cursor, limit=2)
     assert [run['run_id'] for run in page] == ['old']
     assert not more and total == 3
+
+
+@pytest.mark.asyncio
+async def test_skill_pins_belong_to_run_and_survive_store_restart(store, db_engine):
+    for run_id in ('old', 'new'):
+        await store.create_run(
+            run_id=run_id,
+            event_id=run_id,
+            binding_id='binding',
+            runner_id='runner',
+            workspace_id='ws',
+            conversation_id='same',
+            status='running',
+        )
+    assert await store.pin_skill_revision('old', 'ws', 'demo', 'v1') == 'v1'
+    restarted = RunLedgerStore(db_engine)
+    assert await restarted.pin_skill_revision('old', 'ws', 'demo', 'v2') == 'v1'
+    assert await restarted.pin_skill_revision('new', 'ws', 'demo', 'v2') == 'v2'
+    assert (await restarted.get_run('old'))['metadata']['activated_skills'] == {'demo': 'v1'}
+    with pytest.raises(ValueError, match='current Workspace'):
+        await restarted.pin_skill_revision('old', 'other', 'demo', 'v2')
+    await store.finalize_run(run_id='old', status='completed')
+    with pytest.raises(ValueError, match='active run'):
+        await restarted.pin_skill_revision('old', 'ws', 'demo', 'v2')
+
+
+@pytest.mark.asyncio
+async def test_activation_uses_new_revision_only_for_new_run(store, db_engine, tmp_path):
+    from types import SimpleNamespace
+    from unittest.mock import Mock
+    from langbot.pkg.api.http.context import ExecutionContext
+    from langbot.pkg.skill.repository import SkillRepository
+    from langbot.pkg.skill.manager import SkillManager
+    from langbot.pkg.provider.tools.loaders.skill_authoring import SkillToolLoader
+
+    context = ExecutionContext(instance_uuid='instance', workspace_uuid='ws', placement_generation=1)
+
+    async def binding(workspace_uuid, **kwargs):
+        return context
+
+    app = SimpleNamespace(
+        logger=Mock(),
+        instance_config=SimpleNamespace(data={'skills': {'root': str(tmp_path / 'skills')}}),
+        workspace_service=SimpleNamespace(get_execution_binding=binding),
+        persistence_mgr=SimpleNamespace(get_db_engine=lambda: db_engine),
+    )
+    app.skill_repository = SkillRepository(app)
+    app.skill_mgr = SkillManager(app)
+    loader = SkillToolLoader(app)
+    first = await app.skill_repository.create_skill(context, {'name': 'demo', 'instructions': 'version one'})
+    await app.skill_mgr.reload_skills(context)
+
+    def query(run_id):
+        return SimpleNamespace(
+            instance_uuid='instance',
+            workspace_uuid='ws',
+            placement_generation=1,
+            variables={'_pipeline_bound_skills': ['demo']},
+            _execution_context=context,
+            _agent_run_session={'run_id': run_id},
+        )
+
+    for name in ('old', 'new'):
+        await store.create_run(
+            run_id=name,
+            event_id=name,
+            binding_id='binding',
+            runner_id='runner',
+            workspace_id='ws',
+            conversation_id='same',
+            status='running',
+        )
+    old = query('old')
+    assert 'version one' in (await loader._invoke_activate_skill({'skill_name': 'demo'}, old))['content']
+    second = await app.skill_repository.update_skill(
+        context, 'demo', {'instructions': 'version two'}, base_revision=first['revision']
+    )
+    await app.skill_mgr.reload_skills(context)
+    assert 'version two' in (await loader._invoke_activate_skill({'skill_name': 'demo'}, query('new')))['content']
+    assert 'version one' in (await loader._invoke_activate_skill({'skill_name': 'demo'}, old))['content']
+    await app.skill_repository.delete_skill(context, 'demo')
+    await app.skill_mgr.reload_skills(context)
+    assert 'version one' in (await loader._invoke_activate_skill({'skill_name': 'demo'}, query('old')))['content']
+    assert (await store.get_run('new'))['metadata']['activated_skills']['demo'] == second['revision']
+    revoked = query('old')
+    revoked.variables['_pipeline_bound_skills'] = []
+    with pytest.raises(ValueError, match='not authorized'):
+        await loader._invoke_activate_skill({'skill_name': 'demo'}, revoked)
diff --git a/tests/unit_tests/api/http/service/test_tenant_resource_isolation.py b/tests/unit_tests/api/http/service/test_tenant_resource_isolation.py
index 623822379..d0c58c36a 100644
--- a/tests/unit_tests/api/http/service/test_tenant_resource_isolation.py
+++ b/tests/unit_tests/api/http/service/test_tenant_resource_isolation.py
@@ -294,7 +294,7 @@ async def test_cross_workspace_parent_references_are_rejected(tenant_services):
             {
                 'event_bindings': [
                     {
-                        'event_pattern': 'message.*',
+                        'event_pattern': 'message.received',
                         'target_type': 'pipeline',
                         'target_uuid': 'pipeline-b',
                     }
diff --git a/tests/unit_tests/api/http/test_entitlement_context.py b/tests/unit_tests/api/http/test_entitlement_context.py
index 600eb1082..56232c999 100644
--- a/tests/unit_tests/api/http/test_entitlement_context.py
+++ b/tests/unit_tests/api/http/test_entitlement_context.py
@@ -76,3 +76,21 @@ def test_execution_context_preserves_entitlement_revision():
     )
 
     assert ExecutionContext.from_request(request).entitlement_revision == 11
+
+
+@pytest.mark.asyncio
+async def test_handler_entitlement_denial_is_403_not_internal_error():
+    from langbot.pkg.api.http.controller.group import AuthType
+    from langbot.pkg.cloud.entitlements import EntitlementFeatureUnavailableError
+
+    router = _router(SimpleNamespace(mode='cloud'))
+    router.path = '/policy'
+    router.name = 'policy'
+
+    @router.route('', methods=['GET'], auth_type=AuthType.NONE)
+    async def denied():
+        raise EntitlementFeatureUnavailableError('platform.telegram', entitlement_revision=2)
+
+    response = await router.quart_app.test_client().get('/policy')
+    assert response.status_code == 403
+    assert (await response.get_json())['code'] == 'entitlement_unavailable'
diff --git a/tests/unit_tests/api/http/test_route_tenant_scope.py b/tests/unit_tests/api/http/test_route_tenant_scope.py
index 4b4cb96e6..26e5ae271 100644
--- a/tests/unit_tests/api/http/test_route_tenant_scope.py
+++ b/tests/unit_tests/api/http/test_route_tenant_scope.py
@@ -10,6 +10,7 @@ from sqlalchemy.ext.asyncio import create_async_engine
 
 from langbot.pkg.api.http.controller import group
 from langbot.pkg.persistence.mgr import PersistenceManager, PersistenceMode
+from langbot.pkg.entity.persistence.metadata import WorkspaceMetadata
 
 
 pytestmark = pytest.mark.asyncio
@@ -20,6 +21,8 @@ async def test_authenticated_route_does_not_hold_database_session_during_externa
     release = asyncio.Event()
     observations: list[bool] = []
     engine = create_async_engine('sqlite+aiosqlite:///:memory:')
+    async with engine.begin() as connection:
+        await connection.run_sync(WorkspaceMetadata.__table__.create)
     persistence = PersistenceManager(object(), mode=PersistenceMode.CLOUD_RUNTIME)
     persistence.db = SimpleNamespace(get_engine=lambda: engine)
 
@@ -60,7 +63,7 @@ async def test_authenticated_route_does_not_hold_database_session_during_externa
         )
     )
     try:
-        await entered.wait()
+        await asyncio.wait_for(entered.wait(), timeout=5)
         assert observations == [True]
         release.set()
         response = await request
diff --git a/tests/unit_tests/api/service/test_agent_service.py b/tests/unit_tests/api/service/test_agent_service.py
index 600c3e6d3..d5fc567b8 100644
--- a/tests/unit_tests/api/service/test_agent_service.py
+++ b/tests/unit_tests/api/service/test_agent_service.py
@@ -13,6 +13,7 @@ from langbot.pkg.api.http.service.agent import (
     PIPELINE_EVENT_PATTERNS,
     AgentService,
 )
+from langbot.pkg.api.http.service.model import LLMModelsService
 
 
 pytestmark = pytest.mark.asyncio
@@ -82,6 +83,7 @@ def _compiled_update_values(statement):
 
 def _make_app():
     app = SimpleNamespace()
+    app.llm_model_service = LLMModelsService(app)
     app.persistence_mgr = SimpleNamespace(
         execute_async=AsyncMock(),
         serialize_model=Mock(side_effect=_serialize_agent),
@@ -877,7 +879,9 @@ async def test_event_processor_rejects_invalid_component_and_missing_parameters(
         )
     )
     with pytest.raises(ValueError, match='Required processor parameter'):
-        await service.create_agent(WORKSPACE_UUID, {'kind': 'event_processor', 'component_ref': 'plugin:a/b/c'})
+        await service.create_agent(
+            WORKSPACE_UUID, {'kind': 'event_processor', 'component_ref': 'plugin:a/b/c', 'parameters': {}}
+        )
 
 
 async def test_unavailable_event_processor_can_still_be_renamed():
diff --git a/tests/unit_tests/api/service/test_bot_service.py b/tests/unit_tests/api/service/test_bot_service.py
index 81a180a75..da2bedff6 100644
--- a/tests/unit_tests/api/service/test_bot_service.py
+++ b/tests/unit_tests/api/service/test_bot_service.py
@@ -390,14 +390,14 @@ class TestBotServiceCreateBot:
         async def mock_execute(query):
             nonlocal call_count
             call_count += 1
-            if call_count <= 2:
-                return pipeline_result  # First call: check pipeline
-            elif call_count == 3:
-                return Mock()  # Insert
+            if call_count == 1:
+                return Mock()  # Insert; there is no legacy pipeline lookup
             return bot_result  # Get bot
 
         ap.persistence_mgr.execute_async = AsyncMock(side_effect=mock_execute)
-        ap.persistence_mgr.serialize_model = Mock(return_value={'uuid': 'new-uuid', 'name': 'New Bot'})
+        ap.persistence_mgr.serialize_model = Mock(
+            return_value={'uuid': 'new-uuid', 'name': 'New Bot', 'adapter': 'telegram', 'enable': False}
+        )
 
         service = BotService(ap)
 
@@ -672,8 +672,8 @@ class TestBotServiceDeleteBot:
 class TestBotServiceListEventLogs:
     """Tests for list_event_logs method."""
 
-    async def test_list_event_logs_bot_not_found_raises(self):
-        """Raises Exception when runtime bot not found."""
+    async def test_list_event_logs_inactive_bot_returns_empty(self):
+        """A saved bot without an active runtime has no event logs."""
         # Setup
         ap = SimpleNamespace()
         ap.platform_mgr = SimpleNamespace()
@@ -683,8 +683,7 @@ class TestBotServiceListEventLogs:
         service.get_bot = AsyncMock(return_value={'uuid': 'nonexistent-uuid'})
 
         # Execute & Verify
-        with pytest.raises(Exception, match='Bot not found'):
-            await service.list_event_logs(WORKSPACE_UUID, 'nonexistent-uuid', 0, 10)
+        assert await service.list_event_logs(WORKSPACE_UUID, 'nonexistent-uuid', 0, 10) == ([], 0)
 
     async def test_list_event_logs_returns_logs(self):
         """Returns logs from runtime bot logger."""
diff --git a/tests/unit_tests/api/service/test_monitoring_executions.py b/tests/unit_tests/api/service/test_monitoring_executions.py
new file mode 100644
index 000000000..63fbd32b5
--- /dev/null
+++ b/tests/unit_tests/api/service/test_monitoring_executions.py
@@ -0,0 +1,846 @@
+"""Execution-view tests for the monitoring service.
+
+Covers the unified agent-run + pipeline-query execution list, its rollup
+summary, and single-execution traces.
+"""
+
+from __future__ import annotations
+
+import datetime
+from types import SimpleNamespace
+
+import pytest
+from sqlalchemy.ext.asyncio import AsyncSession, create_async_engine
+from sqlalchemy.orm import sessionmaker
+
+from langbot.pkg.api.http.service.monitoring import MonitoringService
+from langbot.pkg.entity.persistence import agent_interaction as persistence_interaction
+from langbot.pkg.entity.persistence import agent_run as persistence_agent_run
+from langbot.pkg.entity.persistence import monitoring as persistence_monitoring
+from langbot.pkg.entity.persistence import workspace as persistence_workspace  # noqa: F401
+from langbot.pkg.entity.persistence.base import Base
+
+WORKSPACE = 'workspace-1'
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('has_log', [False, True])
+async def test_event_properties_survive_detail_snapshot_fallback(engine, service, has_log):
+    import json
+    from langbot.pkg.entity.persistence.event_log import EventLog
+
+    data = {'member': {'id': 'u1', 'active': False}, 'count': 0, 'custom': {'reason': 'joined'}}
+    async with AsyncSession(engine) as session:
+        session.add(
+            persistence_agent_run.AgentRun(
+                run_id='custom-run',
+                event_id='custom-event',
+                workspace_id=WORKSPACE,
+                runner_id='runner',
+                status='completed',
+                created_at=_dt(10),
+                metadata_json=json.dumps(
+                    {
+                        'event_type': 'group.member_joined',
+                        'input_event': data,
+                        'input': {'text': '', 'contents': [], 'attachments': []},
+                    }
+                ),
+            )
+        )
+        if has_log:
+            session.add(
+                EventLog(
+                    event_id='custom-event',
+                    event_type='group.member_joined',
+                    source='webui',
+                    workspace_id=WORKSPACE,
+                    run_id='custom-run',
+                    input_json='{"text":"","contents":[]}',
+                    created_at=_dt(10),
+                )
+            )
+            session.add(
+                EventLog(
+                    event_id='later-event',
+                    event_type='message.received',
+                    source='webui',
+                    workspace_id=WORKSPACE,
+                    run_id='custom-run',
+                    input_json='{"text":"later"}',
+                    created_at=_dt(10, 1),
+                )
+            )
+        await session.commit()
+    detail = await service.get_execution_detail(WORKSPACE, 'agent', 'custom-run')
+    items = detail['pages']['inputs']['items']
+    assert items[0]['content']['event']['member'] == data['member']
+    assert items[0]['content']['event']['count'] == 0
+    assert items[0]['content']['event']['custom'] == data['custom']
+    if has_log:
+        assert items[1]['content'] == {'text': 'later'}
+
+
+def test_event_content_preserves_conflicting_custom_input_fields():
+    from langbot.pkg.api.http.service.monitoring_execution_details import event_content
+
+    assert event_content({'text': 'normalized'}, {'text': {'custom': False}}) == {
+        'text': 'event',
+        'event': {'text': {'custom': False}},
+        'input': {'text': 'normalized'},
+    }
+
+
+class FakePersistenceManager:
+    """Runs statements against a real in-memory SQLite engine."""
+
+    def __init__(self, engine):
+        self._engine = engine
+        self._factory = sessionmaker(engine, class_=AsyncSession, expire_on_commit=False)
+
+    def get_db_engine(self):
+        return self._engine
+
+    def tenant_uow(self, workspace_uuid):
+        from langbot.pkg.persistence.tenant_uow import TenantUnitOfWork
+
+        return TenantUnitOfWork(self._engine, workspace_uuid)
+
+    async def execute_async(self, statement):
+        # Match PersistenceManager's connection-level result shape.
+        async with self._engine.connect() as connection:
+            return await connection.execute(statement)
+
+    @staticmethod
+    def serialize_model(model, data):
+        out = {}
+        for column in model.__table__.columns:
+            value = getattr(data, column.name)
+            out[column.name] = value.isoformat() if isinstance(value, datetime.datetime) else value
+        return out
+
+
+@pytest.fixture
+async def engine(tmp_path):
+    engine = create_async_engine(f'sqlite+aiosqlite:///{tmp_path / "monitoring_executions.db"}')
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+    yield engine
+    await engine.dispose()
+
+
+@pytest.fixture
+def service(engine):
+    ap = SimpleNamespace(
+        persistence_mgr=FakePersistenceManager(engine),
+        instance_config=None,
+    )
+    return MonitoringService(ap)
+
+
+def _dt(hour: int, minute: int = 0, second: int = 0) -> datetime.datetime:
+    return datetime.datetime(2026, 1, 1, hour, minute, second)
+
+
+async def _seed(engine) -> None:
+    factory = sessionmaker(engine, class_=AsyncSession, expire_on_commit=False)
+    async with factory() as session:
+        session.add_all(
+            [
+                persistence_agent_run.AgentRun(
+                    run_id='run-done',
+                    event_id='evt-1',
+                    agent_id='agent-1',
+                    binding_id='agent_agent-1_default',
+                    runner_id='runner-1',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    status='completed',
+                    created_at=_dt(10),
+                    started_at=_dt(10, 0, 1),
+                    finished_at=_dt(10, 0, 3),
+                    usage_json='{"total_tokens": 42}',
+                ),
+                persistence_agent_run.AgentRun(
+                    run_id='run-failed',
+                    binding_id='agent_agent-1_default',
+                    runner_id='runner-1',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    status='failed',
+                    created_at=_dt(10, 5),
+                ),
+                persistence_agent_run.AgentRun(
+                    run_id='run-debug',
+                    event_id='debug:agent-1:e1',
+                    agent_id='agent-1',
+                    binding_id='debug:agent-1:runner-1',
+                    runner_id='runner-1',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    status='completed',
+                    created_at=_dt(10, 6),
+                    started_at=_dt(10, 6, 0),
+                    finished_at=_dt(10, 6, 1),
+                ),
+                persistence_agent_run.AgentRunEvent(
+                    run_id='run-done',
+                    sequence=1,
+                    type='run.completed',
+                    data_json='{"message": {"role": "assistant", "content": "hi"}}',
+                ),
+                persistence_interaction.AgentInteraction(
+                    interaction_id='int-1',
+                    run_id='run-done',
+                    binding_id='agent_agent-1_default',
+                    runner_id='runner-1',
+                    processor_type='agent',
+                    processor_id='agent-1',
+                    workspace_id=WORKSPACE,
+                    status='pending',
+                    request_json='{}',
+                    callback_token_hash='hash-1',
+                ),
+            ]
+        )
+        for message in [
+            dict(
+                id='msg-user',
+                timestamp=_dt(10, 2),
+                session_id='session-1',
+                message_content='{"chain": [{"type": "Plain", "text": "hello there"}]}',
+                status='success',
+                role='user',
+            ),
+            dict(
+                id='msg-assistant',
+                timestamp=_dt(10, 3),
+                session_id='session-1',
+                message_content='{"chain": [{"type": "Plain", "text": "reply"}]}',
+                status='success',
+                role='assistant',
+            ),
+            dict(
+                id='msg-discarded',
+                timestamp=_dt(10, 4),
+                session_id='session-2',
+                message_content='{"chain": [{"type": "Plain", "text": "ignored"}]}',
+                status='discarded',
+                role='user',
+            ),
+        ]:
+            session.add(
+                persistence_monitoring.MonitoringMessage(
+                    workspace_uuid=WORKSPACE,
+                    bot_id='bot-1',
+                    bot_name='Bot One',
+                    pipeline_id='pipeline-1',
+                    pipeline_name='Pipeline One',
+                    level='info',
+                    **message,
+                )
+            )
+        session.add_all(
+            [
+                persistence_monitoring.MonitoringLLMCall(
+                    id='llm-1',
+                    workspace_uuid=WORKSPACE,
+                    timestamp=_dt(10, 2, 30),
+                    model_name='gpt',
+                    input_tokens=10,
+                    output_tokens=20,
+                    total_tokens=30,
+                    duration=100,
+                    status='success',
+                    bot_id='bot-1',
+                    bot_name='Bot One',
+                    pipeline_id='pipeline-1',
+                    pipeline_name='Pipeline One',
+                    session_id='session-1',
+                    message_id='msg-user',
+                ),
+                persistence_monitoring.MonitoringLLMCall(
+                    id='llm-2',
+                    workspace_uuid=WORKSPACE,
+                    timestamp=_dt(10, 4, 30),
+                    model_name='gpt',
+                    input_tokens=0,
+                    output_tokens=0,
+                    total_tokens=0,
+                    duration=50,
+                    status='success',
+                    bot_id='bot-1',
+                    bot_name='Bot One',
+                    pipeline_id='pipeline-1',
+                    pipeline_name='Pipeline One',
+                    session_id='session-2',
+                    message_id='msg-discarded',
+                ),
+            ]
+        )
+        session.add(
+            persistence_monitoring.MonitoringSession(
+                workspace_uuid=WORKSPACE,
+                bot_id='bot-1',
+                session_id='session-1',
+                bot_name='Bot One',
+                pipeline_id='pipeline-1',
+                pipeline_name='Pipeline One',
+                message_count=2,
+                start_time=_dt(10, 2),
+                last_activity=_dt(10, 5),
+                is_active=True,
+            )
+        )
+        await session.commit()
+
+
+@pytest.mark.asyncio
+async def test_executions_merge_and_order_both_sources(engine, service):
+    await _seed(engine)
+
+    result = await service.get_executions(WORKSPACE, limit=50)
+
+    assert result['total'] == 5
+    ids = [(item['source'], item['id']) for item in result['items']]
+    # Assistant replies are outputs, not executions.
+    assert ('pipeline', 'msg-assistant') not in ids
+    # Newest first across both object types.
+    assert ids == [
+        ('agent', 'run-debug'),
+        ('agent', 'run-failed'),
+        ('pipeline', 'msg-discarded'),
+        ('pipeline', 'msg-user'),
+        ('agent', 'run-done'),
+    ]
+
+    by_id = {item['id']: item for item in result['items']}
+    assert result['summary']['executions']['total'] == 5
+    assert result['summary']['executions']['denominator'] == 4
+    assert result['summary']['tokens']['calls'] == 2
+    assert result['summary']['tokens']['calls_with_usage'] == 1
+    assert by_id['run-done']['status_group'] == 'completed'
+    assert by_id['run-failed']['status_group'] == 'failed'
+    assert by_id['run-debug']['debug'] is True
+    assert by_id['run-done']['debug'] is False
+    assert by_id['run-done']['usage'] == {'total_tokens': 42}
+    assert by_id['run-done']['duration_ms'] == 2000
+    assert by_id['msg-user']['status_group'] == 'completed'
+    assert by_id['msg-discarded']['status_group'] == 'ignored'
+    assert by_id['msg-user']['title'] == 'hello there'
+    assert by_id['msg-user']['target_id'] == 'pipeline-1'
+
+
+@pytest.mark.asyncio
+async def test_executions_source_status_and_paging(engine, service):
+    await _seed(engine)
+
+    only_agents = await service.get_executions(WORKSPACE, source='agent')
+    assert {item['source'] for item in only_agents['items']} == {'agent'}
+    assert only_agents['total'] == 3
+
+    pipeline_only = await service.get_executions(WORKSPACE, source='pipeline')
+    assert {item['source'] for item in pipeline_only['items']} == {'pipeline'}
+    assert pipeline_only['total'] == 2
+
+    # A single-object target selection narrows the union to that source.
+    by_pipeline = await service.get_executions(WORKSPACE, pipeline_ids=['pipeline-1'])
+    assert {item['source'] for item in by_pipeline['items']} == {'pipeline'}
+
+    failed = await service.get_executions(WORKSPACE, statuses=['failed'])
+    assert [item['id'] for item in failed['items']] == ['run-failed']
+
+    # `ignored` only exists for pipeline queries: it must not widen the agent
+    # filter back to "every status".
+    ignored = await service.get_executions(WORKSPACE, statuses=['ignored'])
+    assert [item['id'] for item in ignored['items']] == ['msg-discarded']
+
+    debug_only = await service.get_executions(WORKSPACE, mode='debug')
+    assert [item['id'] for item in debug_only['items']] == ['run-debug']
+
+    first_page = await service.get_executions(WORKSPACE, limit=2)
+    assert len(first_page['items']) == 2
+    assert first_page['has_more'] is True
+    second_page = await service.get_executions(WORKSPACE, limit=2, offset=2)
+    assert first_page['items'][0]['id'] != second_page['items'][0]['id']
+
+
+@pytest.mark.asyncio
+async def test_execution_summary_uses_explicit_denominator(engine, service):
+    await _seed(engine)
+
+    summary = (await service.get_executions(WORKSPACE))['summary']
+    executions = summary['executions']
+
+    # completed = run-done + run-debug + msg-user; failed = run-failed.
+    assert executions['by_source'] == {'agent': 3, 'pipeline': 2}
+    assert executions['completed'] == 3
+    assert executions['failed'] == 1
+    assert executions['ignored'] == 1
+    assert executions['denominator'] == 4
+    assert executions['success_rate'] == 75.0
+    assert executions['debug'] == 1
+    assert executions['real'] == 4
+    assert executions['waiting'] == 1
+    assert executions['p50_duration_ms'] == 1000
+    assert executions['p95_duration_ms'] == 2000
+    # Only run-done and run-debug are completed runs with a start and a finish.
+    assert executions['duration_sample'] == 2
+
+    # Token coverage is explicit: llm-2 succeeded without usage.
+    assert summary['tokens'] == {'total_tokens': 30, 'calls': 2, 'calls_with_usage': 1}
+
+
+@pytest.mark.asyncio
+async def test_overview_metrics_keep_message_rollup(engine, service):
+    await _seed(engine)
+
+    overview = await service.get_overview_metrics(WORKSPACE)
+
+    assert overview['total_messages'] == 3
+    assert overview['llm_calls'] == 2
+    assert overview['model_calls'] == 2
+    assert overview['active_sessions'] == 1
+
+
+@pytest.mark.asyncio
+async def test_execution_detail_for_agent_and_pipeline(engine, service):
+    await _seed(engine)
+
+    agent_detail = await service.get_execution_detail(WORKSPACE, 'agent', 'run-done')
+    assert agent_detail['run']['run_id'] == 'run-done'
+    assert agent_detail['events'][0]['type'] == 'run.completed'
+
+    pipeline_detail = await service.get_execution_detail(WORKSPACE, 'pipeline', 'msg-user')
+    assert pipeline_detail['message']['id'] == 'msg-user'
+    assert [call['id'] for call in pipeline_detail['llm_calls']] == ['llm-1']
+    assert pipeline_detail['tool_calls'] == []
+
+    with pytest.raises(ValueError):
+        await service.get_execution_detail(WORKSPACE, 'agent', 'missing-run')
+
+    with pytest.raises(ValueError):
+        await service.get_execution_detail(WORKSPACE, 'pipeline', 'missing-message')
+
+
+@pytest.mark.asyncio
+async def test_explicit_deliveries_context_and_tenant_isolation(engine, service):
+    import sqlalchemy as sa
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_monitoring.MonitoringMessage)
+            .where(persistence_monitoring.MonitoringMessage.id == 'msg-assistant')
+            .values(parent_message_id='msg-user')
+        )
+        await session.commit()
+    detail = await service.get_execution_detail(WORKSPACE, 'auto', 'msg-assistant')
+    assert detail['row']['id'] == 'msg-user'
+    assert [m['id'] for m in detail['pages']['deliveries']['items']] == ['msg-assistant']
+    assert detail['pages']['inputs']['items'][0]['id'] == 'msg-user'
+    assert detail['pages']['conversation']['items']
+    with pytest.raises(ValueError, match='not found'):
+        await service.get_execution_detail('another-workspace', 'auto', 'msg-assistant')
+
+
+@pytest.mark.asyncio
+async def test_legacy_replies_are_context_not_guessed_deliveries(engine, service):
+    await _seed(engine)
+    detail = await service.get_execution_detail(WORKSPACE, 'pipeline', 'msg-user')
+    assert detail['legacy_context'] is True
+    assert detail['pages']['deliveries']['items'] == []
+    assert 'msg-assistant' in [m['id'] for m in detail['pages']['conversation']['items']]
+
+
+@pytest.mark.asyncio
+async def test_fanout_steering_orphans_and_paged_trace(engine, service):
+    import json
+    import sqlalchemy as sa
+    from langbot.pkg.entity.persistence.event_log import EventLog
+    from langbot.pkg.entity.persistence.transcript import Transcript
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_agent_run.AgentRun)
+            .where(persistence_agent_run.AgentRun.run_id == 'run-done')
+            .values(
+                conversation_id='shared',
+                thread_id='thread-1',
+                metadata_json=json.dumps({'processor_type': 'event_processor'}),
+            )
+        )
+        session.add_all(
+            [
+                EventLog(
+                    event_id='evt-1',
+                    event_type='message.received',
+                    source='platform',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    input_json='{"text":"original input"}',
+                    actor_name='Alice',
+                    created_at=_dt(10),
+                ),
+                EventLog(
+                    event_id='steer',
+                    event_type='message.received',
+                    source='platform',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    run_id='run-done',
+                    input_json='{"text":"additional input"}',
+                    created_at=_dt(10, 1),
+                ),
+                EventLog(
+                    event_id='unrouted',
+                    event_type='group.member.joined',
+                    source='platform',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    metadata_json='{"status":"ignored"}',
+                    created_at=_dt(10, 7),
+                ),
+                persistence_agent_run.AgentRun(
+                    run_id='sibling',
+                    event_id='evt-1',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    runner_id='plugin',
+                    status='completed',
+                    created_at=_dt(10),
+                ),
+                Transcript(
+                    transcript_id='output',
+                    event_id='generated',
+                    conversation_id='shared',
+                    thread_id='thread-1',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    role='assistant',
+                    content='generated only',
+                    run_id='run-done',
+                    seq=1,
+                ),
+                Transcript(
+                    transcript_id='other-thread',
+                    event_id='other',
+                    conversation_id='shared',
+                    thread_id='thread-2',
+                    workspace_id=WORKSPACE,
+                    bot_id='bot-1',
+                    role='user',
+                    content='private other thread',
+                    seq=2,
+                ),
+            ]
+        )
+        session.add_all(
+            [
+                persistence_agent_run.AgentRunEvent(run_id='run-done', sequence=i, type='message.delta', data_json='{}')
+                for i in range(2, 206)
+            ]
+        )
+        await session.commit()
+    result = await service.get_executions(WORKSPACE)
+    assert 'unrouted' in [r['id'] for r in result['items']]
+    assert 'steer' not in [r['id'] for r in result['items']]
+    ignored = await service.get_executions(WORKSPACE, statuses=['ignored'])
+    assert 'unrouted' in [r['id'] for r in ignored['items']]
+    detail = await service.get_execution_detail(WORKSPACE, 'agent', 'run-done')
+    assert detail['row']['target_kind'] == 'event_processor'
+    assert detail['row']['user_name'] == 'Alice'
+    assert [i['id'] for i in detail['pages']['inputs']['items']] == ['evt-1', 'steer']
+    assert detail['pages']['outputs']['items'][0]['content'] == 'generated only'
+    assert [i['id'] for i in detail['pages']['related']['items']] == ['sibling']
+    assert 'other-thread' not in [i['id'] for i in detail['pages']['conversation']['items']]
+    assert len(detail['pages']['events']['items']) == 100
+    assert detail['pages']['events']['has_more']
+    final = await service.get_execution_detail(WORKSPACE, 'agent', 'run-done', section='events', offset=200)
+    assert len(final['pages']['events']['items']) == 5
+    assert not final['pages']['events']['has_more']
+    orphan = await service.get_execution_detail(WORKSPACE, 'event', 'unrouted')
+    assert orphan['pages']['deliveries']['items'] == []
+    assert orphan['pages']['conversation']['items'] == []
+
+
+@pytest.mark.asyncio
+async def test_pipeline_runner_is_not_counted_twice(engine, service):
+    import sqlalchemy as sa
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_monitoring.MonitoringMessage)
+            .where(persistence_monitoring.MonitoringMessage.id == 'msg-user')
+            .values(run_id='run-done')
+        )
+        await session.commit()
+    result = await service.get_executions(WORKSPACE)
+    assert result['total'] == 4
+    assert result['summary']['executions']['total'] == 4
+    assert 'run-done' not in [r['id'] for r in result['items']]
+    detail = await service.get_execution_detail(WORKSPACE, 'pipeline', 'msg-user')
+    assert detail['row']['duration_ms'] == 2000
+    assert detail['events'][0]['type'] == 'run.completed'
+
+
+@pytest.mark.asyncio
+async def test_event_log_fanout_is_idempotent_and_cannot_cross_tenants(engine):
+    from langbot.pkg.agent.runner.event_log_store import EventLogStore
+
+    store = EventLogStore(engine)
+    await store.append_event('event-id', 'message.received', 'platform', workspace_id=WORKSPACE, bot_id='bot')
+    await store.append_event(
+        'event-id',
+        'message.received',
+        'platform',
+        workspace_id=WORKSPACE,
+        bot_id='bot',
+        run_id='first-run',
+        input_json={'text': 'input', 'attachments': []},
+    )
+    await store.append_event(
+        'event-id', 'message.received', 'platform', workspace_id=WORKSPACE, bot_id='bot', run_id='second-run'
+    )
+    event = await store.get_event('event-id')
+    assert event['run_id'] == 'first-run'
+    assert event['input_json']['text'] == 'input'
+    with pytest.raises(Exception):
+        await store.append_event('event-id', 'message.received', 'platform', workspace_id='other', bot_id='bot')
+
+
+@pytest.mark.asyncio
+async def test_processor_filter_scopes_each_kind_and_legacy_calls(engine, service):
+    import json
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        session.add(
+            persistence_agent_run.AgentRun(
+                run_id='event-run',
+                workspace_id=WORKSPACE,
+                runner_id='event-runner',
+                status='completed',
+                created_at=_dt(11),
+                metadata_json=json.dumps({'processor_type': 'event_processor', 'processor_id': 'event-processor-1'}),
+            )
+        )
+        session.add(
+            persistence_monitoring.MonitoringLLMCall(
+                id='event-call',
+                bot_id='bot-1',
+                bot_name='Bot One',
+                pipeline_name='Event Processor',
+                session_id='session-event',
+                input_tokens=6,
+                output_tokens=6,
+                duration=1,
+                timestamp=_dt(11),
+                workspace_uuid=WORKSPACE,
+                model_name='model',
+                message_id='event-run',
+                pipeline_id='',
+                status='success',
+                total_tokens=12,
+            )
+        )
+        await session.commit()
+    for processor, expected in [
+        ('agent-1', {'run-done', 'run-debug'}),
+        ('pipeline-1', {'msg-user', 'msg-discarded'}),
+        ('event-processor-1', {'event-run'}),
+        ('missing', set()),
+    ]:
+        result = await service.get_executions(WORKSPACE, pipeline_ids=[processor])
+        assert {row['id'] for row in result['items']} == expected
+        assert result['total'] == result['summary']['executions']['total'] == len(expected)
+    calls, total = await service.get_llm_calls(WORKSPACE, pipeline_ids=['event-processor-1'])
+    assert total == 1
+    assert [row['id'] for row in calls] == ['event-call']
+    other = await service.get_executions('another-workspace', pipeline_ids=['event-processor-1'])
+    assert other['total'] == 0
+
+
+@pytest.mark.asyncio
+async def test_execution_filters_also_scope_cards_calls_and_traffic(engine, service):
+    import sqlalchemy as sa
+    from langbot.pkg.api.http.service.monitoring_traffic import get_traffic_series
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_monitoring.MonitoringLLMCall)
+            .where(persistence_monitoring.MonitoringLLMCall.id == 'llm-1')
+            .values(message_id='run-failed', pipeline_id='')
+        )
+        await session.commit()
+    failed = await service.get_executions(WORKSPACE, statuses=['failed'])
+    assert failed['total'] == failed['summary']['executions']['total'] == 1
+    assert failed['summary']['executions']['waiting'] == 0
+    calls, total = await service.get_llm_calls(WORKSPACE, execution_statuses=['failed'])
+    assert total == 1 and calls[0]['id'] == 'llm-1'
+    # A successful model call may belong to a failed execution.
+    assert calls[0]['status'] == 'success'
+    overview = await service.get_overview_metrics(WORKSPACE, execution_statuses=['failed'])
+    assert overview['llm_calls'] == 1 and overview['total_messages'] == 0
+    traffic = await get_traffic_series(service.ap, WORKSPACE, execution_statuses=['failed'])
+    assert sum(p['llm_calls'] for p in traffic['points']) == 1
+    assert sum(p['messages'] for p in traffic['points']) == 0
+
+    debug = await service.get_executions(WORKSPACE, mode='debug')
+    assert debug['total'] == debug['summary']['executions']['total'] == 1
+    real = await service.get_executions(WORKSPACE, mode='real')
+    assert real['total'] == real['summary']['executions']['total'] == 4
+    traffic = await get_traffic_series(service.ap, WORKSPACE, mode='debug')
+    assert sum(p['messages'] for p in traffic['points']) == 0
+
+
+@pytest.mark.asyncio
+async def test_pipeline_debug_is_consistent_across_rows_rollups_and_telemetry(engine, service):
+    import sqlalchemy as sa
+    from langbot.pkg.api.http.service.monitoring_traffic import get_traffic_series
+
+    await _seed(engine)
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_monitoring.MonitoringMessage)
+            .where(persistence_monitoring.MonitoringMessage.id == 'msg-user')
+            .values(bot_id='websocket-proxy-bot')
+        )
+        await session.execute(
+            sa.update(persistence_monitoring.MonitoringLLMCall)
+            .where(persistence_monitoring.MonitoringLLMCall.id == 'llm-1')
+            .values(message_id='msg-user', pipeline_id='pipeline-1')
+        )
+        await session.execute(
+            sa.update(persistence_agent_run.AgentRun)
+            .where(persistence_agent_run.AgentRun.run_id == 'run-done')
+            .values(bot_id='websocket-proxy-bot')
+        )
+        await session.commit()
+    debug = await service.get_executions(WORKSPACE, mode='debug')
+    ids = {r['id'] for r in debug['items']}
+    assert ids == {'msg-user', 'run-done', 'run-debug'}
+    assert all(r['debug'] for r in debug['items'])
+    assert debug['summary']['executions']['total'] == 3
+    assert debug['summary']['executions']['debug'] == 3
+    real = await service.get_executions(WORKSPACE, mode='real')
+    assert not ids.intersection(r['id'] for r in real['items'])
+    assert real['summary']['executions']['debug'] == 0
+    filtered = await service.get_executions(WORKSPACE, pipeline_ids=['pipeline-1'], mode='debug')
+    assert [r['id'] for r in filtered['items']] == ['msg-user']
+    calls, total = await service.get_llm_calls(WORKSPACE, pipeline_ids=['pipeline-1'], mode='debug')
+    assert total == 1 and calls[0]['id'] == 'llm-1'
+    real_rows, real_calls = await service.get_llm_calls(WORKSPACE, pipeline_ids=['pipeline-1'], mode='real')
+    assert real_calls == 1 and real_rows[0]['id'] == 'llm-2'
+    traffic = await get_traffic_series(service.ap, WORKSPACE, mode='debug', pipeline_ids=['pipeline-1'])
+    assert sum(p['llm_calls'] for p in traffic['points']) == 1
+
+
+@pytest.mark.asyncio
+async def test_inflight_snapshot_is_scoped_and_keeps_terminal_transition(engine, service):
+    import sqlalchemy as sa
+    from unittest.mock import AsyncMock
+
+    service._get_execution_summary = AsyncMock(side_effect=AssertionError('No dashboard aggregates'))
+    service._get_token_coverage = AsyncMock(side_effect=AssertionError('No dashboard aggregates'))
+    async with AsyncSession(engine) as session:
+        for workspace, run_id in [(WORKSPACE, 'flight-visible'), ('other-workspace', 'flight-secret')]:
+            session.add(
+                persistence_agent_run.AgentRun(
+                    run_id=run_id,
+                    workspace_id=workspace,
+                    runner_id='runner',
+                    status='running',
+                    created_at=_dt(1),
+                )
+            )
+        session.add(persistence_agent_run.AgentRunEvent(run_id='flight-visible', sequence=1, type='tool.started'))
+        await session.commit()
+    # SSE loaders execute after the HTTP handler's tenant scope has exited.
+    # Cloud persistence rejects implicit execute_async calls in that context.
+    from contextlib import asynccontextmanager
+
+    original_execute = service.ap.persistence_mgr.execute_async
+    unscoped_execute = AsyncMock(side_effect=AssertionError('Cloud persistence requires an explicit tenant scope'))
+
+    @asynccontextmanager
+    async def tenant_scope(workspace):
+        assert workspace == WORKSPACE
+        service.ap.persistence_mgr.execute_async = original_execute
+        try:
+            yield
+        finally:
+            service.ap.persistence_mgr.execute_async = unscoped_execute
+
+    service.ap.persistence_mgr.tenant_scope = tenant_scope
+    service.ap.persistence_mgr.execute_async = unscoped_execute
+    snapshot = await service.get_inflight_snapshot(WORKSPACE)
+    rows = {r['id']: r for r in snapshot['items']}
+    assert 'flight-secret' not in rows
+    assert rows['flight-visible']['progress_event'] == 'tool.started'
+    async with AsyncSession(engine) as session:
+        await session.execute(
+            sa.update(persistence_agent_run.AgentRun)
+            .where(persistence_agent_run.AgentRun.run_id == 'flight-visible')
+            .values(status='completed')
+        )
+        await session.commit()
+    next_snapshot = await service.get_inflight_snapshot(WORKSPACE, [rows['flight-visible']])
+    assert next(r for r in next_snapshot['items'] if r['id'] == 'flight-visible')['status_group'] == 'completed'
+
+
+@pytest.mark.asyncio
+async def test_inflight_viewers_share_work_and_cleanup():
+    import asyncio
+    from unittest.mock import AsyncMock
+    from langbot.pkg.utils.inflight import InflightHub
+
+    hub = InflightHub()
+    loader = AsyncMock(return_value={'items': []})
+    first = hub.watch('one', loader)
+    second = hub.watch('one', loader)
+    assert (await anext(first))['kind'] == 'snapshot'
+    assert (await anext(second))['kind'] == 'snapshot'
+    assert loader.await_count == 1
+    other_loader = AsyncMock(return_value={'items': ['private']})
+    other = hub.watch('two', other_loader)
+    assert (await anext(other))['data']['items'] == ['private']
+    hub.track('one', ['run'])
+    hub.notify_run('run', 'tool.started')
+    loader.return_value = {'items': ['updated']}
+    frame = await asyncio.wait_for(anext(first), 2)
+    assert frame['data']['items'] == ['updated']
+    assert (await anext(second)) == frame
+    await first.aclose()
+    assert 'one' in hub.channels
+    await second.aclose()
+    assert 'one' not in hub.channels and 'run' not in hub.run_workspaces
+    await other.aclose()
+    assert not hub.channels
+
+
+@pytest.mark.asyncio
+async def test_inflight_snapshot_failure_is_logged_and_recovers(caplog):
+    import asyncio
+    from unittest.mock import AsyncMock
+    from langbot.pkg.utils.inflight import InflightHub
+
+    hub = InflightHub()
+    snapshot = {'items': [], 'active_total': 0, 'truncated': False}
+    loader = AsyncMock(side_effect=[RuntimeError('snapshot database failure'), snapshot])
+    stream = hub.watch('workspace', loader)
+    try:
+        assert await asyncio.wait_for(anext(stream), 2) == {'kind': 'unavailable'}
+        errors = [
+            record for record in caplog.records if record.message == 'Failed to load in-flight monitoring snapshot'
+        ]
+        assert len(errors) == 1
+        assert errors[0].exc_info[0] is RuntimeError
+        hub.notify('workspace')
+        assert await asyncio.wait_for(anext(stream), 2) == {'kind': 'snapshot', 'data': snapshot}
+    finally:
+        await stream.aclose()
+    assert not hub.channels
diff --git a/tests/unit_tests/api/service/test_monitoring_sessions.py b/tests/unit_tests/api/service/test_monitoring_sessions.py
index 7451c9124..868f51589 100644
--- a/tests/unit_tests/api/service/test_monitoring_sessions.py
+++ b/tests/unit_tests/api/service/test_monitoring_sessions.py
@@ -183,6 +183,31 @@ async def test_analysis_fails_closed_and_scopes_statistics(service):
     assert result['session']['bot_id'] == 'b'
 
 
+async def test_single_message_analysis_without_tool_calls(service):
+    from langbot.pkg.api.http.controller.groups.monitoring import validate_monitoring_window
+    from werkzeug.exceptions import BadRequest
+
+    await service.record_session_start(context('a'), session_id='person_42', **resource('a'))
+    await service.record_message(context('a'), session_id='person_42', message_content='hi', **resource('a'))
+    row = (await service.ap.persistence_mgr.execute_async(sa.select(models.MonitoringMessage.timestamp))).one()
+    timestamp = row.timestamp
+    validate_monitoring_window(timestamp, timestamp, allow_point_window=True)
+    result = await service.get_session_analysis(
+        context(), 'person_42', bot_id='a', start_time=timestamp, end_time=timestamp
+    )
+    assert result['found'] is True
+    assert result['message_stats']['total'] == 1
+    assert result['tool_calls'] == []
+    for start, end in [
+        (timestamp + dt.timedelta(seconds=1), timestamp),
+        (timestamp, timestamp + dt.timedelta(days=366)),
+    ]:
+        with pytest.raises(BadRequest):
+            validate_monitoring_window(start, end, allow_point_window=True)
+    with pytest.raises(BadRequest):
+        validate_monitoring_window(timestamp, timestamp)
+
+
 async def test_activity_requires_bot_and_upsert_counts_racing_first_queries(service):
     for _ in range(2):
         await service.record_session_start(context('a'), session_id='person_42', **resource('a'))
diff --git a/tests/unit_tests/api/service/test_monitoring_tenancy.py b/tests/unit_tests/api/service/test_monitoring_tenancy.py
index 2e2b4216f..e745757c1 100644
--- a/tests/unit_tests/api/service/test_monitoring_tenancy.py
+++ b/tests/unit_tests/api/service/test_monitoring_tenancy.py
@@ -423,3 +423,21 @@ async def test_cleanup_commits_sqlite_delete_before_vacuum(tmp_path):
         assert remaining == 3
     finally:
         await engine.dispose()
+
+
+async def test_cloud_cleanup_uses_plan_retention_and_isolates_workspaces(service):
+    from unittest.mock import AsyncMock
+
+    service.ap.deployment = SimpleNamespace(mode='cloud')
+    service.ap.entitlement_resolver = SimpleNamespace(resolve=AsyncMock(return_value=SimpleNamespace(limits={'chat_history.retention_days': 7})))
+    old = datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None) - datetime.timedelta(days=10)
+    for workspace in (WORKSPACE_A, WORKSPACE_B):
+        await service.ap.persistence_mgr.execute_async(sqlalchemy.insert(MonitoringMessage).values(
+            id=f'policy-{workspace}', workspace_uuid=workspace, timestamp=old,
+            bot_id='bot', bot_name='Bot', pipeline_id='pipeline', pipeline_name='Pipeline',
+            message_content='history', session_id='session', status='success', level='info',
+        ))
+    deleted = await service.cleanup_expired_records(_context(WORKSPACE_A), retention_days=90)
+    assert deleted['monitoring_messages'] == 1
+    rows = await service.ap.persistence_mgr.execute_async(sqlalchemy.select(MonitoringMessage.workspace_uuid))
+    assert rows.scalars().all() == [WORKSPACE_B]
diff --git a/tests/unit_tests/api/service/test_pipeline_migration_resources.py b/tests/unit_tests/api/service/test_pipeline_migration_resources.py
index aa169a569..c0b865ff6 100644
--- a/tests/unit_tests/api/service/test_pipeline_migration_resources.py
+++ b/tests/unit_tests/api/service/test_pipeline_migration_resources.py
@@ -103,20 +103,28 @@ async def test_valid_local_resources_and_real_descriptor_options(local):
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize('resource', ['model', 'kb', 'mcp', 'tool'])
-async def test_foreign_or_unresolved_resource_is_denied(local, resource):
+async def test_foreign_or_unresolved_model_is_denied(local):
     env, _, config, plan = local
-    if resource == 'model':
-        config['model']['primary'] = 'model_foreign'
-    elif resource == 'kb':
+    config['model']['primary'] = 'model_foreign'
+    async with env.pm.tenant_scope(WS):
+        with pytest.raises(env.m.MigrationError, match='runner_resource_unavailable'):
+            await env.svc._verify_runtime(ExecutionContext.from_request(context()), plan)
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('resource', ['kb', 'mcp', 'tool'])
+async def test_foreign_or_unresolved_reference_is_pruned(local, resource):
+    env, _, config, plan = local
+    if resource == 'kb':
         config['knowledge-bases'] = ['kb_foreign']
     elif resource == 'mcp':
         config['mcp-resources'][0]['server_uuid'] = 'mcp_foreign'
     else:
         config['tools'] = ['foreign_tool']
     async with env.pm.tenant_scope(WS):
-        with pytest.raises(env.m.MigrationError, match='runner_resource_unavailable'):
-            await env.svc._verify_runtime(ExecutionContext.from_request(context()), plan)
+        await env.svc._verify_runtime(ExecutionContext.from_request(context()), plan)
+    key = {'kb': 'knowledge-bases', 'mcp': 'mcp-resources', 'tool': 'tools'}[resource]
+    assert config[key] == []
 
 
 @pytest.mark.asyncio
diff --git a/tests/unit_tests/api/service/test_provider_cascade.py b/tests/unit_tests/api/service/test_provider_cascade.py
index 1fd58e98d..971dafc3b 100644
--- a/tests/unit_tests/api/service/test_provider_cascade.py
+++ b/tests/unit_tests/api/service/test_provider_cascade.py
@@ -15,6 +15,7 @@ from langbot.pkg.api.http.controller.groups.provider.providers import ModelProvi
 from langbot.pkg.api.http.service.provider import ModelProviderService
 from langbot.pkg.entity.persistence.model import CodexCredential, EmbeddingModel, LLMModel, ModelProvider, RerankModel
 from langbot.pkg.entity.persistence.user import User
+from langbot.pkg.entity.persistence.metadata import WorkspaceMetadata
 from langbot.pkg.entity.persistence.workspace import Workspace
 from langbot.pkg.persistence.mgr import PersistenceManager, PersistenceMode
 from langbot.pkg.provider.modelmgr.modelmgr import ModelManager
@@ -51,7 +52,7 @@ async def deletion(tmp_path):
             }
 
     async with engine.begin() as conn:
-        for table in (User, Workspace, ModelProvider, CodexCredential, *MODEL_TYPES):
+        for table in (User, Workspace, WorkspaceMetadata, ModelProvider, CodexCredential, *MODEL_TYPES):
             await conn.run_sync(table.__table__.create)
         for workspace in contexts:
             await conn.execute(
diff --git a/tests/unit_tests/api/service/test_session_context_reset.py b/tests/unit_tests/api/service/test_session_context_reset.py
new file mode 100644
index 000000000..ece96c7bd
--- /dev/null
+++ b/tests/unit_tests/api/service/test_session_context_reset.py
@@ -0,0 +1,148 @@
+import contextlib
+import datetime
+import json
+from types import SimpleNamespace
+
+import pytest
+import sqlalchemy as sa
+from sqlalchemy.ext.asyncio import create_async_engine
+
+from langbot.pkg.api.http.service.monitoring import MonitoringService
+from langbot.pkg.agent.runner.context_reset import get_reset_generation
+from langbot.pkg.agent.runner.transcript_store import TranscriptStore
+from langbot.pkg.entity.persistence.base import Base
+from langbot.pkg.entity.persistence.monitoring import MonitoringSession
+from langbot.pkg.entity.persistence.runner_state import RunnerState
+from langbot.pkg.entity.persistence.agent_run import AgentRun
+from langbot.pkg.entity.persistence.transcript import Transcript
+
+
+class Persistence:
+    def __init__(self, engine):
+        self.engine = engine
+
+    def get_db_engine(self):
+        return self.engine
+
+    @contextlib.asynccontextmanager
+    async def tenant_uow(self, workspace):
+        async with self.engine.begin() as conn:
+            self.conn = conn
+            yield
+
+    async def execute_async(self, statement):
+        return await self.conn.execute(statement)
+
+
+@pytest.fixture
+async def context_service(tmp_path):
+    engine = create_async_engine(f'sqlite+aiosqlite:///{tmp_path / "reset.db"}')
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+        await conn.execute(
+            sa.insert(MonitoringSession).values(
+                workspace_uuid='w',
+                bot_id='b',
+                session_id='person_u',
+                bot_name='Bot',
+                pipeline_id='p',
+                pipeline_name='Pipeline',
+                start_time=datetime.datetime.utcnow(),
+                last_activity=datetime.datetime.utcnow(),
+            )
+        )
+        for workspace, bot, conversation, scope in [
+            ('w', 'b', 'person_u', 'conversation'),
+            ('w', 'b2', 'person_u', 'conversation'),
+            ('w2', 'b', 'person_u', 'conversation'),
+            ('w', 'b', 'person_other', 'conversation'),
+            ('w', 'b', 'person_u', 'actor'),
+        ]:
+            await conn.execute(
+                sa.insert(RunnerState).values(
+                    workspace_id=workspace,
+                    bot_id=bot,
+                    conversation_id=conversation,
+                    scope=scope,
+                    scope_key=f'{workspace}:{bot}:{conversation}:{scope}',
+                    state_key='checkpoint',
+                    runner_id='runner',
+                    binding_identity='binding',
+                    value_json=json.dumps('old'),
+                )
+            )
+    yield MonitoringService(SimpleNamespace(persistence_mgr=Persistence(engine))), engine
+    await engine.dispose()
+
+
+@pytest.mark.asyncio
+async def test_reset_preserves_logs_and_isolates_context(context_service):
+    service, engine = context_service
+    store = TranscriptStore(engine)
+    for bot in ['b', 'b2']:
+        await store.append_transcript(None, 'event', 'person_u', 'user', bot_id=bot, workspace_id='w', content='old')
+    await service.reset_session_context('w', 'b', 'person_u')
+    assert await get_reset_generation(engine, 'w', 'b', 'person_u')
+    async with engine.connect() as conn:
+        assert (await conn.execute(sa.select(sa.func.count()).select_from(Transcript))).scalar() == 2
+        assert (await conn.execute(sa.select(sa.func.count()).select_from(MonitoringSession))).scalar() == 1
+        assert (
+            await conn.execute(
+                sa.select(sa.func.count()).select_from(RunnerState).where(RunnerState.runner_id == 'runner')
+            )
+        ).scalar() == 4
+    page = await store.page_transcript('person_u', bot_id='b', workspace_id='w')
+    assert not page[0]
+    assert (await store.page_transcript('person_u', bot_id='b2', workspace_id='w'))[0]
+    await store.append_transcript(None, 'new', 'person_u', 'user', bot_id='b', workspace_id='w', content='fresh')
+    assert len((await store.page_transcript('person_u', bot_id='b', workspace_id='w'))[0]) == 1
+    await service.reset_session_context('w', 'b', 'person_u')
+    assert not (await store.page_transcript('person_u', bot_id='b', workspace_id='w'))[0]
+
+
+@pytest.mark.asyncio
+async def test_reset_refuses_wrong_scope_and_running_tasks(context_service):
+    service, engine = context_service
+    for workspace, bot in [('other', 'b'), ('w', 'other')]:
+        with pytest.raises(LookupError):
+            await service.reset_session_context(workspace, bot, 'person_u')
+    async with engine.begin() as conn:
+        await conn.execute(
+            sa.insert(AgentRun).values(
+                run_id='run',
+                event_id='evt',
+                agent_id='agent',
+                binding_id='binding',
+                runner_id='runner',
+                workspace_id='w',
+                bot_id='b',
+                conversation_id='person_u',
+                status='running',
+            )
+        )
+    with pytest.raises(RuntimeError):
+        await service.reset_session_context('w', 'b', 'person_u')
+    assert await get_reset_generation(engine, 'w', 'b', 'person_u') is None
+
+
+@pytest.mark.asyncio
+async def test_reset_starts_new_external_runner_conversation(context_service):
+    from unit_tests.agent.test_state_store import FakeBinding, FakeEventEnvelope, make_descriptor
+    from langbot.pkg.agent.runner.persistent_state_store import PersistentStateStore
+
+    service, engine = context_service
+    store = PersistentStateStore(engine)
+    event = FakeEventEnvelope(workspace_id='w', bot_id='b', conversation_id='person_u')
+    binding = FakeBinding()
+    descriptor = make_descriptor()
+    await store.apply_update_from_event(
+        event, binding, descriptor, 'conversation', 'external.conversation_id', 'remote-old', None
+    )
+    assert (await store.build_snapshot_from_event(event, binding, descriptor))['conversation'][
+        'external.conversation_id'
+    ] == 'remote-old'
+    await service.reset_session_context('w', 'b', 'person_u')
+    snapshot = await store.build_snapshot_from_event(event, binding, descriptor)
+    assert snapshot['conversation']['external.conversation_id'] == await get_reset_generation(
+        engine, 'w', 'b', 'person_u'
+    )
diff --git a/tests/unit_tests/api/service/test_space_service.py b/tests/unit_tests/api/service/test_space_service.py
index a05754af5..8dd1543e7 100644
--- a/tests/unit_tests/api/service/test_space_service.py
+++ b/tests/unit_tests/api/service/test_space_service.py
@@ -1025,3 +1025,45 @@ class TestSpaceServiceCreditsCache:
         assert result == 500
         assert 'test@example.com' in service._credits_cache
         assert service._credits_cache['test@example.com'][0] == 500
+
+
+@pytest.mark.parametrize('status', [200, 503])
+async def test_assistant_recommendation_uses_dedicated_endpoint_without_fallback(status):
+    service = SpaceService(SimpleNamespace(instance_config=SimpleNamespace(data={})))
+    service.get_model_selection = AsyncMock()
+    service._resolve_recommended_model = AsyncMock(return_value={'uuid': 'local', 'name': 'deepseek-v4-flash'})
+    response = MagicMock(status=status)
+    _set_response_body(
+        response,
+        {
+            'code': 0,
+            'data': {
+                'model': {
+                    'uuid': 'upstream',
+                    'model_id': 'deepseek-v4-flash',
+                    'category': 'chat',
+                    'provider': 'deepseek',
+                    'status': 'active',
+                }
+            },
+        },
+    )
+    session = MagicMock()
+    session.get.return_value.__aenter__ = AsyncMock(return_value=response)
+    session.get.return_value.__aexit__ = AsyncMock(return_value=None)
+    context = object()
+    with patch('langbot.pkg.api.http.service.space.httpclient.get_session', return_value=session):
+        if status == 200:
+            assert await service.get_recommended_assistant_model(context) == {
+                'uuid': 'local',
+                'name': 'deepseek-v4-flash',
+            }
+            args = service._resolve_recommended_model.await_args.args
+            assert args[0] is context
+            assert args[1].model_id == 'deepseek-v4-flash'
+        else:
+            with pytest.raises(ValueError, match='No recommended assistant model'):
+                await service.get_recommended_assistant_model(context)
+            service._resolve_recommended_model.assert_not_awaited()
+    service.get_model_selection.assert_not_awaited()
+    session.get.assert_called_once_with('https://space.langbot.app/api/v1/models/assistant/recommendation')
diff --git a/tests/unit_tests/api/service/test_user_service.py b/tests/unit_tests/api/service/test_user_service.py
index 12b440b32..add2ea7b3 100644
--- a/tests/unit_tests/api/service/test_user_service.py
+++ b/tests/unit_tests/api/service/test_user_service.py
@@ -357,7 +357,7 @@ class TestUserServiceAuthenticate:
         service = UserService(ap)
 
         # Execute & Verify
-        with pytest.raises(ValueError, match='用户不存在'):
+        with pytest.raises(ValueError, match='Resource not found'):
             await service.authenticate('nonexistent@example.com', 'password')
 
     async def test_authenticate_space_user_without_password_raises_error(self):
@@ -377,7 +377,7 @@ class TestUserServiceAuthenticate:
         service = UserService(ap)
 
         # Execute & Verify
-        with pytest.raises(ValueError, match='请使用 LangBot 账号登录'):
+        with pytest.raises(ValueError, match='Please sign in with your LangBot Account'):
             await service.authenticate('space@example.com', 'password')
 
 
@@ -418,7 +418,6 @@ class TestUserServiceGenerateJwtToken:
         assert token is not None
 
 
-
 class TestUserServiceVerifyJwtToken:
     """Tests for verify_jwt_token method."""
 
diff --git a/tests/unit_tests/api/service/test_wizard_regressions.py b/tests/unit_tests/api/service/test_wizard_regressions.py
new file mode 100644
index 000000000..e79ac43ef
--- /dev/null
+++ b/tests/unit_tests/api/service/test_wizard_regressions.py
@@ -0,0 +1,20 @@
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+import sqlalchemy
+
+from langbot.pkg.entity.persistence.metadata import Metadata, WorkspaceMetadata
+from langbot.pkg.api.http.service.bot import BotService
+
+
+def test_metadata_supports_long_wizard_json():
+    for model in (Metadata, WorkspaceMetadata):
+        assert isinstance(model.__table__.c.value.type, sqlalchemy.Text)
+
+
+@pytest.mark.asyncio
+async def test_saved_bot_without_runtime_has_empty_logs():
+    service = BotService(SimpleNamespace(platform_mgr=SimpleNamespace(get_bot_by_uuid=AsyncMock(return_value=None))))
+    service.get_bot = AsyncMock(return_value={'uuid': 'bot'})
+    assert await service.list_event_logs('workspace', 'bot', 0, 20) == ([], 0)
diff --git a/tests/unit_tests/api/test_bots_controller_errors.py b/tests/unit_tests/api/test_bots_controller_errors.py
index 12920a83b..1d8730c61 100644
--- a/tests/unit_tests/api/test_bots_controller_errors.py
+++ b/tests/unit_tests/api/test_bots_controller_errors.py
@@ -106,6 +106,7 @@ async def test_runtime_error_redacts_persisted_secrets_on_partial_update():
 
     ap = SimpleNamespace(
         persistence_mgr=SimpleNamespace(execute_async=AsyncMock(return_value=SimpleNamespace(rowcount=1))),
+        sess_mgr=SimpleNamespace(session_list=[]),
         platform_mgr=SimpleNamespace(
             remove_bot=AsyncMock(),
             load_bot=AsyncMock(side_effect=ValueError('Invalid app_secret: persisted-secret-value')),
@@ -113,7 +114,12 @@ async def test_runtime_error_redacts_persisted_secrets_on_partial_update():
     )
     service = BotService(ap)
     service.get_bot = AsyncMock(
-        return_value={'uuid': 'bot-1', 'adapter_config': {'app_secret': 'persisted-secret-value'}}
+        return_value={
+            'uuid': 'bot-1',
+            'adapter': 'aiocqhttp',
+            'enable': True,
+            'adapter_config': {'app_secret': 'persisted-secret-value'},
+        }
     )
     with pytest.raises(BotApplyError) as captured:
         await service.update_bot('workspace-test', 'bot-1', {'enable': True})
diff --git a/tests/unit_tests/api/test_diagnostics_management_http.py b/tests/unit_tests/api/test_diagnostics_management_http.py
deleted file mode 100644
index 5c7186d6c..000000000
--- a/tests/unit_tests/api/test_diagnostics_management_http.py
+++ /dev/null
@@ -1,194 +0,0 @@
-"""Real Quart registration boundaries with content-canary payloads."""
-
-import asyncio
-import json
-import re
-from types import SimpleNamespace
-from unittest.mock import AsyncMock
-
-import pytest
-import quart
-
-from langbot.pkg.api.http.controller.group import AuthType, RouterGroup
-from langbot.pkg.telemetry import diagnostics as d
-
-
-class Recorder:
-    enabled = True
-
-    def __init__(self):
-        self.events = []
-
-    def emit(self, kind, operation, outcome, **fields):
-        # A real transport projects the exception class, never its message.
-        error = fields.pop('error', None)
-        if error is not None:
-            fields['error_type'] = type(error).__name__
-        self.events.append(dict(kind=kind, operation=operation, outcome=outcome, **fields))
-
-
-class Routes(RouterGroup):
-    path = '/management'
-    name = 'management'
-
-    async def initialize(self):
-        @self.route('/ok/<identifier>', auth_type=AuthType.NONE)
-        async def success(identifier):
-            self.ap.seen.append(d.current_span())
-            return self.success({'secret': identifier})
-
-        @self.route('/business', auth_type=AuthType.NONE)
-        async def business():
-            return self.fail('private-error-code', 'private-error-message')
-
-        @self.route('/auth')
-        async def authenticated():
-            raise AssertionError('must not run')
-
-        @self.route('/error', auth_type=AuthType.NONE)
-        async def error():
-            raise ValueError('private-exception-message')
-
-        @self.route('/cancel', auth_type=AuthType.NONE)
-        async def cancel():
-            raise asyncio.CancelledError('private-cancel-message')
-
-        @self.route('/stream', auth_type=AuthType.NONE)
-        async def stream():
-            async def body():
-                self.ap.seen.append(d.current_span())
-                yield b'private-stream-chunk'
-                self.ap.seen.append(d.current_span())
-
-            return quart.Response(body())
-
-
-async def setup(manager=True):
-    app = quart.Quart(__name__)
-    ap = SimpleNamespace(seen=[])
-    if manager is True:
-        manager = Recorder()
-    if manager is not None:
-        ap.diagnostics = manager
-    routes = Routes(ap, app)
-    await routes.initialize()
-    return app, ap, routes
-
-
-@pytest.mark.asyncio
-async def test_http_success_uses_code_identity_not_path_or_payload():
-    app, ap, _ = await setup()
-    response = await app.test_client().get(
-        '/management/ok/private-id?token=private-query', headers={'Authorization': 'private-token'}
-    )
-    assert (await response.get_json())['data']['secret'] == 'private-id'
-    assert [e['outcome'] for e in ap.diagnostics.events] == ['started', 'succeeded']
-    event = ap.diagnostics.events[-1]
-    assert event['source'] == 'http'
-    assert re.fullmatch(r'[A-Za-z_][A-Za-z0-9_.:-]{0,127}', event['operation'])
-    assert ap.seen[0] is not None
-    assert d.current_span() is None
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'path,status,outcome', [('business', 200, 'failed'), ('auth', 401, 'rejected'), ('error', 500, 'failed')]
-)
-async def test_http_business_auth_and_exception_outcomes(path, status, outcome):
-    app, ap, _ = await setup()
-    response = await app.test_client().get('/management/' + path)
-    assert response.status_code == status
-    assert ap.diagnostics.events[-1]['outcome'] == outcome
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-async def test_http_cancellation_propagates():
-    app, ap, _ = await setup()
-    async with app.test_request_context('/management/cancel'):
-        with pytest.raises(asyncio.CancelledError):
-            await app.full_dispatch_request()
-    assert ap.diagnostics.events[-1]['outcome'] == 'cancelled'
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_http_auth_cancellation_not_reinterpreted_as_api_key():
-    app, ap, routes = await setup()
-    routes._authenticate_support_admin = AsyncMock(return_value=None)
-    routes._authenticate_account = AsyncMock(side_effect=asyncio.CancelledError())
-    async with app.test_request_context('/management/auth', headers={'Authorization': 'Bearer private-token'}):
-        with pytest.raises(asyncio.CancelledError):
-            await app.full_dispatch_request()
-    assert ap.diagnostics.events[-1]['outcome'] == 'cancelled'
-
-
-@pytest.mark.asyncio
-async def test_http_stream_has_parent_context_without_consumer_leak():
-    app, ap, _ = await setup()
-    async with app.test_request_context('/management/stream'):
-        response = await app.full_dispatch_request()
-    assert [e['outcome'] for e in ap.diagnostics.events] == ['started']
-    assert d.current_span() is None
-    async with response.response as body:
-        iterator = body.__aiter__()
-        assert await anext(iterator) == b'private-stream-chunk'
-        assert d.current_span() is None
-        assert ap.seen[-1] is not None
-        with pytest.raises(StopAsyncIteration):
-            await anext(iterator)
-    assert ap.seen[0] is ap.seen[1]
-    assert ap.diagnostics.events[-1]['outcome'] == 'succeeded'
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'manager', [None, SimpleNamespace(enabled=False, emit=lambda *a, **kw: pytest.fail('disabled emission'))]
-)
-async def test_absent_disabled_manager_preserves_result_without_context(manager):
-    app, ap, _ = await setup(manager)
-    response = await app.test_client().get('/management/ok/private-id')
-    assert response.status_code == 200
-    assert ap.seen == [None]
-
-
-@pytest.mark.asyncio
-async def test_broken_diagnostic_emit_never_masks_operation():
-    class Broken(Recorder):
-        def emit(self, *args, **kwargs):
-            raise RuntimeError('diagnostics broken')
-
-    app, _, _ = await setup(Broken())
-    response = await app.test_client().get('/management/ok/private-id')
-    assert response.status_code == 200
-    async with app.test_request_context('/management/cancel'):
-        with pytest.raises(asyncio.CancelledError):
-            await app.full_dispatch_request()
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_anonymous_handlers_have_distinct_stable_code_operations():
-    app, ap, routes = await setup()
-
-    @routes.route('/items/<identifier>', auth_type=AuthType.NONE, methods=['GET'])
-    async def _(identifier):
-        return routes.success()
-
-    @routes.route('/items/<identifier>', auth_type=AuthType.NONE, methods=['POST'])
-    async def _(identifier):
-        return routes.success()
-
-    @routes.route('/other', auth_type=AuthType.NONE)
-    async def _():
-        return routes.success()
-
-    await app.test_client().get('/management/items/private-id')
-    await app.test_client().post('/management/items/private-id')
-    await app.test_client().get('/management/other')
-    operations = [e['operation'] for e in ap.diagnostics.events if e['outcome'] == 'started']
-    assert len(set(operations)) == 3
-    assert all(re.fullmatch(r'[A-Za-z_][A-Za-z0-9_.:-]{0,127}', op) for op in operations)
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
diff --git a/tests/unit_tests/api/test_diagnostics_management_inventory.py b/tests/unit_tests/api/test_diagnostics_management_inventory.py
deleted file mode 100644
index 7a97fd175..000000000
--- a/tests/unit_tests/api/test_diagnostics_management_inventory.py
+++ /dev/null
@@ -1,47 +0,0 @@
-"""Verify identities across every source-declared management registration."""
-
-import ast
-from pathlib import Path
-import re
-from types import SimpleNamespace
-
-from langbot.pkg.api.management_diagnostics import operation_id
-
-
-def test_every_http_registration_has_a_distinct_wire_safe_operation():
-    root = Path(__file__).resolve().parents[3] / 'src/langbot/pkg/api/http/controller/groups'
-    identities = []
-    for path in root.rglob('*.py'):
-        module = 'langbot.pkg.api.http.controller.groups.' + '.'.join(path.relative_to(root).with_suffix('').parts)
-        for cls in ast.walk(ast.parse(path.read_text())):
-            if not isinstance(cls, ast.ClassDef):
-                continue
-            prefix = ''
-            for decorator in cls.decorator_list:
-                if (
-                    isinstance(decorator, ast.Call)
-                    and isinstance(decorator.func, ast.Attribute)
-                    and decorator.func.attr == 'group_class'
-                ):
-                    prefix = ast.literal_eval(decorator.args[1])
-            for fn in ast.walk(cls):
-                if not isinstance(fn, ast.AsyncFunctionDef):
-                    continue
-                for decorator in fn.decorator_list:
-                    if not (
-                        isinstance(decorator, ast.Call)
-                        and isinstance(decorator.func, ast.Attribute)
-                        and decorator.func.attr == 'route'
-                    ):
-                        continue
-                    rule = prefix + ast.literal_eval(decorator.args[0])
-                    methods = next(
-                        (ast.literal_eval(k.value) for k in decorator.keywords if k.arg == 'methods'), ['GET']
-                    )
-                    operation = operation_id(
-                        'http', SimpleNamespace(__module__=module, __name__=fn.name), rule=rule, methods=methods
-                    )
-                    assert re.fullmatch(r'[A-Za-z_][A-Za-z0-9_.:-]{0,127}', operation), operation
-                    identities.append(operation)
-    assert len(identities) >= 200  # Guard against accidentally scanning an empty/subset tree.
-    assert len(set(identities)) == len(identities)
diff --git a/tests/unit_tests/api/test_diagnostics_management_isolation.py b/tests/unit_tests/api/test_diagnostics_management_isolation.py
deleted file mode 100644
index c55eb7bb2..000000000
--- a/tests/unit_tests/api/test_diagnostics_management_isolation.py
+++ /dev/null
@@ -1,431 +0,0 @@
-"""CORE-DIAG-6: management ownership barriers under a foreign live span."""
-
-import asyncio
-from types import SimpleNamespace as NS
-from unittest.mock import AsyncMock, patch
-from uuid import uuid4
-
-import pytest
-import quart
-
-from langbot.pkg.api import management_diagnostics as md
-from langbot.pkg.api.http.context import (
-    ExecutionContext,
-    PrincipalContext,
-    PrincipalType,
-    RequestContext,
-    WorkspaceContext,
-)
-from langbot.pkg.api.http.controller.group import AuthType
-from langbot.pkg.api.http.controller.groups.pipelines.embed import EmbedRouterGroup
-from langbot.pkg.api.http.controller.groups.pipelines.websocket_chat import WebSocketChatRouterGroup
-from langbot.pkg.api.http.service.agent import AgentService
-from langbot.pkg.api.mcp.server import LangBotMCPServer
-from langbot.pkg.telemetry import diagnostics as d
-from tests.unit_tests.api.test_diagnostics_management_http import Recorder, Routes
-
-
-MODES = ['enabled', 'disabled', 'absent', 'mismatched', 'policy_off', 'beta_off', 'broken']
-
-
-def app(identity):
-    ap = NS(instance_config=NS(data={'space': {'url': 'https://example.invalid'}}))
-    ap.diagnostics = d.DiagnosticsManager(ap, version='4.11.0b2', instance_id=identity)
-    return ap
-
-
-def owners(mode):
-    a, b = app('instance-A'), app('instance-B')
-    b_manager = b.diagnostics
-    if mode == 'disabled':
-        b.instance_config.data['space']['disable_telemetry'] = True
-        b.diagnostics = d.DiagnosticsManager(b, version='4.11.0b2', instance_id='instance-B')
-        b_manager = b.diagnostics
-    elif mode == 'absent':
-        del b.diagnostics
-    elif mode == 'mismatched':
-        b.diagnostics = a.diagnostics
-    elif mode in {'policy_off', 'beta_off'}:
-        # Also test an attached structural producer whose enabled flag stays true.
-        b.diagnostics = Recorder()
-        b.instance_config.data['space']['disable_telemetry' if mode == 'policy_off' else 'disable_beta_diagnostics'] = (
-            True
-        )
-    elif mode == 'broken':
-        b.diagnostics = NS(enabled=True, emit=lambda *a, **kw: (_ for _ in ()).throw(RuntimeError('broken')))
-    ca = ExecutionContext('instance-A', str(uuid4()), 1)
-    cb = ExecutionContext('instance-B', str(uuid4()), 1)
-    d.privacy.code_value('operation', 'http.isolation.parent')
-    parent = d.Span(a.diagnostics, 'api', 'http.isolation.parent', {'workspace_uuid': ca.workspace_uuid})
-    return a, b, b_manager, ca, cb, parent
-
-
-@d.observe('event', 'platform.target2yiri', source='platform', stage='convert')
-async def converter(native):
-    return native
-
-
-async def nested(b, cb, seen):
-    md.workspace(cb)
-    seen.append(d.current_span())
-    with md.scope(b, 'websocket.isolation.inner', source='websocket'):
-        md.workspace(cb)
-        assert await converter(42) == 42
-    # Both an inner management boundary and an ownerless converter must be safe.
-    assert await converter(43) == 43
-    return 44
-
-
-def assert_isolated(a, b, b_manager, ca, cb, parent, mode):
-    assert parent.fields['workspace_uuid'] == ca.workspace_uuid
-    assert parent.outcome is None
-    assert len(a.diagnostics.pending) == 1  # Only the caller's own start event.
-    assert not b_manager.pending if mode != 'enabled' else b_manager.pending
-    if isinstance(getattr(b, 'diagnostics', None), Recorder):
-        assert b.diagnostics.events == []
-    if mode == 'enabled':
-        assert all(e['instance_id'] == 'instance-B' for e in b_manager.pending)
-        assert all(e['trace_id'] != parent.fields['trace_id'] for e in b_manager.pending)
-        assert all(e['workspace_uuid'] != ca.workspace_uuid for e in b_manager.pending)
-        assert any(e['workspace_uuid'] == cb.workspace_uuid for e in b_manager.pending)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-@pytest.mark.parametrize('shape', ['decorator', 'scope'])
-async def test_nested_management_masks_foreign_parent_and_restores_caller(mode, shape):
-    a, b, manager, ca, cb, parent = owners(mode)
-    seen = []
-
-    @md.observe('http.isolation.endpoint', source='http', ap=b)
-    async def endpoint():
-        return await nested(b, cb, seen)
-
-    with parent.activate():
-        if shape == 'decorator':
-            assert await endpoint() == 44
-        else:
-            with md.scope(b, 'websocket.isolation.outer', source='websocket'):
-                assert await nested(b, cb, seen) == 44
-        assert d.current_span() is parent
-    assert d.current_span() is None
-    assert (seen[0] is not None) == (mode == 'enabled')
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-async def test_actual_agent_debug_never_mutates_foreign_workspace(mode):
-    a, b, manager, ca, cb, parent = owners(mode)
-    service = AgentService(b)
-    seen = []
-
-    async def missing(*args):
-        await nested(b, cb, seen)
-        return None
-
-    service.get_agent = missing
-    with parent.activate():
-        with pytest.raises(ValueError, match='^Agent not found$'):
-            await service.debug_agent(cb, 'private-agent', {})
-        assert d.current_span() is parent
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-@pytest.mark.parametrize('termination', ['exhaust', 'close', 'error', 'cancel'])
-async def test_real_http_body_keeps_barrier_during_advancement_and_cleanup(mode, termination):
-    a, b, manager, ca, cb, parent = owners(mode)
-    web = quart.Quart(__name__)
-    routes = Routes(b, web)
-    seen, closed = [], []
-    failure = asyncio.CancelledError('private-cancel') if termination == 'cancel' else ValueError('private-error')
-
-    @routes.route('/isolated', auth_type=AuthType.NONE)
-    async def endpoint():
-        await nested(b, cb, seen)
-
-        async def stream():
-            try:
-                await nested(b, cb, seen)
-                yield b'one'
-                if termination in {'cancel', 'error'}:
-                    raise failure
-                await nested(b, cb, seen)
-            finally:
-                await nested(b, cb, seen)
-                closed.append(True)
-
-        return quart.Response(stream())
-
-    # Construct with no caller parent; the later body consumer has app A's span.
-    async with web.test_request_context('/management/isolated'):
-        response = await web.full_dispatch_request()
-    with parent.activate():
-        try:
-            async with response.response as body:
-                iterator = body.__aiter__()
-                assert await anext(iterator) == b'one'
-                assert d.current_span() is parent
-                if termination != 'close':
-                    with pytest.raises((StopAsyncIteration, type(failure))) as caught:
-                        await anext(iterator)
-                    if termination in {'cancel', 'error'}:
-                        assert caught.value is failure
-                    else:
-                        assert isinstance(caught.value, StopAsyncIteration)
-        finally:
-            assert d.current_span() is parent
-    assert closed == [True]
-    assert all((span is not None) == (mode == 'enabled') for span in seen)
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-async def test_real_mcp_registered_tool_masks_foreign_parent(mode):
-    a, b, manager, ca, cb, parent = owners(mode)
-    seen = []
-
-    async def get_bot(*args, **kwargs):
-        return {'value': await nested(b, cb, seen)}
-
-    b.bot_service = NS(get_bot=get_bot)
-    server = LangBotMCPServer(b)
-    with parent.activate(), patch('langbot.pkg.api.mcp.server._authorized', return_value=cb):
-        assert await server.mcp.call_tool('get_bot', {'bot_uuid': 'private-bot'})
-        assert d.current_span() is parent
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-@pytest.mark.parametrize('group_class', [WebSocketChatRouterGroup, EmbedRouterGroup])
-async def test_real_websocket_receive_masks_foreign_parent(mode, group_class):
-    a, b, manager, ca, cb, parent = owners(mode)
-    group = group_class(b, quart.Quart(__name__))
-    connection = NS(is_active=True, connection_id='private-id', send_queue=asyncio.Queue())
-    seen = []
-
-    async def handle(*args, **kwargs):
-        await nested(b, cb, seen)
-
-    adapter = NS(handle_websocket_message=handle)
-    if group_class is WebSocketChatRouterGroup:
-        group._revalidate_websocket_authorization = AsyncMock(return_value=cb)
-    else:
-        group._resolve_connected_bot = AsyncMock(return_value=NS(execution_context=cb))
-
-    async def receive():
-        connection.is_active = False
-        return '{"type":"message","text":"private-prompt"}'
-
-    with (
-        parent.activate(),
-        patch('quart.websocket', NS(receive=receive)),
-        patch(
-            'langbot.pkg.api.http.controller.groups.pipelines.websocket_chat.ws_connection_manager.update_activity',
-            new=AsyncMock(),
-        ),
-    ):
-        await group._handle_receive(connection, adapter, NS(execution_context=cb), 'private-token')
-        assert d.current_span() is parent
-    assert seen
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.parametrize('kind', ['execution', 'request'])
-def test_workspace_rejects_foreign_instance_or_foreign_active_span(kind):
-    a, b, _, ca, cb, parent = owners('enabled')
-    context = cb
-    if kind == 'request':
-        context = RequestContext(
-            cb.instance_uuid,
-            1,
-            'request',
-            'api_key',
-            PrincipalContext(PrincipalType.API_KEY),
-            WorkspaceContext(cb.workspace_uuid, None, None, frozenset()),
-        )
-    with md.scope(a, 'http.isolation.annotation', source='http'):
-        span = d.current_span()
-        md.workspace(context)
-        assert 'workspace_uuid' not in span.fields
-        md.workspace(ca)
-        assert span.fields['workspace_uuid'] == ca.workspace_uuid
-        with parent.activate():
-            before = dict(parent.fields)
-            md.workspace(ca)
-            md.workspace(context)
-            assert parent.fields == before
-
-
-def test_structural_recorder_workspace_and_scope_exception_identity():
-    recorder = Recorder()
-    ap = NS(diagnostics=recorder)
-    ctx = ExecutionContext('instance', str(uuid4()), 1)
-    error = ValueError('private-error')
-    with pytest.raises(ValueError) as caught:
-        with md.scope(ap, 'http.isolation.recorder', source='http'):
-            md.workspace(ctx)
-            assert d.current_span().fields['workspace_uuid'] == ctx.workspace_uuid
-            raise error
-    assert caught.value is error
-    assert recorder.events[-1]['outcome'] == 'failed'
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES[1:])
-@pytest.mark.parametrize('termination', ['success', 'error', 'cancel'])
-async def test_nested_off_boundary_restores_outer_annotation_and_error(mode, termination):
-    a, b, manager, ca, cb, parent = owners(mode)
-    error = asyncio.CancelledError('private-cancel') if termination == 'cancel' else ValueError('private-error')
-
-    @md.observe('http.isolation.off', source='http', ap=b)
-    async def endpoint():
-        await nested(b, cb, [])
-        md.outcome('failed')
-        if termination != 'success':
-            raise error
-        return 42
-
-    with parent.activate():
-        with md.scope(a, 'http.isolation.outer', source='http'):
-            outer = d.current_span()
-            if termination == 'success':
-                assert await endpoint() == 42
-            else:
-                with pytest.raises(type(error)) as caught:
-                    await endpoint()
-                assert caught.value is error
-            assert d.current_span() is outer
-            md.workspace(ca)
-            assert outer.fields['workspace_uuid'] == ca.workspace_uuid
-            assert outer.outcome is None
-        assert d.current_span() is parent
-    assert not manager.pending
-    assert [e['operation'] for e in a.diagnostics.pending] == [
-        'http.isolation.parent',
-        'http.isolation.outer',
-        'http.isolation.outer',
-    ]
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-@pytest.mark.parametrize('group_class', [WebSocketChatRouterGroup, EmbedRouterGroup])
-async def test_websocket_send_cancel_restores_foreign_parent(mode, group_class):
-    a, b, manager, ca, cb, parent = owners(mode)
-    group = group_class(b, quart.Quart(__name__))
-    connection = NS(is_active=False, send_queue=asyncio.Queue())
-    await connection.send_queue.put({'text': 'private-answer'})
-    error = asyncio.CancelledError('private-cancel')
-
-    async def send(payload):
-        await nested(b, cb, [])
-        raise error
-
-    with parent.activate(), patch('quart.websocket', NS(send=send)):
-        with pytest.raises(asyncio.CancelledError) as caught:
-            await group._handle_send(connection)
-        assert caught.value is error
-        assert d.current_span() is parent
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', MODES)
-async def test_sync_http_body_exhaustion_has_owned_context_without_consumer_leak(mode):
-    a, b, manager, ca, cb, parent = owners(mode)
-    seen, closed = [], []
-
-    def stream():
-        try:
-            for value in (b'one', b'two'):
-                md.workspace(cb)
-                seen.append(d.current_span())
-                yield value
-        finally:
-            seen.append(d.current_span())
-            closed.append(True)
-
-    @md.observe('http.isolation.sync_body', source='http', ap=b, http=True)
-    async def endpoint():
-        return quart.Response(stream())
-
-    response = await endpoint()
-    with parent.activate():
-        async with response.response as body:
-            chunks = []
-            async for chunk in body:
-                assert d.current_span() is parent
-                chunks.append(chunk)
-        assert d.current_span() is parent
-    assert chunks == [b'one', b'two']
-    assert closed == [True]
-    assert all((span is not None) == (mode == 'enabled') for span in seen)
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('mode', ['enabled', 'disabled', 'absent', 'mismatched'])
-@pytest.mark.parametrize('failure_stage', ['enter', 'iterator', 'advance', 'exit', None])
-async def test_custom_http_body_protocol_and_cancellation_identity(mode, failure_stage):
-    from quart.wrappers.response import IterableBody
-
-    a, b, manager, ca, cb, parent = owners(mode)
-    error = asyncio.CancelledError('private-body-cancel')
-    seen = []
-
-    def visit(stage):
-        md.workspace(cb)
-        seen.append((stage, d.current_span()))
-        if stage == failure_stage:
-            raise error
-
-    class Body(IterableBody):
-        def __init__(self):
-            pass
-
-        async def __aenter__(self):
-            visit('enter')
-            return self
-
-        def __aiter__(self):
-            visit('iterator')
-            return self
-
-        async def __anext__(self):
-            visit('advance')
-            raise StopAsyncIteration
-
-        async def __aexit__(self, *args):
-            visit('exit')
-
-    @md.observe('http.isolation.body_protocol', source='http', ap=b, http=True)
-    async def endpoint():
-        response = quart.Response()
-        response.response = Body()
-        return response
-
-    response = await endpoint()
-
-    async def consume():
-        async with response.response as body:
-            assert d.current_span() is parent
-            async for _ in body:
-                pytest.fail('empty body yielded')
-
-    with parent.activate():
-        if failure_stage is None:
-            await consume()
-        else:
-            with pytest.raises(asyncio.CancelledError) as caught:
-                await consume()
-            assert caught.value is error
-        assert d.current_span() is parent
-    assert all((span is not None) == (mode == 'enabled') for _, span in seen)
-    assert [s for s, _ in seen].count('exit') == (0 if failure_stage in {'enter', 'iterator'} else 1)
-    assert_isolated(a, b, manager, ca, cb, parent, mode)
diff --git a/tests/unit_tests/api/test_diagnostics_management_lifecycle.py b/tests/unit_tests/api/test_diagnostics_management_lifecycle.py
deleted file mode 100644
index df68d44bf..000000000
--- a/tests/unit_tests/api/test_diagnostics_management_lifecycle.py
+++ /dev/null
@@ -1,163 +0,0 @@
-"""HTTP streaming termination, fail-open behavior and real producer privacy."""
-
-import asyncio
-import json
-import re
-from types import SimpleNamespace
-from unittest.mock import AsyncMock
-
-import pytest
-import quart
-
-from langbot.pkg.agent.runner import errors as runner_errors
-from langbot.pkg.api import management_diagnostics as md
-from langbot.pkg.api.http.controller.group import AuthType
-from langbot.pkg.api.http.controller.groups.agent_debug_stream import debug_stream_response
-from langbot.pkg.api.mcp.mount import MCPMount
-from langbot.pkg.telemetry import diagnostics as d
-from tests.unit_tests.api.test_diagnostics_management_http import Recorder, setup
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('termination', ['close', 'cancel', 'error'])
-async def test_stream_termination_preserves_cleanup_and_error(termination):
-    app, ap, routes = await setup()
-    closed = []
-
-    @routes.route('/end', auth_type=AuthType.NONE)
-    async def end():
-        async def body():
-            try:
-                yield b'one'
-                if termination == 'error':
-                    raise ValueError('private-error')
-                if termination == 'cancel':
-                    raise asyncio.CancelledError('private-cancel')
-                yield b'two'
-            finally:
-                closed.append(d.current_span())
-
-        return quart.Response(body())
-
-    async with app.test_request_context('/management/end'):
-        response = await app.full_dispatch_request()
-    if termination == 'close':
-        async with response.response as body:
-            assert await anext(body.__aiter__()) == b'one'
-    else:
-        error = ValueError if termination == 'error' else asyncio.CancelledError
-        with pytest.raises(error):
-            async with response.response as body:
-                iterator = body.__aiter__()
-                assert await anext(iterator) == b'one'
-                await anext(iterator)
-    assert len(closed) == 1 and closed[0] is not None
-    expected = 'failed' if termination == 'error' else 'cancelled'
-    assert [e['outcome'] for e in ap.diagnostics.events] == ['started', expected]
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_debug_ndjson_error_marks_http_business_failure():
-    app, ap, routes = await setup()
-    service = SimpleNamespace(debug_agent=AsyncMock(side_effect=runner_errors.RunnerNotFoundError('private-error')))
-
-    @routes.route('/ndjson', auth_type=AuthType.NONE)
-    async def ndjson():
-        return debug_stream_response(service, object(), 'private-agent', {'text': 'private-text'})
-
-    response = await app.test_client().get('/management/ndjson')
-    assert response.status_code == 200
-    assert json.loads(await response.get_data())['kind'] == 'error'
-    assert ap.diagnostics.events[-1]['outcome'] == 'failed'
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-async def test_terminal_diagnostic_failure_preserves_success_and_stream():
-    class BrokenFinish(Recorder):
-        def emit(self, kind, operation, outcome, **fields):
-            if outcome != 'started':
-                raise RuntimeError('private-diagnostic-error')
-            super().emit(kind, operation, outcome, **fields)
-
-    app, _, _ = await setup(BrokenFinish())
-    response = await app.test_client().get('/management/stream')
-    assert await response.get_data() == b'private-stream-chunk'
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_real_manager_status_privacy_and_no_network_in_request():
-    app, ap, _ = await setup(None)
-    ap.instance_config = SimpleNamespace(data={'space': {'url': 'https://example.invalid'}})
-    ap.diagnostics = d.DiagnosticsManager(ap, version='4.11.0-beta.2', instance_id='instance-test', capacity=20)
-    ap.diagnostics.credentials = AsyncMock(side_effect=AssertionError('must not await credentials'))
-    for path in ['ok/private-identifier', 'business', 'auth', 'cancel']:
-        async with app.test_request_context('/management/' + path):
-            try:
-                await app.full_dispatch_request()
-            except asyncio.CancelledError:
-                pass
-    events = list(ap.diagnostics.pending)
-    assert len(events) == 8
-    assert {e['outcome'] for e in events} == {'started', 'succeeded', 'failed', 'rejected', 'cancelled'}
-    assert all(re.fullmatch(r'[A-Za-z_][A-Za-z0-9_.:-]{0,127}', e['operation']) for e in events)
-    assert 'private-' not in json.dumps(events)
-    ap.diagnostics.credentials.assert_not_awaited()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'config,version',
-    [
-        ({}, '4.11.0'),
-        ({'disable_telemetry': True}, '4.11.0-beta.2'),
-        ({'disable_beta_diagnostics': True}, '4.11.0-beta.2'),
-    ],
-)
-async def test_real_manager_disabled_modes_do_not_install_context(config, version):
-    app, ap, _ = await setup(None)
-    ap.instance_config = SimpleNamespace(data={'space': config})
-    ap.diagnostics = d.DiagnosticsManager(ap, version=version, instance_id='instance-test')
-    response = await app.test_client().get('/management/ok/private-id')
-    assert response.status_code == 200
-    assert ap.seen == [None]
-    assert not ap.diagnostics.pending
-
-
-@pytest.mark.asyncio
-async def test_mcp_mount_observes_auth_without_headers_or_body():
-    ap = SimpleNamespace(
-        diagnostics=Recorder(), apikey_service=SimpleNamespace(authenticate_api_key=AsyncMock(return_value=None))
-    )
-    mount = MCPMount(ap)
-    send = AsyncMock()
-    fallback = AsyncMock()
-    await mount.wrap(fallback)(
-        {'type': 'http', 'path': '/mcp/private-path', 'headers': [(b'x-api-key', b'private-token')]}, AsyncMock(), send
-    )
-    assert send.call_args_list[0].args[0]['status'] == 401
-    assert [e['outcome'] for e in ap.diagnostics.events] == ['started', 'rejected']
-    assert ap.diagnostics.events[-1]['operation'] == 'mcp.request'
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-    fallback.assert_not_awaited()
-
-
-@pytest.mark.asyncio
-async def test_debug_boundary_source_inherits_without_duplicate_run():
-    recorder = Recorder()
-    ap = SimpleNamespace(diagnostics=recorder)
-
-    @md.observe('http.test.debug', source='webui_debug', ap=ap)
-    async def debug():
-        @d.observe('run', 'runner.run', source='agent', ap=ap)
-        async def run():
-            return 'private-content'
-
-        return await run()
-
-    assert await debug() == 'private-content'
-    assert len([e for e in recorder.events if e['kind'] == 'run' and e['outcome'] == 'started']) == 1
-    assert all(e['source'] == 'webui_debug' and e['attributes']['synthetic'] for e in recorder.events)
-    assert 'private-' not in json.dumps(recorder.events)
diff --git a/tests/unit_tests/api/test_diagnostics_management_transports.py b/tests/unit_tests/api/test_diagnostics_management_transports.py
deleted file mode 100644
index 4b78cd91a..000000000
--- a/tests/unit_tests/api/test_diagnostics_management_transports.py
+++ /dev/null
@@ -1,173 +0,0 @@
-"""Management MCP, WebSocket and debug execution boundaries."""
-
-import asyncio
-import json
-from types import SimpleNamespace
-from unittest.mock import AsyncMock, patch
-
-import pytest
-import quart
-
-from langbot.pkg.api.http.controller.groups.pipelines.websocket_chat import WebSocketChatRouterGroup
-from langbot.pkg.api.http.controller.groups.pipelines.embed import EmbedRouterGroup
-from langbot.pkg.api.mcp.server import LangBotMCPServer
-from langbot.pkg.telemetry import diagnostics as d
-from tests.unit_tests.api.test_diagnostics_management_http import Recorder
-
-
-@pytest.mark.asyncio
-async def test_mcp_real_registered_tool_has_source_and_never_content():
-    recorder = Recorder()
-    seen = []
-
-    async def get_bot(*args, **kwargs):
-        seen.append(d.current_span())
-        return {'secret': 'private-result'}
-
-    ap = SimpleNamespace(diagnostics=recorder, bot_service=SimpleNamespace(get_bot=get_bot))
-    server = LangBotMCPServer(ap)
-    with patch('langbot.pkg.api.mcp.server._authorized', return_value=object()):
-        result = await server.mcp.call_tool('get_bot', {'bot_uuid': 'private-bot'})
-    assert result
-    assert [e['outcome'] for e in recorder.events] == ['started', 'succeeded']
-    assert recorder.events[-1]['operation'] == 'mcp.get_bot'
-    assert recorder.events[-1]['source'] == 'mcp'
-    assert seen[0] is not None
-    assert 'private-' not in json.dumps(recorder.events)
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_mcp_tool_permission_rejection_is_observed():
-    server = LangBotMCPServer(SimpleNamespace(diagnostics=Recorder()))
-    with pytest.raises(Exception):
-        await server.mcp.call_tool('list_bots', {})
-    assert server.ap.diagnostics.events[-1]['outcome'] in {'rejected', 'failed'}
-    assert server.ap.diagnostics.events[-1]['source'] == 'mcp'
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'group_class,path',
-    [
-        (WebSocketChatRouterGroup, '/api/v1/pipelines/private-pipeline/ws/connect'),
-        (EmbedRouterGroup, '/api/v1/embed/11111111-1111-4111-8111-111111111111/ws/connect?session_id=private-session'),
-    ],
-)
-async def test_real_websocket_session_auth_rejection_is_content_free(group_class, path):
-    app = quart.Quart(__name__)
-    ap = SimpleNamespace(diagnostics=Recorder())
-    group = group_class(ap, app)
-    group._authenticate_websocket = AsyncMock(side_effect=ValueError('private-token'))
-    if group_class is EmbedRouterGroup:
-        group._resolve_bot = AsyncMock(return_value=(object(), 'private-pipeline'))
-    await group.initialize()
-    async with app.test_client().websocket(path) as socket:
-        frame = json.loads(await socket.receive())
-        assert frame['type'] == 'error'
-    terminal = [e for e in ap.diagnostics.events if e['outcome'] != 'started']
-    assert terminal
-    assert terminal[-1]['outcome'] == 'rejected'
-    assert terminal[-1]['source'] == 'websocket'
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('group_class', [WebSocketChatRouterGroup, EmbedRouterGroup])
-async def test_websocket_received_message_boundaries_do_not_capture_frames(group_class):
-    app = quart.Quart(__name__)
-    ap = SimpleNamespace(diagnostics=Recorder())
-    group = group_class(ap, app)
-    connection = SimpleNamespace(is_active=True, connection_id='private-id', send_queue=asyncio.Queue())
-    adapter = SimpleNamespace(handle_websocket_message=AsyncMock())
-    if group_class is WebSocketChatRouterGroup:
-        group._revalidate_websocket_authorization = AsyncMock(return_value=object())
-        args = (connection, adapter, object(), 'private-token')
-    else:
-        group._resolve_connected_bot = AsyncMock(return_value=object())
-        args = (connection, adapter, object(), 'private-pipeline')
-
-    async def receive():
-        connection.is_active = False
-        return json.dumps({'type': 'message', 'text': 'private-prompt'})
-
-    with (
-        patch('quart.websocket', SimpleNamespace(receive=receive)),
-        patch(
-            'langbot.pkg.api.http.controller.groups.pipelines.websocket_chat.ws_connection_manager.update_activity',
-            new=AsyncMock(),
-        ),
-    ):
-        await group._handle_receive(*args)
-    adapter.handle_websocket_message.assert_awaited_once()
-    assert [e['outcome'] for e in ap.diagnostics.events] == ['started', 'succeeded']
-    assert ap.diagnostics.events[-1]['operation'].endswith('.message')
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_debug_service_marks_synthetic_source_before_validation():
-    from langbot.pkg.api.http.service.agent import AgentService
-
-    ap = SimpleNamespace(diagnostics=Recorder())
-    service = AgentService(ap)
-    seen = []
-
-    async def get_agent(*args):
-        seen.append(d.current_span())
-        return None
-
-    service.get_agent = get_agent
-    with pytest.raises(ValueError):
-        await service.debug_agent(object(), 'private-id', {'text': 'private-prompt'})
-    assert seen[0].fields['source'] == 'webui_debug'
-    assert ap.diagnostics.events[-1]['operation'] == 'http.agent.debug_agent'
-    assert ap.diagnostics.events[-1]['source'] == 'webui_debug'
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('group_class', [WebSocketChatRouterGroup, EmbedRouterGroup])
-@pytest.mark.parametrize(
-    'frame,expected', [('private-invalid-json', 'rejected'), ('{"type":"private-unknown"}', 'skipped')]
-)
-async def test_websocket_invalid_frames_have_finite_outcomes(group_class, frame, expected):
-    app = quart.Quart(__name__)
-    ap = SimpleNamespace(diagnostics=Recorder())
-    group = group_class(ap, app)
-    connection = SimpleNamespace(is_active=True, connection_id='private-id', send_queue=asyncio.Queue())
-
-    async def receive():
-        connection.is_active = False
-        return frame
-
-    with (
-        patch('quart.websocket', SimpleNamespace(receive=receive)),
-        patch(
-            'langbot.pkg.api.http.controller.groups.pipelines.websocket_chat.ws_connection_manager.update_activity',
-            new=AsyncMock(),
-        ),
-    ):
-        await group._handle_receive(connection, object(), object(), 'private-token')
-    assert ap.diagnostics.events[-1]['outcome'] == expected
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('group_class', [WebSocketChatRouterGroup, EmbedRouterGroup])
-async def test_websocket_send_boundary_preserves_payload_and_cancellation(group_class):
-    app = quart.Quart(__name__)
-    ap = SimpleNamespace(diagnostics=Recorder())
-    group = group_class(ap, app)
-    connection = SimpleNamespace(is_active=False, send_queue=asyncio.Queue())
-    await connection.send_queue.put({'text': 'private-answer'})
-    send = AsyncMock(side_effect=asyncio.CancelledError('private-error'))
-    with patch('quart.websocket', SimpleNamespace(send=send)):
-        with pytest.raises(asyncio.CancelledError):
-            await group._handle_send(connection)
-    assert json.loads(send.call_args.args[0]) == {'text': 'private-answer'}
-    assert ap.diagnostics.events[-1]['outcome'] == 'cancelled'
-    assert ap.diagnostics.events[-1]['operation'].endswith('.send')
-    assert 'private-' not in json.dumps(ap.diagnostics.events)
-    assert d.current_span() is None
diff --git a/tests/unit_tests/api/test_extensions_runtime_fence.py b/tests/unit_tests/api/test_extensions_runtime_fence.py
index 12f36917e..02bff861f 100644
--- a/tests/unit_tests/api/test_extensions_runtime_fence.py
+++ b/tests/unit_tests/api/test_extensions_runtime_fence.py
@@ -99,7 +99,11 @@ async def test_extensions_route_redacts_plugin_secrets_without_mutating_runtime_
 
 @pytest.mark.asyncio
 async def test_extensions_parallel_reads_open_explicit_child_task_scopes():
+    from langbot.pkg.entity.persistence.metadata import WorkspaceMetadata
+
     engine = create_async_engine('sqlite+aiosqlite:///:memory:')
+    async with engine.begin() as connection:
+        await connection.run_sync(WorkspaceMetadata.__table__.create)
     account = SimpleNamespace(uuid='account-a', user='owner@example.com')
     ap = SimpleNamespace()
     ap.persistence_mgr = PersistenceManager(ap)
diff --git a/tests/unit_tests/api/test_management_assistant.py b/tests/unit_tests/api/test_management_assistant.py
index a3f993e6b..8e6fc78b7 100644
--- a/tests/unit_tests/api/test_management_assistant.py
+++ b/tests/unit_tests/api/test_management_assistant.py
@@ -74,7 +74,7 @@ async def assistant(tmp_path):
     ap = SimpleNamespace(
         persistence_mgr=SimpleNamespace(execute_async=execute),
         logger=logging.getLogger('assistant-test'),
-        space_service=SimpleNamespace(get_recommended_chat_model=AsyncMock(return_value={'uuid': 'model'})),
+        space_service=SimpleNamespace(get_recommended_assistant_model=AsyncMock(return_value={'uuid': 'model'})),
         model_mgr=SimpleNamespace(get_model_by_uuid=AsyncMock(return_value=model)),
         pipeline_service=SimpleNamespace(create_pipeline=AsyncMock(return_value='created-pipeline')),
     )
@@ -234,3 +234,250 @@ async def test_model_switch_preserves_history_and_rejects_invalid_selection(assi
     ap.model_mgr.get_model_by_uuid.assert_awaited_with(
         next_provider.invoke_llm.call_args.kwargs['execution_context'], 'second'
     )
+
+
+@pytest.mark.asyncio
+async def test_recommendation_resolved_before_any_turn(assistant):
+    service, ap, provider = assistant
+    assert await service.recommended_model(context()) == {'uuid': 'model', 'name': 'test-model'}
+    provider.invoke_llm.assert_not_awaited()
+    model = ap.model_mgr.get_model_by_uuid.return_value
+    model.model_entity.abilities = []
+    with pytest.raises(AssistantError, match='model_unavailable'):
+        await service.recommended_model(context())
+
+
+@pytest.mark.asyncio
+async def test_stream_selected_model_deltas_tools_and_approval(assistant):
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+    import copy
+
+    service, ap, old_provider = assistant
+    events = []
+    inputs = []
+
+    async def stream(**kwargs):
+        inputs.append(kwargs)
+        if len(inputs) == 1:
+            yield MessageChunk(role='assistant', content='Create ')
+            yield MessageChunk(role='assistant', content='this?')
+            call = proposal().tool_calls[0].model_dump()
+            call['function']['arguments'] = '{"name":"Demo",'
+            yield MessageChunk(role='assistant', tool_calls=[call])
+            call['function']['arguments'] = '"description":"Test draft"}'
+            yield MessageChunk(
+                role='assistant', tool_calls=[call], is_final=True, provider_specific_fields={'signature': 'kept'}
+            )
+        else:
+            yield MessageChunk(role='assistant', content='Done')
+            yield MessageChunk(role='assistant', content='', all_content='Done', is_final=True)
+
+    chosen = SimpleNamespace(
+        provider=SimpleNamespace(invoke_llm_stream=stream),
+        model_entity=SimpleNamespace(name='manual-model', abilities=['func_call'], extra_args={}),
+    )
+    ap.model_mgr.get_model_by_uuid.return_value = chosen
+
+    async def emit(event):
+        events.append(copy.deepcopy(event))
+
+    ctx = context()
+    cid = (await service.create(ctx))['uuid']
+    pending = await service.turn(ctx, cid, 0, text='Create', model_uuid='manual', on_event=emit)
+    assert pending['status'] == 'approval'
+    assert pending['model_uuid'] == 'manual'
+    assert inputs[0]['model'] is chosen
+    ap.space_service.get_recommended_assistant_model.assert_not_awaited()
+    old_provider.invoke_llm.assert_not_awaited()
+    assert [e['data']['text'] for e in events if e['kind'] == 'text'] == ['Create ', 'Create this?']
+    ap.pipeline_service.create_pipeline.assert_not_awaited()
+    events.clear()
+    complete = await service.turn(ctx, cid, 1, approved=True, on_event=emit)
+    assert complete['status'] == 'ready'
+    assert complete['messages'][-1]['content'] == 'Done'
+    ap.pipeline_service.create_pipeline.assert_awaited_once()
+    assert inputs[1]['messages'][2].provider_specific_fields == {'signature': 'kept'}
+    tool_start = next(i for i, e in enumerate(events) if e['kind'] == 'phase' and e['data']['phase'] == 'tool')
+    result = next(
+        i
+        for i, e in enumerate(events)
+        if e['kind'] == 'snapshot' and any(m['role'] == 'tool' for m in e['data']['messages'])
+    )
+    assert tool_start < result
+    with pytest.raises(AssistantError, match='stale_turn'):
+        await service.turn(ctx, cid, 1, approved=True, on_event=emit)
+
+
+@pytest.mark.asyncio
+async def test_truncated_stream_never_executes_proposed_write(assistant):
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+
+    service, ap, provider = assistant
+
+    async def stream(**kwargs):
+        yield MessageChunk(role='assistant', tool_calls=proposal().tool_calls)
+
+    provider.invoke_llm_stream = stream
+    cid = (await service.create(context()))['uuid']
+    result = await service.turn(context(), cid, 0, text='Create', on_event=AsyncMock())
+    assert result['status'] == 'failed'
+    assert result['error'] == 'stream_interrupted'
+    ap.pipeline_service.create_pipeline.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_cancelled_stream_persists_non_replayable_state(assistant):
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+
+    service, ap, provider = assistant
+    started = asyncio.Event()
+
+    async def stream(**kwargs):
+        yield MessageChunk(role='assistant', content='Hello')
+        started.set()
+        await asyncio.Event().wait()
+
+    provider.invoke_llm_stream = stream
+    ctx = context()
+    cid = (await service.create(ctx))['uuid']
+    task = asyncio.create_task(service.turn(ctx, cid, 0, text='Hello', on_event=AsyncMock()))
+    await asyncio.wait_for(started.wait(), 5)
+    task.cancel()
+    with pytest.raises(asyncio.CancelledError):
+        await task
+    saved = await service.get(ctx, cid)
+    assert saved['status'] == 'failed'
+    assert saved['error'] == 'result_unknown'
+    with pytest.raises(AssistantError, match='stale_turn'):
+        await service.turn(ctx, cid, 0, text='Hello')
+
+
+@pytest.mark.asyncio
+async def test_http_stream_emits_before_provider_completion_and_restores_tenant(assistant):
+    import contextlib
+    import json
+    import quart
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+    from langbot.pkg.api.http.controller.groups.assistant import assistant_stream_response, TurnInput
+
+    service, ap, provider = assistant
+    released = asyncio.Event()
+    scoped = []
+
+    @contextlib.asynccontextmanager
+    async def tenant_scope(workspace):
+        scoped.append(workspace)
+        try:
+            yield
+        finally:
+            scoped.pop()
+
+    async def stream(**kwargs):
+        assert scoped == ['workspace-a']
+        yield MessageChunk(role='assistant', content='First')
+        await released.wait()
+        yield MessageChunk(role='assistant', content=' second', is_final=True)
+
+    ap.persistence_mgr.tenant_scope = tenant_scope
+    provider.invoke_llm_stream = stream
+    ctx = context()
+    cid = (await service.create(ctx))['uuid']
+    app = quart.Quart(__name__)
+
+    @app.post('/stream')
+    async def route():
+        return assistant_stream_response(service, ctx, cid, TurnInput(revision=0, text='Hi'))
+
+    frames = []
+    async with app.test_client().request('/stream', method='POST') as connection:
+        while not any(frame['kind'] == 'text' for frame in frames):
+            chunk = await asyncio.wait_for(connection.receive(), 5)
+            frames.extend(json.loads(line) for line in chunk.decode().splitlines() if line.strip())
+        assert frames[-1]['data']['text'] == 'First'
+        assert not released.is_set()
+        released.set()
+        while not any(frame['kind'] == 'completed' for frame in frames):
+            chunk = await asyncio.wait_for(connection.receive(), 5)
+            frames.extend(json.loads(line) for line in chunk.decode().splitlines() if line.strip())
+    assert frames[-1]['data']['messages'][-1]['content'] == 'First second'
+    assert scoped == []
+
+
+@pytest.mark.asyncio
+async def test_background_run_survives_http_disconnect_and_finishes(assistant):
+    import json
+    import quart
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+    from langbot.pkg.api.http.controller.groups.assistant import assistant_stream_response, TurnInput
+
+    service, ap, provider = assistant
+    released = asyncio.Event()
+
+    async def stream(**kwargs):
+        yield MessageChunk(role='assistant', content='Background')
+        await released.wait()
+        yield MessageChunk(role='assistant', content=' done', is_final=True)
+
+    provider.invoke_llm_stream = stream
+    ctx = context()
+    cid = (await service.create(ctx))['uuid']
+    app = quart.Quart(__name__)
+
+    @app.post('/stream')
+    async def route():
+        return assistant_stream_response(service, ctx, cid, TurnInput(revision=0, text='Hi'))
+
+    async with app.test_client().request('/stream', method='POST') as connection:
+        while True:
+            chunk = await asyncio.wait_for(connection.receive(), 5)
+            if any(json.loads(line)['kind'] == 'text' for line in chunk.decode().splitlines() if line.strip()):
+                break
+        task = service._runs[cid]['task']
+        await connection.disconnect()
+    assert not task.done()
+    assert (await service.progress(ctx, cid))['progress']['text'] == 'Background'
+    released.set()
+    await asyncio.wait_for(task, 5)
+    completed = await service.get(ctx, cid)
+    assert completed['status'] == 'ready'
+    assert completed['messages'][-1]['content'] == 'Background done'
+
+
+@pytest.mark.asyncio
+async def test_stop_is_private_revision_scoped_and_does_not_stop_other_chat(assistant):
+    from langbot_plugin.api.entities.builtin.provider.message import MessageChunk
+    from langbot.pkg.api.http.controller.groups.assistant import TurnInput
+
+    service, ap, provider = assistant
+    released = asyncio.Event()
+
+    async def stream(**kwargs):
+        yield MessageChunk(role='assistant', content='Partial')
+        await released.wait()
+        yield MessageChunk(role='assistant', content=' done', is_final=True)
+
+    provider.invoke_llm_stream = stream
+    ctx = context()
+    ids = [(await service.create(ctx))['uuid'] for _ in range(2)]
+    tasks = []
+    for cid in ids:
+        queue = service.start_run(ctx, cid, TurnInput(revision=0, text='Background task'))
+        tasks.append(service._runs[cid]['task'])
+        while (await asyncio.wait_for(queue.get(), 5))['kind'] != 'text':
+            pass
+    for other in (context(account='bob'), context(workspace='workspace-b')):
+        assert await service.list_conversations(other) == []
+        with pytest.raises(AssistantError, match='conversation_not_found'):
+            await service.stop(other, ids[0], 1)
+    with pytest.raises(AssistantError, match='stale_turn'):
+        await service.stop(ctx, ids[0], 0)
+    result = await service.stop(ctx, ids[0], 1)
+    assert result['status'] == 'failed'
+    assert result['error'] == 'stopped'
+    assert result['messages'][-1]['content'] == 'Partial'
+    assert not tasks[1].done()
+    assert len(await service.list_conversations(ctx)) == 2
+    assert (await service.stop(ctx, ids[0], 1))['error'] == 'stopped'
+    released.set()
+    await asyncio.wait_for(tasks[1], 5)
+    assert (await service.get(ctx, ids[1]))['status'] == 'ready'
diff --git a/tests/unit_tests/api/test_mcp_server.py b/tests/unit_tests/api/test_mcp_server.py
index 228b23ffa..3615f8286 100644
--- a/tests/unit_tests/api/test_mcp_server.py
+++ b/tests/unit_tests/api/test_mcp_server.py
@@ -65,3 +65,15 @@ async def test_mcp_server_exposes_bot_event_route_tools():
     assert 'test_bot_event_route' not in tool_names
     assert 'list_processors' in tool_names
     assert 'list_agents' not in tool_names
+    assert 'get_workspace_message_language' in tool_names
+    language_tool = next(tool for tool in tools if tool.name == 'set_workspace_message_language')
+    assert set(language_tool.inputSchema['properties']['language']['enum']) == {
+        'en_US',
+        'zh_Hans',
+        'zh_Hant',
+        'ja_JP',
+        'es_ES',
+        'ru_RU',
+        'th_TH',
+        'vi_VN',
+    }
diff --git a/tests/unit_tests/api/test_pipeline_service_defaults.py b/tests/unit_tests/api/test_pipeline_service_defaults.py
index dd7687d87..20fa685ac 100644
--- a/tests/unit_tests/api/test_pipeline_service_defaults.py
+++ b/tests/unit_tests/api/test_pipeline_service_defaults.py
@@ -8,6 +8,7 @@ import pytest
 
 from langbot.pkg.agent.runner.descriptor import RunnerDescriptor
 from langbot.pkg.api.http.service.pipeline import PipelineService
+from langbot.pkg.api.http.service.model import LLMModelsService
 
 
 class FakeLogger:
@@ -54,6 +55,7 @@ async def test_default_pipeline_config_uses_first_installed_runner_schema():
         logger=FakeLogger(),
         runner_registry=FakeRegistry([custom_agent, local_agent]),
     )
+    ap.llm_model_service = LLMModelsService(ap)
 
     config = await PipelineService(ap).get_default_pipeline_config('workspace-test')
 
diff --git a/tests/unit_tests/api/test_slack_setup.py b/tests/unit_tests/api/test_slack_setup.py
new file mode 100644
index 000000000..2a25effd6
--- /dev/null
+++ b/tests/unit_tests/api/test_slack_setup.py
@@ -0,0 +1,153 @@
+import asyncio
+import hashlib
+import hmac
+import json
+import time
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+
+from langbot.pkg.api.http.service.slack_setup import SlackSetupService, SlackSetupError, callback_url
+
+
+def context(workspace='one', principal='user'):
+    return SimpleNamespace(instance_uuid='local', workspace_uuid=workspace, placement_generation=1, principal=principal)
+
+
+def service():
+    ap = SimpleNamespace(
+        bot_service=SimpleNamespace(get_bot=AsyncMock(return_value={'adapter': 'slack-omni'})),
+        workspace_service=SimpleNamespace(get_execution_binding=AsyncMock()),
+        task_mgr=SimpleNamespace(
+            create_user_task=lambda coro, **kwargs: SimpleNamespace(task=asyncio.create_task(coro))
+        ),
+    )
+    obj = SlackSetupService(ap)
+    obj.api = AsyncMock(
+        side_effect=[
+            {'app_id': 'A123', 'credentials': {'client_id': 'id', 'client_secret': 'secret', 'signing_secret': 'sign'}},
+            {'ok': True},
+            {'app_id': 'A123', 'token_type': 'bot', 'access_token': 'bot-token'},
+        ]
+    )
+    return obj
+
+
+async def start(obj, **kwargs):
+    result = await obj.start(
+        context(),
+        bot_uuid='bot',
+        access_token='access',
+        refresh_token='refresh',
+        webhook_url='https://example.com/bots/bot',
+        **kwargs,
+    )
+    sid = result['session_id']
+    await obj.sessions[sid]['task']
+    return sid
+
+
+@pytest.mark.asyncio
+async def test_webhook_setup_oauth_and_scope():
+    obj = service()
+    sid = await start(obj)
+    assert obj.status(context(), sid)['status'] == 'waiting'
+    assert 'state=' + sid in obj.status(context(), sid)['authorize_url']
+    for ctx in (context('two'), context(principal='other')):
+        with pytest.raises(SlackSetupError):
+            obj.status(ctx, sid)
+    calls = obj.api.await_args_list
+    assert 'event_subscriptions' not in json.loads(calls[0].kwargs['manifest'])['settings']
+    assert json.loads(calls[1].kwargs['manifest'])['settings']['event_subscriptions']['request_url'].endswith(
+        '/bots/bot'
+    )
+    await obj.finish(sid, 'code')
+    assert obj.status(context(), sid)['config']['bot_token'] == 'bot-token'
+    assert 'credentials' not in obj.sessions[sid]
+    with pytest.raises(SlackSetupError):
+        await obj.finish(sid, 'replay')
+    obj.discard(sid)
+    assert not obj.sessions
+
+
+@pytest.mark.asyncio
+async def test_socket_manifest_needs_no_public_webhook():
+    obj = service()
+    sid = await start(
+        obj, socket_mode=True, redirect_url='http://localhost:5300/api/v1/platform/adapters/slack/setup/callback'
+    )
+    config = json.loads(obj.api.await_args_list[1].kwargs['manifest'])
+    assert config['settings']['socket_mode_enabled']
+    assert 'request_url' not in config['settings']['event_subscriptions']
+    obj.discard(sid)
+
+
+@pytest.mark.asyncio
+async def test_expired_token_rotates_once():
+    obj = service()
+    obj.api.side_effect = [
+        SlackSetupError('token_expired'),
+        {'token': 'new-access', 'refresh_token': 'new-refresh'},
+        {'app_id': 'A123', 'credentials': {'client_id': 'id', 'client_secret': 'secret', 'signing_secret': 'sign'}},
+        {'ok': True},
+    ]
+    sid = await start(obj)
+    assert obj.api.await_args_list[1].args == ('tooling.tokens.rotate',)
+    assert obj.api.await_args_list[2].args == ('apps.manifest.create', 'new-access')
+    assert 'refresh' not in obj.sessions[sid]
+    obj.discard(sid)
+
+
+@pytest.mark.asyncio
+async def test_provisioning_error_is_visible_without_secrets():
+    obj = service()
+    obj.api.side_effect = RuntimeError('sensitive upstream text')
+    sid = await start(obj)
+    assert obj.status(context(), sid) == {'status': 'error', 'error': 'setup_failed'}
+    obj.discard(sid)
+
+
+@pytest.mark.asyncio
+async def test_denied_authorization():
+    obj = service()
+    sid = await start(obj)
+    with pytest.raises(SlackSetupError):
+        await obj.finish(sid, '', 'access_denied')
+    assert obj.status(context(), sid)['error'] == 'authorization_denied'
+    assert 'credentials' not in obj.sessions[sid]
+    obj.discard(sid)
+
+
+@pytest.mark.asyncio
+async def test_setup_webhook_only_accepts_signed_verification():
+    obj = service()
+    sid = await start(obj)
+    body = json.dumps({'type': 'url_verification', 'challenge': 'test'}).encode()
+    timestamp = str(int(time.time()))
+    signature = 'v0=' + hmac.new(b'sign', b'v0:' + timestamp.encode() + b':' + body, hashlib.sha256).hexdigest()
+    req = SimpleNamespace(
+        get_data=AsyncMock(return_value=body),
+        get_json=AsyncMock(return_value=json.loads(body)),
+        headers={'X-Slack-Request-Timestamp': timestamp, 'X-Slack-Signature': signature},
+    )
+    assert await obj.challenge('bot', req) == {'challenge': 'test'}
+    req.headers['X-Slack-Signature'] = 'forged'
+    assert await obj.challenge('bot', req) is None
+    obj.discard(sid)
+
+
+@pytest.mark.parametrize(
+    'url',
+    [
+        'http://example.com/bots/bot',
+        'https://example.com/bots/other',
+        'https://user:pass@example.com/bots/bot',
+        'https://localhost/bots/bot',
+        'https://127.0.0.1/bots/bot',
+        'https://example.com/bots/bot?key=secret',
+    ],
+)
+def test_invalid_callback_url(url):
+    with pytest.raises(SlackSetupError):
+        callback_url(url, 'bot')
diff --git a/tests/unit_tests/box/test_attachment_host_capabilities.py b/tests/unit_tests/box/test_attachment_host_capabilities.py
new file mode 100644
index 000000000..27aabc7de
--- /dev/null
+++ b/tests/unit_tests/box/test_attachment_host_capabilities.py
@@ -0,0 +1,142 @@
+from unittest.mock import AsyncMock, Mock
+
+import pytest
+
+from langbot.pkg.box.service import BoxService
+from langbot_plugin.box.client import BoxRuntimeClient
+from tests.unit_tests.box.test_box_service import make_app, make_query
+
+
+@pytest.mark.asyncio
+async def test_host_without_openat_uses_sandbox_attachment_channel(monkeypatch, tmp_path):
+    monkeypatch.setattr('os.supports_dir_fd', set())
+    service = BoxService(make_app(Mock()), client=Mock(spec=BoxRuntimeClient))
+    service._available = True
+    service._tenant_workspace = Mock(return_value=str(tmp_path))
+    service._read_outbox_host = Mock(side_effect=AssertionError('POSIX host reader must not run'))
+    service._read_outbox_via_exec = AsyncMock(return_value=[])
+    service._clear_outbox = AsyncMock()
+    query = make_query()
+    assert service._host_query_dir(service.OUTBOX_SUBDIR, query) is None
+    assert await service.collect_outbound_attachments(query) == []
+    service._read_outbox_via_exec.assert_awaited_once_with(query)
+    service._clear_outbox.assert_awaited_once_with(query, None)
+
+
+@pytest.mark.asyncio
+async def test_outbox_exec_preserves_workspace_and_full_attachment(monkeypatch, tmp_path):
+    import base64
+    import json
+    from langbot_plugin.box.models import BoxExecutionResult, BoxExecutionStatus
+
+    monkeypatch.setattr('os.supports_dir_fd', set())
+    client = Mock(spec=BoxRuntimeClient)
+    content = base64.b64encode(b'file content' * 1000).decode()
+    client.execute = AsyncMock(
+        return_value=BoxExecutionResult(
+            session_id='person_test_user',
+            backend_name='docker',
+            status=BoxExecutionStatus.COMPLETED,
+            exit_code=0,
+            stdout=json.dumps([{'name': 'report.txt', 'b64': content}]),
+            stderr='',
+            duration_ms=1,
+        )
+    )
+    service = BoxService(make_app(Mock(), [str(tmp_path)]), client=client, output_limit_chars=100)
+    service._available = True
+    service._tenant_workspace = Mock(return_value=str(tmp_path))
+    service._clear_outbox = AsyncMock()
+    query = make_query()
+    attachments = await service.collect_outbound_attachments(query)
+    assert attachments == [{'type': 'File', 'name': 'report.txt', 'base64': content}]
+    call = client.execute.await_args
+    assert call.kwargs['action_context'].workspace_uuid == query.workspace_uuid
+    assert call.kwargs['action_context'].instance_uuid == query.instance_uuid
+    assert call.args[0].host_path == str(tmp_path)
+    assert call.args[0].session_id == 'person_test_user'
+    service._clear_outbox.assert_awaited_once()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('failure', ['rpc', 'command', 'invalid_json'])
+async def test_outbox_read_failure_keeps_files(monkeypatch, failure):
+    from langbot_plugin.box.errors import BoxValidationError
+    from langbot_plugin.box.models import BoxExecutionResult, BoxExecutionStatus
+
+    monkeypatch.setattr('os.supports_dir_fd', set())
+    client = Mock(spec=BoxRuntimeClient)
+    client.execute = AsyncMock(
+        return_value=BoxExecutionResult(
+            session_id='person_test_user',
+            backend_name='docker',
+            status=BoxExecutionStatus.COMPLETED,
+            exit_code=1 if failure == 'command' else 0,
+            stdout='not json',
+            stderr='',
+            duration_ms=1,
+        )
+    )
+    if failure == 'rpc':
+        client.execute.side_effect = BoxValidationError('missing context')
+    service = BoxService(make_app(Mock()), client=client)
+    service._available = True
+    service._clear_outbox = AsyncMock()
+    with pytest.raises(BoxValidationError):
+        await service.collect_outbound_attachments(make_query())
+    service._clear_outbox.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('oversized', [False, True])
+async def test_real_outbox_reader_handles_image_above_old_256k_limit(tmp_path, monkeypatch, oversized):
+    import asyncio
+    import base64
+    import sys
+    from langbot_plugin.box.errors import BoxValidationError
+    from langbot_plugin.box.models import BoxExecutionResult, BoxExecutionStatus
+
+    monkeypatch.setattr('os.supports_dir_fd', set())
+    service = BoxService(make_app(Mock()), client=Mock(spec=BoxRuntimeClient))
+    service._available = True
+    service.OUTBOX_MOUNT_DIR = tmp_path.as_posix()
+    if oversized:
+        service._ATTACHMENT_MAX_BYTES = 300000
+    folder = tmp_path / 'query-42'
+    folder.mkdir()
+    payload = b'\x89PNG\r\n\x1a\n' + b'x' * (343154 - 8)
+    (folder / 'sunset.png').write_bytes(payload)
+
+    async def execute_script(spec, query):
+        script = spec['cmd'].split('\n', 1)[1].rsplit('\nLBPY', 1)[0]
+        process = await asyncio.create_subprocess_exec(
+            sys.executable,
+            '-c',
+            script,
+            stdout=asyncio.subprocess.PIPE,
+            stderr=asyncio.subprocess.PIPE,
+        )
+        stdout, stderr = await process.communicate()
+        return BoxExecutionResult(
+            session_id='person_test_user',
+            backend_name='test',
+            status=BoxExecutionStatus.COMPLETED,
+            exit_code=process.returncode,
+            stdout=stdout.decode(),
+            stderr=stderr.decode(),
+            duration_ms=1,
+        )
+
+    service._execute_spec_payload_result = execute_script
+    service._clear_outbox = AsyncMock()
+    if oversized:
+        with pytest.raises(BoxValidationError):
+            await service.collect_outbound_attachments(make_query())
+        service._clear_outbox.assert_not_awaited()
+        assert (folder / 'sunset.png').exists()
+    else:
+        attachments = await service.collect_outbound_attachments(make_query())
+        assert len(attachments) == 1
+        assert attachments[0]['type'] == 'Image'
+        assert base64.b64decode(attachments[0]['base64'].split(',', 1)[1]) == payload
+        service._clear_outbox.assert_awaited_once()
diff --git a/tests/unit_tests/box/test_box_service.py b/tests/unit_tests/box/test_box_service.py
index 4ae7e3517..1d680ba7a 100644
--- a/tests/unit_tests/box/test_box_service.py
+++ b/tests/unit_tests/box/test_box_service.py
@@ -931,20 +931,21 @@ async def test_truncate_stderr_independently():
 
 
 @pytest.mark.asyncio
-async def test_profile_default_provides_defaults():
+@pytest.mark.parametrize('profile_name,network', [('default', BoxNetworkMode.ON), ('offline', BoxNetworkMode.OFF)])
+async def test_profile_default_provides_defaults(profile_name, network):
     """When tool call omits network/image, profile defaults are used."""
     logger = Mock()
     backend = FakeBackend(logger)
     runtime = BoxRuntime(logger=logger, backends=[backend], session_ttl_sec=300)
-    service = BoxService(make_app(logger), client=_InProcessBoxRuntimeClient(logger, runtime))
+    service = BoxService(make_app(logger, profile=profile_name), client=_InProcessBoxRuntimeClient(logger, runtime))
     await service.initialize()
 
     result = await service.execute_tool({'command': 'echo hi'}, make_query(30))
 
     assert result['ok'] is True
     spec = backend.start_specs[0]
-    profile = BUILTIN_PROFILES['default']
-    assert spec.network == BoxNetworkMode.OFF
+    profile = BUILTIN_PROFILES[profile_name]
+    assert spec.network == network
     assert spec.image == profile.image
     assert spec.timeout_sec == profile.timeout_sec
 
@@ -959,14 +960,14 @@ async def test_profile_unlocked_field_can_be_overridden():
     await service.initialize()
 
     result = await service.execute_spec_payload(
-        {'cmd': 'echo hi', 'timeout_sec': 60, 'network': 'on', 'session_id': 'person_test_user'},
+        {'cmd': 'echo hi', 'timeout_sec': 60, 'network': 'off', 'session_id': 'person_test_user'},
         make_query(31),
     )
 
     assert result['ok'] is True
     spec = backend.start_specs[0]
     assert spec.timeout_sec == 60
-    assert spec.network == BoxNetworkMode.ON
+    assert spec.network == BoxNetworkMode.OFF
 
 
 @pytest.mark.asyncio
@@ -1034,10 +1035,15 @@ def test_unknown_profile_raises_error():
 def test_builtin_profiles_are_consistent():
     """Basic sanity check on all built-in profiles."""
     assert 'default' in BUILTIN_PROFILES
+    assert 'offline' in BUILTIN_PROFILES
     assert 'offline_readonly' in BUILTIN_PROFILES
     assert 'network_basic' in BUILTIN_PROFILES
     assert 'network_extended' in BUILTIN_PROFILES
 
+    default = BUILTIN_PROFILES['default']
+    assert default.network == BoxNetworkMode.ON
+    assert BUILTIN_PROFILES['offline'] == default.model_copy(update={'name': 'offline', 'network': BoxNetworkMode.OFF})
+
     offline = BUILTIN_PROFILES['offline_readonly']
     assert offline.network == BoxNetworkMode.OFF
     assert offline.host_path_mode == BoxHostMountMode.READ_ONLY
@@ -1837,7 +1843,7 @@ class TestBuildSkillExecutionMounts:
             assert mounts == [
                 {
                     'host_path': live_dir,
-                    'mount_path': '/workspace/.skills/alive',
+                    'mount_path': '/opt/langbot-artifacts/' + '1' * 64,
                     'mode': 'ro',
                     'content_digest': 'sha256:' + '1' * 64,
                     'manifest_path': manifest_path,
@@ -2027,7 +2033,8 @@ class TestInboundOutboundRoundTrip:
 
         calls = []
 
-        async def fake_client_execute(spec):
+        async def fake_client_execute(spec, *, action_context=None):
+            assert action_context.workspace_uuid == query.workspace_uuid
             cmd = spec.cmd
             calls.append(cmd)
             if 'os.scandir' in cmd:
@@ -2070,7 +2077,8 @@ class TestInboundOutboundRoundTrip:
 
         calls = []
 
-        async def fake_client_execute(spec):
+        async def fake_client_execute(spec, *, action_context=None):
+            assert action_context.workspace_uuid == query.workspace_uuid
             cmd = spec.cmd
             calls.append(cmd)
             if 'os.scandir' in cmd:
@@ -2388,3 +2396,31 @@ async def test_execution_cannot_override_runner_bound_session():
         await service.execute_spec_payload({'cmd': 'true', 'session_id': 'other-box'}, make_query())
 
     assert backend.start_calls == []
+
+
+@pytest.mark.asyncio
+async def test_execute_tool_preserves_trusted_read_only_mounts(tmp_path):
+    logger = Mock()
+    backend = FakeBackend(logger)
+    runtime = BoxRuntime(logger=logger, backends=[backend], session_ttl_sec=300)
+    service = BoxService(
+        make_app(logger, allowed_mount_roots=[str(tmp_path)]),
+        client=_InProcessBoxRuntimeClient(logger, runtime),
+    )
+    package = tmp_path / 'published'
+    package.mkdir()
+    await service.initialize()
+    try:
+        result = await service.execute_tool(
+            {'command': 'cat /workspace/.skills/demo/SKILL.md'},
+            make_query(),
+            read_only_mounts=[{'host_path': str(package), 'mount_path': '/workspace/.skills/demo', 'mode': 'ro'}],
+        )
+        assert result['exit_code'] == 0
+        assert len(backend.start_specs) == 1
+        mount = backend.start_specs[0].extra_mounts[0]
+        assert mount.host_path == str(package)
+        assert mount.mount_path == '/workspace/.skills/demo'
+        assert mount.mode == BoxHostMountMode.READ_ONLY
+    finally:
+        await service.shutdown()
diff --git a/tests/unit_tests/box/test_runner_box.py b/tests/unit_tests/box/test_runner_box.py
index 1f462d2cc..74e4262e5 100644
--- a/tests/unit_tests/box/test_runner_box.py
+++ b/tests/unit_tests/box/test_runner_box.py
@@ -12,11 +12,12 @@ from langbot.pkg.box.runner import RunnerBoxService, RunBoxBinding, exported_mes
 def service():
     sessions = {}
 
-    async def create(context, spec):
+    async def create(context, spec, **kwargs):
         sessions.setdefault(spec['session_id'], spec)
         return sessions[spec['session_id']]
 
     box = SimpleNamespace(
+        ap=SimpleNamespace(),
         enabled=True,
         managed_admission_required=False,
         _ATTACHMENT_MAX_TOTAL_BYTES=1000000,
@@ -36,7 +37,7 @@ async def test_acquire_reuses_key_and_bind_is_explicit_and_run_scoped():
         api.acquire('workspace', {'reuse_key': 'global'}),
     )
     assert first == second
-    q1, q2 = SimpleNamespace(), SimpleNamespace()
+    q1, q2 = SimpleNamespace(variables={}), SimpleNamespace(variables={})
     a = await api.bind('workspace', q1, 'run-a', first['id'])
     b = await api.bind('workspace', q2, 'run-b', first['id'])
     assert a['box_id'] == b['box_id']
@@ -53,7 +54,7 @@ async def test_concurrent_binding_cannot_switch_boxes():
     api = service()
     a = await api.acquire('ws', {'reuse_key': 'a'})
     b = await api.acquire('ws', {'reuse_key': 'b'})
-    q = SimpleNamespace()
+    q = SimpleNamespace(variables={})
     results = await asyncio.gather(
         api.bind('ws', q, 'run', a['id']), api.bind('ws', q, 'run', b['id']), return_exceptions=True
     )
diff --git a/tests/unit_tests/cloud/test_plan_policy.py b/tests/unit_tests/cloud/test_plan_policy.py
new file mode 100644
index 000000000..20b5b36e8
--- /dev/null
+++ b/tests/unit_tests/cloud/test_plan_policy.py
@@ -0,0 +1,44 @@
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+
+from langbot.pkg.cloud.entitlements import EntitlementFeatureUnavailableError, EntitlementUnavailableError
+from langbot.pkg.cloud.plan_policy import require_workspace_platform, workspace_history_retention_days
+
+
+def cloud_app():
+    snapshots = {
+        'restricted': SimpleNamespace(features={'platform.telegram': False}, limits={'chat_history.retention_days': 7}, entitlement_revision=2),
+        'open': SimpleNamespace(features={}, limits={'chat_history.retention_days': 90}, entitlement_revision=3),
+    }
+    return SimpleNamespace(deployment=SimpleNamespace(mode='cloud'), entitlement_resolver=SimpleNamespace(resolve=AsyncMock(side_effect=lambda ws: snapshots[ws])))
+
+
+@pytest.mark.asyncio
+async def test_platform_policy_is_scoped_and_covers_aliases():
+    ap = cloud_app()
+    for adapter in ['telegram', 'telegram-omni', 'telegram-eba']:
+        with pytest.raises(EntitlementFeatureUnavailableError):
+            await require_workspace_platform(ap, 'restricted', adapter)
+        await require_workspace_platform(ap, 'open', adapter)
+    await require_workspace_platform(ap, 'restricted', 'discord')
+
+
+@pytest.mark.asyncio
+async def test_retention_uses_workspace_entitlement():
+    ap = cloud_app()
+    assert await workspace_history_retention_days(ap, 'restricted', 100) == 7
+    assert await workspace_history_retention_days(ap, 'open', 100) == 90
+
+
+@pytest.mark.asyncio
+async def test_oss_and_unavailable_projection():
+    ap = SimpleNamespace(deployment=SimpleNamespace(mode='local'))
+    await require_workspace_platform(ap, 'any', 'telegram')
+    assert await workspace_history_retention_days(ap, 'any', 60) == 60
+    ap.deployment.mode = 'cloud'
+    with pytest.raises(EntitlementUnavailableError):
+        await require_workspace_platform(ap, 'any', 'telegram')
+    with pytest.raises(EntitlementUnavailableError):
+        await workspace_history_retention_days(ap, 'any')
diff --git a/tests/unit_tests/core/test_app_resource_maintenance.py b/tests/unit_tests/core/test_app_resource_maintenance.py
index 8ba094d1f..c29af0228 100644
--- a/tests/unit_tests/core/test_app_resource_maintenance.py
+++ b/tests/unit_tests/core/test_app_resource_maintenance.py
@@ -32,7 +32,11 @@ async def _wait_forever() -> None:
     await asyncio.Event().wait()
 
 
-async def test_resource_maintenance_waits_and_shares_workspace_discovery() -> None:
+async def test_resource_maintenance_waits_and_shares_workspace_discovery(monkeypatch) -> None:
+    async def idle_leases(_ap):
+        await _wait_forever()
+
+    monkeypatch.setattr('langbot.pkg.utils.execution_lease.maintain_execution_leases', idle_leases)
     stop = asyncio.Event()
     completed = asyncio.Event()
     discovery_calls = 0
diff --git a/tests/unit_tests/persistence/test_tenant_uow.py b/tests/unit_tests/persistence/test_tenant_uow.py
index 0c374f201..ee4c0b570 100644
--- a/tests/unit_tests/persistence/test_tenant_uow.py
+++ b/tests/unit_tests/persistence/test_tenant_uow.py
@@ -980,6 +980,18 @@ async def test_scoped_sql_structure_allows_only_the_production_vocabulary(statem
     _validate_scoped_statement_call((statement,), {})
 
 
+async def test_monitoring_json_casts_are_limited_to_journal_metadata() -> None:
+    from langbot.pkg.entity.persistence.agent_run import AgentRun
+    from langbot.pkg.entity.persistence.event_log import EventLog
+
+    for column in (AgentRun.metadata_json, EventLog.metadata_json, EventLog.input_json):
+        statement = sa.select(sa.cast(column, sa.JSON)['status'].as_string())
+        _validate_scoped_statement_call((statement,), {})
+    for column in (AgentRun.authorization_json, sa.literal('{}')):
+        with pytest.raises(ScopedSessionTransactionError, match='casts'):
+            _validate_scoped_statement_call((sa.select(sa.cast(column, sa.JSON)),), {})
+
+
 async def test_scoped_session_executes_sqlite_strftime() -> None:
     engine = create_async_engine('sqlite+aiosqlite:///:memory:')
     try:
diff --git a/tests/unit_tests/pipeline/conftest.py b/tests/unit_tests/pipeline/conftest.py
index 22d81e2b4..07567c1f7 100644
--- a/tests/unit_tests/pipeline/conftest.py
+++ b/tests/unit_tests/pipeline/conftest.py
@@ -88,6 +88,12 @@ class MockApplication:
     def _create_mock_persistence_manager(self):
         persistence_mgr = AsyncMock()
         persistence_mgr.execute_async = AsyncMock()
+        result = Mock(scalar_one_or_none=Mock(return_value=None))
+        uow = AsyncMock()
+        uow.execute = AsyncMock(return_value=result)
+        boundary = AsyncMock()
+        boundary.__aenter__.return_value = uow
+        persistence_mgr.tenant_uow = Mock(return_value=boundary)
         return persistence_mgr
 
     def _create_mock_query_pool(self):
diff --git a/tests/unit_tests/pipeline/test_banwords.py b/tests/unit_tests/pipeline/test_banwords.py
index 2dcaf2cf4..a33a0a2e1 100644
--- a/tests/unit_tests/pipeline/test_banwords.py
+++ b/tests/unit_tests/pipeline/test_banwords.py
@@ -82,7 +82,7 @@ async def test_invalid_sensitive_word_regex_still_blocks():
     result = await filt.process(Mock(), 'any message')
 
     assert result.level == entities.ResultLevel.BLOCK
-    assert result.user_notice == '内容检查规则执行失败，请联系管理员'
+    assert result.user_notice == 'Content checking failed. Please contact the administrator.'
     assert 'rejected' in result.console_notice.lower() or 'invalid' in result.console_notice.lower()
 
 
@@ -97,7 +97,7 @@ async def test_oversized_word_list_is_blocked():
 
     assert result.level == entities.ResultLevel.BLOCK
     assert result.replacement == ''
-    assert result.user_notice == '内容检查规则执行失败，请联系管理员'
+    assert result.user_notice == 'Content checking failed. Please contact the administrator.'
     assert 'at most 256 regex patterns are allowed' in result.console_notice.lower()
 
 
diff --git a/tests/unit_tests/pipeline/test_chat_handler.py b/tests/unit_tests/pipeline/test_chat_handler.py
index 439c41491..ce4be91af 100644
--- a/tests/unit_tests/pipeline/test_chat_handler.py
+++ b/tests/unit_tests/pipeline/test_chat_handler.py
@@ -263,6 +263,57 @@ class TestChatMessageHandlerReal:
 class TestChatHandlerStreaming:
     """Tests for streaming behavior."""
 
+    @pytest.mark.asyncio
+    @pytest.mark.parametrize('with_tool_call', [False, True])
+    async def test_terminal_delta_and_snapshot_close_card_once(
+        self, fake_app, mock_event_ctx, set_runner, with_tool_call
+    ):
+        from tests.factories import text_query
+        from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+        from langbot_plugin.api.entities.builtin.platform.message import File, MessageChain
+
+        fake_app.plugin_connector.emit_event = AsyncMock(return_value=mock_event_ctx)
+        query = text_query('stream test')
+        query.adapter = Mock()
+        query.adapter.is_stream_output_supported = AsyncMock(return_value=True)
+        query.adapter.create_message_card = AsyncMock()
+        attachments = MessageChain([File(url='https://example.com/file.txt', name='file.txt')])
+
+        class StreamRunner:
+            def __init__(self, app, config):
+                pass
+
+            async def run(self, query):
+                if with_tool_call:
+                    yield MessageChunk(
+                        role='assistant',
+                        is_final=True,
+                        tool_calls=[
+                            {'id': 'call-1', 'type': 'function', 'function': {'name': 'search', 'arguments': '{}'}}
+                        ],
+                    )
+                yield MessageChunk(role='assistant', content='hello')
+                yield MessageChunk(role='assistant', content='hello world', is_final=True)
+                completed = Message(role='assistant', content='hello world')
+                completed.attachments = attachments
+                yield completed
+
+        set_runner(StreamRunner)
+        handler = get_chat_handler().ChatMessageHandler(fake_app)
+        handler._ensure_conversation_for_history = AsyncMock()
+        delivered = []
+        async for stage_result in handler.handle(query):
+            assert stage_result.result_type == get_entities().ResultType.CONTINUE
+            delivered.append(query.resp_messages[-1])
+        assert [result.content for result in delivered if result.content] == ['hello', 'hello world']
+        if with_tool_call:
+            assert delivered[0].tool_calls
+            assert not delivered[0].is_final
+        assert sum(result.is_final for result in delivered) == 1
+        assert delivered[-1].attachments is attachments
+        assert len({result.resp_message_id for result in delivered}) == 1
+        query.adapter.create_message_card.assert_awaited_once()
+
     @pytest.mark.asyncio
     async def test_streaming_chunks_collected(self, fake_app, mock_event_ctx, set_runner):
         """Streaming produces multiple results."""
diff --git a/tests/unit_tests/pipeline/test_legacy_config_migration.py b/tests/unit_tests/pipeline/test_legacy_config_migration.py
index d054b58ee..b3ea09e17 100644
--- a/tests/unit_tests/pipeline/test_legacy_config_migration.py
+++ b/tests/unit_tests/pipeline/test_legacy_config_migration.py
@@ -1,4 +1,4 @@
-"""Behavioral tests for the pure, fail-closed legacy migration planner."""
+"""Behavioral tests for the pure legacy migration planner (relaxed v5 semantics)."""
 
 import copy
 import importlib
@@ -11,15 +11,15 @@ import pytest
 
 FIXTURES = json.loads((Path(__file__).parents[2] / 'fixtures/pipeline_migration/synthetic_legacy.json').read_text())
 TARGETS = {
-    'local-agent': ('LocalAgent', '0.1.10', None),
-    'dify-service-api': ('DifyAgent', '0.1.10', None),
-    'coze-api': ('CozeAgent', '0.1.10', None),
-    'dashscope-app-api': ('DashScopeAgent', '0.1.10', None),
-    'n8n-service-api': ('N8nAgent', '0.1.10', None),
-    'langflow-api': ('LangflowAgent', '0.1.10', None),
-    'deerflow-api': ('DeerFlowAgent', '0.1.10', None),
-    'tbox-app-api': ('TboxAgent', '0.1.8', None),
-    'weknora-api': ('WeKnoraAgent', '0.1.10', None),
+    'local-agent': ('LocalAgent', '0.2.1', None),
+    'dify-service-api': ('DifyAgent', '0.2.3', None),
+    'coze-api': ('CozeAgent', '0.2.0', None),
+    'dashscope-app-api': ('DashScopeAgent', '0.2.2', None),
+    'n8n-service-api': ('N8nAgent', '0.2.0', None),
+    'langflow-api': ('LangflowAgent', '0.2.2', None),
+    'deerflow-api': ('DeerFlowAgent', '0.2.1', None),
+    'tbox-app-api': ('TboxAgent', '0.2.1', None),
+    'weknora-api': ('WeKnoraAgent', '0.2.1', None),
 }
 
 
@@ -47,6 +47,21 @@ def assert_block(result, code, field=None):
     assert expected in result['blockers']
 
 
+def assert_repaired(result, code, field=None):
+    """Relaxed planner (v5): non-critical problems are repaired and reported as warnings."""
+    assert result['state'] == 'ready', result['blockers']
+    assert result['blockers'] == []
+    assert result['config'] is not None
+    codes = {(w['code'], w.get('field')) for w in result['warnings']}
+    if field is None:
+        assert code in {c for c, _ in codes}
+    else:
+        assert (code, field) in codes
+    public = {key: value for key, value in result.items() if key != 'config'}
+    assert 'SECRET' not in json.dumps(public)
+    assert 'secret-' not in json.dumps(public)
+
+
 @pytest.mark.parametrize('runner', TARGETS)
 def test_recognizes_all_nine_exact_official_identities(runner):
     result = plan(source_for(runner))
@@ -140,11 +155,22 @@ def test_malformed_envelopes_are_safe_blockers(field, value):
     assert_block(plan(source), 'invalid_type', 'ai' if field == 'ai' else 'ai.runner')
 
 
-@pytest.mark.parametrize('current', [None, '', False, {}, 'plugin:langbot-team/DeerFlowAgent/default'])
-def test_mixed_selections_block_even_empty_current_id(current):
+@pytest.mark.parametrize('current', [None, '', False, {}])
+def test_mixed_selection_with_unusable_current_id_migrates_legacy(current):
     source = source_for()
     source['ai']['runner']['id'] = current
-    assert_block(plan(source), 'mixed_runner_selection', 'ai.runner')
+    result = plan(source)
+    assert_repaired(result, 'migration.mixed_runner_selection', 'ai.runner')
+    assert result['config']['ai']['runner']['id'] == 'plugin:langbot-team/DeerFlowAgent/default'
+    assert 'runner' not in result['config']['ai']['runner']
+
+
+def test_mixed_selection_prefers_valid_current_id():
+    source = source_for()
+    source['ai']['runner']['id'] = 'plugin:langbot-team/DeerFlowAgent/default'
+    result = plan(source)
+    assert result['state'] == 'already_current'
+    assert {'code': 'migration.mixed_runner_selection', 'field': 'ai.runner'} in result['warnings']
 
 
 @pytest.mark.parametrize('value', [None, False, 0, '', [], 'secret'])
@@ -154,33 +180,59 @@ def test_malformed_active_runner_section_is_not_defaulted(value):
     assert_block(plan(source), 'invalid_type', 'ai.deerflow-api')
 
 
-def test_existing_runner_config_is_not_overwritten():
+def test_existing_runner_config_is_replaced_with_warning():
     source = source_for()
-    source['ai']['runner_config'] = {'plugin:custom/Runner/default': {'secret': 'value'}}
-    assert_block(plan(source), 'mixed_runner_config', 'ai.runner_config')
-
-
-def test_unknown_selection_is_not_echoed_as_public_legacy_id():
-    source = source_for()
-    source['ai']['runner']['runner'] = 'secret-value'
+    source['ai']['runner_config'] = {'plugin:custom/Runner/default': {'secret': 'SECRET-value'}}
     result = plan(source)
-    assert result['state'] == 'not_legacy'
-    assert 'secret-value' not in json.dumps(result)
+    assert_repaired(result, 'migration.runner_config_replaced', 'ai.runner_config')
+    assert list(result['config']['ai']['runner_config']) == [result['target_runner_id']]
+
+
+@pytest.mark.parametrize('name', ['my-runner', 'custom_runner'])
+def test_custom_legacy_runner_is_unsupported_not_blocked(name):
+    source = source_for()
+    source['ai']['runner']['runner'] = name
+    result = plan(source)
+    assert result['state'] == 'unsupported_runner'
+    assert result['config'] is None
+    assert result['blockers'] == []
+    assert result['changed_paths'] == []
+    assert {'code': 'legacy.custom_runner_unsupported', 'field': 'ai.runner.runner'} in result['warnings']
+
+
+def test_custom_runner_diagnostics_never_echo_section_values():
+    source = source_for()
+    source['ai']['runner']['runner'] = 'my-runner'
+    source['ai']['my-runner'] = {'api-key': 'SECRET-KEY'}
+    result = plan(source)
+    assert result['state'] == 'unsupported_runner'
+    assert 'SECRET' not in json.dumps(result)
+
+
+def test_leftover_custom_section_does_not_block_builtin_migration():
+    source = source_for()
+    source['ai']['my-runner'] = {'api-key': 'SECRET-KEY'}
+    result = plan(source)
+    assert_repaired(result, 'migration.unknown_section_archived', 'ai')
+    assert 'my-runner' not in result['config']['ai']
 
 
 @pytest.mark.parametrize('value', [None, False, -1, 1.5, '0'])
-def test_expiry_is_strict_nonnegative_integer(value):
+def test_malformed_expiry_is_defaulted(value):
     source = source_for()
     source['ai']['runner']['expire-time'] = value
-    assert_block(plan(source), 'invalid_expiry', 'ai.runner.expire-time')
+    result = plan(source)
+    assert result['state'] == 'ready'
+    assert result['config']['ai']['runner']['expire-time'] == 0
 
 
-def test_unknown_active_field_name_and_value_never_leak():
+def test_unknown_active_field_is_dropped_without_leaking():
     source = source_for()
     source['ai']['deerflow-api']['secret-field-name'] = 'secret-field-value'
     result = plan(source)
-    assert_block(result, 'unknown_field', 'ai.deerflow-api')
-    assert 'secret-field' not in json.dumps(result)
+    assert_repaired(result, 'migration.unknown_field_dropped', 'ai.deerflow-api')
+    assert 'secret-field' not in json.dumps({k: v for k, v in result.items() if k != 'config'})
+    assert 'secret-field-name' not in json.dumps(result['config']['ai'])
 
 
 @pytest.mark.parametrize('value', [float('nan'), float('inf'), {1: 'secret'}, ('tuple',), b'secret'])
@@ -198,10 +250,12 @@ def test_cyclic_input_blocks_without_recursion_error():
 
 @pytest.mark.parametrize('runner', TARGETS)
 @pytest.mark.parametrize('value', [None, False, 0, '', [], {}])
-def test_unknown_active_keys_block_even_when_falsey(runner, value):
+def test_unknown_active_keys_are_dropped_even_when_falsey(runner, value):
     source = source_for(runner)
     source['ai'][runner]['secret-unknown-key'] = value
-    assert_block(plan(source), 'unknown_field', f'ai.{runner}')
+    result = plan(source)
+    assert_repaired(result, 'migration.unknown_field_dropped', f'ai.{runner}')
+    assert 'secret-unknown-key' not in selected_config(result)
 
 
 @pytest.mark.parametrize('runner', TARGETS)
@@ -217,8 +271,9 @@ def test_unknown_active_envelope_keys_are_not_echoed(container):
     target = source['ai'] if container == 'ai' else source['ai']['runner']
     target['secret-unknown-key'] = 'secret-value'
     result = plan(source)
-    assert_block(result, 'unknown_field', 'ai' if container == 'ai' else 'ai.runner')
-    assert 'secret-' not in json.dumps(result)
+    code = 'migration.unknown_section_archived' if container == 'ai' else 'migration.unknown_field_dropped'
+    assert_repaired(result, code, 'ai' if container == 'ai' else 'ai.runner')
+    assert 'secret-' not in json.dumps(result['config']['ai'])
 
 
 @pytest.mark.parametrize(
@@ -237,17 +292,29 @@ def test_unknown_active_envelope_keys_are_not_echoed(container):
         {'mcp_resources': [False]},
     ],
 )
-def test_malformed_extension_preferences_block_without_expanding_access(preferences):
+def test_malformed_extension_preferences_are_ignored_without_expanding_access(preferences):
     result = plan(source_for(), preferences)
-    assert result['state'] == 'blocked'
-    assert any(item['code'] == 'invalid_extension_preferences' for item in result['blockers'])
-    assert 'secret' not in json.dumps(result)
+    assert result['state'] == 'ready'
+    assert any(item['code'] == 'invalid_extension_preferences' for item in result['warnings'])
+    assert result['extensions_preferences'] is None
+    assert 'secret' not in json.dumps({k: v for k, v in result.items() if k != 'config'})
 
 
-def test_explicit_plugin_denylist_is_not_silently_broadened():
-    preferences = {'enable_all_plugins': False, 'plugins': []}
-    assert_block(plan(source_for(), preferences), 'extensions.runner_excluded', 'extensions_preferences.plugins')
-    assert preferences == {'enable_all_plugins': False, 'plugins': []}
+def test_explicit_plugin_allowlist_gets_only_the_target_runner_added():
+    preferences = {'enable_all_plugins': False, 'plugins': [{'author': 'other', 'name': 'Plugin'}]}
+    result = plan(source_for(), preferences)
+    assert_repaired(result, 'extensions.runner_added', 'extensions_preferences.plugins')
+    assert result['extensions_preferences'] == {
+        'enable_all_plugins': False,
+        'plugins': [{'author': 'other', 'name': 'Plugin'}, {'author': 'langbot-team', 'name': 'DeerFlowAgent'}],
+    }
+    assert preferences == {'enable_all_plugins': False, 'plugins': [{'author': 'other', 'name': 'Plugin'}]}
+
+
+def test_enable_all_plugins_needs_no_extension_patch():
+    result = plan(source_for(), {'enable_all_plugins': True})
+    assert result['state'] == 'ready'
+    assert result['extensions_preferences'] is None
 
 
 @pytest.mark.parametrize('field', ['api-key', 'auth-header'])
@@ -274,29 +341,50 @@ def test_deerflow_option_trim_drift_is_not_repaired(field, value):
 
 
 @pytest.mark.parametrize('field', ['thinking-enabled', 'plan-mode', 'subagent-enabled'])
-@pytest.mark.parametrize('value', [None, 0, 1, 'false', []])
-def test_deerflow_requires_real_booleans(field, value):
+@pytest.mark.parametrize(
+    'value,expected,code',
+    [
+        (None, False, 'migration.invalid_value_defaulted'),
+        ([], False, 'migration.invalid_value_defaulted'),
+        (0, False, 'migration.value_coerced'),
+        (1, True, 'migration.value_coerced'),
+        ('false', False, 'migration.value_coerced'),
+        ('true', True, 'migration.value_coerced'),
+    ],
+)
+def test_deerflow_booleans_are_coerced_or_defaulted(field, value, expected, code):
     source = source_for()
     source['ai']['deerflow-api'][field] = value
-    assert_block(plan(source), 'invalid_type', f'ai.deerflow-api.{field}')
+    result = plan(source)
+    assert_repaired(result, code, f'ai.deerflow-api.{field}')
+    assert selected_config(result)[field] is expected
 
 
-@pytest.mark.parametrize('field', ['timeout', 'max-concurrent-subagents', 'recursion-limit'])
-@pytest.mark.parametrize('value', [None, False, 1.5, '3'])
-def test_deerflow_requires_real_integers_without_coercion(field, value):
+@pytest.mark.parametrize(
+    'value,expected,code',
+    [
+        (None, 300, 'migration.invalid_value_defaulted'),
+        (False, 300, 'migration.invalid_value_defaulted'),
+        (1.5, 300, 'migration.invalid_value_defaulted'),
+        ('3', 3, 'migration.value_coerced'),
+        (0, 300, 'migration.invalid_value_defaulted'),
+        (-1, 300, 'migration.invalid_value_defaulted'),
+    ],
+)
+def test_deerflow_timeout_is_coerced_or_defaulted(value, expected, code):
     source = source_for()
-    source['ai']['deerflow-api'][field] = value
-    assert_block(plan(source), 'invalid_type', f'ai.deerflow-api.{field}')
+    source['ai']['deerflow-api']['timeout'] = value
+    result = plan(source)
+    assert_repaired(result, code, 'ai.deerflow-api.timeout')
+    assert selected_config(result)['timeout'] == expected
 
 
-@pytest.mark.parametrize('field', ['timeout', 'max-concurrent-subagents', 'recursion-limit'])
+@pytest.mark.parametrize('field', ['max-concurrent-subagents', 'recursion-limit'])
 @pytest.mark.parametrize('value', [0, -1, 1])
 def test_deerflow_does_not_invent_numeric_clamps(field, value):
     source = source_for()
     source['ai']['deerflow-api'][field] = value
-    result = plan(source)
-    assert result['state'] == 'ready'
-    assert result['config']['ai']['runner_config'][result['target_runner_id']][field] == value
+    assert selected_config(plan(source))[field] == value
 
 
 @pytest.mark.parametrize('value', [None, False, 0, [], ''])
@@ -333,7 +421,7 @@ def test_deerflow_native_runtime_defaults_are_materialized_not_ui_defaults():
         'mcp-resources': [],
         'mcp-resource-agent-read-enabled': False,
     }
-    assert 'expire-time' not in result['config']['ai']['runner']
+    assert result['config']['ai']['runner']['expire-time'] == 0
 
 
 def test_deerflow_missing_endpoint_blocks_instead_of_using_schema_localhost():
@@ -361,11 +449,21 @@ def test_ready_public_projection_contains_no_secrets_or_credentials_in_urls():
     assert 'SECRET' not in json.dumps(public)
 
 
-@pytest.mark.parametrize('value', [None, 0, 'false'])
-def test_remove_think_flag_is_not_coerced(value):
+@pytest.mark.parametrize(
+    'value,expected,code',
+    [
+        (None, False, 'migration.invalid_value_defaulted'),
+        (0, False, 'migration.value_coerced'),
+        ('false', False, 'migration.value_coerced'),
+        ('true', True, 'migration.value_coerced'),
+    ],
+)
+def test_remove_think_flag_is_coerced_or_defaulted(value, expected, code):
     source = source_for('tbox-app-api')
     source['output']['misc']['remove-think'] = value
-    assert_block(plan(source), 'invalid_type', 'output.misc.remove-think')
+    result = plan(source)
+    assert_repaired(result, code, 'output.misc.remove-think')
+    assert selected_config(result)['remove-think'] is expected
 
 
 @pytest.mark.parametrize(
@@ -401,83 +499,130 @@ def test_local_rounds_are_never_translated_into_transcript_item_counts(rounds):
 
 
 @pytest.mark.parametrize(
-    'field,value,code',
+    'field,value,code,expected',
     [
         (
             'model',
             {'primary': 'model', 'fallbacks': [], 'reasoning': {'SECRET-model': 'invalid-level'}},
-            'local.reasoning_value',
+            'local.reasoning_dropped',
+            {'primary': 'model', 'fallbacks': [], 'reasoning': {}},
+        ),
+        (
+            'prompt',
+            [{'role': 'system', 'content': 42}],
+            'local.prompt_normalized',
+            [{'role': 'system', 'content': 'You are a helpful assistant.'}],
+        ),
+        (
+            'prompt',
+            [{'role': 'system', 'content': 'text', 'SECRET-key': 'SECRET-value'}],
+            'local.prompt_normalized',
+            [{'role': 'system', 'content': 'text'}],
+        ),
+        (
+            'prompt',
+            [{'role': 'system', 'content': [{'type': 'text', 'text': 42}]}],
+            'local.prompt_normalized',
+            [{'role': 'system', 'content': 'You are a helpful assistant.'}],
         ),
-        ('prompt', [{'role': 'system', 'content': 42}], 'local.prompt_shape'),
-        ('prompt', [{'role': 'system', 'content': 'text', 'SECRET-key': 'SECRET-value'}], 'local.prompt_shape'),
-        ('prompt', [{'role': 'system', 'content': [{'type': 'text', 'text': 42}]}], 'local.prompt_shape'),
     ],
 )
-def test_local_unsupported_behaviors_have_specific_safe_blockers(field, value, code):
+def test_local_unsupported_behaviors_are_repaired_without_leaking(field, value, code, expected):
     source = source_for('local-agent')
     source['ai']['local-agent'][field] = value
     result = plan(source)
-    assert_block(result, code, f'ai.local-agent.{field}')
-    assert 'SECRET' not in json.dumps(result)
+    assert_repaired(result, code, f'ai.local-agent.{field}' + ('.reasoning' if field == 'model' else ''))
+    assert selected_config(result)[field] == expected
+    assert 'SECRET' not in json.dumps(result['config']['ai'])
 
 
 @pytest.mark.parametrize(
-    'field,value',
+    'field,value,expected',
     [
-        ('enable-all-tools', None),
-        ('enable-all-tools', 0),
-        ('tools', None),
-        ('tools', [42]),
-        ('mcp-resources', None),
-        ('mcp-resources', [False]),
-        ('mcp-resource-agent-read-enabled', 'false'),
+        ('enable-all-tools', None, True),
+        ('enable-all-tools', 0, False),
+        ('tools', None, []),
+        ('tools', [42], []),
+        ('mcp-resources', None, []),
+        ('mcp-resources', [False], []),
+        ('mcp-resource-agent-read-enabled', 'false', False),
     ],
 )
-def test_local_host_policy_malformed_values_block(field, value):
+def test_local_host_policy_malformed_values_are_repaired(field, value, expected):
     source = source_for('local-agent')
     source['ai']['local-agent'][field] = value
-    assert_block(plan(source), 'invalid_type', f'ai.local-agent.{field}')
+    result = plan(source)
+    assert result['state'] == 'ready'
+    assert selected_config(result)[field] == expected
+    codes = {(w['code'], w['field']) for w in result['warnings']}
+    assert codes & {
+        ('migration.value_coerced', f'ai.local-agent.{field}'),
+        ('migration.invalid_value_defaulted', f'ai.local-agent.{field}'),
+    }
 
 
 @pytest.mark.parametrize(
     'runner,field',
     [
         ('dify-service-api', 'api-key'),
-        ('dify-service-api', 'base-prompt'),
-        ('coze-api', 'timeout'),
-        ('coze-api', 'api-base'),
         ('langflow-api', 'flow-id'),
         ('n8n-service-api', 'webhook-url'),
         ('tbox-app-api', 'app-id'),
-        ('weknora-api', 'app-type'),
     ],
 )
-def test_defaultless_native_fields_are_not_filled_from_target_schema(runner, field):
+def test_missing_required_credentials_and_endpoints_still_block(runner, field):
     source = source_for(runner)
     del source['ai'][runner][field]
     assert_block(plan(source), 'missing_field', f'ai.{runner}.{field}')
 
 
-@pytest.mark.parametrize('value', [None, False, 0, [], {}])
-@pytest.mark.parametrize('runner', [r for r in TARGETS if r not in ('local-agent', 'n8n-service-api')])
+@pytest.mark.parametrize(
+    'runner,field',
+    [
+        ('dify-service-api', 'base-prompt'),
+        ('coze-api', 'timeout'),
+        ('coze-api', 'api-base'),
+        ('weknora-api', 'app-type'),
+    ],
+)
+def test_missing_optional_fields_are_filled_with_410_defaults(runner, field):
+    source = source_for(runner)
+    del source['ai'][runner][field]
+    result = plan(source)
+    assert_repaired(result, 'migration.missing_default_filled', f'ai.{runner}.{field}')
+
+
+@pytest.mark.parametrize('value', [None, False, [], {}])
+@pytest.mark.parametrize('runner', [r for r in TARGETS if r not in ('local-agent', 'n8n-service-api', 'deerflow-api')])
 def test_external_api_keys_require_strings_even_if_target_coerces(runner, value):
     source = source_for(runner)
     source['ai'][runner]['api-key'] = value
     assert_block(plan(source), 'invalid_type', f'ai.{runner}.api-key')
 
 
+@pytest.mark.parametrize('runner', [r for r in TARGETS if r not in ('local-agent', 'n8n-service-api')])
+def test_numeric_external_api_key_is_stringified(runner):
+    source = source_for(runner)
+    source['ai'][runner]['api-key'] = 0
+    result = plan(source)
+    assert_repaired(result, 'migration.value_coerced', f'ai.{runner}.api-key')
+    assert selected_config(result)['api-key'] == '0'
+
+
 @pytest.mark.parametrize(
-    'section',
+    'section,expected',
     [
-        {'auto_save_history': False, 'auto-save-history': True},
-        {'auto_save_history': True, 'auto-save-history': False},
+        ({'auto_save_history': False, 'auto-save-history': True}, True),
+        ({'auto_save_history': True, 'auto-save-history': False}, False),
     ],
 )
-def test_coze_actual_underscore_history_key_missing_null_or_conflict_blocks(section):
+def test_coze_history_alias_conflict_uses_canonical_field(section, expected):
     source = source_for('coze-api')
     source['ai']['coze-api'].pop('auto_save_history')
     source['ai']['coze-api'].update(section)
-    assert_block(plan(source), 'coze.history_alias', 'ai.coze-api.auto_save_history')
+    result = plan(source)
+    assert_repaired(result, 'migration.alias_conflict_resolved', 'ai.coze-api.auto_save_history')
+    assert selected_config(result)['auto-save-history'] is expected
 
 
 @pytest.mark.parametrize('value', [False, True])
@@ -497,27 +642,36 @@ def test_coze_custom_endpoint_not_replaced_with_schema_region():
     assert 'secret-' not in json.dumps(result)
 
 
-@pytest.mark.parametrize('underscore,hyphen', [('one', 'two')])
-def test_dashscope_seed_hyphen_is_not_actual_native_runtime_key(underscore, hyphen):
+def test_dashscope_seed_alias_conflict_is_resolved_with_warning():
     source = source_for('dashscope-app-api')
     section = source['ai']['dashscope-app-api']
     section.pop('references_quote')
-    if underscore is not None:
-        section['references_quote'] = underscore
-    section['references-quote'] = hyphen
-    assert_block(plan(source), 'dashscope.references_alias', 'ai.dashscope-app-api.references_quote')
+    section['references_quote'] = 'one'
+    section['references-quote'] = 'two'
+    assert_repaired(plan(source), 'migration.alias_conflict_resolved', 'ai.dashscope-app-api.references-quote')
 
 
 @pytest.mark.parametrize('axis', ['input', 'output'])
-@pytest.mark.parametrize('underscore,hyphen', [(None, 'text'), ('chat', 'text'), ('text', 'chat')])
-def test_langflow_conflicting_or_nondefault_ui_only_alias_blocks(axis, underscore, hyphen):
+@pytest.mark.parametrize(
+    'underscore,hyphen,code',
+    [
+        (None, 'text', 'migration.alias_repaired'),
+        ('chat', 'text', 'migration.alias_conflict_resolved'),
+        ('text', 'chat', 'migration.alias_conflict_resolved'),
+    ],
+)
+def test_langflow_io_aliases_use_canonical_hyphen_field(axis, underscore, hyphen, code):
     source = source_for('langflow-api')
     section = source['ai']['langflow-api']
     section.pop(f'{axis}_type')
     if underscore is not None:
         section[f'{axis}_type'] = underscore
     section[f'{axis}-type'] = hyphen
-    assert_block(plan(source), 'langflow.io_alias', f'ai.langflow-api.{axis}_type')
+    result = plan(source)
+    assert result['state'] == 'ready'
+    assert selected_config(result)[f'{axis}-type'] == hyphen
+    if underscore is not None:
+        assert {'code': code, 'field': f'ai.langflow-api.{axis}_type'} in result['warnings']
 
 
 @pytest.mark.parametrize(
@@ -537,12 +691,13 @@ def test_langflow_conflicting_or_nondefault_ui_only_alias_blocks(axis, underscor
         0,
     ],
 )
-def test_langflow_tweaks_require_strict_object_json_without_fallback_loss(raw):
+def test_invalid_langflow_tweaks_are_defaulted_without_leaking(raw):
     source = source_for('langflow-api')
     source['ai']['langflow-api']['tweaks'] = raw
     result = plan(source)
-    assert_block(result, 'langflow.invalid_tweaks', 'ai.langflow-api.tweaks')
-    assert 'broken-secret' not in json.dumps(result)
+    assert_repaired(result, 'langflow.tweaks_invalid_defaulted', 'ai.langflow-api.tweaks')
+    assert selected_config(result)['tweaks'] == {}
+    assert 'broken-secret' not in json.dumps({k: v for k, v in result.items() if k != 'config'})
 
 
 @pytest.mark.parametrize('raw', [None, '', '{}', '{"nested":{"token":"SECRET","values":[null,false,0]}}'])
@@ -616,19 +771,20 @@ def test_n8n_secrets_remain_raw_and_not_public_even_for_inactive_modes(mode):
 
 
 @pytest.mark.parametrize(
-    'runner,field,value',
+    'runner,field,value,expected',
     [
-        ('dify-service-api', 'app-type', 'unknown'),
-        ('dashscope-app-api', 'app-type', 'chat'),
-        ('weknora-api', 'app-type', 'workflow'),
-        ('n8n-service-api', 'auth-type', 'SECRET-unknown'),
+        ('dify-service-api', 'app-type', 'unknown', 'chat'),
+        ('dashscope-app-api', 'app-type', 'chat', 'agent'),
+        ('weknora-api', 'app-type', 'workflow', 'chat'),
+        ('n8n-service-api', 'auth-type', 'SECRET-unknown', 'none'),
     ],
 )
-def test_invalid_enum_values_are_not_echoed(runner, field, value):
+def test_invalid_enum_values_are_defaulted_and_not_echoed(runner, field, value, expected):
     source = source_for(runner)
     source['ai'][runner][field] = value
     result = plan(source)
-    assert_block(result, 'invalid_value', f'ai.{runner}.{field}')
+    assert_repaired(result, 'migration.invalid_value_defaulted', f'ai.{runner}.{field}')
+    assert selected_config(result)[field] == expected
     assert 'SECRET' not in json.dumps(result)
 
 
@@ -818,7 +974,7 @@ def test_dashscope_unambiguous_seed_alias_is_repaired(value):
 
 @pytest.mark.parametrize('mode', ['agent', 'chat'])
 @pytest.mark.parametrize('value', ['ABSENT', None, '', ' ', 'saved-agent'])
-def test_weknora_agent_absence_is_not_null_or_empty(mode, value):
+def test_weknora_agent_absence_is_defaulted(mode, value):
     source = source_for('weknora-api')
     section = source['ai']['weknora-api']
     section['app-type'] = mode
@@ -826,9 +982,8 @@ def test_weknora_agent_absence_is_not_null_or_empty(mode, value):
     if value != 'ABSENT':
         section['agent-id'] = value
     selected = selected_config(plan(source))
-    assert selected['agent-id'] == (
-        ('builtin-quick-answer' if mode == 'chat' else 'builtin-smart-reasoning') if value == 'ABSENT' else value
-    )
+    default = 'builtin-quick-answer' if mode == 'chat' else 'builtin-smart-reasoning'
+    assert selected['agent-id'] == (default if value in ('ABSENT', None) else value)
     assert selected['knowledge-base-ids'] == []
     assert selected['knowledge-bases'] == []
 
@@ -914,18 +1069,22 @@ def test_n8n_ignore_and_native_latin1_are_selected():
     assert selected['basic-encoding'] == 'latin1'
 
 
-def test_weknora_blank_remote_kb_is_rejected():
+def test_weknora_blank_remote_kb_is_dropped():
     source = source_for('weknora-api')
     source['ai']['weknora-api']['knowledge-base-ids'] = [' ']
-    assert_block(plan(source), 'invalid_value', 'ai.weknora-api.knowledge-base-ids')
+    result = plan(source)
+    assert_repaired(result, 'migration.value_coerced', 'ai.weknora-api.knowledge-base-ids')
+    assert selected_config(result)['knowledge-base-ids'] == []
 
 
 @pytest.mark.parametrize('runner', ['coze-api', 'n8n-service-api', 'weknora-api'])
 @pytest.mark.parametrize('value', [0, -1])
-def test_external_target_requires_positive_timeout(runner, value):
+def test_nonpositive_external_timeout_is_defaulted(runner, value):
     source = source_for(runner)
     source['ai'][runner]['timeout'] = value
-    assert_block(plan(source), 'invalid_value', f'ai.{runner}.timeout')
+    result = plan(source)
+    assert_repaired(result, 'migration.invalid_value_defaulted', f'ai.{runner}.timeout')
+    assert selected_config(result)['timeout'] == 120
 
 
 @pytest.mark.parametrize(
@@ -975,6 +1134,7 @@ def test_ready_all9_nested_candidates_are_detached_and_no_io(runner, monkeypatch
         'target_runner_id',
         'target_plugin',
         'config',
+        'extensions_preferences',
         'changed_paths',
         'blockers',
         'warnings',
@@ -989,13 +1149,14 @@ def test_empty_config_is_not_legacy():
         'target_runner_id': None,
         'target_plugin': None,
         'config': None,
+        'extensions_preferences': None,
         'changed_paths': [],
         'blockers': [],
         'warnings': [],
     }
 
 
-def test_default_local_agent_blocks_instead_of_inventing_round_translation():
+def test_minimal_local_agent_is_completed_with_defaults():
     source = {
         'ai': {
             'runner': {'runner': 'local-agent', 'expire-time': 0},
@@ -1004,15 +1165,25 @@ def test_default_local_agent_blocks_instead_of_inventing_round_translation():
     }
     original = copy.deepcopy(source)
     result = planner().plan_legacy_pipeline(source)
-    assert planner().PLANNER_VERSION == '4'
-    assert result['state'] == 'blocked'
+    assert planner().PLANNER_VERSION == '5'
     assert result['target_runner_id'] == 'plugin:langbot-team/LocalAgent/default'
-    assert result['target_plugin'] == {'author': 'langbot-team', 'name': 'LocalAgent', 'version': '0.1.10'}
-    assert {'code': 'missing_field', 'field': 'ai.local-agent.prompt'} in result['blockers']
-    assert result['config'] is None
+    assert result['target_plugin'] == {'author': 'langbot-team', 'name': 'LocalAgent', 'version': '0.2.1'}
+    selected = selected_config(result)
+    assert selected['model'] == {'primary': 'model-id', 'fallbacks': [], 'reasoning': {}}
+    assert 'max-round' not in selected
+    assert selected['prompt'] == [{'role': 'system', 'content': 'You are a helpful assistant.'}]
+    assert {'code': 'migration.new_defaults', 'field': 'ai.local-agent'} in result['warnings']
     assert source == original
 
 
+def test_local_agent_without_primary_model_still_blocks():
+    source = source_for('local-agent')
+    del source['ai']['local-agent']['model']
+    result = plan(source)
+    assert result['state'] == 'blocked'
+    assert result['config'] is None
+
+
 @pytest.mark.parametrize('template', ['{global}', '{launcher_type}_{launcher_id}', '{sender_id}', '{project}'])
 def test_local_box_reuse_templates_are_preserved_for_plugin(template):
     source = source_for('local-agent')
@@ -1025,8 +1196,11 @@ def test_local_box_reuse_templates_are_preserved_for_plugin(template):
 @pytest.mark.parametrize(
     'template', ['', ' ', '{}', '{0}', '{sender_id', '{sender_id!r}', '{query_id:04}', '{actor.id}', '{actor[id]}']
 )
-def test_invalid_box_templates_are_reported_before_migration(template):
+def test_invalid_box_templates_are_defaulted(template):
     source = source_for('local-agent')
     source['ai']['local-agent']['box-session-id-template'] = template
     result = plan(source)
-    assert_block(result, 'local.box_template_invalid', 'ai.local-agent.box-session-id-template')
+    assert_repaired(result, 'local.box_template_defaulted', 'ai.local-agent.box-session-id-template')
+    assert selected_config(result)['box-session-id-template'] == '{launcher_type}_{launcher_id}'
+
+
diff --git a/tests/unit_tests/pipeline/test_migration_current_shape.py b/tests/unit_tests/pipeline/test_migration_current_shape.py
index f6189633d..547149007 100644
--- a/tests/unit_tests/pipeline/test_migration_current_shape.py
+++ b/tests/unit_tests/pipeline/test_migration_current_shape.py
@@ -57,11 +57,13 @@ def test_box_reuse_template_is_preserved_for_runner(template):
     assert result['config']['ai']['runner_config'][result['target_runner_id']]['box-session-id-template'] == template
 
 
-def test_empty_box_template_requires_correction():
+def test_empty_box_template_is_defaulted_with_warning():
     source = {'ai': {'runner': {'runner': 'local-agent'}, 'local-agent': copy.deepcopy(FIXTURES['local-agent'])}}
     source['ai']['local-agent']['box-session-id-template'] = ''
     result = plan_legacy_pipeline(source)
-    assert result['state'] == 'blocked'
-    assert {'code': 'local.box_template_invalid', 'field': 'ai.local-agent.box-session-id-template'} in result[
-        'blockers'
+    assert result['state'] == 'ready'
+    selected = result['config']['ai']['runner_config'][result['target_runner_id']]
+    assert selected['box-session-id-template'] == '{launcher_type}_{launcher_id}'
+    assert {'code': 'local.box_template_defaulted', 'field': 'ai.local-agent.box-session-id-template'} in result[
+        'warnings'
     ]
diff --git a/tests/unit_tests/pipeline/test_stream_results.py b/tests/unit_tests/pipeline/test_stream_results.py
new file mode 100644
index 000000000..7ddd7af3c
--- /dev/null
+++ b/tests/unit_tests/pipeline/test_stream_results.py
@@ -0,0 +1,112 @@
+import pytest
+
+from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+from langbot_plugin.api.entities.builtin.platform.message import File, MessageChain
+from langbot.pkg.pipeline.process.stream_results import coalesce_stream_results
+
+
+async def source(*items):
+    for item in items:
+        yield item
+
+
+@pytest.mark.asyncio
+async def test_terminal_snapshot_is_delivered_once_with_attachments():
+    text = '<think>reasoning</think>answer'
+    completed = Message(role='assistant', content=text)
+    completed.attachments = MessageChain([File(url='https://example.com/report.txt', name='report.txt')])
+    results = [
+        item
+        async for item in coalesce_stream_results(
+            source(
+                MessageChunk(role='assistant', content='answer', all_content=text, is_final=True),
+                completed,
+            )
+        )
+    ]
+    assert results == [completed]
+    assert results[0].attachments is completed.attachments
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('with_snapshot', [False, True])
+async def test_progressive_chunks_are_not_buffered(with_snapshot):
+    first = MessageChunk(role='assistant', content='hel', all_content='hel')
+    final = MessageChunk(role='assistant', content='lo', all_content='hello', is_final=True)
+    completed = Message(role='assistant', content='hello')
+    stream = coalesce_stream_results(source(first, final, *([completed] if with_snapshot else [])))
+    assert await anext(stream) == first
+    assert [item async for item in stream] == [completed if with_snapshot else final]
+
+
+@pytest.mark.asyncio
+async def test_completed_only_and_distinct_messages_are_preserved():
+    final = MessageChunk(role='assistant', content='first', is_final=True)
+    other = Message(role='assistant', content='second')
+    results = [item async for item in coalesce_stream_results(source(final, other))]
+    assert [item.content for item in results] == ['first', 'second']
+    assert [item async for item in coalesce_stream_results(source(other))] == [other]
+
+
+@pytest.mark.asyncio
+async def test_blank_terminal_delta_uses_completed_snapshot():
+    completed = Message(role='assistant', content='answer')
+    results = [
+        item
+        async for item in coalesce_stream_results(
+            source(
+                MessageChunk(role='assistant', is_final=True),
+                completed,
+            )
+        )
+    ]
+    assert results == [completed]
+
+
+@pytest.mark.asyncio
+async def test_failures_propagate_without_successful_final():
+    async def failing():
+        yield MessageChunk(role='assistant', content='partial')
+        raise RuntimeError('provider failed')
+
+    stream = coalesce_stream_results(failing())
+    assert (await anext(stream)).content == 'partial'
+    with pytest.raises(RuntimeError, match='provider failed'):
+        await anext(stream)
+
+
+@pytest.mark.asyncio
+async def test_cumulative_stream_ends_with_last_snapshot_even_when_final_is_blank():
+    results = [
+        item
+        async for item in coalesce_stream_results(
+            source(
+                MessageChunk(role='assistant', content='hel'),
+                MessageChunk(role='assistant', content='hello'),
+                MessageChunk(role='assistant', is_final=True),
+            )
+        )
+    ]
+    assert [item.all_content for item in results] == ['hel', 'hello', 'hello']
+    assert results[-1].is_final
+
+
+@pytest.mark.asyncio
+async def test_multi_round_completion_keeps_one_transcript_and_attachments():
+    first = '<think>plan</think>I will create a file.'
+    last = '<think>file created</think>Here is your file.'
+    completed = Message(role='assistant', content=last)
+    completed.attachments = MessageChain([File(name='report.txt', base64='aGVsbG8=')])
+    results = [
+        item
+        async for item in coalesce_stream_results(
+            source(
+                MessageChunk(role='assistant', content=first, is_final=False),
+                MessageChunk(role='assistant', content=first + last, is_final=True),
+                completed,
+            )
+        )
+    ]
+    assert len(results) == 2
+    assert results[-1].content == first + last
+    assert results[-1].attachments is completed.attachments
diff --git a/tests/unit_tests/platform/onebot_mock.py b/tests/unit_tests/platform/onebot_mock.py
new file mode 100644
index 000000000..6295bae75
--- /dev/null
+++ b/tests/unit_tests/platform/onebot_mock.py
@@ -0,0 +1,107 @@
+"""Independent OneBot v11 wire peer; no LangBot transport/converter imports.
+
+Contract: https://github.com/botuniverse/onebot-11 (api/public.md,
+communication/{ws,ws-reverse,http}.md and event/*.md).
+Only test data is served; no real QQ account or vendor service is contacted.
+"""
+
+import asyncio
+import json
+
+
+class OneBotMock:
+    def __init__(self):
+        self.ready = asyncio.Event()
+        self.requests = []
+        self.send = None
+        self.response = None
+        self.user = {'user_id': 10001, 'nickname': 'Alice', 'sex': 'unknown', 'age': 20}
+        self.group = {'group_id': 20001, 'group_name': 'Mock group', 'member_count': 2, 'max_member_count': 200}
+        self.member = self.user | {'group_id': 20001, 'card': 'Card', 'role': 'admin', 'title': 'Title', 'join_time': 1}
+
+    async def run(self, send, receive):
+        self.send = send
+        self.ready.set()
+        while True:
+            request = json.loads(await receive())
+            self.requests.append(request)
+            if self.response is not None:
+                response = dict(self.response)
+            else:
+                response = self.handle(request)
+            await send(json.dumps(response | {'echo': request.get('echo')}))
+
+    def handle(self, request):
+        action, params = request['action'], request.get('params', {})
+        for key in ('user_id', 'group_id', 'message_id', 'duration'):
+            if key in params:
+                assert type(params[key]) is int, (action, key, params[key])
+        data = None
+        if action in ('send_msg', 'send_private_msg', 'send_group_msg'):
+            assert isinstance(params['message'], (list, str))
+            data = {'message_id': 123}
+        elif action == 'get_msg':
+            data = self.message('private', 'friend')
+        elif action == 'get_login_info':
+            data = {'user_id': 42, 'nickname': 'Mock bot'}
+        elif action == 'get_stranger_info':
+            data = self.user
+        elif action == 'get_friend_list':
+            data = [self.user | {'remark': 'Friend'}]
+        elif action == 'get_group_info':
+            data = self.group
+        elif action == 'get_group_list':
+            data = [self.group]
+        elif action == 'get_group_member_info':
+            data = self.member
+        elif action == 'get_group_member_list':
+            data = [self.member]
+        elif action == 'get_status':
+            data = {'online': True, 'good': True}
+        elif action == 'get_version_info':
+            data = {'app_name': 'onebot-test-peer', 'app_version': '1', 'protocol_version': 'v11'}
+        elif action in ('can_send_image', 'can_send_record'):
+            data = {'yes': True}
+        elif action in ('get_image', 'get_record'):
+            data = {'file': '/mock/media'}
+        elif action == 'get_forward_msg':
+            data = {'message': []}
+        elif action == 'get_group_honor_info':
+            data = {'group_id': 20001}
+        elif action not in {
+            'delete_msg',
+            'set_group_name',
+            'set_group_ban',
+            'set_group_kick',
+            'set_group_leave',
+            'set_friend_add_request',
+            'set_group_add_request',
+            'set_group_card',
+            'set_group_special_title',
+            'set_group_admin',
+            'set_group_whole_ban',
+        }:
+            return {'status': 'failed', 'retcode': 1404, 'data': None}
+        return {'status': 'ok', 'retcode': 0, 'data': data}
+
+    async def emit(self, payload):
+        await self.ready.wait()
+        await self.send(json.dumps({'time': 1700000000, 'self_id': 42} | payload))
+
+    def message(self, kind, subtype, message=None):
+        result = {
+            'post_type': 'message',
+            'message_type': kind,
+            'sub_type': subtype,
+            'time': 1700000000,
+            'self_id': 42,
+            'message_id': 9,
+            'user_id': 10001,
+            'message': message if message is not None else [{'type': 'text', 'data': {'text': 'hello'}}],
+            'raw_message': 'hello',
+            'font': 0,
+            'sender': self.user,
+        }
+        if kind == 'group':
+            result.update(group_id=20001, anonymous=None, sender=self.member)
+        return result
diff --git a/tests/unit_tests/platform/test_aiocqhttp_eba_adapter.py b/tests/unit_tests/platform/test_aiocqhttp_eba_adapter.py
index 584d39d91..58473bf00 100644
--- a/tests/unit_tests/platform/test_aiocqhttp_eba_adapter.py
+++ b/tests/unit_tests/platform/test_aiocqhttp_eba_adapter.py
@@ -34,7 +34,7 @@ class DummyLogger(AbstractEventLogger):
 
 
 def make_adapter() -> AiocqhttpAdapter:
-    return AiocqhttpAdapter({'host': '127.0.0.1', 'port': 2280, 'access-token': ''}, DummyLogger())
+    return AiocqhttpAdapter({'host': '127.0.0.1', 'port': 2280, 'access-token': 'test-secret'}, DummyLogger())
 
 
 def manifest() -> dict:
@@ -48,13 +48,40 @@ def manifest() -> dict:
         / 'aiocqhttp'
         / 'manifest.yaml'
     )
-    return yaml.safe_load(manifest_path.read_text())
+    return yaml.safe_load(manifest_path.read_text(encoding='utf-8'))
 
 
 def onebot_event(payload: dict) -> aiocqhttp.Event:
     return aiocqhttp.Event.from_payload(payload)
 
 
+@pytest.mark.parametrize('meta_type', ['heartbeat', 'lifecycle', 'future_transport_metadata'])
+async def test_protocol_metadata_never_reaches_business_listeners(meta_type):
+    adapter = make_adapter()
+    listener = AsyncMock()
+    adapter.register_listener(platform_events.EBAEvent, listener)
+    payload = {'post_type': 'meta_event', 'meta_event_type': meta_type, 'self_id': 42, 'time': 1700000000}
+    assert await AiocqhttpEventConverter.target2yiri(onebot_event(payload)) is None
+    await adapter._receive_event(payload)
+    listener.assert_not_awaited()
+    assert adapter.bot_account_id == '42'
+
+    # Genuine platform-specific user interactions still reach business routing.
+    await adapter._receive_event(
+        {
+            'post_type': 'notice',
+            'notice_type': 'notify',
+            'sub_type': 'poke',
+            'self_id': 42,
+            'time': 1700000001,
+            'user_id': 10001,
+            'target_id': 42,
+        }
+    )
+    listener.assert_awaited_once()
+    assert listener.await_args.args[0].type == 'platform.specific'
+
+
 def test_aiocqhttp_supported_events_match_manifest():
     assert make_adapter().get_supported_events() == manifest()['spec']['supported_events']
 
@@ -255,7 +282,7 @@ async def test_aiocqhttp_event_converter_maps_private_and_group_messages():
 
     assert isinstance(private_event, platform_events.MessageReceivedEvent)
     assert private_event.type == 'message.received'
-    assert private_event.adapter_name == 'aiocqhttp-omni'
+    assert private_event.adapter_name == 'onebot-v11'
     assert private_event.chat_type == platform_entities.ChatType.PRIVATE
     assert private_event.chat_id == 10001
     assert private_event.sender.nickname == 'Alice'
diff --git a/tests/unit_tests/platform/test_botmgr_execution_trace.py b/tests/unit_tests/platform/test_botmgr_execution_trace.py
new file mode 100644
index 000000000..1d805f77f
--- /dev/null
+++ b/tests/unit_tests/platform/test_botmgr_execution_trace.py
@@ -0,0 +1,565 @@
+"""RuntimeBot ingress tracing: one inbound event owns one v2 execution chain."""
+
+from __future__ import annotations
+
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, Mock
+import asyncio
+
+import pytest
+
+from langbot.pkg.api.http.context import ExecutionContext
+
+TEST_CONTEXT = ExecutionContext(
+    instance_uuid='instance-test',
+    workspace_uuid='workspace-test',
+    placement_generation=1,
+    bot_uuid='bot-1',
+)
+
+
+class FakeTelemetryManager:
+    def __init__(self, config=None):
+        self.telemetry_config = (
+            {'url': 'https://space.example.test', 'execution_trace': 'all'} if config is None else config
+        )
+        self.sent: list[dict] = []
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+
+def make_bot(event_bindings: list[dict], config=None, agent=None):
+    from langbot.pkg.platform.botmgr import RuntimeBot
+    from langbot.pkg.telemetry.execution import ExecutionCounters
+
+    manager = FakeTelemetryManager(config)
+    counters = ExecutionCounters(manager)
+    bot = object.__new__(RuntimeBot)
+    bot.bot_entity = SimpleNamespace(
+        uuid='bot-1',
+        adapter='aiocqhttp',
+        workspace_uuid=TEST_CONTEXT.workspace_uuid,
+        event_bindings=event_bindings,
+        plugin_processors=[],
+    )
+    bot.execution_context = TEST_CONTEXT
+    bot.workspace_uuid = TEST_CONTEXT.workspace_uuid
+    bot.placement_generation = TEST_CONTEXT.placement_generation
+    bot.logger = SimpleNamespace(info=AsyncMock(), warning=AsyncMock(), error=AsyncMock(), debug=AsyncMock())
+    bot.adapter = FakeAdapter()
+    bot.ap = SimpleNamespace(
+        workspace_service=SimpleNamespace(get_execution_binding=AsyncMock()),
+        telemetry=SimpleNamespace(execution=counters),
+        agent_service=SimpleNamespace(get_agent=AsyncMock(return_value=agent)),
+        pipeline_service=SimpleNamespace(get_pipeline=AsyncMock(return_value=None)),
+    )
+    return bot, manager, counters
+
+
+def message_received_event():
+    from langbot_plugin.api.entities.builtin.platform import entities, events, message
+
+    return events.MessageReceivedEvent(
+        message_id='message-1',
+        message_chain=message.MessageChain([message.Plain(text='hello')]),
+        sender=entities.User(id='user-1', nickname='QA User'),
+        chat_type=entities.ChatType.PRIVATE,
+        chat_id='user-1',
+    )
+
+
+class FakeAdapter:
+    async def get_supported_apis(self):
+        return []
+
+
+async def flushed_records(counters, manager) -> list[dict]:
+    """Flush buffered telemetry and return everything the event emitted.
+
+    Records are batched and uploaded by ``ExecutionCounters.flush()``; there is
+    no per-trace direct send any more.
+    """
+
+    await counters.flush()
+    assert manager.sent, 'expected the flushed telemetry batch'
+    return [record for payload in manager.sent for record in payload.get('records', [])]
+
+
+def execution_nodes(records: list[dict]) -> list[dict]:
+    return [record for record in records if record.get('event_type') == 'execution_node']
+
+
+def execution_chains(records: list[dict]) -> list[dict]:
+    return [record for record in records if record.get('event_type') == 'execution_chain']
+
+
+@pytest.mark.asyncio
+async def test_route_miss_emits_nothing():
+    bot, manager, counters = make_bot([])
+
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+
+    await counters.flush()
+    await counters.shutdown()
+    assert manager.sent == []
+    assert counters.records == []
+    assert counters.dropped == 0
+    assert counters.registry == {}
+
+
+@pytest.mark.asyncio
+async def test_unavailable_route_target_carries_route_identity():
+    bot, manager, counters = make_bot(
+        [
+            {
+                'id': 'agent-binding',
+                'enabled': True,
+                'event_pattern': 'message.received',
+                'target_type': 'agent',
+                'target_uuid': 'agent-1',
+            }
+        ]
+    )
+
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+
+    nodes = execution_nodes(await flushed_records(counters, manager))
+    assert [(node['family'], node['outcome'], node['mode']) for node in nodes] == [
+        ('event_route', 'failed', 'agent'),
+        ('platform_event', 'success', 'none'),
+    ]
+    assert nodes[0]['route_ref'] == 'agent:agent-1'
+    assert nodes[1]['route_ref'] == ''
+
+
+@pytest.mark.asyncio
+async def test_discarded_route_is_only_logged_locally():
+    bot, manager, counters = make_bot(
+        [
+            {
+                'id': 'discard-binding',
+                'enabled': True,
+                'event_pattern': '*',
+                'target_type': 'discard',
+            }
+        ]
+    )
+
+    await bot._handle_platform_event(SimpleNamespace(type='platform.member.joined'), FakeAdapter())
+
+    await counters.flush()
+    assert manager.sent == []
+    assert counters.records == []
+    assert bot.logger.info.await_args.kwargs['metadata']['status'] == 'discarded'
+
+
+@pytest.mark.asyncio
+async def test_telemetry_opt_out_emits_nothing():
+    bot, manager, counters = make_bot([], config={'url': 'https://space.example.test', 'disable_telemetry': True})
+
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+
+    assert manager.sent == []
+    assert counters.records == []
+    assert counters.registry == {}
+
+
+@pytest.mark.asyncio
+async def test_route_trace_scope_is_restored_after_dispatch():
+    from langbot.pkg.telemetry import trace as trace_mod
+
+    bot, _, counters = make_bot([])
+    assert trace_mod.current() is None
+
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+
+    assert trace_mod.current() is None
+    buffered = list(counters.records)
+    # A later record outside the ingress must not join the closed chain, and an
+    # observation without an owning chain is not buffered at all.
+    from langbot.pkg.telemetry.execution import record
+
+    record(bot.ap, TEST_CONTEXT, family='platform_api', operation='send_message', outcome='success')
+    assert counters.records == buffered
+    assert counters.registry == {}
+
+
+@pytest.mark.asyncio
+async def test_mock_adapter_route_miss_is_not_reported():
+    """Mock adapter objects used by other tests must not break the ingress path."""
+    bot, manager, counters = make_bot([])
+    adapter = Mock()
+
+    await bot._handle_platform_event(message_received_event(), adapter)
+
+    await counters.flush()
+    assert manager.sent == []
+
+
+def business_route(**overrides):
+    return {
+        'id': 'route',
+        'enabled': True,
+        'event_pattern': '*',
+        'target_type': 'agent',
+        'target_uuid': 'agent-1',
+        **overrides,
+    }
+
+
+async def assert_no_upload(counters, manager):
+    assert counters.records == []
+    assert counters.task is None
+    await counters.flush()
+    await counters.shutdown()
+    assert manager.sent == []
+    assert counters.health() == {
+        'records_buffered': 0,
+        'records_sent': 0,
+        'records_dropped': 0,
+        'executions_in_flight': 0,
+    }
+
+
+@pytest.mark.parametrize('edition', ['community', 'cloud'])
+@pytest.mark.parametrize('mode', ['all', 'sampled', 'failures', 'off'])
+@pytest.mark.parametrize(
+    'routes',
+    [
+        [],
+        [business_route(enabled=False)],
+        [business_route(event_pattern='group.member_joined')],
+        [business_route(filters=[{'field': 'chat_id', 'operator': 'eq', 'value': 'another-chat'}])],
+        [business_route(target_type='discard', priority=10), business_route(priority=1)],
+    ],
+)
+async def test_unrouted_events_never_upload_in_any_mode(monkeypatch, edition, mode, routes):
+    from langbot.pkg.utils import constants
+
+    monkeypatch.setattr(constants, 'edition', edition)
+    bot, manager, counters = make_bot(routes, config={'url': 'https://space.example.test', 'execution_trace': mode})
+    bot._dispatch_eba_message_to_pipeline = AsyncMock()
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    await assert_no_upload(counters, manager)
+    # Normal local route diagnostics still exist, including explicit discard.
+    assert bot.logger.info.await_args.kwargs['metadata']['status'] in {'not_matched', 'discarded'}
+
+
+@pytest.mark.parametrize('mode', ['all', 'sampled', 'failures'])
+async def test_matched_target_failure_is_retained_even_after_root_success(mode):
+    bot, manager, counters = make_bot(
+        [business_route()],
+        config={
+            'url': 'https://space.example.test',
+            'execution_trace': mode,
+            'execution_trace_sample': 100000,
+        },
+    )
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    records = await flushed_records(counters, manager)
+    (chain,) = execution_chains(records)
+    assert chain['outcome'] == 'failed'
+    assert chain['error'] == 'processor_not_found'
+    if mode == 'failures':
+        assert execution_nodes(records) == []
+    await counters.shutdown()
+
+
+def install_processor(bot, patterns=None):
+    bot.bot_entity.plugin_processors = [{'processor_uuid': 'processor', 'enabled': True}]
+    bot.ap.agent_service.get_agent = AsyncMock(
+        return_value={
+            'kind': 'event_processor',
+            'component_ref': 'plugin:test/runner/default',
+        }
+    )
+    bot.ap.runner_registry = SimpleNamespace(
+        get=AsyncMock(
+            return_value=SimpleNamespace(
+                usages=['event'],
+                supported_event_patterns=['*'] if patterns is None else patterns,
+            )
+        )
+    )
+    bot._agent_product_to_binding = Mock(return_value=SimpleNamespace())
+    bot._eba_event_to_agent_envelope = Mock(return_value=SimpleNamespace())
+
+    async def run(*args, **kwargs):
+        from langbot.pkg.telemetry.execution import record
+
+        record(bot.ap, TEST_CONTEXT, family='runner', operation='execute', outcome='success')
+        if False:
+            yield None
+
+    bot.ap.agent_run_orchestrator = SimpleNamespace(run=run)
+
+
+@pytest.mark.parametrize('primary', ['miss', 'discard'])
+@pytest.mark.parametrize('primary_first', [True, False])
+async def test_matching_subscription_admits_chain_regardless_of_primary_order(primary, primary_first):
+    from langbot.pkg.telemetry.execution import record
+
+    routes = [] if primary == 'miss' else [business_route(target_type='discard')]
+    bot, manager, counters = make_bot(routes)
+    install_processor(bot)
+    bot.bot_entity.plugin_processors *= 2  # Duplicate subscriptions still execute once.
+    original = bot._dispatch_eba_event_to_processor
+    first_done = asyncio.Event()
+    run_calls = []
+
+    async def ordered_dispatch(event, adapter, event_binding=None, agent=None, execution_id=None):
+        is_primary = event_binding is None
+        if is_primary != primary_first:
+            await first_done.wait()
+        result = await original(event, adapter, event_binding, agent, execution_id)
+        if not is_primary:
+            run_calls.append(result)
+        if is_primary == primary_first:
+            first_done.set()
+        return result
+
+    async def discard_side_effect(*args, **kwargs):
+        # A discarded message can still trigger a webhook/adapter side effect.
+        # A concurrent admitted subscriber must not expose this branch's node.
+        record(bot.ap, TEST_CONTEXT, family='platform_api', operation='discard_side_effect', outcome='success')
+        await counters.flush()
+
+    bot._dispatch_eba_event_to_processor = ordered_dispatch
+    bot._dispatch_eba_message_to_pipeline = discard_side_effect
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    records = await flushed_records(counters, manager)
+    (chain,) = execution_chains(records)
+    nodes = execution_nodes(records)
+    assert len(run_calls) == 1
+    assert chain['stage_count'] == 3
+    assert {node['family'] for node in nodes} == {'runner', 'event_route', 'platform_event'}
+    assert all(node['outcome'] == 'success' for node in nodes)
+    (root,) = [node for node in nodes if node['root']]
+    assert root['family'] == 'platform_event'
+    assert all(node['parent_node_id'] == root['node_id'] for node in nodes if not node['root'])
+    assert all(node['event_id'] == chain['event_id'] for node in nodes)
+    await counters.shutdown()
+
+
+@pytest.mark.parametrize('kind', ['disabled', 'pattern_miss', 'no_patterns', 'missing', 'declaration_error'])
+async def test_unmatched_subscription_cannot_admit_event(kind):
+    bot, manager, counters = make_bot([])
+    install_processor(bot)
+    if kind == 'disabled':
+        bot.bot_entity.plugin_processors[0]['enabled'] = False
+    elif kind in {'pattern_miss', 'no_patterns'}:
+        install_processor(bot, [] if kind == 'no_patterns' else ['group.member_joined'])
+    elif kind == 'missing':
+        bot.ap.agent_service.get_agent.return_value = None
+    else:
+        bot.ap.runner_registry.get.side_effect = RuntimeError('unavailable')
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    await assert_no_upload(counters, manager)
+
+
+@pytest.mark.parametrize('failure', [RuntimeError('runner failed'), TimeoutError('timeout'), asyncio.CancelledError()])
+async def test_matched_subscription_keeps_failed_timeout_and_cancelled_execution(failure):
+    bot, manager, counters = make_bot([], config={'url': 'https://space.example.test', 'execution_trace': 'failures'})
+    install_processor(bot)
+
+    async def run(*args, **kwargs):
+        raise failure
+        yield
+
+    bot.ap.agent_run_orchestrator.run = run
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    (chain,) = execution_chains(await flushed_records(counters, manager))
+    assert chain['outcome'] == 'failed'
+    assert chain['error']
+    await counters.shutdown()
+
+
+@pytest.mark.parametrize('processor', ['agent', 'pipeline', 'invalid', 'expired', 'duplicate'])
+async def test_interaction_admission_requires_valid_callback(processor):
+    from langbot_plugin.api.entities.builtin.platform import events
+
+    bot, manager, counters = make_bot([])
+    callbacks = SimpleNamespace(
+        consume_callback=AsyncMock(return_value={'processor_type': processor, 'processor_id': 'target'}),
+        acknowledge_submission=AsyncMock(),
+    )
+    if processor in {'expired', 'duplicate'}:
+        callbacks.consume_callback.side_effect = ValueError(processor)
+    bot.ap.agent_run_orchestrator = SimpleNamespace(interaction_manager=callbacks)
+    bot._resume_agent_interaction = AsyncMock(side_effect=ValueError('target unavailable'))
+    event = events.PlatformSpecificEvent(platform='test', action='interaction.submitted', data={})
+    await bot._handle_platform_event(event, FakeAdapter())
+    if processor in {'agent', 'pipeline'}:
+        (chain,) = execution_chains(await flushed_records(counters, manager))
+        assert chain['outcome'] == 'failed'
+        callbacks.acknowledge_submission.assert_awaited_once()
+        await counters.shutdown()
+    else:
+        callbacks.acknowledge_submission.assert_not_awaited()
+        await assert_no_upload(counters, manager)
+
+
+async def test_rejected_event_does_not_contaminate_next_matched_event_or_monitoring():
+    from langbot.pkg.telemetry import trace
+
+    bot, manager, counters = make_bot([])
+    bot._persist_monitoring_ingress = AsyncMock()
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    assert bot._persist_monitoring_ingress.await_count == 2
+    assert counters.records == []
+    assert trace.current() is None
+    assert trace.node_stack() == ()
+    assert counters.registry == {}
+    bot.bot_entity.event_bindings = [business_route()]
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    assert len(execution_chains(await flushed_records(counters, manager))) == 1
+    await counters.shutdown()
+
+
+async def test_pending_gate_blocks_nested_and_rpc_records_and_failure_fallback():
+    import contextvars
+    from langbot.pkg.telemetry import execution, trace
+
+    bot, manager, counters = make_bot([])
+    with pytest.raises(RuntimeError):
+        with execution.ingress(bot.ap, context=TEST_CONTEXT, execution_id='pending', require_route=True, debug=True):
+            with execution.ingress(bot.ap, context=TEST_CONTEXT, debug=True):
+                execution.bind_trace(bot.ap, 'run-alias')
+                execution.record(bot.ap, TEST_CONTEXT, family='tool', operation='nested', outcome='failed')
+
+                async def rpc():
+                    execution.record(
+                        bot.ap,
+                        TEST_CONTEXT,
+                        execution_id='run-alias',
+                        family='platform_api',
+                        operation='rpc',
+                        outcome='failed',
+                    )
+                    await counters.flush()
+
+                await contextvars.Context().run(asyncio.create_task, rpc())
+            assert counters.records == []
+            raise RuntimeError('failed before a route matched')
+    assert trace.current() is None
+    await assert_no_upload(counters, manager)
+
+
+@pytest.mark.parametrize('origin', ['api', 'webui'])
+async def test_independent_debug_entry_still_reports(origin):
+    from langbot.pkg.telemetry import execution
+
+    bot, manager, counters = make_bot([], config={'url': 'https://space.example.test', 'execution_trace': 'failures'})
+    with execution.ingress(
+        bot.ap,
+        context=TEST_CONTEXT,
+        execution_id='debug',
+        debug=True,
+        origin=origin,
+        synthetic_event='message.received',
+    ):
+        execution.record(bot.ap, TEST_CONTEXT, family='runner', operation='execute', outcome='success')
+    (chain,) = execution_chains(await flushed_records(counters, manager))
+    assert chain['debug'] is True
+    assert chain['origin'] == origin
+    await counters.shutdown()
+
+
+@pytest.mark.parametrize('edition', ['community', 'cloud'])
+async def test_admission_does_not_override_telemetry_opt_out(monkeypatch, edition):
+    from langbot.pkg.utils import constants
+
+    monkeypatch.setattr(constants, 'edition', edition)
+    bot, manager, counters = make_bot(
+        [business_route()],
+        config={
+            'url': 'https://space.example.test',
+            'disable_telemetry': True,
+        },
+    )
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    await assert_no_upload(counters, manager)
+
+
+async def test_unrelated_broken_subscription_does_not_poison_matched_delivery():
+    bot, manager, counters = make_bot([business_route(target_type='pipeline')])
+    bot._dispatch_eba_message_to_pipeline = AsyncMock()
+    bot.bot_entity.plugin_processors = [{'processor_uuid': 'missing'}]
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    records = await flushed_records(counters, manager)
+    (chain,) = execution_chains(records)
+    assert chain['outcome'] == 'success'
+    assert chain['error'] == ''
+    assert all(node['outcome'] == 'success' for node in execution_nodes(records))
+    bot.logger.error.assert_awaited_once()
+    await counters.shutdown()
+
+
+async def test_pending_event_cancelled_during_dispatch_never_uploads():
+    bot, manager, counters = make_bot([])
+    started = asyncio.Event()
+
+    async def waiting_dispatch(*args, **kwargs):
+        started.set()
+        await asyncio.Event().wait()
+
+    bot._dispatch_eba_event_to_processor = waiting_dispatch
+    task = asyncio.create_task(bot._handle_platform_event(message_received_event(), FakeAdapter()))
+    try:
+        await asyncio.wait_for(started.wait(), timeout=5)
+        await counters.flush()
+    finally:
+        task.cancel()
+        with pytest.raises(asyncio.CancelledError):
+            await task
+    await assert_no_upload(counters, manager)
+
+
+async def test_retry_of_matched_batch_cannot_include_intervening_unrouted_event():
+    bot, manager, counters = make_bot([business_route()])
+    await bot._handle_platform_event(message_received_event(), FakeAdapter())
+    expected = list(counters.records)
+    original_send = manager.send
+    manager.send = AsyncMock(return_value=False)
+    assert await counters.flush() is False
+    assert counters.records == expected
+    bot.bot_entity.event_bindings = []
+    ignored = message_received_event()
+    ignored.message_id = 'ignored-message'
+    await bot._handle_platform_event(ignored, FakeAdapter())
+    assert counters.records == expected
+    manager.send = original_send
+    await counters.shutdown()
+    sent = [record for batch in manager.sent for record in batch['records']]
+    assert sent == expected
+    assert counters.dropped == 0
+
+
+async def test_valid_interaction_success_keeps_acknowledgement_and_execution():
+    from langbot.pkg.telemetry.execution import record
+    from langbot_plugin.api.entities.builtin.platform import events
+
+    bot, manager, counters = make_bot([])
+
+    async def acknowledge(*args):
+        record(bot.ap, TEST_CONTEXT, family='platform_api', operation='ack', outcome='success')
+
+    async def resume(*args):
+        record(bot.ap, TEST_CONTEXT, family='runner', operation='execute', outcome='success')
+
+    bot.ap.agent_run_orchestrator = SimpleNamespace(
+        interaction_manager=SimpleNamespace(
+            consume_callback=AsyncMock(return_value={'processor_type': 'agent'}),
+            acknowledge_submission=acknowledge,
+        )
+    )
+    bot._resume_agent_interaction = resume
+    event = events.PlatformSpecificEvent(platform='test', action='interaction.submitted', data={})
+    await bot._handle_platform_event(event, FakeAdapter())
+    records = await flushed_records(counters, manager)
+    (chain,) = execution_chains(records)
+    assert chain['outcome'] == 'success'
+    assert {node['family'] for node in execution_nodes(records)} == {'platform_event', 'platform_api', 'runner'}
+    await counters.shutdown()
diff --git a/tests/unit_tests/platform/test_botmgr_tenancy.py b/tests/unit_tests/platform/test_botmgr_tenancy.py
index ac100bd46..9687ff8bf 100644
--- a/tests/unit_tests/platform/test_botmgr_tenancy.py
+++ b/tests/unit_tests/platform/test_botmgr_tenancy.py
@@ -38,6 +38,7 @@ def _runtime(application, workspace_uuid: str, bot_uuid: str) -> RuntimeBot:
         uuid=bot_uuid,
         workspace_uuid=workspace_uuid,
         name='Same Name',
+        adapter='aiocqhttp',
         enable=True,
         event_bindings=[],
     )
@@ -243,7 +244,8 @@ async def test_reload_stops_and_drops_existing_platform_runtimes():
 @pytest.mark.asyncio
 @pytest.mark.parametrize(
     ('saved_adapter', 'registered_adapter'),
-    [('probe', 'probe')] + [(f'{name}-eba', f'{name}-omni') for name in sorted(OMNI_ADAPTER_NAMES)],
+    [('probe', 'probe'), ('aiocqhttp-eba', 'onebot-v11'), ('aiocqhttp-omni', 'onebot-v11')]
+    + [(f'{name}-eba', f'{name}-omni') for name in sorted(OMNI_ADAPTER_NAMES)],
 )
 async def test_cloud_startup_reuses_validated_platform_binding(saved_adapter, registered_adapter):
     class TenantUow:
@@ -277,7 +279,7 @@ async def test_cloud_startup_reuses_validated_platform_binding(saved_adapter, re
         description='',
         adapter=saved_adapter,
         adapter_config={},
-        enable=False,
+        enable=True,
         event_bindings=[],
     )
     workspace_service = SimpleNamespace(
@@ -326,6 +328,7 @@ def test_runtime_bot_rejects_workspace_mismatch():
         uuid=BOT_A,
         workspace_uuid=WORKSPACE_A,
         name='Bot',
+        adapter='aiocqhttp',
         enable=True,
         event_bindings=[],
     )
@@ -385,6 +388,7 @@ async def test_platform_callback_carries_scope_without_holding_database_session(
         uuid=BOT_A,
         workspace_uuid=WORKSPACE_A,
         name='Bot',
+        adapter='aiocqhttp',
         enable=True,
         event_bindings=[],
     )
diff --git a/tests/unit_tests/platform/test_dingtalk_api.py b/tests/unit_tests/platform/test_dingtalk_api.py
index 450128f3c..91803baac 100644
--- a/tests/unit_tests/platform/test_dingtalk_api.py
+++ b/tests/unit_tests/platform/test_dingtalk_api.py
@@ -46,7 +46,7 @@ def test_dingtalk_card_param_map_stringifies_unregistered_structures():
 
 
 async def test_create_card_embeds_layout_config_as_template_parameter(monkeypatch):
-    response = type('Response', (), {'status_code': 200})()
+    response = type('Response', (), {'status_code': 200, 'json': lambda self: {}})()
     post = AsyncMock(return_value=response)
 
     @asynccontextmanager
@@ -78,6 +78,7 @@ async def test_create_card_embeds_layout_config_as_template_parameter(monkeypatc
         'cardParamMap': {
             'content': 'hello',
             'config': '{"autoLayout": true}',
+            'autoLayout': 'true',
         }
     }
     assert original_params == {'content': 'hello'}
diff --git a/tests/unit_tests/platform/test_dingtalk_eba_adapter.py b/tests/unit_tests/platform/test_dingtalk_eba_adapter.py
index 81ac54c51..74d5a1c04 100644
--- a/tests/unit_tests/platform/test_dingtalk_eba_adapter.py
+++ b/tests/unit_tests/platform/test_dingtalk_eba_adapter.py
@@ -3,7 +3,7 @@ from __future__ import annotations
 import json
 import pathlib
 from types import SimpleNamespace
-from unittest.mock import AsyncMock, patch
+from unittest.mock import AsyncMock, Mock, patch
 
 import pytest
 import yaml
@@ -38,7 +38,7 @@ class DummyLogger(AbstractEventLogger):
 class DummyDingTalkClient(DingTalkClient):
     def __init__(self, *args, **kwargs):
         self._message_handlers = {}
-        self.client = SimpleNamespace(register_callback_handler=AsyncMock())
+        self.client = SimpleNamespace(register_callback_handler=Mock(), register_all_event_handler=Mock())
         self.markdown_card = kwargs.get('markdown_card', True)
         self.access_token = ''
         self.send_message = AsyncMock()
@@ -243,7 +243,7 @@ async def test_dingtalk_message_converter_maps_inbound_components():
     assert isinstance(chain[0], platform_message.Source)
     assert isinstance(chain[1], platform_message.At)
     assert isinstance(chain[2], platform_message.Plain)
-    assert chain[2].text == ' hello'
+    assert chain[2].text == '@LangBot hello'
     assert isinstance(chain[3], platform_message.File)
     assert isinstance(chain[4], platform_message.Quote)
     assert str(chain[4].origin) == 'quoted text'
@@ -287,8 +287,8 @@ async def test_dingtalk_event_converter_maps_card_feedback():
     assert feedback.session_id == 'space-1'
     assert feedback.message_id == 'bot-msg-1'
 
-    assert isinstance(callback, platform_events.PlatformSpecificEvent)
-    assert callback.action == 'card.callback'
+    assert isinstance(callback, platform_events.DingTalkCardActionEvent)
+    assert callback.action == 'open_details'
 
 
 @pytest.mark.asyncio
@@ -522,3 +522,40 @@ async def test_dingtalk_adapter_dispatches_native_input_submission():
     assert calls[0].data['callback_token'] == 'callback-token'
     assert calls[0].data['values'] == {'score': 42.5}
     assert result['message_id'] not in adapter.interaction_callback_contexts
+
+
+def test_dingtalk_subscription_member_and_group_mapping():
+    from dingtalk_stream.frames import EventMessage
+    from langbot.pkg.platform.adapters.dingtalk.subscriptions import subscription_events
+
+    message = EventMessage()
+    message.headers.event_type = 'chat_add_member'
+    message.data = {'OpenConversationId': 'group', 'UserId': ['a', 'b']}
+    converted = list(subscription_events(message))
+    assert [event.member.id for event in converted] == ['a', 'b']
+    assert all(event.type == 'group.member_joined' for event in converted)
+    message.headers.event_type = 'chat_update_owner'
+    message.data = {'OpenConversationId': 'group', 'Owner': 'new-owner'}
+    event = list(subscription_events(message))[0]
+    assert event.changed_fields == ['owner_id']
+    assert event.group.owner_id == 'new-owner'
+    message.headers.event_type = 'chat_add_member'
+    message.data = {'chatId': 'group', 'unionId': ['union-id']}
+    assert list(subscription_events(message))[0].type == 'platform.specific'
+
+
+@pytest.mark.asyncio
+async def test_dingtalk_subscription_retry_does_not_repeat_successful_members():
+    from dingtalk_stream.frames import EventMessage
+    from langbot.pkg.platform.adapters.dingtalk.subscriptions import DingTalkSubscriptionHandler
+
+    message = EventMessage()
+    message.headers.event_id = 'event-1'
+    message.headers.event_type = 'chat_add_member'
+    message.data = {'OpenConversationId': 'group', 'UserId': ['a', 'b']}
+    adapter = SimpleNamespace(_dispatch_eba_event=AsyncMock(side_effect=[None, RuntimeError('retry'), None]))
+    handler = DingTalkSubscriptionHandler(adapter)
+    assert (await handler.process(message))[0] == 500
+    assert (await handler.process(message))[0] == 200
+    assert (await handler.process(message))[0] == 200
+    assert [call.args[0].member.id for call in adapter._dispatch_eba_event.await_args_list] == ['a', 'b', 'b']
diff --git a/tests/unit_tests/platform/test_lark_eba_adapter.py b/tests/unit_tests/platform/test_lark_eba_adapter.py
index 827a2f4f6..9db1d3c28 100644
--- a/tests/unit_tests/platform/test_lark_eba_adapter.py
+++ b/tests/unit_tests/platform/test_lark_eba_adapter.py
@@ -154,6 +154,7 @@ def make_adapter(config: dict | None = None) -> LarkAdapter:
             },
             DummyLogger(),
         )
+    adapter.message_converter = LarkMessageConverter()
     adapter.api_client = DummyAPIClient()
     adapter.bot = DummyWSClient()
     return adapter
@@ -301,6 +302,7 @@ async def test_lark_event_converter_maps_group_and_private_message():
 @pytest.mark.asyncio
 async def test_lark_adapter_dispatches_and_caches_message_event():
     adapter = make_adapter()
+    adapter._get_bot_open_id = AsyncMock(return_value='bot-open-id')
     calls: list[platform_events.Event] = []
 
     async def listener(event, adapter):
@@ -704,7 +706,9 @@ async def test_lark_streaming_card_uses_strictly_increasing_sequences():
     final_request = adapter.api_client.cardkit.v1.card.aupdate.call_args.args[0]
     assert final_request.request_body.sequence == 3
     final_card = json.loads(final_request.request_body.card.data)
-    assert final_card['body']['elements'] == [{'tag': 'markdown', 'content': 'answer'}]
+    assert final_card['body']['elements'][0] == {'tag': 'markdown', 'content': 'answer'}
+    assert 'header' not in final_card
+    assert final_card['body']['elements'][-1]['columns'][1]['elements'][0]['content'] == '[LangBot](https://langbot.app?utm_source=feishu&utm_medium=bot_card&utm_campaign=langbot)'
     assert not final_card['config'].get('streaming_mode', False)
     assert 'response-1' not in adapter.card_id_dict
     assert 'stream-card-1' not in adapter.card_sequence_dict
@@ -877,9 +881,9 @@ async def test_lark_webhook_dispatches_native_form_submission():
 
     response = await adapter.handle_unified_webhook('bot-1', '', request)
 
-    assert response['toast'] == {'type': 'success', 'content': 'Submitted / 已提交'}
+    assert response['toast'] == {'type': 'success', 'content': 'Submitted'}
     assert response['card']['type'] == 'raw'
-    assert response['card']['data']['elements'][0]['text']['content'] == '**Submitted / 已提交**'
+    assert response['card']['data']['elements'][0]['text']['content'] == '**Submitted**'
     assert len(calls) == 1
     assert calls[0].action == 'interaction.submitted'
     assert calls[0].data['callback_token'] == 'callback-token'
@@ -908,4 +912,42 @@ async def test_lark_webhook_cardkit_submission_uses_async_card_update_only():
 
     response = await adapter.handle_unified_webhook('bot-1', '', request)
 
-    assert response == {'toast': {'type': 'success', 'content': 'Submitted / 已提交'}}
+    assert response == {'toast': {'type': 'success', 'content': 'Submitted'}}
+
+
+def test_lark_stream_card_separates_multiple_reasoning_rounds():
+    from langbot.pkg.platform.adapters.lark.stream_card import build_card, streaming_text
+    content = '<think>first thought</think>first answer<think>second thought</think>final answer'
+    preview = streaming_text(content)
+    assert '<think>' not in preview and '</think>' not in preview
+    card = build_card(content)
+    panel, answer, footer = card['body']['elements']
+    assert panel['tag'] == 'collapsible_panel'
+    assert panel['expanded'] is False
+    assert panel['elements'][0]['content'] == 'first thought\n\nsecond thought'
+    assert answer['content'] == 'first answer\n\nfinal answer'
+    assert 'AI' in footer['columns'][0]['elements'][0]['content']
+    assert footer['columns'][1]['elements'][0]['icon']['tag'] == 'custom_icon'
+
+
+def test_lark_live_preview_is_bounded_but_final_reasoning_is_complete():
+    from langbot.pkg.platform.adapters.lark.stream_card import build_card, streaming_text
+    thought = 'reasoning ' * 200
+    content = '<think>' + thought + '</thi'
+    assert len(streaming_text(content)) < 450
+    assert '</thi' not in streaming_text(content)
+    card = build_card(content)
+    assert card['body']['elements'][0]['elements'][0]['content'] == thought.strip()
+    assert build_card('plain answer')['body']['elements'][0]['content'] == 'plain answer'
+
+
+@pytest.mark.asyncio
+async def test_lark_stream_card_creation_is_async_and_branded():
+    adapter = make_adapter()
+    await adapter.create_card_id('response-1')
+    request = adapter.api_client.cardkit.v1.card.acreate.await_args.args[0]
+    card = json.loads(request.request_body.data)
+    assert 'header' not in card
+    assert card['body']['elements'][-1]['columns'][1]['elements'][0]['content'] == '[LangBot](https://langbot.app?utm_source=feishu&utm_medium=bot_card&utm_campaign=langbot)'
+    assert card['config']['streaming_mode'] is True
+    assert card['body']['elements'][0]['element_id'] == 'streaming_txt'
diff --git a/tests/unit_tests/platform/test_native_stream_completion.py b/tests/unit_tests/platform/test_native_stream_completion.py
new file mode 100644
index 000000000..d24140b92
--- /dev/null
+++ b/tests/unit_tests/platform/test_native_stream_completion.py
@@ -0,0 +1,85 @@
+"""Exercise terminal runner snapshots against native adapter delivery methods."""
+
+import datetime
+import time
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+import telegram
+
+from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+from langbot_plugin.api.entities.builtin.platform.message import MessageChain, Plain
+from langbot.pkg.pipeline.process.stream_results import coalesce_stream_results
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('platform', ['lark', 'dingtalk', 'telegram', 'discord', 'qqofficial', 'wecombot'])
+async def test_native_stream_receives_one_terminal_snapshot(platform):
+    module = __import__(f'tests.unit_tests.platform.test_{platform}_eba_adapter', fromlist=['make_adapter'])
+    adapter = module.make_adapter()
+    object.__setattr__(adapter, 'reply_message', AsyncMock())
+    source = SimpleNamespace(source_platform_object=SimpleNamespace(incoming_message={}))
+    if platform == 'lark':
+        adapter.card_id_dict['response'] = 'card'
+        adapter.card_sequence_dict['card'] = 0
+        adapter._replace_streaming_card = AsyncMock()
+        source.source_platform_object = None
+    elif platform == 'dingtalk':
+        adapter.card_instance_id_dict['response'] = ('card', 'id')
+    elif platform == 'telegram':
+        source.source_platform_object = telegram.Update(
+            1,
+            message=telegram.Message(
+                1,
+                datetime.datetime.now(datetime.timezone.utc),
+                telegram.Chat(1, 'private'),
+                text='hi',
+            ),
+        )
+        adapter.bot = SimpleNamespace(send_message_draft=AsyncMock(), send_message=AsyncMock())
+        adapter.msg_stream_id['response'] = ('private', 1, False)
+    elif platform == 'discord':
+        adapter._stream_buffer['response'] = {
+            'channel': SimpleNamespace(send=AsyncMock(return_value=SimpleNamespace(content='hello', edit=AsyncMock()))),
+            'sent_message': None,
+            'last_content': '',
+            'chunk_count': 0,
+            'updated_at': time.time(),
+        }
+        channel = adapter._stream_buffer['response']['channel']
+    elif platform == 'qqofficial':
+        source = await module.QQOfficialEventConverter().target2yiri(module.qq_event())
+        await adapter.create_message_card('response', source)
+    else:
+        source = await module.WecomBotEventConverter().target2yiri(module.wecombot_event())
+
+    async def runner():
+        yield MessageChunk(role='assistant', content='hello', all_content='hello', is_final=True)
+        yield Message(role='assistant', content='hello')
+
+    count = 0
+    async for result in coalesce_stream_results(runner()):
+        chunk = MessageChunk(role='assistant', content=result.content, resp_message_id='response', is_final=True)
+        await adapter.reply_message_chunk(source, chunk, MessageChain([Plain(text=result.content)]), is_final=True)
+        count += 1
+    assert count == 1
+    adapter.reply_message.assert_not_awaited()
+    if platform == 'lark':
+        adapter._replace_streaming_card.assert_awaited_once_with('card', 'hello', finished=True)
+        assert not adapter.card_id_dict
+    elif platform == 'dingtalk':
+        adapter.bot.send_card_message.assert_awaited_once_with('card', 'id', 'hello', True)
+        assert not adapter.card_instance_id_dict
+    elif platform == 'telegram':
+        adapter.bot.send_message.assert_awaited_once()
+        assert not adapter.msg_stream_id
+    elif platform == 'discord':
+        channel.send.assert_awaited_once_with('hello')
+        assert not adapter._stream_buffer
+    elif platform == 'qqofficial':
+        assert len([call for call in adapter.bot.sent if call[0] == 'stream']) == 1
+        assert not adapter._stream_ctx
+    else:
+        adapter.bot.push_stream_chunk.assert_awaited_once_with('msg-1', 'hello', is_final=True)
+        adapter.bot.reply_text.assert_not_awaited()
diff --git a/tests/unit_tests/platform/test_omni_adapter_names.py b/tests/unit_tests/platform/test_omni_adapter_names.py
index 5b8aa038c..7651b26bf 100644
--- a/tests/unit_tests/platform/test_omni_adapter_names.py
+++ b/tests/unit_tests/platform/test_omni_adapter_names.py
@@ -13,13 +13,17 @@ from langbot.pkg.platform.botmgr import PlatformManager
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize('name', sorted(OMNI_ADAPTER_NAMES))
-async def test_saved_adapter_ids_are_normalized_across_bot_api(name):
-    saved = {'uuid': 'bot', 'adapter': f'{name}-eba', 'adapter_config': {}}
+@pytest.mark.parametrize(
+    'saved_name,public_name',
+    [(f'{name}-eba', f'{name}-omni') for name in sorted(OMNI_ADAPTER_NAMES)]
+    + [('aiocqhttp-eba', 'onebot-v11'), ('aiocqhttp-omni', 'onebot-v11'), ('onebot-v11', 'onebot-v11')],
+)
+async def test_saved_adapter_ids_are_normalized_across_bot_api(saved_name, public_name):
+    saved = {'uuid': 'bot', 'adapter': saved_name, 'adapter_config': {}}
     component = SimpleNamespace(
-        metadata=SimpleNamespace(name=f'{name}-omni'),
+        metadata=SimpleNamespace(name=public_name),
         spec={'config': [{'type': 'webhook-url'}]},
-        to_plain_dict=lambda: {'metadata': {'name': f'{name}-omni'}},
+        to_plain_dict=lambda: {'metadata': {'name': public_name}},
     )
     result = SimpleNamespace(all=lambda: [saved], first=lambda: saved)
     application = SimpleNamespace(
@@ -30,18 +34,20 @@ async def test_saved_adapter_ids_are_normalized_across_bot_api(name):
         ),
     )
     service = BotService(application)
-    assert (await service.get_bots('workspace'))[0]['adapter'] == f'{name}-omni'
-    assert (await service.get_bot('workspace', 'bot'))['adapter'] == f'{name}-omni'
+    assert (await service.get_bots('workspace'))[0]['adapter'] == public_name
+    assert (await service.get_bot('workspace', 'bot'))['adapter'] == public_name
     assert service._adapter_declares_webhook_url(saved['adapter'])
-    assert (await service._prepare_bot_data('workspace', saved, include_uuid=True))['adapter'] == f'{name}-omni'
-    assert saved['adapter'] == f'{name}-eba'
+    assert (await service._prepare_bot_data('workspace', saved, include_uuid=True))['adapter'] == public_name
+    assert saved['adapter'] == saved_name
     manager = PlatformManager(application)
     manager.adapter_components = [component]
     assert manager.get_available_adapter_manifest_by_name(saved['adapter']) is component
-    assert manager.get_available_adapter_info_by_name(saved['adapter'])['metadata']['name'] == f'{name}-omni'
+    assert manager.get_available_adapter_info_by_name(saved['adapter'])['metadata']['name'] == public_name
 
 
-@pytest.mark.parametrize('name', ['telegram', 'custom-eba', 'my-telegram-eba', 'telegram-omni', 'websocket'])
+@pytest.mark.parametrize(
+    'name', ['aiocqhttp', 'onebot-v11', 'telegram', 'custom-eba', 'my-telegram-eba', 'telegram-omni', 'websocket']
+)
 def test_legacy_and_custom_adapter_ids_are_unchanged(name):
     assert canonical_adapter_name(name) == name
 
@@ -49,14 +55,23 @@ def test_legacy_and_custom_adapter_ids_are_unchanged(name):
 def test_discovered_manifests_expose_only_omni_names():
     root = Path(__file__).resolve().parents[3] / 'src/langbot/pkg/platform/adapters'
     for name in OMNI_ADAPTER_NAMES:
-        metadata = yaml.safe_load((root / name / 'manifest.yaml').read_text())['metadata']
+        metadata = yaml.safe_load((root / name / 'manifest.yaml').read_text(encoding='utf-8'))['metadata']
         assert metadata['name'] == f'{name}-omni'
 
 
 @pytest.mark.asyncio
-@pytest.mark.parametrize('disabled_name', ['telegram-eba', 'telegram-omni'])
-async def test_old_disabled_adapter_config_still_hides_omni(disabled_name):
-    component = SimpleNamespace(metadata=SimpleNamespace(name='telegram-omni'))
+@pytest.mark.parametrize(
+    'disabled_name,public_name',
+    [
+        ('telegram-eba', 'telegram-omni'),
+        ('telegram-omni', 'telegram-omni'),
+        ('aiocqhttp-eba', 'onebot-v11'),
+        ('aiocqhttp-omni', 'onebot-v11'),
+        ('onebot-v11', 'onebot-v11'),
+    ],
+)
+async def test_old_disabled_adapter_config_still_hides_omni(disabled_name, public_name):
+    component = SimpleNamespace(metadata=SimpleNamespace(name=public_name))
     application = SimpleNamespace(
         storage_mgr=SimpleNamespace(storage_provider=SimpleNamespace(delete_dir_recursive=AsyncMock())),
         instance_config=SimpleNamespace(data={'system': {'disabled_adapters': [disabled_name]}}),
diff --git a/tests/unit_tests/platform/test_omni_mainline_parity.py b/tests/unit_tests/platform/test_omni_mainline_parity.py
index 7e1b8c5eb..656cf84fa 100644
--- a/tests/unit_tests/platform/test_omni_mainline_parity.py
+++ b/tests/unit_tests/platform/test_omni_mainline_parity.py
@@ -39,7 +39,13 @@ def chain(text):
 def test_omni_config_contains_mainline_options_and_help(name):
     legacy = yaml.safe_load((PLATFORM / 'sources' / f'{name}.yaml').read_text())['spec']
     omni = yaml.safe_load((PLATFORM / 'adapters' / name / 'manifest.yaml').read_text())['spec']
-    assert {c['name'] for c in legacy['config']} <= {c['name'] for c in omni['config']}
+    retired = {'lark': {'bot_name'}, 'aiocqhttp': {'host', 'port'}}.get(name, set())
+    if name == 'aiocqhttp':
+        # The new adapter uses the Core HTTP port or an outbound WS URL.
+        fields = {c['name'] for c in omni['config']}
+        assert {'connection_mode', 'ws_url'} <= fields
+        assert not {'host', 'port'} & fields
+    assert {c['name'] for c in legacy['config']} - retired <= {c['name'] for c in omni['config']}
     assert omni['help_links'] == legacy['help_links']
 
 
@@ -216,9 +222,9 @@ async def test_wecombot_reply_uploads_media_instead_of_placeholder(kind, stream)
         reply_text=AsyncMock(),
         push_stream_chunk=AsyncMock(return_value=True),
         upload_media=AsyncMock(return_value={'media_id': 'media'}),
-        reply_image=AsyncMock(),
-        reply_voice=AsyncMock(),
-        reply_file=AsyncMock(),
+        reply_image=AsyncMock(return_value={'errcode': 0}),
+        reply_voice=AsyncMock(return_value={'errcode': 0}),
+        reply_file=AsyncMock(return_value={'errcode': 0}),
     )
     adapter = WecomBotAdapter.model_construct(config={}, bot=bot)
     event = WecomBotEvent({'message_id': 'message', 'req_id': 'request'})
diff --git a/tests/unit_tests/platform/test_onebot_protocol.py b/tests/unit_tests/platform/test_onebot_protocol.py
new file mode 100644
index 000000000..6de37879b
--- /dev/null
+++ b/tests/unit_tests/platform/test_onebot_protocol.py
@@ -0,0 +1,294 @@
+"""Wire-level conformance tests against an independent OneBot v11 peer."""
+
+import asyncio
+from contextlib import asynccontextmanager
+
+import pytest
+from quart import Quart
+from websockets.asyncio.server import serve
+from aiocqhttp.exceptions import ActionFailed
+from langbot_plugin.api.entities.builtin.platform import events, message
+
+from langbot.pkg.api.http.controller.groups.webhooks import WebhookRouterGroup
+from tests.unit_tests.platform.onebot_mock import OneBotMock
+from tests.unit_tests.platform.test_aiocqhttp_eba_adapter import make_adapter
+from tests.unit_tests.platform.test_onebot_transport import bind
+
+
+@asynccontextmanager
+async def session(mode):
+    adapter, peer = make_adapter(), OneBotMock()
+    _, ap = bind(adapter)
+    tasks = []
+    try:
+        if mode == 'forward_ws':
+
+            async def endpoint(ws):
+                assert ws.request.headers['Authorization'] == 'Bearer secret'
+                task = asyncio.create_task(peer.run(ws.send, ws.recv))
+                tasks.append(task)
+                await asyncio.gather(task, return_exceptions=True)
+
+            async with serve(endpoint, '127.0.0.1', 0) as server:
+                adapter.config.update(
+                    connection_mode=mode, ws_url=f'ws://127.0.0.1:{server.sockets[0].getsockname()[1]}'
+                )
+                tasks.append(asyncio.create_task(adapter.run_async()))
+                await asyncio.wait_for(peer.ready.wait(), 3)
+                # Establish account metadata and synchronize through a real API.
+                await peer.emit({'post_type': 'meta_event', 'meta_event_type': 'lifecycle', 'sub_type': 'connect'})
+                for _ in range(300):
+                    if adapter.bot.connection:
+                        break
+                    await asyncio.sleep(0.01)
+                yield adapter, peer
+                await adapter.kill()
+        else:
+            app = Quart(__name__)
+            await WebhookRouterGroup(ap, app).initialize()
+            headers = {'Authorization': 'Bearer secret', 'X-Self-ID': '42', 'X-Client-Role': 'Universal'}
+            async with app.test_client().websocket('/bots/bot-a/onebot/v11/ws', headers=headers) as ws:
+                tasks.append(asyncio.create_task(peer.run(ws.send, ws.receive)))
+                await asyncio.wait_for(peer.ready.wait(), 3)
+                for _ in range(300):
+                    if adapter.bot.connection:
+                        break
+                    await asyncio.sleep(0.01)
+                yield adapter, peer
+    finally:
+        await adapter.kill()
+        for task in tasks:
+            task.cancel()
+        results = await asyncio.gather(*tasks, return_exceptions=True)
+        for result in results:
+            if isinstance(result, AssertionError):
+                raise result
+
+
+@pytest.fixture(params=['forward_ws', 'reverse_ws'])
+async def wire(request):
+    async with session(request.param) as connected:
+        async with asyncio.timeout(15):
+            yield connected
+
+
+@pytest.mark.parametrize(
+    'kind,subtype',
+    [
+        ('private', 'friend'),
+        ('private', 'group'),
+        ('private', 'other'),
+        ('group', 'normal'),
+        ('group', 'anonymous'),
+        ('group', 'notice'),
+    ],
+)
+@pytest.mark.parametrize('encoding', ['array', 'cq'])
+async def test_messages_and_quoted_reply(wire, kind, subtype, encoding):
+    adapter, peer = wire
+    received = asyncio.Queue()
+
+    async def callback(event, source):
+        await source.reply_message(event, message.MessageChain([message.Plain(text='reply')]), quote_origin=True)
+        await received.put(event)
+
+    adapter.register_listener(events.MessageReceivedEvent, callback)
+    body = (
+        'hi &amp; &#91;x&#93;[CQ:at,qq=42]'
+        if encoding == 'cq'
+        else [
+            {'type': 'text', 'data': {'text': 'hi & [x]'}},
+            {'type': 'at', 'data': {'qq': '42'}},
+        ]
+    )
+    await peer.emit(peer.message(kind, subtype, body))
+    event = await asyncio.wait_for(received.get(), 3)
+    assert str(event.chat_id) == ('20001' if kind == 'group' else '10001')
+    assert event.message_chain.get_first(message.Plain).text == 'hi & [x]'
+    assert str(event.message_chain.get_first(message.At).target) == '42'
+    request = peer.requests[-1]
+    assert request['action'] == 'send_msg'
+    assert request['params']['message_type'] == kind
+    assert request['params']['message'][0] == {'type': 'reply', 'data': {'id': '9'}}
+
+
+NOTICES = [
+    ({'notice_type': 'group_upload', 'file': {'id': 'f', 'name': 'a.txt', 'size': 1, 'busid': 0}}, 'platform.specific'),
+    ({'notice_type': 'group_admin', 'sub_type': 'set'}, 'platform.specific'),
+    ({'notice_type': 'group_admin', 'sub_type': 'unset'}, 'platform.specific'),
+    ({'notice_type': 'group_increase', 'sub_type': 'approve'}, 'group.member_joined'),
+    ({'notice_type': 'group_increase', 'sub_type': 'invite'}, 'group.member_joined'),
+    ({'notice_type': 'group_increase', 'sub_type': 'invite', 'user_id': 42}, 'bot.invited_to_group'),
+    ({'notice_type': 'group_decrease', 'sub_type': 'leave'}, 'group.member_left'),
+    ({'notice_type': 'group_decrease', 'sub_type': 'kick'}, 'group.member_left'),
+    ({'notice_type': 'group_decrease', 'sub_type': 'kick_me', 'user_id': 42}, 'bot.removed_from_group'),
+    ({'notice_type': 'group_ban', 'sub_type': 'ban', 'duration': 60}, 'group.member_banned'),
+    ({'notice_type': 'group_ban', 'sub_type': 'lift_ban', 'duration': 0}, 'platform.specific'),
+    ({'notice_type': 'group_ban', 'sub_type': 'ban', 'duration': 60, 'user_id': 42}, 'bot.muted'),
+    ({'notice_type': 'group_ban', 'sub_type': 'lift_ban', 'duration': 0, 'user_id': 42}, 'bot.unmuted'),
+    ({'notice_type': 'friend_add'}, 'friend.added'),
+    ({'notice_type': 'group_recall', 'message_id': 9}, 'message.deleted'),
+    ({'notice_type': 'friend_recall', 'message_id': 9}, 'message.deleted'),
+    ({'notice_type': 'notify', 'sub_type': 'poke', 'target_id': 42}, 'platform.specific'),
+    ({'notice_type': 'notify', 'sub_type': 'lucky_king', 'target_id': 42}, 'platform.specific'),
+    ({'notice_type': 'notify', 'sub_type': 'honor', 'honor_type': 'talkative'}, 'platform.specific'),
+]
+
+
+@pytest.mark.parametrize('payload,expected', NOTICES)
+async def test_notices(wire, payload, expected):
+    adapter, peer = wire
+    received = asyncio.Queue()
+
+    async def callback(event, source):
+        await received.put(event)
+
+    adapter.register_listener(events.EBAEvent, callback)
+    base = {'post_type': 'notice', 'user_id': 10001}
+    if payload['notice_type'].startswith('group_') or payload['notice_type'] == 'notify':
+        base.update(group_id=20001, operator_id=10002)
+    await peer.emit(base | payload)
+    event = await asyncio.wait_for(received.get(), 3)
+    assert event.type == expected
+    assert event.timestamp == 1700000000
+    if payload['notice_type'] == 'friend_recall':
+        assert str(event.operator.id) == '10001'
+        assert str(event.chat_id) == '10001'
+    if expected == 'platform.specific':
+        assert event.data['notice_type'] == payload['notice_type']
+
+
+@pytest.mark.parametrize(
+    'kind,subtype,expected',
+    [
+        ('friend', None, 'friend.request_received'),
+        ('group', 'invite', 'bot.invited_to_group'),
+        ('group', 'add', 'platform.specific'),
+    ],
+)
+async def test_requests(wire, kind, subtype, expected):
+    adapter, peer = wire
+    received = asyncio.Queue()
+
+    async def callback(event, source):
+        if kind == 'friend':
+            await source.approve_friend_request(event.request_id, False, 'no')
+        elif subtype == 'invite':
+            await source.approve_group_invite(event.request_id, False)
+        await received.put(event)
+
+    adapter.register_listener(events.EBAEvent, callback)
+    payload = {
+        'post_type': 'request',
+        'request_type': kind,
+        'user_id': 10001,
+        'comment': 'please',
+        'flag': 'opaque-flag',
+    }
+    if subtype:
+        payload.update(sub_type=subtype, group_id=20001)
+    await peer.emit(payload)
+    event = await asyncio.wait_for(received.get(), 3)
+    assert event.type == expected
+    if expected != 'platform.specific':
+        assert peer.requests[-1]['params']['flag'] == 'opaque-flag'
+        assert peer.requests[-1]['params']['approve'] is False
+
+
+async def test_api_surface(wire):
+    adapter, peer = wire
+    chain = message.MessageChain([message.Plain(text='test')])
+    for target in ('person', 'private', 'group'):
+        assert str((await adapter.send_message(target, '10001', chain)).message_id) == '123'
+    await adapter.delete_message('person', 10001, 9)
+    assert str((await adapter.get_message('person', 10001, 9)).message_id) == '9'
+    for target in ('person', 'private', 'group'):
+        assert str((await adapter.forward_message('person', 10001, 9, target, 10001)).message_id) == '123'
+    assert (await adapter.get_group_info('20001')).name == 'Mock group'
+    assert len(await adapter.get_group_list()) == 1
+    assert len(await adapter.get_group_member_list('20001')) == 1
+    assert (await adapter.get_group_member_info('20001', '10001')).display_name == 'Card'
+    assert (await adapter.get_user_info('10001')).nickname == 'Alice'
+    assert len(await adapter.get_friend_list()) == 1
+    await adapter.set_group_name('20001', 'new')
+    await adapter.mute_member('20001', '10001', 60)
+    assert peer.requests[-1]['params']['duration'] == 60
+    await adapter.unmute_member('20001', '10001')
+    assert peer.requests[-1]['params']['duration'] == 0
+    await adapter.kick_member('20001', '10001')
+    await adapter.leave_group('20001')
+    await adapter.approve_friend_request('flag', True, 'remark')
+    await adapter.approve_group_invite('flag')
+    for action, params in [
+        ('get_login_info', {}),
+        ('get_status', {}),
+        ('get_version_info', {}),
+        ('get_group_honor_info', {'group_id': 20001, 'type': 'all'}),
+        ('set_group_card', {'group_id': 20001, 'user_id': 10001, 'card': 'card'}),
+        ('set_group_special_title', {'group_id': 20001, 'user_id': 10001, 'special_title': 't'}),
+        ('set_group_admin', {'group_id': 20001, 'user_id': 10001, 'enable': True}),
+        ('set_group_whole_ban', {'group_id': 20001, 'enable': True}),
+        ('get_forward_msg', {'id': 'f'}),
+        ('get_record', {'file': 'f', 'out_format': 'mp3'}),
+        ('get_image', {'file': 'f'}),
+        ('can_send_image', {}),
+        ('can_send_record', {}),
+    ]:
+        result = await adapter.call_platform_api(action, params)
+        assert isinstance(result, dict)
+        assert peer.requests[-1]['action'] == action
+        assert peer.requests[-1]['params'] == params
+
+
+async def test_metadata_self_messages_and_recovery(wire):
+    adapter, peer = wire
+    received = asyncio.Queue()
+
+    async def callback(event, source):
+        await received.put(event)
+
+    adapter.register_listener(events.EBAEvent, callback)
+    for subtype in ('connect', 'enable', 'disable'):
+        await peer.emit({'post_type': 'meta_event', 'meta_event_type': 'lifecycle', 'sub_type': subtype})
+    await peer.emit(
+        {
+            'post_type': 'meta_event',
+            'meta_event_type': 'heartbeat',
+            'status': {'online': True, 'good': True},
+            'interval': 5000,
+        }
+    )
+    await peer.emit(peer.message('private', 'friend') | {'user_id': 42})
+    # An invalid business event must not tear down the transport or block later events.
+    await peer.emit(peer.message('private', 'friend', [{'type': 'face', 'data': {'id': 'invalid'}}]))
+    await peer.emit(peer.message('private', 'friend'))
+    event = await asyncio.wait_for(received.get(), 3)
+    assert event.type == 'message.received'
+    assert received.empty()
+    assert adapter.bot_account_id == '42'
+    results = await asyncio.gather(*(adapter.bot.get_login_info() for _ in range(32)))
+    assert all(item['user_id'] == 42 for item in results)
+    assert len({item['echo'] for item in peer.requests}) == len(peer.requests)
+    assert not adapter.bot.connection.pending
+
+
+@pytest.mark.parametrize(
+    'status,code,valid',
+    [
+        ('ok', 0, True),
+        ('async', 1, True),
+        ('failed', 100, False),
+        ('failed', 1400, False),
+        ('failed', 1404, False),
+        ('ok', 1, False),
+        ('async', 0, False),
+    ],
+)
+async def test_response_contract(wire, status, code, valid):
+    adapter, peer = wire
+    peer.response = {'status': status, 'retcode': code, 'data': None}
+    if valid:
+        assert await adapter.bot.call_action('delete_msg', message_id=9) is None
+    else:
+        with pytest.raises(ActionFailed):
+            await adapter.bot.call_action('delete_msg', message_id=9)
diff --git a/tests/unit_tests/platform/test_onebot_transport.py b/tests/unit_tests/platform/test_onebot_transport.py
new file mode 100644
index 000000000..114cb450a
--- /dev/null
+++ b/tests/unit_tests/platform/test_onebot_transport.py
@@ -0,0 +1,329 @@
+import asyncio
+import json
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+from quart import Quart
+import quart_cors
+from quart.testing.connections import WebsocketResponseError
+from aiocqhttp.exceptions import ActionFailed, ApiNotAvailable
+from websockets.asyncio.server import serve
+
+from langbot.pkg.platform.adapters.aiocqhttp.transport import OneBotConnection
+from langbot.pkg.api.http.controller.groups.webhooks import WebhookRouterGroup
+from tests.unit_tests.platform.test_aiocqhttp_eba_adapter import make_adapter
+
+
+def bind(adapter):
+    runtime = SimpleNamespace(
+        adapter=adapter, enable=True, workspace_uuid='workspace-a', bot_entity=SimpleNamespace(uuid='bot-a')
+    )
+    ap = SimpleNamespace(platform_mgr=SimpleNamespace(resolve_public_bot=AsyncMock(return_value=runtime)))
+    runtime.ap = ap
+    adapter.bind_runtime(runtime)
+    adapter.config['access-token'] = 'secret'
+    return runtime, ap
+
+
+async def test_echo_out_of_order_and_event_api_does_not_deadlock():
+    incoming, outgoing = asyncio.Queue(), asyncio.Queue()
+    handled = asyncio.Event()
+
+    async def event(data):
+        assert await connection.call_action('get_group_info', group_id=1) == {'name': 'group'}
+        handled.set()
+
+    connection = OneBotConnection(outgoing.put, incoming.get, AsyncMock(), event, AsyncMock())
+    task = asyncio.create_task(connection.run())
+    first = asyncio.create_task(connection.call_action('first'))
+    second = asyncio.create_task(connection.call_action('second'))
+    a, b = json.loads(await outgoing.get()), json.loads(await outgoing.get())
+    for request in (b, a):
+        await incoming.put(
+            json.dumps({'echo': request['echo'], 'status': 'ok', 'retcode': 0, 'data': request['action']})
+        )
+    assert await first == 'first'
+    assert await second == 'second'
+    await incoming.put(json.dumps({'post_type': 'message', 'self_id': 42}))
+    request = json.loads(await outgoing.get())
+    await incoming.put(json.dumps({'echo': request['echo'], 'status': 'ok', 'retcode': 0, 'data': {'name': 'group'}}))
+    await asyncio.wait_for(handled.wait(), 1)
+    task.cancel()
+    await asyncio.gather(task, return_exceptions=True)
+    assert not connection.pending
+
+
+async def test_timeout_disconnect_and_failure():
+    outgoing = asyncio.Queue()
+    connection = OneBotConnection(outgoing.put, AsyncMock(), AsyncMock(), AsyncMock(), AsyncMock(), timeout=0.01)
+    with pytest.raises(TimeoutError):
+        await connection.call_action('timeout')
+    assert not connection.pending
+    await outgoing.get()
+    connection.timeout = 2
+    action = asyncio.create_task(connection.call_action('failure'))
+    request = json.loads(await outgoing.get())
+    connection.pending[request['echo']].set_result({'status': 'failed', 'retcode': 100})
+    with pytest.raises(ActionFailed):
+        await action
+    action = asyncio.create_task(connection.call_action('disconnect'))
+    await outgoing.get()
+    await connection.close()
+    with pytest.raises(ApiNotAvailable):
+        await action
+
+
+@pytest.mark.parametrize('payload', ['[]', 'broken', '{"post_type":"message","self_id":99}'])
+async def test_invalid_frames_and_account_mismatch(payload):
+    incoming = asyncio.Queue()
+    await incoming.put(payload)
+    connection = OneBotConnection(AsyncMock(), incoming.get, AsyncMock(), AsyncMock(), AsyncMock(), self_id='42')
+    with pytest.raises(ValueError):
+        await connection.run()
+    assert connection.closed
+
+
+async def test_main_port_reverse_ws_auth_and_placement():
+    adapter = make_adapter()
+    runtime, ap = bind(adapter)
+    app = Quart(__name__)
+    quart_cors.cors(app, allow_origin='*')
+    await WebhookRouterGroup(ap, app).initialize()
+    client = app.test_client()
+    headers = {'Authorization': 'Bearer secret', 'X-Self-ID': '42', 'X-Client-Role': 'Universal'}
+    for changes, status in [({'Authorization': 'Bearer wrong'}, 401), ({'X-Client-Role': 'Event'}, 400)]:
+        with pytest.raises(WebsocketResponseError) as error:
+            async with client.websocket('/bots/bot-a/onebot/v11/ws', headers=headers | changes):
+                await asyncio.sleep(0.01)
+        assert error.value.response.status_code == status
+    async with client.websocket('/bots/bot-a/onebot/v11/ws', headers=headers | {'X-Workspace-Id': 'attacker'}) as ws:
+        await ws.send(json.dumps({'post_type': 'meta_event', 'meta_event_type': 'heartbeat', 'self_id': 42}))
+        for _ in range(100):
+            if adapter.bot.connection:
+                break
+            await asyncio.sleep(0.001)
+        action = asyncio.create_task(adapter.bot.get_login_info())
+        request = json.loads(await ws.receive())
+        await ws.send(json.dumps({'echo': request['echo'], 'status': 'ok', 'retcode': 0, 'data': {'user_id': 42}}))
+        assert await asyncio.wait_for(action, 1) == {'user_id': 42}
+        ap.platform_mgr.resolve_public_bot.return_value = None
+        with pytest.raises(RuntimeError, match='placement'):
+            await adapter.bot.get_login_info()
+    assert runtime.workspace_uuid == 'workspace-a'
+    await adapter.kill()
+
+
+async def test_forward_ws_round_trip_and_shutdown():
+    connected = asyncio.Event()
+
+    async def endpoint(socket):
+        assert socket.request.headers['Authorization'] == 'Bearer secret'
+        await socket.send(json.dumps({'post_type': 'meta_event', 'meta_event_type': 'heartbeat', 'self_id': 42}))
+        connected.set()
+        async for raw in socket:
+            request = json.loads(raw)
+            await socket.send(
+                json.dumps({'echo': request['echo'], 'status': 'ok', 'retcode': 0, 'data': {'message_id': 7}})
+            )
+
+    async with serve(endpoint, '127.0.0.1', 0) as server:
+        adapter = make_adapter()
+        bind(adapter)
+        adapter.config.update(
+            connection_mode='forward_ws', ws_url=f'ws://127.0.0.1:{server.sockets[0].getsockname()[1]}'
+        )
+        task = asyncio.create_task(adapter.run_async())
+        await asyncio.wait_for(connected.wait(), 2)
+        for _ in range(100):
+            if adapter.bot.connection:
+                break
+            await asyncio.sleep(0.001)
+        assert await adapter.bot.send_private_msg(user_id=1, message='hello') == {'message_id': 7}
+        await adapter.kill()
+        await asyncio.gather(task, return_exceptions=True)
+        assert adapter.bot.connection is None
+
+
+async def test_idle_placement_revocation_closes_connection():
+    adapter = make_adapter()
+    _, ap = bind(adapter)
+    closed = asyncio.Event()
+
+    async def close():
+        closed.set()
+
+    incoming = asyncio.Queue()
+    task = asyncio.create_task(adapter.serve_connection(AsyncMock(), incoming.get, close, '42'))
+    for _ in range(100):
+        if adapter.bot.connection:
+            break
+        await asyncio.sleep(0.001)
+    with pytest.raises(RuntimeError, match='active connection'):
+        await adapter.serve_connection(AsyncMock(), incoming.get, close, '42')
+    ap.platform_mgr.resolve_public_bot.return_value = None
+    await asyncio.wait_for(closed.wait(), 7)
+    with pytest.raises(RuntimeError, match='placement'):
+        await task
+    assert adapter.bot.connection is None
+
+
+async def test_forward_reconnects_after_endpoint_disconnect():
+    count = 0
+    reconnected = asyncio.Event()
+
+    async def endpoint(socket):
+        nonlocal count
+        count += 1
+        if count == 1:
+            await socket.close()
+        else:
+            reconnected.set()
+            await socket.wait_closed()
+
+    async with serve(endpoint, '127.0.0.1', 0) as server:
+        adapter = make_adapter()
+        bind(adapter)
+        adapter.config.update(
+            connection_mode='forward_ws', ws_url=f'ws://127.0.0.1:{server.sockets[0].getsockname()[1]}'
+        )
+        task = asyncio.create_task(adapter.run_async())
+        await asyncio.wait_for(reconnected.wait(), 4)
+        await adapter.kill()
+        await asyncio.gather(task, return_exceptions=True)
+        assert count == 2
+
+
+def test_legacy_manifest_keeps_independent_adapter():
+    from langbot.pkg.discover.engine import ComponentDiscoveryEngine
+    from langbot.pkg.platform.adapters.aiocqhttp.adapter import AiocqhttpAdapter as NewAdapter
+    from langbot.pkg.platform.sources.aiocqhttp import AiocqhttpAdapter as LegacyAdapter
+
+    engine = ComponentDiscoveryEngine(SimpleNamespace())
+    legacy = engine.load_component_manifest('pkg/platform/sources/aiocqhttp.yaml')
+    modern = engine.load_component_manifest('pkg/platform/adapters/aiocqhttp/manifest.yaml')
+    assert legacy.get_python_component_class() is LegacyAdapter
+    assert modern.get_python_component_class() is NewAdapter
+    assert modern.metadata.name == 'onebot-v11'
+    assert LegacyAdapter is not NewAdapter
+    fields = {item['name']: item for item in legacy.spec['config']}
+    assert fields['port']['default'] == 2280
+    assert fields['access-token']['required'] is False
+    assert 'connection_mode' not in fields
+
+
+def test_configuration_rejects_invalid_credentials_and_url():
+    from langbot.pkg.platform.adapters.aiocqhttp.adapter import AiocqhttpAdapter
+    from tests.unit_tests.platform.test_aiocqhttp_eba_adapter import DummyLogger
+
+    with pytest.raises(ValueError, match='token'):
+        AiocqhttpAdapter({}, DummyLogger())
+    for url in ('https://example.com', 'ws://user:password@example.com', ''):
+        with pytest.raises(ValueError, match='ws_url'):
+            AiocqhttpAdapter({'access-token': 'secret', 'connection_mode': 'forward_ws', 'ws_url': url}, DummyLogger())
+
+
+@pytest.mark.parametrize('mode', ['reverse_ws', 'forward_ws', 'legacy'])
+async def test_message_event_to_reply_roundtrip(mode):
+    """Exercise real converters and callbacks against a simulated protocol peer."""
+    from langbot.pkg.platform.sources.aiocqhttp import AiocqhttpAdapter as LegacyAdapter
+    from langbot_plugin.api.entities.builtin.platform import events, message
+    from tests.unit_tests.platform.test_aiocqhttp_eba_adapter import DummyLogger
+
+    adapter = LegacyAdapter({'access-token': 'secret'}, DummyLogger()) if mode == 'legacy' else make_adapter()
+    finished = asyncio.Event()
+    received = []
+
+    async def callback(event, source):
+        received.append(str(event.message_chain))
+        await source.send_message('person', '10001', message.MessageChain([message.Plain(text='roundtrip reply')]))
+        finished.set()
+
+    adapter.register_listener(events.FriendMessage, callback)
+
+    async def protocol_peer(send, receive):
+        await send(
+            json.dumps(
+                {
+                    'time': 1700000000,
+                    'self_id': 42,
+                    'post_type': 'message',
+                    'message_type': 'private',
+                    'sub_type': 'friend',
+                    'message_id': 9,
+                    'user_id': 10001,
+                    'message': [{'type': 'text', 'data': {'text': 'roundtrip incoming'}}],
+                    'raw_message': 'roundtrip incoming',
+                    'font': 0,
+                    'sender': {'user_id': 10001, 'nickname': 'Protocol test'},
+                }
+            )
+        )
+        while not finished.is_set():
+            request = json.loads(await asyncio.wait_for(receive(), 3))
+            action = request['action']
+            if action == 'send_private_msg':
+                assert str(request['params']['user_id']) == '10001'
+                assert 'roundtrip reply' in json.dumps(request['params']['message'])
+                data = {'message_id': 10}
+            elif action == 'get_stranger_info':
+                data = {'user_id': 10001, 'nickname': 'Protocol test'}
+            else:
+                raise AssertionError(f'Unexpected action: {action}')
+            await send(json.dumps({'echo': request['echo'], 'status': 'ok', 'retcode': 0, 'data': data}))
+            if action == 'send_private_msg':
+                await asyncio.wait_for(finished.wait(), 3)
+
+    if mode == 'forward_ws':
+        bind(adapter)
+
+        async def endpoint(ws):
+            assert ws.request.headers['Authorization'] == 'Bearer secret'
+            await protocol_peer(ws.send, ws.recv)
+
+        async with serve(endpoint, '127.0.0.1', 0) as server:
+            adapter.config.update(connection_mode=mode, ws_url=f'ws://127.0.0.1:{server.sockets[0].getsockname()[1]}')
+            task = asyncio.create_task(adapter.run_async())
+            try:
+                await asyncio.wait_for(finished.wait(), 5)
+            finally:
+                await adapter.kill()
+                await asyncio.gather(task, return_exceptions=True)
+    else:
+        if mode == 'legacy':
+            app, path = adapter.bot._server_app, '/ws/'
+        else:
+            _, ap = bind(adapter)
+            app, path = Quart(__name__), '/bots/bot-a/onebot/v11/ws'
+            await WebhookRouterGroup(ap, app).initialize()
+        headers = {'Authorization': 'Bearer secret', 'X-Self-ID': '42', 'X-Client-Role': 'Universal'}
+        async with app.test_client().websocket(path, headers=headers) as ws:
+            await protocol_peer(ws.send, ws.receive)
+        await adapter.kill()
+    assert received == ['roundtrip incoming']
+
+
+@pytest.mark.parametrize(
+    'name,path',
+    [
+        ('aiocqhttp', 'pkg/platform/sources/aiocqhttp.yaml'),
+        ('onebot-v11', 'pkg/platform/adapters/aiocqhttp/manifest.yaml'),
+        ('aiocqhttp-omni', 'pkg/platform/adapters/aiocqhttp/manifest.yaml'),
+        ('aiocqhttp-eba', 'pkg/platform/adapters/aiocqhttp/manifest.yaml'),
+    ],
+)
+async def test_public_url_only_available_for_new_onebot_adapter(name, path):
+    from langbot.pkg.discover.engine import ComponentDiscoveryEngine
+    from langbot.pkg.api.http.service.bot import BotService
+
+    ap = SimpleNamespace(
+        instance_config=SimpleNamespace(data={'api': {'webhook_prefix': 'https://bot.example.com'}}),
+        platform_mgr=SimpleNamespace(get_bot_by_uuid=AsyncMock(return_value=None)),
+    )
+    ap.discover = ComponentDiscoveryEngine(ap)
+    ap.discover.load_component_manifest(path)
+    service = BotService(ap)
+    service.get_bot = AsyncMock(return_value={'uuid': 'bot-a', 'adapter': name, 'adapter_config': {}})
+    result = await service.get_runtime_bot_info('workspace-a', 'bot-a')
+    expected = 'https://bot.example.com/bots/bot-a' if name != 'aiocqhttp' else None
+    assert result['adapter_runtime_values']['webhook_full_url'] == expected
diff --git a/tests/unit_tests/platform/test_openclaw_weixin_typing.py b/tests/unit_tests/platform/test_openclaw_weixin_typing.py
new file mode 100644
index 000000000..61efe4135
--- /dev/null
+++ b/tests/unit_tests/platform/test_openclaw_weixin_typing.py
@@ -0,0 +1,138 @@
+import asyncio
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+
+from langbot.libs.openclaw_weixin_api.typing_indicator import TypingIndicator
+from langbot.pkg.platform.processing_indicator import processing_indicator
+
+
+def indicator():
+    client = SimpleNamespace(
+        get_config=AsyncMock(return_value=SimpleNamespace(ret=0, typing_ticket='ticket')),
+        send_typing=AsyncMock(),
+        stop_typing=AsyncMock(),
+    )
+    manager = TypingIndicator(client, lambda peer: 'context')
+    manager.INTERVAL = 0.005
+    return manager, client
+
+
+async def until(predicate):
+    async with asyncio.timeout(1):
+        while not predicate():
+            await asyncio.sleep(0.001)
+
+
+@pytest.mark.asyncio
+async def test_keepalive_shared_lease_and_cached_ticket():
+    manager, client = indicator()
+    async with manager.processing('peer'):
+        async with manager.processing('peer'):
+            await until(lambda: client.send_typing.await_count >= 2)
+            assert len(manager.leases) == 1
+        client.stop_typing.assert_not_awaited()
+    client.get_config.assert_awaited_once_with('peer', 'context')
+    client.stop_typing.assert_awaited_once_with('peer', 'ticket')
+    assert not manager.leases
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('failure', ['error', 'cancel'])
+async def test_processing_failure_stops_typing(failure):
+    manager, client = indicator()
+
+    async def work():
+        async with manager.processing('peer'):
+            await until(lambda: client.send_typing.await_count)
+            if failure == 'error':
+                raise ValueError('processing failed')
+            await asyncio.Event().wait()
+
+    task = asyncio.create_task(work())
+    await until(lambda: client.send_typing.await_count)
+    if failure == 'cancel':
+        task.cancel()
+    with pytest.raises(ValueError if failure == 'error' else asyncio.CancelledError):
+        await task
+    client.stop_typing.assert_awaited_once()
+    assert not manager.leases
+
+
+@pytest.mark.asyncio
+async def test_ticket_failure_is_backed_off_without_interrupting_work():
+    manager, client = indicator()
+    client.get_config.side_effect = OSError('offline')
+    async with manager.processing('peer'):
+        await asyncio.sleep(0.03)
+    client.get_config.assert_awaited_once()
+    client.send_typing.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_shutdown_and_peer_isolation():
+    manager, client = indicator()
+    async with manager.processing('one'), manager.processing('two'):
+        await until(lambda: client.send_typing.await_count >= 2)
+        await manager.close()
+        assert client.stop_typing.await_count == 2
+        count = client.send_typing.await_count
+        async with manager.processing('three'):
+            await asyncio.sleep(0.01)
+        assert client.send_typing.await_count == count
+    assert not manager.leases
+
+
+@pytest.mark.asyncio
+async def test_stop_refreshes_expired_ticket():
+    manager, client = indicator()
+    async with manager.processing('peer'):
+        await until(lambda: client.send_typing.await_count)
+        manager.cache['peer'] = ('old', 0, 600)
+        client.get_config.return_value.typing_ticket = 'fresh'
+    client.stop_typing.assert_awaited_once_with('peer', 'fresh')
+
+
+@pytest.mark.asyncio
+async def test_shutdown_does_not_cancel_stop_already_in_progress():
+    manager, client = indicator()
+    stopping = asyncio.Event()
+    release = asyncio.Event()
+
+    async def stop(*args):
+        stopping.set()
+        await release.wait()
+
+    client.stop_typing.side_effect = stop
+
+    async def work():
+        async with manager.processing('peer'):
+            await until(lambda: client.send_typing.await_count)
+
+    task = asyncio.create_task(work())
+    await asyncio.wait_for(stopping.wait(), 1)
+    shutdown = asyncio.create_task(manager.close())
+    await asyncio.sleep(0)
+    assert not shutdown.done()
+    release.set()
+    await asyncio.wait_for(asyncio.gather(task, shutdown), 1)
+    client.stop_typing.assert_awaited_once()
+    assert not manager.leases
+
+
+@pytest.mark.asyncio
+async def test_optional_adapter_hook():
+    async with processing_indicator(object(), 'person', 'peer'):
+        pass
+    from langbot.pkg.platform.sources.openclaw_weixin import OpenClawWeixinAdapter
+
+    manager, client = indicator()
+    adapter = OpenClawWeixinAdapter.model_construct(client=client)
+    adapter._typing_indicator = manager
+    async with processing_indicator(adapter, 'group', 'peer'):
+        await asyncio.sleep(0)
+    client.get_config.assert_not_awaited()
+    async with processing_indicator(adapter, 'person', 'peer'):
+        await until(lambda: client.send_typing.await_count)
+    client.stop_typing.assert_awaited_once()
diff --git a/tests/unit_tests/platform/test_plugin_subscriptions.py b/tests/unit_tests/platform/test_plugin_subscriptions.py
index 44819f934..64f56aa57 100644
--- a/tests/unit_tests/platform/test_plugin_subscriptions.py
+++ b/tests/unit_tests/platform/test_plugin_subscriptions.py
@@ -1,6 +1,7 @@
 """Independent bot subscriptions must not change primary route delivery."""
 
 import asyncio
+import contextvars
 from types import SimpleNamespace
 from unittest.mock import AsyncMock
 
@@ -8,12 +9,18 @@ import pytest
 from langbot_plugin.api.entities.builtin.platform.events import MemberJoinedEvent
 
 from langbot.pkg.platform.botmgr import RuntimeBot
+from langbot.pkg.persistence.tenant_uow import PersistenceScope, PersistenceScopeBoundary
 
 
 def make_bot(subscriptions):
     bot = object.__new__(RuntimeBot)
-    bot.bot_entity = SimpleNamespace(uuid='bot', event_bindings=[], plugin_processors=subscriptions)
+    bot.bot_entity = SimpleNamespace(
+        uuid='bot', adapter='aiocqhttp', event_bindings=[], plugin_processors=subscriptions
+    )
     bot.execution_context = 'workspace'
+    bot.workspace_uuid = 'workspace'
+    bot.placement_generation = 1
+    bot.ap = SimpleNamespace(workspace_service=SimpleNamespace(get_execution_binding=AsyncMock()))
     bot._record_adapter_event = AsyncMock()
     bot._record_event_route_trace = AsyncMock()
     bot.logger = SimpleNamespace(error=AsyncMock())
@@ -33,14 +40,14 @@ async def test_slow_failed_and_duplicate_subscriptions_do_not_block_primary_rout
     started = []
     release = asyncio.Event()
 
-    async def subscriber(event, adapter, uuid):
+    async def subscriber(event, adapter, uuid, execution_id=None):
         started.append(uuid)
         if uuid == 'slow':
             await release.wait()
         if uuid == 'failed':
             raise ValueError('plugin error')
 
-    async def primary(event, adapter):
+    async def primary(event, adapter, execution_id=None):
         started.append('primary')
 
     bot._dispatch_plugin_subscription = subscriber
@@ -105,3 +112,47 @@ async def test_missing_processor_is_logged_without_unscoped_lookup():
     await bot._dispatch_plugin_subscription(MemberJoinedEvent(), None, 'missing')
     bot.ap.agent_service.get_agent.assert_awaited_once_with('workspace', 'missing')
     assert bot._record_event_route_trace.await_args.kwargs['status'] == 'failed'
+
+
+@pytest.mark.parametrize('primary_fails', [False, True])
+async def test_cloud_deliveries_own_their_scope_and_preserve_parent(primary_fails):
+    bot = make_bot([{'processor_uuid': 'plugin'}])
+    bot.workspace_uuid = 'workspace'
+    active_scope = contextvars.ContextVar('test_event_scope', default=None)
+    active_transaction = contextvars.ContextVar('test_event_transaction', default=None)
+
+    def tenant_scope(workspace):
+        return PersistenceScopeBoundary(
+            PersistenceScope.workspace(workspace),
+            active_scope=active_scope,
+            active_transaction=active_transaction,
+        )
+
+    bot.ap = SimpleNamespace(
+        workspace_service=SimpleNamespace(get_execution_binding=AsyncMock()),
+        persistence_mgr=SimpleNamespace(mode=SimpleNamespace(value='cloud_runtime'), tenant_scope=tenant_scope),
+    )
+    seen = []
+
+    async def delivery(event, adapter, *args, **kwargs):
+        state = active_scope.get()
+        assert state.owner_task is asyncio.current_task()
+        assert state.scope == PersistenceScope.workspace('workspace')
+        seen.append(state)
+        await asyncio.sleep(0)
+        assert active_scope.get() is state
+        if primary_fails and not args:
+            raise ValueError('primary failed')
+
+    bot._dispatch_eba_event_to_processor = delivery
+    bot._dispatch_plugin_subscription = delivery
+    async with tenant_scope('workspace'):
+        parent = active_scope.get()
+        await bot._handle_platform_event(MemberJoinedEvent(), None)
+        assert active_scope.get() is parent
+        assert parent.owner_task is asyncio.current_task()
+    assert active_scope.get() is None
+    assert len(seen) == 2
+    assert seen[0] is not seen[1]
+    assert all(state is not parent and state.depth == 0 for state in seen)
+    assert bot.logger.error.await_count == int(primary_fails)
diff --git a/tests/unit_tests/platform/test_routing_rules.py b/tests/unit_tests/platform/test_routing_rules.py
index 1a339c936..c35788a39 100644
--- a/tests/unit_tests/platform/test_routing_rules.py
+++ b/tests/unit_tests/platform/test_routing_rules.py
@@ -38,6 +38,7 @@ class TestEventRouteTrace:
         bot = object.__new__(RuntimeBot)
         bot.bot_entity = SimpleNamespace(
             uuid='bot-1',
+            adapter='aiocqhttp',
             workspace_uuid=TEST_CONTEXT.workspace_uuid,
             event_bindings=event_bindings,
         )
@@ -170,7 +171,9 @@ class TestEventRouteTrace:
             agent_service=SimpleNamespace(get_agent=AsyncMock(side_effect=[malformed_agent, valid_agent])),
             agent_run_orchestrator=SimpleNamespace(run=fake_run),
         )
-        event = SimpleNamespace(type='platform.member.joined')
+        from langbot_plugin.api.entities.builtin.platform.events import EBAEvent
+
+        event = EBAEvent.model_construct(type='platform.member.joined')
 
         failed = await bot._dispatch_eba_event_to_processor(event, Mock())
         delivered = await bot._dispatch_eba_event_to_processor(event, Mock())
@@ -701,7 +704,10 @@ async def test_installed_event_processor_never_receives_unbound_events():
     from langbot_plugin.api.entities.builtin.platform.events import MemberJoinedEvent
 
     bot = TestEventRouteTrace._make_bot([])
-    bot.ap = SimpleNamespace(plugin_connector=SimpleNamespace(emit_event=AsyncMock()))
+    bot.ap = SimpleNamespace(
+        workspace_service=active_workspace_service(),
+        plugin_connector=SimpleNamespace(emit_event=AsyncMock()),
+    )
     bot._record_adapter_event = AsyncMock()
     await bot._handle_platform_event(MemberJoinedEvent(), Mock())
     bot.ap.plugin_connector.emit_event.assert_not_called()
@@ -767,6 +773,7 @@ async def test_processor_outputs_require_explicit_platform_actions(kind, output_
     """Draining runner results must not send text, duplicate replies, or stream cards."""
     from langbot_plugin.api.entities.builtin.platform import entities, events, message
     from langbot_plugin.api.entities.builtin.provider import message as provider_message
+    from datetime import datetime
 
     from langbot.pkg.agent.runner.platform_tools import execute_platform_tool, freeze_platform_context
 
@@ -783,6 +790,10 @@ async def test_processor_outputs_require_explicit_platform_actions(kind, output_
     completed = []
 
     async def run(envelope, binding, adapter_context):
+        if kind == 'event_processor':
+            restored = message.MessageChain.model_validate(envelope.data['message_chain'])
+            assert restored.source.id == 'source-1'
+            assert restored.source.time == datetime(2026, 10, 8, 12)
         assert envelope.delivery.supports_streaming is False
         assert binding.delivery_policy.enable_streaming is False
         assert binding.delivery_policy.enable_reply is False
@@ -823,7 +834,9 @@ async def test_processor_outputs_require_explicit_platform_actions(kind, output_
     )
     event = events.MessageReceivedEvent(
         message_id='message-1',
-        message_chain=message.MessageChain([message.Plain(text='hello')]),
+        message_chain=message.MessageChain(
+            [message.Source(id='source-1', time=datetime(2026, 10, 8, 12)), message.Plain(text='hello')]
+        ),
         sender=entities.User(id='user-1', nickname='QA'),
         chat_type=entities.ChatType.PRIVATE,
         chat_id='user-1',
diff --git a/tests/unit_tests/platform/test_slack_eba_adapter.py b/tests/unit_tests/platform/test_slack_eba_adapter.py
index 0b8f7deac..8c6e7decd 100644
--- a/tests/unit_tests/platform/test_slack_eba_adapter.py
+++ b/tests/unit_tests/platform/test_slack_eba_adapter.py
@@ -190,7 +190,7 @@ async def test_slack_event_converter_maps_private_group_and_platform_specific():
     assert isinstance(group_event.message_chain[1], platform_message.At)
 
     assert isinstance(platform_event, platform_events.PlatformSpecificEvent)
-    assert platform_event.action == 'slack.file_share'
+    assert platform_event.action == 'slack.message'
 
 
 @pytest.mark.asyncio
@@ -212,7 +212,7 @@ async def test_slack_adapter_dispatches_eba_and_legacy_and_caches_group_event():
     assert len(eba_calls) == 1
     assert len(legacy_calls) == 1
     received = eba_calls[0]
-    assert await adapter.get_message('group', 'C-1', 'evt-1') == received
+    assert await adapter.get_message('group', 'C-1', '1710003600.000100') == received
     assert (await adapter.get_group_info('C-1')).id == 'C-1'
     assert (await adapter.get_group_member_info('C-1', 'U-1')).user.id == 'U-1'
 
@@ -225,7 +225,7 @@ async def test_slack_send_reply_platform_api_and_unsupported():
     reply_result = await adapter.reply_message(
         source_event, platform_message.MessageChain([platform_message.Plain(text='reply')])
     )
-    assert reply_result.message_id == 'evt-1'
+    assert reply_result.message_id == '1.2'
     assert ('person', 'U-1', 'reply') in adapter.bot.sent
 
     await adapter.send_message(
@@ -235,6 +235,7 @@ async def test_slack_send_reply_platform_api_and_unsupported():
 
     assert await adapter.call_platform_api('get_mode', {}) == {
         'webhook': True,
+        'socket_mode': False,
         'bot_account_id': 'B-1',
     }
     assert await adapter.call_platform_api('auth_test', {}) == {'ok': True, 'user_id': 'B-1'}
diff --git a/tests/unit_tests/platform/test_slack_events.py b/tests/unit_tests/platform/test_slack_events.py
new file mode 100644
index 000000000..733bdf8ab
--- /dev/null
+++ b/tests/unit_tests/platform/test_slack_events.py
@@ -0,0 +1,200 @@
+import asyncio
+import json
+import time
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+from urllib.parse import urlencode
+
+import pytest
+from slack_sdk.signature import SignatureVerifier
+
+from langbot.libs.slack_api.api import SlackClient
+from langbot.libs.slack_api.slackevent import SlackEvent
+from langbot.pkg.platform.adapters.slack.event_converter import SlackEventConverter
+from langbot.pkg.api.http.service.slack_setup import manifest
+from langbot.libs.slack_api.media import send_chain
+from langbot_plugin.api.entities.builtin.platform import message as pm
+
+
+def client():
+    return SlackClient('test', 'secret', SimpleNamespace(error=AsyncMock(), warning=AsyncMock()), True)
+
+
+async def test_thread_text_and_file_use_the_same_channel_and_thread():
+    bot = client()
+    bot.client = SimpleNamespace(
+        chat_postMessage=AsyncMock(return_value={'ok': True, 'ts': '2.1'}),
+        files_upload_v2=AsyncMock(return_value={'ok': True, 'files': [{'id': 'F1'}]}),
+    )
+    await send_chain(
+        bot,
+        'channel',
+        'D1',
+        pm.MessageChain(
+            [
+                pm.Plain(text='reply'),
+                pm.File(name='test.txt', base64='dGVzdA=='),
+            ]
+        ),
+        thread_ts='1.1',
+    )
+    bot.client.chat_postMessage.assert_awaited_once_with(channel='D1', text='reply', thread_ts='1.1')
+    assert bot.client.files_upload_v2.call_args.kwargs['thread_ts'] == '1.1'
+    assert bot.client.files_upload_v2.call_args.kwargs['channel'] == 'D1'
+
+
+async def test_http_backpressure_does_not_mark_event_seen():
+    bot = client()
+    for index in range(100):
+        assert bot._enqueue({'index': index}, str(index))
+    assert not bot._enqueue({'event_id': 'overflow'}, 'overflow')
+    assert 'overflow' not in bot.socket_seen
+    await bot.close_socket()
+
+
+@pytest.mark.parametrize(
+    ('raw', 'expected'),
+    [
+        (
+            {'type': 'message', 'subtype': 'message_changed', 'message': {'ts': '1.1', 'text': 'new', 'user': 'U1'}},
+            'message.edited',
+        ),
+        ({'type': 'message', 'subtype': 'message_deleted', 'deleted_ts': '1.1'}, 'message.deleted'),
+        (
+            {'type': 'reaction_added', 'user': 'U1', 'reaction': 'thumbsup', 'item': {'type': 'message', 'ts': '1.1'}},
+            'message.reaction',
+        ),
+        (
+            {
+                'type': 'reaction_removed',
+                'user': 'U1',
+                'reaction': 'thumbsup',
+                'item': {'type': 'message', 'ts': '1.1'},
+            },
+            'message.reaction',
+        ),
+        ({'type': 'member_joined_channel', 'user': 'U1'}, 'group.member_joined'),
+        ({'type': 'member_left_channel', 'user': 'U1'}, 'group.member_left'),
+        ({'type': 'member_joined_channel', 'user': 'UBOT', 'inviter': 'U1'}, 'bot.invited_to_group'),
+        ({'type': 'member_left_channel', 'user': 'UBOT'}, 'bot.removed_from_group'),
+        ({'type': 'channel_rename', 'channel': {'id': 'C1', 'name': 'new'}}, 'group.info_updated'),
+        ({'type': 'group_archive'}, 'group.info_updated'),
+        ({'type': 'message', 'subtype': 'channel_topic', 'topic': 'new'}, 'group.info_updated'),
+        ({'type': 'app_home_opened', 'user': 'U1', 'tab': 'home'}, 'platform.specific'),
+        ({'type': 'app_uninstalled'}, 'platform.specific'),
+        ({'type': 'tokens_revoked', 'tokens': {'bot': ['UBOT']}}, 'platform.specific'),
+    ],
+)
+async def test_standard_events(raw, expected):
+    event = await SlackEventConverter(bot_user_id='UBOT').target2yiri(
+        SlackEvent({'event_id': 'delivery', 'event': {'channel': 'C1', 'channel_type': 'channel', **raw}})
+    )
+    assert event.type == expected
+    if expected.startswith('message.'):
+        assert event.message_id == '1.1'
+        assert event.chat_id == 'C1'
+    if expected == 'message.reaction':
+        assert event.is_add == (raw['type'] == 'reaction_added')
+    if expected == 'message.deleted':
+        assert event.operator is None
+    if expected == 'message.edited':
+        assert 'new' in str(event.new_content)
+    if raw['type'] == 'channel_rename':
+        assert event.group.name == 'new'
+        assert event.changed_fields == ['name']
+
+
+async def test_channel_messages_are_not_duplicate_mentions_or_new_messages_on_edit():
+    bot = client()
+    bot._handle_message = AsyncMock()
+    await bot.handle_payload({'event': {'type': 'message', 'channel_type': 'channel', 'text': '<@BOT> hello'}})
+    bot._handle_message.assert_not_awaited()
+    await bot.handle_payload({'event': {'type': 'app_mention', 'channel': 'C1', 'text': '<@BOT> hello'}})
+    assert bot._handle_message.await_count == 1
+    await bot.handle_payload({'event': {'type': 'message', 'channel_type': 'im', 'subtype': 'message_changed'}})
+    assert bot._handle_message.call_args.args[1] == 'event'
+
+
+@pytest.mark.parametrize('kind', ['block_actions', 'view_submission', 'view_closed', 'shortcut', 'message_action'])
+async def test_interactions_dispatched_and_sanitized(kind):
+    bot = client()
+    handler = AsyncMock()
+    bot.on_message('event')(handler)
+    payload = {'type': kind, 'token': 'secret', 'response_url': 'secret-url', 'user': {'id': 'U1'}}
+    await bot.handle_payload(payload)
+    event = await SlackEventConverter().target2yiri(handler.call_args.args[0])
+    assert event.action == f'slack.{kind}'
+    assert 'token' not in event.data and 'response_url' not in event.data
+
+
+async def test_http_signature_fast_ack_form_and_retry_dedup():
+    bot = client()
+    gate = asyncio.Event()
+
+    async def process(_):
+        await gate.wait()
+
+    bot.handle_payload = AsyncMock(side_effect=process)
+
+    def request(body, mime):
+        stamp = str(int(time.time()))
+        return SimpleNamespace(
+            get_data=AsyncMock(return_value=body),
+            mimetype=mime,
+            headers={
+                'X-Slack-Request-Timestamp': stamp,
+                'X-Slack-Signature': SignatureVerifier('secret').generate_signature(timestamp=stamp, body=body),
+            },
+        )
+
+    body = urlencode({'payload': json.dumps({'type': 'block_actions'})}).encode()
+    try:
+        async with bot.app.app_context():
+            assert await bot._handle_callback_internal(request(body, 'application/x-www-form-urlencoded')) == ('', 200)
+            invalid = request(body, 'application/x-www-form-urlencoded')
+            invalid.headers['X-Slack-Signature'] = 'invalid'
+            assert (await bot._handle_callback_internal(invalid))[1] == 401
+            event_body = json.dumps({'event_id': 'E1', 'event': {'type': 'app_home_opened'}}).encode()
+            await bot._handle_callback_internal(request(event_body, 'application/json'))
+            await bot._handle_callback_internal(request(event_body, 'application/json'))
+            assert bot.socket_queue.qsize() == 2
+    finally:
+        gate.set()
+        await bot.close_socket()
+
+
+async def test_socket_interactivity_ack_and_delivery():
+    bot = client()
+    socket = SimpleNamespace(send_socket_mode_response=AsyncMock())
+    payload = {'type': 'view_submission'}
+    request = SimpleNamespace(type='interactive', envelope_id='E1', payload=payload)
+    await bot._socket_request(socket, request)
+    await bot._socket_request(socket, request)
+    assert bot.socket_queue.qsize() == 1
+    socket.send_socket_mode_response.assert_awaited()
+    assert await bot.socket_queue.get() == payload
+
+
+@pytest.mark.parametrize('socket', [True, False])
+def test_setup_complete_manifest(socket):
+    result = manifest(
+        'LangBot', 'https://example.com/bots/1', 'https://example.com/callback', events=True, socket_mode=socket
+    )
+    settings = result['settings']
+    assert {
+        'message.channels',
+        'message.groups',
+        'message.mpim',
+        'reaction_added',
+        'reaction_removed',
+        'member_joined_channel',
+        'member_left_channel',
+        'app_home_opened',
+        'app_uninstalled',
+        'tokens_revoked',
+    } <= set(settings['event_subscriptions']['bot_events'])
+    assert 'reactions:read' in result['oauth_config']['scopes']['bot']
+    assert settings['interactivity']['is_enabled']
+    assert ('request_url' in settings['interactivity']) is not socket
+    assert ('request_url' in settings['event_subscriptions']) is not socket
+    assert 'interactivity' not in manifest('LangBot', '', '', events=False, socket_mode=socket)['settings']
diff --git a/tests/unit_tests/platform/test_slack_media.py b/tests/unit_tests/platform/test_slack_media.py
new file mode 100644
index 000000000..3238c3e7c
--- /dev/null
+++ b/tests/unit_tests/platform/test_slack_media.py
@@ -0,0 +1,207 @@
+import base64
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, patch
+
+import pytest
+from slack_sdk.web.async_slack_response import AsyncSlackResponse
+
+from langbot.libs.slack_api.api import SlackClient
+from langbot.libs.slack_api.media import attachment_bytes, download_media, send_chain
+from langbot.libs.slack_api.slackevent import SlackEvent
+from langbot.pkg.platform.adapters.slack.message_converter import SlackMessageConverter
+from langbot_plugin.api.entities.builtin.platform import message as pm
+
+
+def sdk_response(data):
+    return AsyncSlackResponse(
+        client=None,
+        http_verb='POST',
+        api_url='https://slack.com/api/test',
+        req_args={},
+        data=data,
+        headers={},
+        status_code=200,
+    )
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('target_type,target_id', [('person', 'U123'), ('channel', 'C123')])
+async def test_real_sdk_responses_for_text_and_attachment(target_type, target_id):
+    client = SlackClient('test-token', '', SimpleNamespace(error=AsyncMock()), unified_mode=True)
+    text_data = {'ok': True, 'channel': 'C123', 'message': {'ts': '1.2', 'bot_id': 'B123'}}
+    file_data = {'ok': True, 'files': [{'id': 'F123'}]}
+    client.client = SimpleNamespace(
+        chat_postMessage=AsyncMock(return_value=sdk_response(text_data)),
+        conversations_open=AsyncMock(return_value=sdk_response({'ok': True, 'channel': {'id': 'D123'}})),
+        files_upload_v2=AsyncMock(return_value=sdk_response(file_data)),
+    )
+    result = await send_chain(
+        client,
+        target_type,
+        target_id,
+        pm.MessageChain(
+            [
+                pm.Plain(text='picture'),
+                pm.File(name='image.png', base64='dGVzdA=='),
+            ]
+        ),
+    )
+    assert result['responses'] == [text_data, file_data]
+    assert client.client.chat_postMessage.await_count == 1
+    assert client.client.files_upload_v2.await_count == 1
+
+
+def bot():
+    return SimpleNamespace(
+        bot_token='test-token',
+        client=SimpleNamespace(
+            conversations_open=AsyncMock(return_value={'ok': True, 'channel': {'id': 'D123'}}),
+            files_upload_v2=AsyncMock(return_value={'ok': True, 'files': [{'id': 'F123'}]}),
+        ),
+        send_message_to_one=AsyncMock(return_value={'ok': True}),
+        send_message_to_channel=AsyncMock(return_value={'ok': True}),
+    )
+
+
+@pytest.mark.asyncio
+async def test_svg_base64_is_uploaded_to_dm_not_sent_as_filename():
+    client = bot()
+    svg = b'<svg xmlns="http://www.w3.org/2000/svg"/>'
+    await send_chain(
+        client,
+        'person',
+        'U123',
+        pm.MessageChain(
+            [
+                pm.Plain(text='Here is the picture'),
+                pm.File(name='moon.svg', base64=base64.b64encode(svg).decode()),
+            ]
+        ),
+    )
+    client.send_message_to_one.assert_awaited_once_with('Here is the picture', 'U123')
+    client.client.conversations_open.assert_awaited_once_with(users=['U123'])
+    client.client.files_upload_v2.assert_awaited_once_with(
+        channel='D123', file=svg, filename='moon.svg', title='moon.svg'
+    )
+
+
+@pytest.mark.asyncio
+async def test_multiple_attachments_to_channel_preserve_bytes_and_text_order(tmp_path):
+    client = bot()
+    path = tmp_path / 'notes.txt'
+    path.write_bytes(b'hello')
+    await send_chain(
+        client,
+        'channel',
+        'C123',
+        pm.MessageChain(
+            [
+                pm.Image(base64='data:image/png;base64,' + base64.b64encode(b'png bytes').decode()),
+                pm.File(path=str(path)),
+                pm.Plain(text='done'),
+            ]
+        ),
+    )
+    calls = client.client.files_upload_v2.await_args_list
+    assert [c.kwargs['file'] for c in calls] == [b'png bytes', b'hello']
+    assert [c.kwargs['filename'] for c in calls] == ['attachment.png', 'notes.txt']
+    client.client.conversations_open.assert_not_awaited()
+    client.send_message_to_channel.assert_awaited_once_with('done', 'C123')
+
+
+@pytest.mark.asyncio
+async def test_upload_failure_is_not_reported_as_success():
+    client = bot()
+    client.client.files_upload_v2.return_value = {'ok': False, 'error': 'missing_scope'}
+    with pytest.raises(RuntimeError, match='missing_scope'):
+        await send_chain(
+            client,
+            'channel',
+            'C123',
+            pm.MessageChain(
+                [
+                    pm.File(name='test.txt', base64='dGVzdA=='),
+                ]
+            ),
+        )
+
+
+@pytest.mark.asyncio
+async def test_missing_attachment_data_fails_instead_of_faking_delivery():
+    with pytest.raises(ValueError, match='no base64'):
+        await attachment_bytes(pm.File(name='missing.txt'))
+
+
+@pytest.mark.asyncio
+async def test_local_bytes_take_priority_over_private_url():
+    data, name = await attachment_bytes(
+        pm.File(name='file.txt', url='https://files.slack.com/private', base64='dGVzdA==')
+    )
+    assert (data, name) == (b'test', 'file.txt')
+
+
+@pytest.mark.asyncio
+async def test_received_multiple_files_keep_type_and_authenticated_bytes():
+    event = SlackEvent(
+        {
+            'event': {
+                'channel_type': 'im',
+                'text': 'attachments',
+                'files': [
+                    {'id': 'F1', 'name': 'a.png', 'mimetype': 'image/png', 'url_private': 'https://files.slack.com/a'},
+                    {
+                        'id': 'F2',
+                        'name': 'b.pdf',
+                        'mimetype': 'application/pdf',
+                        'url_private': 'https://files.slack.com/b',
+                    },
+                ],
+            }
+        }
+    )
+    with patch(
+        'langbot.pkg.platform.adapters.slack.message_converter.download_media',
+        new=AsyncMock(return_value=(b'data', 'application/octet-stream')),
+    ) as download:
+        chain = await SlackMessageConverter.target2yiri(event, 'test-token')
+    assert isinstance(chain[1], pm.Image)
+    assert chain[1].base64 == 'data:image/png;base64,ZGF0YQ=='
+    assert isinstance(chain[2], pm.File) and chain[2].name == 'b.pdf'
+    assert chain[2].base64 == 'ZGF0YQ=='
+    assert chain[3].text == 'attachments'
+    assert download.await_count == 2
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    'url,authenticated',
+    [
+        ('https://files.slack.com/file', True),
+        ('https://example.com/file', False),
+        ('https://files.slack.com.evil.test/file', False),
+    ],
+)
+async def test_download_credentials_are_scoped_to_slack(url, authenticated):
+    response = SimpleNamespace(status=200, headers={'Content-Type': 'image/png'})
+    context = AsyncMock()
+    context.__aenter__.return_value = response
+    with (
+        patch('langbot.libs.slack_api.media.httpclient.get_session') as session,
+        patch('langbot.libs.slack_api.media.httpclient.read_limited', new=AsyncMock(return_value=b'png')),
+    ):
+        session.return_value.get.return_value = context
+        assert await download_media(url, 'test-token') == (b'png', 'image/png')
+        kwargs = session.return_value.get.call_args.kwargs
+        assert bool(kwargs['headers']) == authenticated
+        if authenticated:
+            assert kwargs['allow_redirects'] is False
+
+
+@pytest.mark.asyncio
+async def test_download_http_failure_is_visible():
+    context = AsyncMock()
+    context.__aenter__.return_value = SimpleNamespace(status=403)
+    with patch('langbot.libs.slack_api.media.httpclient.get_session') as session:
+        session.return_value.get.return_value = context
+        with pytest.raises(ValueError, match='HTTP 403'):
+            await download_media('https://files.slack.com/file', 'test-token')
diff --git a/tests/unit_tests/platform/test_slack_socket.py b/tests/unit_tests/platform/test_slack_socket.py
new file mode 100644
index 000000000..80e13b08b
--- /dev/null
+++ b/tests/unit_tests/platform/test_slack_socket.py
@@ -0,0 +1,63 @@
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+import asyncio
+
+import pytest
+
+from langbot.libs.slack_api.api import SlackClient
+
+
+@pytest.mark.asyncio
+async def test_socket_ack_dedup_and_backpressure():
+    bot = SlackClient('test', '', SimpleNamespace(warning=AsyncMock(), error=AsyncMock()), unified_mode=True)
+    client = SimpleNamespace(send_socket_mode_response=AsyncMock())
+    req = SimpleNamespace(type='events_api', envelope_id='envelope', payload={'event_id': 'event'})
+    await bot._socket_request(client, req)
+    await bot._socket_request(client, req)
+    assert client.send_socket_mode_response.await_count == 2
+    assert bot.socket_queue.qsize() == 1
+    for i in range(99):
+        bot.socket_queue.put_nowait({'id': i})
+    req.payload = {'event_id': 'another'}
+    await bot._socket_request(client, req)
+    assert client.send_socket_mode_response.await_count == 2
+    assert 'another' not in bot.socket_seen
+    await bot.close_socket()
+    assert bot.socket_queue.empty()
+
+
+@pytest.mark.asyncio
+async def test_bot_events_do_not_loop():
+    bot = SlackClient('test', '', SimpleNamespace(), unified_mode=True)
+    bot._handle_message = AsyncMock()
+    await bot.handle_payload({'event': {'bot_id': 'other-bot', 'channel_type': 'im'}})
+    bot._handle_message.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_socket_lifecycle_closes_sdk_and_worker(monkeypatch):
+    connected = asyncio.Event()
+    client = SimpleNamespace(
+        socket_mode_request_listeners=[], connect=AsyncMock(side_effect=connected.set), close=AsyncMock()
+    )
+    monkeypatch.setattr('slack_sdk.socket_mode.aiohttp.SocketModeClient', lambda **kwargs: client)
+    bot = SlackClient('test', '', SimpleNamespace(info=AsyncMock()), unified_mode=True)
+    task = asyncio.create_task(bot.run_socket('app-token'))
+    await asyncio.wait_for(connected.wait(), 1)
+    await bot.close_socket()
+    await asyncio.wait_for(task, 1)
+    client.close.assert_awaited_once()
+    assert bot.socket_worker is None
+
+
+@pytest.mark.asyncio
+async def test_socket_connect_failure_cleans_resources(monkeypatch):
+    client = SimpleNamespace(
+        socket_mode_request_listeners=[], connect=AsyncMock(side_effect=ValueError('invalid_auth')), close=AsyncMock()
+    )
+    monkeypatch.setattr('slack_sdk.socket_mode.aiohttp.SocketModeClient', lambda **kwargs: client)
+    bot = SlackClient('test', '', SimpleNamespace(info=AsyncMock()), unified_mode=True)
+    with pytest.raises(ValueError):
+        await bot.run_socket('app-token')
+    client.close.assert_awaited_once()
+    assert bot.socket_worker is None
diff --git a/tests/unit_tests/platform/test_telegram_eba_adapter.py b/tests/unit_tests/platform/test_telegram_eba_adapter.py
index 8fa5f7617..eca17ca4a 100644
--- a/tests/unit_tests/platform/test_telegram_eba_adapter.py
+++ b/tests/unit_tests/platform/test_telegram_eba_adapter.py
@@ -402,7 +402,10 @@ async def test_telegram_converter_maps_non_message_updates():
             }
         }
     )
-    reaction_event = await TelegramEventConverter.target2yiri(reaction, make_adapter().bot, 'test_bot')
+    adapter = make_adapter()
+    reaction_handler = next(h for h in adapter.application.handlers[0] if isinstance(h, MessageReactionHandler))
+    assert reaction_handler.check_update(reaction) is True
+    reaction_event = await TelegramEventConverter.target2yiri(reaction, adapter.bot, 'test_bot')
     assert isinstance(reaction_event, platform_events.MessageReactionEvent)
     assert reaction_event.reaction == '👍'
     assert reaction_event.is_add is True
@@ -471,6 +474,7 @@ async def test_telegram_converter_maps_bot_status_events():
     )
     muted_event = await TelegramEventConverter.target2yiri(muted, make_adapter().bot, 'test_bot')
     assert isinstance(muted_event, platform_events.BotMutedEvent)
+    assert muted_event.duration is None
 
     unmuted = make_update(
         {
@@ -489,6 +493,30 @@ async def test_telegram_converter_maps_bot_status_events():
     unmuted_event = await TelegramEventConverter.target2yiri(unmuted, make_adapter().bot, 'test_bot')
     assert isinstance(unmuted_event, platform_events.BotUnmutedEvent)
 
+    partial_unmute = make_update(
+        {
+            'my_chat_member': {
+                **base_member,
+                'old_chat_member': restricted_member,
+                'new_chat_member': {**restricted_member, 'can_send_messages': True},
+            }
+        }
+    )
+    partial_event = await TelegramEventConverter.target2yiri(partial_unmute, make_adapter().bot, 'test_bot')
+    assert isinstance(partial_event, platform_events.BotUnmutedEvent)
+
+    removed = make_update(
+        {
+            'my_chat_member': {
+                **base_member,
+                'old_chat_member': restricted_member,
+                'new_chat_member': {'user': restricted_member['user'], 'status': 'kicked', 'until_date': 0},
+            }
+        }
+    )
+    removed_event = await TelegramEventConverter.target2yiri(removed, make_adapter().bot, 'test_bot')
+    assert isinstance(removed_event, platform_events.BotRemovedFromGroupEvent)
+
 
 @pytest.mark.asyncio
 async def test_telegram_reply_message_sends_text_image_and_file_components():
@@ -627,3 +655,152 @@ def test_runtime_bot_maps_telegram_eba_events_to_plugin_events():
         plugin_event = RuntimeBot._eba_event_to_plugin_event(platform_event)
         assert isinstance(plugin_event, plugin_event_type)
         assert plugin_event.model_dump()['event_name'] == plugin_event_type.__name__
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('with_text', [True, False])
+async def test_stream_completion_delivers_attachments_once(with_text):
+    adapter = make_adapter()
+    bot = SimpleNamespace(
+        send_message=AsyncMock(return_value=SimpleNamespace(message_id=900)),
+        edit_message_text=AsyncMock(),
+        send_photo=AsyncMock(return_value=SimpleNamespace(message_id=901)),
+        send_document=AsyncMock(return_value=SimpleNamespace(message_id=902)),
+    )
+    object.__setattr__(adapter, 'bot', bot)
+    adapter.msg_stream_id['stream'] = ('message', 999, True)
+    source = platform_events.MessageReceivedEvent(
+        message_id=88, source_platform_object=make_update({'message': base_message_payload(message_id=88)})
+    )
+    parts = [platform_message.Plain(text='done')] if with_text else []
+    parts.extend(
+        [
+            platform_message.Image(base64='data:image/png;base64,' + base64.b64encode(b'image').decode()),
+            platform_message.File(
+                name='result.txt', base64='data:text/plain;base64,' + base64.b64encode(b'result').decode()
+            ),
+        ]
+    )
+    chunk = SimpleNamespace(resp_message_id='stream', msg_sequence=2, tool_calls=None)
+    chain = platform_message.MessageChain(parts)
+    await adapter.reply_message_chunk(source, chunk, chain, is_final=False)
+    bot.send_photo.assert_not_awaited()
+    bot.send_document.assert_not_awaited()
+    await adapter.reply_message_chunk(source, chunk, chain, quote_origin=True, is_final=True)
+    await adapter.reply_message_chunk(source, chunk, chain, quote_origin=True, is_final=True)
+    bot.send_photo.assert_awaited_once()
+    bot.send_document.assert_awaited_once()
+    assert bot.send_document.await_args.kwargs['document'].filename == 'result.txt'
+    assert bot.send_photo.await_args.kwargs['reply_to_message_id'] == 88
+    assert 'stream' not in adapter.msg_stream_id
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('reply', [False, True])
+async def test_telegram_send_returns_message_result(reply):
+    adapter = make_adapter()
+    bot = SimpleNamespace(send_message=AsyncMock(return_value=SimpleNamespace(message_id=901)))
+    object.__setattr__(adapter, 'bot', bot)
+    chain = platform_message.MessageChain([platform_message.Plain(text='hello')])
+    if reply:
+        source = platform_events.MessageReceivedEvent(
+            message_id=88, source_platform_object=make_update({'message': base_message_payload(message_id=88)})
+        )
+        result = await adapter.reply_message(source, chain, quote_origin=True)
+        assert bot.send_message.await_args.kwargs['reply_to_message_id'] == 88
+    else:
+        result = await adapter.send_message('group', '-100123#7', chain)
+        assert bot.send_message.await_args.kwargs['message_thread_id'] == 7
+    assert isinstance(result, platform_events.MessageResult)
+    assert result.message_id == 901
+    assert result.raw == {'message_ids': [901]}
+
+
+@pytest.mark.parametrize('until, expected', [(0, None), (1060, 60), (900, 0)])
+def test_restriction_duration_uses_event_time(until, expected):
+    from langbot.pkg.platform.adapters.telegram.event_converter import _restriction_duration
+
+    assert (
+        _restriction_duration(
+            datetime.datetime.fromtimestamp(until, datetime.UTC),
+            datetime.datetime.fromtimestamp(1000, datetime.UTC),
+        )
+        == expected
+    )
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('new_reaction', [[], [{'type': 'emoji', 'emoji': '❤'}]])
+async def test_reaction_removal_keeps_removed_emoji(new_reaction):
+    update = make_update(
+        {
+            'message_reaction': {
+                'chat': {'id': -1001, 'type': 'supergroup', 'title': 'Group'},
+                'message_id': 77,
+                'date': 1000,
+                'user': {'id': 3, 'is_bot': False, 'first_name': 'Reactor'},
+                'old_reaction': [{'type': 'emoji', 'emoji': '👍'}, *new_reaction],
+                'new_reaction': new_reaction,
+            }
+        }
+    )
+    event = await TelegramEventConverter.target2yiri(update, make_adapter().bot, 'test_bot')
+    assert event.reaction == '👍'
+    assert event.is_add is False
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('reply', [False, True])
+async def test_rejected_photo_falls_back_without_resending_other_attachments(reply):
+    adapter = make_adapter()
+    bot = SimpleNamespace(
+        send_photo=AsyncMock(side_effect=telegram.error.BadRequest('Image_process_failed')),
+        send_document=AsyncMock(return_value=SimpleNamespace(message_id=902)),
+    )
+    object.__setattr__(adapter, 'bot', bot)
+    chain = platform_message.MessageChain(
+        [
+            platform_message.File(name='result.txt', base64='data:text/plain;base64,cmVzdWx0'),
+            platform_message.Image(base64='data:image/png;base64,aW1hZ2U='),
+        ]
+    )
+    if reply:
+        source = platform_events.MessageReceivedEvent(
+            message_id=88, source_platform_object=make_update({'message': base_message_payload(message_id=88)})
+        )
+        result = await adapter.reply_message(source, chain, quote_origin=True)
+    else:
+        result = await adapter.send_message('private', '123', chain)
+    assert result.message_id == 902
+    assert bot.send_document.await_count == 2
+    assert [call.kwargs['document'].filename for call in bot.send_document.await_args_list] == ['result.txt', 'image']
+    assert bot.send_document.await_args.kwargs['document'].input_file_content == b'image'
+    if reply:
+        assert bot.send_document.await_args.kwargs['reply_to_message_id'] == 88
+
+
+@pytest.mark.asyncio
+async def test_attachment_only_completion_removes_thinking_placeholder():
+    adapter = make_adapter()
+    object.__setattr__(
+        adapter,
+        'bot',
+        SimpleNamespace(
+            send_document=AsyncMock(return_value=SimpleNamespace(message_id=902)),
+            delete_message=AsyncMock(),
+        ),
+    )
+    adapter.msg_stream_id['stream'] = ('message', 999, False)
+    source = platform_events.MessageReceivedEvent(
+        message_id=88, source_platform_object=make_update({'message': base_message_payload(message_id=88)})
+    )
+    await adapter.reply_message_chunk(
+        source,
+        SimpleNamespace(resp_message_id='stream', msg_sequence=1, tool_calls=None),
+        platform_message.MessageChain(
+            [platform_message.File(name='result.txt', base64='data:text/plain;base64,cmVzdWx0')]
+        ),
+        is_final=True,
+    )
+    adapter.bot.delete_message.assert_awaited_once_with(chat_id=123, message_id=999)
+    assert 'stream' not in adapter.msg_stream_id
diff --git a/tests/unit_tests/platform/test_websocket_session_isolation.py b/tests/unit_tests/platform/test_websocket_session_isolation.py
index 18dad8a29..d6c123695 100644
--- a/tests/unit_tests/platform/test_websocket_session_isolation.py
+++ b/tests/unit_tests/platform/test_websocket_session_isolation.py
@@ -596,7 +596,8 @@ def test_widget_sends_stable_session_id_to_all_conversation_endpoints():
     assert 'window.sessionStorage' in widget
     assert 'window.localStorage' not in widget
     assert 'session_id=' in widget
-    assert widget.count('encodeURIComponent(state.sessionId)') >= 3
+    assert widget.count('encodeURIComponent(state.sessionId)') >= 2
+    assert 'endpoint.searchParams.set("session_id", state.sessionId)' in widget
     assert 'loadHistory(true)' in widget
     assert 'messageVersion !== state.messageVersion' in widget
     assert 'scheduleHistoryReload();' in widget
diff --git a/tests/unit_tests/platform/test_wecombot_eba_adapter.py b/tests/unit_tests/platform/test_wecombot_eba_adapter.py
index 1952505da..4e8fccc9b 100644
--- a/tests/unit_tests/platform/test_wecombot_eba_adapter.py
+++ b/tests/unit_tests/platform/test_wecombot_eba_adapter.py
@@ -241,7 +241,7 @@ async def test_wecombot_event_converter_maps_private_and_group_messages_to_eba()
     assert private_event.adapter_name == 'wecombot-omni'
     assert private_event.chat_type == platform_entities.ChatType.PRIVATE
     assert private_event.chat_id == 'user-1'
-    assert str(private_event.message_chain) == 'hello'
+    assert str(private_event.message_chain) == '@EBA Bot hello'
 
     assert isinstance(group_event, platform_events.MessageReceivedEvent)
     assert group_event.chat_type == platform_entities.ChatType.GROUP
@@ -302,7 +302,9 @@ async def test_wecombot_send_reply_feedback_and_platform_api_use_underlying_clie
 
     source_event = await WecomBotEventConverter().target2yiri(wecombot_event())
     await adapter.reply_message(source_event, message)
-    adapter.bot.reply_text.assert_awaited_once_with('req-1', 'hello')
+    adapter.bot.push_stream_chunk.assert_awaited_once_with('msg-1', 'hello', is_final=True)
+    adapter.bot.reply_text.assert_not_awaited()
+    adapter.bot.push_stream_chunk.reset_mock()
 
     await adapter.reply_message_chunk(source_event, None, message, is_final=True)
     adapter.bot.push_stream_chunk.assert_awaited_once_with('msg-1', 'hello', is_final=True)
@@ -328,3 +330,170 @@ async def test_wecombot_webhook_mode_rejects_proactive_send():
         await adapter.send_message(
             'person', 'user-1', platform_message.MessageChain([platform_message.Plain(text='hi')])
         )
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('kind', ['enter_chat', 'template_card_event'])
+async def test_business_events_preserve_identity_payload_and_deduplicate(kind):
+    from langbot.libs.wecom_ai_bot_api.ws_client import WecomBotWsClient
+
+    adapter = make_adapter()
+    listener = AsyncMock()
+    adapter.listeners[platform_events.EBAEvent] = listener
+    client = WecomBotWsClient('bot-id', 'secret', DummyLogger())
+    client.on_message('event')(adapter._handle_native_event)
+    client.on_message('template_card_event')(adapter._handle_interaction_event)
+    body = {
+        'msgid': 'business-event-1',
+        'msgtype': 'event',
+        'chattype': 'single',
+        'from': {'userid': 'visitor-1', 'alias': 'Visitor'},
+        'create_time': 1700000000,
+        'event': {
+            'eventtype': kind,
+            'template_card_event': {
+                'task_id': 'business-card',
+                'event_key': 'approve',
+                'card_type': 'button_interaction',
+                'selected_items': [{'question_key': 'choice', 'option_ids': ['yes']}],
+            },
+        },
+    }
+    frame = {'headers': {'req_id': 'callback-1'}, 'body': body}
+    await client._handle_event_callback(frame)
+    await client._handle_event_callback(frame)
+    listener.assert_awaited_once()
+    event = listener.await_args.args[0]
+    assert event.type == 'wecombot.' + kind
+    assert event.user.id == 'visitor-1'
+    assert event.chat_id == 'visitor-1'
+    assert event.timestamp == 1700000000
+    assert event.data['req_id'] == 'callback-1'
+    assert event.data['event'] == body['event']
+    assert type(event).model_validate(event.model_dump()).type == event.type
+    if kind == 'template_card_event':
+        assert event.event_key == 'approve'
+        assert event.selected_items[0]['option_ids'] == ['yes']
+
+
+@pytest.mark.asyncio
+async def test_managed_card_callback_does_not_start_business_event():
+    adapter = make_adapter()
+    listener = AsyncMock()
+    adapter.listeners[platform_events.EBAEvent] = listener
+    await adapter._handle_interaction_event(
+        WecomBotEvent(
+            {
+                'type': 'single',
+                'msgtype': 'event',
+                'userid': 'user-1',
+                'event': {
+                    'eventtype': 'template_card_event',
+                    'template_card_event': {
+                        'task_id': 'lbi-managed',
+                        'event_key': 'lbi:token:a:0',
+                    },
+                },
+            }
+        )
+    )
+    listener.assert_awaited_once()
+    event = listener.await_args.args[0]
+    assert event.type == 'platform.specific'
+    assert event.action == 'interaction.submitted'
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('webhook', [False, True])
+async def test_enter_chat_reply_uses_welcome_callback_and_checks_deadline(webhook):
+    import time
+
+    adapter = make_adapter(webhook)
+    adapter.bot.reply_welcome = AsyncMock(return_value={'ok': True})
+    context = {
+        'event_type': 'wecombot.enter_chat',
+        'data': {
+            'timestamp': time.time(),
+            'message_id': 'entry-1',
+            'data': {'req_id': 'request-1'},
+        },
+    }
+    message = platform_message.MessageChain([platform_message.Plain(text='Welcome')])
+    await adapter.reply_to_event(context=context, target_type='person', target_id='user-1', message=message)
+    adapter.bot.reply_welcome.assert_awaited_once_with('entry-1' if webhook else 'request-1', 'Welcome')
+    context['data']['timestamp'] -= 10
+    with pytest.raises(ValueError, match='5 seconds'):
+        await adapter.reply_to_event(context=context, target_type='person', target_id='user-1', message=message)
+
+
+@pytest.mark.asyncio
+async def test_file_attachment_is_uploaded_and_delivered_after_stream():
+    adapter = make_adapter()
+    adapter.bot.upload_media = AsyncMock(return_value={'media_id': 'file-1'})
+    adapter.bot.reply_file = AsyncMock(return_value={'errcode': 0})
+    source = await adapter.event_converter.target2yiri(wecombot_event())
+    message = platform_message.MessageChain(
+        [
+            platform_message.Plain(text='Here is your file'),
+            platform_message.File(name='report.txt', base64='aGVsbG8='),
+        ]
+    )
+    await adapter.reply_message_chunk(source, None, message, is_final=True)
+    adapter.bot.upload_media.assert_awaited_once_with(b'hello', 'report.txt', media_type='file')
+    adapter.bot.reply_file.assert_awaited_once_with('req-1', 'file-1')
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('failure', ['upload', 'reply', 'timeout'])
+async def test_file_delivery_errors_are_not_silent(failure):
+    adapter = make_adapter()
+    adapter.bot.upload_media = AsyncMock(return_value=None if failure == 'upload' else {'media_id': 'file-1'})
+    adapter.bot.reply_file = AsyncMock(return_value=None if failure == 'timeout' else {'errcode': 40001})
+    source = await adapter.event_converter.target2yiri(wecombot_event())
+    with pytest.raises(RuntimeError, match='WeCom file'):
+        await adapter.reply_message(
+            source,
+            platform_message.MessageChain(
+                [
+                    platform_message.File(name='report.txt', base64='aGVsbG8='),
+                ]
+            ),
+        )
+
+
+@pytest.mark.asyncio
+async def test_multi_round_file_completion_finalizes_once_without_text_fallback():
+    from langbot.pkg.pipeline.process.stream_results import coalesce_stream_results
+    from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+
+    adapter = make_adapter()
+    adapter.bot.upload_media = AsyncMock(return_value={'media_id': 'file-1'})
+    adapter.bot.reply_file = AsyncMock(return_value={'errcode': 0})
+    source = await adapter.event_converter.target2yiri(wecombot_event())
+    prefix = '<think>plan</think>Creating your file.'
+    final = '<think>done</think>Your file is ready.'
+    completed = Message(role='assistant', content=final)
+    completed.attachments = platform_message.MessageChain(
+        [
+            platform_message.File(name='report.txt', base64='aGVsbG8='),
+        ]
+    )
+
+    async def results():
+        yield MessageChunk(role='assistant', content=prefix)
+        yield MessageChunk(role='assistant', content=prefix + final, tool_calls=[], is_final=True)
+        yield completed
+
+    async for result in coalesce_stream_results(results()):
+        await adapter.reply_message_chunk(
+            source,
+            None,
+            result.get_content_platform_message_chain(),
+            is_final=isinstance(result, Message) or result.is_final,
+        )
+    assert sum(call.kwargs['is_final'] for call in adapter.bot.push_stream_chunk.await_args_list) == 1
+    adapter.bot.reply_text.assert_not_awaited()
+    adapter.bot.reply_file.assert_awaited_once_with('req-1', 'file-1')
+    final_text = adapter.bot.push_stream_chunk.await_args_list[-1].args[1]
+    assert final_text.count('<think>') == 1
+    assert final_text.count('Your file is ready.') == 1
diff --git a/tests/unit_tests/platform/test_wecombot_stream_empty_final.py b/tests/unit_tests/platform/test_wecombot_stream_empty_final.py
index e21d163b6..8f18fa866 100644
--- a/tests/unit_tests/platform/test_wecombot_stream_empty_final.py
+++ b/tests/unit_tests/platform/test_wecombot_stream_empty_final.py
@@ -117,3 +117,61 @@ async def test_real_final_chunk_is_sent_and_closes_stream():
     assert frames[-1]['stream']['content'] == 'the answer'
     assert frames[-1]['stream']['finish'] is True
     assert not client.stream_still_open()
+
+
+@pytest.mark.asyncio
+async def test_duplicate_final_is_acknowledged_without_sending_again():
+    client = RecordingClient()
+    client.seed_stream()
+    assert await client.push_stream_chunk('msg-1', 'answer', is_final=True)
+    assert await client.push_stream_chunk('msg-1', 'answer', is_final=True)
+    assert len(client.stream_frames()) == 1
+    assert not await client.push_stream_chunk('different-message', 'answer', is_final=True)
+    assert not await client.push_stream_chunk('msg-1', 'different answer', is_final=True)
+
+
+@pytest.mark.asyncio
+async def test_runner_cumulative_snapshots_reach_wire_once_without_concatenation():
+    from langbot.pkg.pipeline.process.stream_results import coalesce_stream_results
+    from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+    from tests.unit_tests.platform.test_wecombot_eba_adapter import (
+        make_adapter,
+        wecombot_event,
+        WecomBotEventConverter,
+    )
+
+    client = RecordingClient()
+    client.seed_stream()
+    adapter = make_adapter()
+    adapter.bot = client
+    event = await WecomBotEventConverter().target2yiri(wecombot_event())
+    snapshots = ['<think>test', '<think>test</think>\nOK', '<think>test</think>\nOK!']
+
+    async def runner():
+        # LocalAgent's StreamingModelCaller sets content to the accumulated
+        # text, leaves all_content unset, then emits message.completed.
+        for index, text in enumerate(snapshots):
+            yield MessageChunk(role='assistant', content=text, msg_sequence=index + 1, is_final=index == 2)
+        yield Message(role='assistant', content=snapshots[-1])
+
+    async for result in coalesce_stream_results(runner()):
+        final = isinstance(result, Message) or result.is_final
+        if isinstance(result, MessageChunk):
+            result = result.model_copy(update={'content': result.all_content})
+        await adapter.reply_message_chunk(event, result, result.get_content_platform_message_chain(), is_final=final)
+
+    frames = client.stream_frames()
+    from langbot.pkg.platform.adapters.wecombot.stream_text import format_stream_text
+
+    assert [frame['stream']['content'] for frame in frames] == [format_stream_text(text) for text in snapshots]
+    assert [frame['stream']['finish'] for frame in frames] == [False, False, True]
+    assert len(client.replies) == 3  # No extra plain-text reply after completion.
+
+
+@pytest.mark.asyncio
+async def test_revised_snapshot_replaces_previous_text():
+    client = RecordingClient()
+    client.seed_stream()
+    await client.push_stream_chunk('msg-1', '<think>test</think>\nanswer')
+    await client.push_stream_chunk('msg-1', 'answer', is_final=True)
+    assert client.stream_frames()[-1]['stream']['content'] == 'answer'
diff --git a/tests/unit_tests/platform/test_wecombot_stream_text.py b/tests/unit_tests/platform/test_wecombot_stream_text.py
new file mode 100644
index 000000000..fe19d7bab
--- /dev/null
+++ b/tests/unit_tests/platform/test_wecombot_stream_text.py
@@ -0,0 +1,30 @@
+import pytest
+
+from langbot.pkg.platform.adapters.wecombot.stream_text import format_stream_text
+
+
+def test_multiple_rounds_have_one_thinking_section_and_separated_answers():
+    raw = '<think>one</think>Checking.<think>two</think>More checks.<think>three</think>Done.'
+    result = format_stream_text(raw)
+    assert result == '<think>\none\n\ntwo\n\nthree\n</think>\n\nChecking.\n\nMore checks.\n\nDone.'
+    assert format_stream_text(result) == result
+
+
+@pytest.mark.parametrize('suffix', ['<', '<t', '<thi', '<think', '</', '</thi'])
+def test_partial_tags_are_not_shown_in_answer(suffix):
+    result = format_stream_text('<think>one</think>Checking.' + suffix)
+    assert result.endswith('Checking.')
+
+
+def test_later_incomplete_thinking_remains_inside_single_section():
+    result = format_stream_text('<think>one</think>Checking.<think>two')
+    assert result == '<think>\none\n\ntwo\n</think>\n\nChecking.'
+
+
+def test_plain_markdown_is_unchanged():
+    text = '  hello\n\n```python\na < b\n```\n'
+    assert format_stream_text(text) == text
+
+
+def test_initial_partial_thinking_tag_waits_for_next_snapshot():
+    assert format_stream_text('<thi') == ''
diff --git a/tests/unit_tests/platform/test_wecombot_template_card.py b/tests/unit_tests/platform/test_wecombot_template_card.py
index d8b46c538..d27a6650c 100644
--- a/tests/unit_tests/platform/test_wecombot_template_card.py
+++ b/tests/unit_tests/platform/test_wecombot_template_card.py
@@ -369,11 +369,13 @@ async def test_ws_stream_sends_cumulative_snapshots_to_wecom():
 
     assert await client.push_stream_chunk('msg-1', '你', is_final=False)
     assert await client.push_stream_chunk('msg-1', '你好', is_final=False)
+    assert await client.push_stream_chunk('msg-1', '<think>test</think>你好', is_final=False)
     assert await client.push_stream_chunk('msg-1', '你好', is_final=True)
 
     assert sent == [
         ('req-1', 'stream-1', '你', False),
         ('req-1', 'stream-1', '你好', False),
+        ('req-1', 'stream-1', '<think>test</think>你好', False),
         ('req-1', 'stream-1', '你好', True),
     ]
 
@@ -385,6 +387,7 @@ async def test_webhook_stream_queues_cumulative_snapshots_for_followups():
 
     assert await client.push_stream_chunk('msg-1', '你', is_final=False)
     assert await client.push_stream_chunk('msg-1', '你好', is_final=False)
+    assert await client.push_stream_chunk('msg-1', '<think>test</think>你好', is_final=False)
     assert await client.push_stream_chunk('msg-1', '你好', is_final=True)
 
     assert session.queue.qsize() == 1
@@ -566,3 +569,132 @@ def test_action_stage_only_shows_content_after_fields_without_placeholders():
     assert card['sub_title_text'] == '请选择操作'
     assert '{{#$output.' not in card['sub_title_text']
     assert [button['text'] for button in card['button_list']] == ['yes', 'no']
+
+
+@pytest.mark.asyncio
+async def test_webhook_enter_chat_returns_welcome_without_message_stream():
+    from unittest.mock import AsyncMock
+    from langbot.libs.wecom_ai_bot_api.wecombotevent import WecomBotEvent
+
+    logger = types.SimpleNamespace(info=AsyncMock(), debug=AsyncMock(), error=AsyncMock(), warning=AsyncMock())
+    client = WecomBotClient('', '', '', logger, unified_mode=True)
+    client._encrypt_and_reply = AsyncMock(side_effect=lambda reply, nonce: reply)
+    ordinary = AsyncMock()
+    client.on_message('single')(ordinary)
+
+    async def welcome(event):
+        assert isinstance(event, WecomBotEvent)
+        await client.reply_welcome(event.message_id, 'Welcome')
+
+    client.on_message('event')(welcome)
+    result = await client._handle_enter_chat(
+        {
+            'msgid': 'entry-1',
+            'msgtype': 'event',
+            'chattype': 'single',
+            'from': {'userid': 'visitor'},
+            'event': {'eventtype': 'enter_chat'},
+        },
+        'nonce',
+    )
+    assert result == {'msgtype': 'text', 'text': {'content': 'Welcome'}}
+    ordinary.assert_not_awaited()
+    assert not client._welcome_replies
+    with pytest.raises(ValueError, match='expired'):
+        await client.reply_welcome('entry-1', 'too late')
+    await client.close()
+
+
+@pytest.mark.asyncio
+async def test_webhook_unmanaged_card_dispatches_but_expired_dify_does_not():
+    import asyncio
+    from unittest.mock import AsyncMock
+
+    logger = types.SimpleNamespace(info=AsyncMock(), debug=AsyncMock(), error=AsyncMock(), warning=AsyncMock())
+    client = WecomBotClient('', '', '', logger, unified_mode=True)
+    client._encrypt_and_reply = AsyncMock(return_value={})
+    handler = AsyncMock()
+    client.on_message('template_card_event')(handler)
+    for task_id in ['business-card', 'dify-expired']:
+        await client._handle_template_card_event(
+            {
+                'msgid': task_id,
+                'msgtype': 'event',
+                'chattype': 'single',
+                'from': {'userid': 'visitor'},
+                'event': {
+                    'eventtype': 'template_card_event',
+                    'template_card_event': {
+                        'task_id': task_id,
+                        'event_key': 'approve',
+                    },
+                },
+            },
+            'nonce',
+        )
+    await asyncio.gather(*client._dispatch_tasks)
+    handler.assert_awaited_once()
+    assert handler.await_args.args[0].userid == 'visitor'
+    await client.close()
+
+
+@pytest.mark.asyncio
+async def test_ws_welcome_uses_dedicated_command():
+    from unittest.mock import AsyncMock
+
+    client = WecomBotWsClient('bot-id', 'secret', object())
+    client._send_reply = AsyncMock(return_value={})
+    await client.reply_welcome('req-1', 'Welcome')
+    client._send_reply.assert_awaited_once_with(
+        'req-1', {'msgtype': 'text', 'text': {'content': 'Welcome'}}, cmd='aibot_respond_welcome_msg'
+    )
+
+
+@pytest.mark.asyncio
+async def test_expired_custom_managed_card_never_becomes_business_event():
+    from unittest.mock import AsyncMock
+
+    logger = types.SimpleNamespace(info=AsyncMock(), debug=AsyncMock(), error=AsyncMock(), warning=AsyncMock())
+    client = WecomBotWsClient('bot-id', 'secret', logger)
+    handler = AsyncMock()
+    client.on_message('template_card_event')(handler)
+    client._drop_pending_form_task('custom-managed-task', {})
+    body = {
+        'msgtype': 'event',
+        'event': {
+            'eventtype': 'template_card_event',
+            'template_card_event': {
+                'task_id': 'custom-managed-task',
+                'event_key': 'approve',
+            },
+        },
+    }
+    await client._handle_template_card_event_frame({'headers': {'req_id': 'r1'}}, body)
+    handler.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_webhook_welcome_timeout_cleans_pending_reply(monkeypatch):
+    import asyncio
+    from unittest.mock import AsyncMock
+
+    logger = types.SimpleNamespace(info=AsyncMock(), debug=AsyncMock(), error=AsyncMock(), warning=AsyncMock())
+    client = WecomBotClient('', '', '', logger, unified_mode=True)
+    client._encrypt_and_reply = AsyncMock(side_effect=lambda reply, nonce: reply)
+
+    async def timeout(awaitable, timeout):
+        awaitable.cancel()
+        raise asyncio.TimeoutError
+
+    monkeypatch.setattr(asyncio, 'wait_for', timeout)
+    result = await client._handle_enter_chat(
+        {
+            'msgid': 'no-reply',
+            'msgtype': 'event',
+            'event': {'eventtype': 'enter_chat'},
+        },
+        'nonce',
+    )
+    assert result == {}
+    assert not client._welcome_replies
+    await client.close()
diff --git a/tests/unit_tests/platform/test_wecomcs_eba_adapter.py b/tests/unit_tests/platform/test_wecomcs_eba_adapter.py
index e14a32745..e30b44a1c 100644
--- a/tests/unit_tests/platform/test_wecomcs_eba_adapter.py
+++ b/tests/unit_tests/platform/test_wecomcs_eba_adapter.py
@@ -200,7 +200,7 @@ async def test_wecomcs_event_converter_maps_media_and_unknown_messages():
     assert isinstance(file_event.message_chain[1], platform_message.File)
     assert isinstance(voice_event.message_chain[1], platform_message.Voice)
     assert isinstance(unknown_event, platform_events.PlatformSpecificEvent)
-    assert unknown_event.action == 'wecomcs.event'
+    assert unknown_event.action == 'wecomcs.unknown'
 
 
 @pytest.mark.asyncio
diff --git a/tests/unit_tests/plugin/test_certified_plugin_policy.py b/tests/unit_tests/plugin/test_certified_plugin_policy.py
index 0f5d80da2..29d0ae22f 100644
--- a/tests/unit_tests/plugin/test_certified_plugin_policy.py
+++ b/tests/unit_tests/plugin/test_certified_plugin_policy.py
@@ -12,8 +12,16 @@ from langbot_plugin.entities.io.context import PluginExecutionMode
 @pytest.mark.parametrize(
     ('deployment', 'certificate', 'certificate_id', 'force', 'expected_disposition', 'expected_code'),
     [
-        ('cloud', ('valid', 'shared-runtime-v1'), 'issuer', False, 'shared_eligible', 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE'),
-        ('cloud', ('absent', None), None, False, 'rejected', 'CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_REQUIRED'),
+        (
+            'cloud',
+            ('valid', 'shared-runtime-v1'),
+            'issuer',
+            False,
+            'shared_eligible',
+            'CERTIFIED_PLUGIN_SHARED_ELIGIBLE',
+        ),
+        ('cloud', ('absent', None), None, False, 'dedicated_allowed', 'CERTIFIED_PLUGIN_UNSIGNED_DEDICATED'),
+        ('cloud', ('absent', None), None, True, 'dedicated_allowed', 'CERTIFIED_PLUGIN_UNSIGNED_DEDICATED'),
         ('cloud', ('malformed', None), None, False, 'rejected', 'CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_INVALID'),
         (
             'cloud',
@@ -24,7 +32,14 @@ from langbot_plugin.entities.io.context import PluginExecutionMode
             'CERTIFIED_PLUGIN_CLOUD_CERTIFICATE_INVALID',
         ),
         ('oss', ('absent', None), None, False, 'dedicated_allowed', 'CERTIFIED_PLUGIN_OSS_LEGACY_DEDICATED'),
-        ('oss', ('valid', 'shared-runtime-v1'), 'issuer', False, 'shared_eligible', 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE'),
+        (
+            'oss',
+            ('valid', 'shared-runtime-v1'),
+            'issuer',
+            False,
+            'dedicated_allowed',
+            'CERTIFIED_PLUGIN_OSS_UNTRUSTED_DEDICATED',
+        ),
         (
             'oss',
             ('invalid', 'shared-runtime-v1'),
@@ -85,6 +100,9 @@ def test_admission_policy_enforces_certification_matrix(
         certificate=CertificateFacts(
             verification=CertificateVerification(verification),
             runtime_profile=runtime_profile,
+            component_model=(
+                'stateless-v1' if verification == 'valid' and runtime_profile == 'shared-runtime-v1' else None
+            ),
             certificate_id=certificate_id,
         ),
     )
@@ -102,6 +120,92 @@ def test_admission_policy_enforces_certification_matrix(
     )
 
 
+def test_legacy_shared_certificate_without_stateless_contract_is_not_shared() -> None:
+    from langbot.pkg.plugin.certification import (
+        AdmissionDisposition,
+        CertificateFacts,
+        CertificateVerification,
+        DeploymentMode,
+        PluginCertificationFacts,
+        decide_plugin_admission,
+    )
+
+    decision = decide_plugin_admission(
+        deployment=DeploymentMode.CLOUD,
+        facts=PluginCertificationFacts(
+            installation_uuid='00000000-0000-4000-8000-000000000001',
+            artifact_digest='a' * 64,
+            certificate=CertificateFacts(
+                verification=CertificateVerification.VALID,
+                runtime_profile='shared-runtime-v1',
+                component_model=None,
+                certificate_id='legacy-issuer',
+            ),
+        ),
+    )
+
+    assert decision.disposition is AdmissionDisposition.REJECTED
+
+
+def test_oss_never_selects_shared_even_with_a_resolved_valid_certificate() -> None:
+    from langbot.pkg.plugin.certification import (
+        CertificateFacts,
+        CertificateVerification,
+        PluginCertificationFacts,
+        decide_plugin_admission,
+    )
+
+    facts = PluginCertificationFacts(
+        'installation',
+        'a' * 64,
+        CertificateFacts(
+            CertificateVerification.VALID,
+            'shared-runtime-v1',
+            'stateless-v1',
+            'issuer',
+        ),
+    )
+    decision = decide_plugin_admission(deployment='oss', facts=facts)
+    assert decision.runtime_profile == 'dedicated'
+    assert decision.disposition.value == 'dedicated_allowed'
+
+
+def test_oss_reconcile_never_reuses_persisted_shared_placement() -> None:
+    from types import SimpleNamespace
+    from langbot.pkg.plugin.connector import PluginRuntimeConnector
+
+    connector = object.__new__(PluginRuntimeConnector)
+    connector.ap = SimpleNamespace(deployment=SimpleNamespace(mode='oss'))
+    setting = SimpleNamespace(
+        artifact_digest='a' * 64, install_info={'_certification': _complete_persisted_certification()}
+    )
+    assert connector._execution_mode_from_setting(setting) is PluginExecutionMode.DEDICATED
+
+
+def test_oss_does_not_treat_valid_nonshared_signature_as_dedicated_certification() -> None:
+    from langbot.pkg.plugin.certification import (
+        AdmissionDisposition,
+        CertificateFacts,
+        CertificateVerification,
+        PluginCertificationFacts,
+        decide_plugin_admission,
+    )
+
+    decision = decide_plugin_admission(
+        deployment='oss',
+        facts=PluginCertificationFacts(
+            installation_uuid='00000000-0000-4000-8000-000000000001',
+            artifact_digest='a' * 64,
+            certificate=CertificateFacts(
+                verification=CertificateVerification.VALID,
+                runtime_profile=None,
+                certificate_id='issuer',
+            ),
+        ),
+    )
+    assert decision.disposition is AdmissionDisposition.ADMINISTRATOR_FORCE_REQUIRED
+
+
 def test_archive_inspection_preserves_legacy_tuple_and_exposes_certificate_facts() -> None:
     from langbot.pkg.plugin.archive import (
         ArchiveCertificateState,
@@ -181,6 +285,7 @@ def test_log_visibility_policy_only_scopes_valid_shared_certifications(
         certificate=CertificateFacts(
             verification=CertificateVerification(verification),
             runtime_profile='shared-runtime-v1',
+            component_model='stateless-v1',
         ),
     )
 
@@ -195,6 +300,7 @@ def _complete_persisted_certification() -> dict[str, object]:
         'normalized_digest': 'b' * 64,
         'verification': 'valid',
         'certificate_runtime_profile': 'shared-runtime-v1',
+        'certificate_component_model': 'stateless-v1',
         'certificate_id': 'ed25519:trusted-issuer',
         'runtime_profile': 'shared-runtime-v1',
         'admission_code': 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE',
@@ -261,6 +367,7 @@ def test_persisted_certification_selects_shared_execution_only_for_exact_admitte
         'normalized_digest',
         'verification',
         'certificate_runtime_profile',
+        'certificate_component_model',
         'certificate_id',
         'runtime_profile',
         'admission_code',
@@ -303,6 +410,9 @@ def test_persisted_certification_requires_every_shared_admission_fact(missing_fi
         ('certificate_runtime_profile', None),
         ('certificate_runtime_profile', 123),
         ('certificate_runtime_profile', ''),
+        ('certificate_component_model', None),
+        ('certificate_component_model', 123),
+        ('certificate_component_model', ''),
         ('runtime_profile', None),
         ('runtime_profile', 123),
         ('runtime_profile', ''),
diff --git a/tests/unit_tests/plugin/test_connector_methods.py b/tests/unit_tests/plugin/test_connector_methods.py
index 716b0f1bf..c971324ba 100644
--- a/tests/unit_tests/plugin/test_connector_methods.py
+++ b/tests/unit_tests/plugin/test_connector_methods.py
@@ -749,6 +749,7 @@ class TestSetPluginConfig:
         """Config changes are fenced by a new runtime revision."""
         get_connector_module()
         connector = create_mock_connector()
+        connector.ap.deployment.mode = 'cloud'
 
         configure_handler(connector, AsyncMock())
         connector.handler.register_installation_binding = Mock()
@@ -765,6 +766,7 @@ class TestSetPluginConfig:
                     'normalized_digest': 'b' * 64,
                     'verification': 'valid',
                     'certificate_runtime_profile': 'shared-runtime-v1',
+                    'certificate_component_model': 'stateless-v1',
                     'certificate_id': 'ed25519:trusted-issuer',
                     'runtime_profile': 'shared-runtime-v1',
                     'admission_code': 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE',
diff --git a/tests/unit_tests/plugin/test_connector_reconcile.py b/tests/unit_tests/plugin/test_connector_reconcile.py
index fde178fa2..0b54f9384 100644
--- a/tests/unit_tests/plugin/test_connector_reconcile.py
+++ b/tests/unit_tests/plugin/test_connector_reconcile.py
@@ -52,6 +52,7 @@ def mock_archive_admission(connector: PluginRuntimeConnector, digest: str) -> No
                     'normalized_digest': digest,
                     'verification': 'valid',
                     'certificate_runtime_profile': 'shared-runtime-v1',
+                    'certificate_component_model': 'stateless-v1',
                     'certificate_id': 'ed25519:trusted-issuer',
                     'runtime_profile': 'shared-runtime-v1',
                     'admission_code': 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE',
@@ -190,6 +191,7 @@ async def test_reconcile_reload_projects_certified_exact_artifact_to_shared_exec
             'normalized_digest': 'b' * 64,
             'verification': 'valid',
             'certificate_runtime_profile': 'shared-runtime-v1',
+            'certificate_component_model': 'stateless-v1',
             'certificate_id': 'ed25519:trusted-issuer',
             'runtime_profile': 'shared-runtime-v1',
             'admission_code': 'CERTIFIED_PLUGIN_SHARED_ELIGIBLE',
@@ -733,11 +735,13 @@ async def test_apply_dependency_failure_raises_stable_observable_error():
     assert error.installation_uuid == setting.installation_uuid
     assert error.error_code == 'dependency_prepare_failed'
     assert '[dependency_prepare_failed]' in str(error)
-    assert connector._installation_failures[setting.installation_uuid] == {
+    recorded = connector._installation_failures[setting.installation_uuid]
+    assert {key: value for key, value in recorded.items() if key != 'failed_at'} == {
         'installation_uuid': setting.installation_uuid,
         'error_code': 'dependency_prepare_failed',
         'message': 'Plugin dependency installer exited with code 1',
     }
+    assert str(recorded['failed_at']).endswith('Z')
     connector.ap.logger.error.assert_called_once()
 
 
@@ -765,9 +769,12 @@ async def test_shared_reconcile_records_one_failure_without_blocking_other_state
         setting_a.installation_uuid,
         setting_b.installation_uuid,
     }
-    assert connector._installation_failures == {
-        setting_a.installation_uuid: failure,
-    }
+    recorded_failures = connector._installation_failures
+    assert {
+        uuid: {key: value for key, value in record.items() if key != 'failed_at'}
+        for uuid, record in recorded_failures.items()
+    } == {setting_a.installation_uuid: failure}
+    assert str(recorded_failures[setting_a.installation_uuid]['failed_at']).endswith('Z')
     assert set(connector._known_desired_states) == {
         setting_a.installation_uuid,
         setting_b.installation_uuid,
diff --git a/tests/unit_tests/plugin/test_handler_helpers.py b/tests/unit_tests/plugin/test_handler_helpers.py
index dc86b7a5f..b126a9b18 100644
--- a/tests/unit_tests/plugin/test_handler_helpers.py
+++ b/tests/unit_tests/plugin/test_handler_helpers.py
@@ -10,6 +10,27 @@ from __future__ import annotations
 import pytest
 from unittest.mock import Mock, AsyncMock
 from importlib import import_module
+from types import SimpleNamespace
+
+
+@pytest.mark.parametrize(
+    'call_args',
+    [
+        {'num_retries': 0},
+        {'max_retries': 0},
+        {'timeout': 3},
+        {'request_policy': {'stream_idle_timeout': 2}},
+    ],
+)
+def test_llm_bridge_keeps_model_and_call_policy_layers_separate(call_args):
+    from langbot.pkg.provider.modelmgr.request_policy import RequestPolicy
+
+    model_args = {'temperature': 0.2, 'request_policy': {'retries': 2, 'read_timeout': 90}}
+    model = SimpleNamespace(model_entity=SimpleNamespace(extra_args=model_args))
+    merged = get_handler_module()._merge_model_extra_args(model, call_args, llm_policy=True)
+    assert RequestPolicy.resolve(model_args, merged) == RequestPolicy.resolve(model_args, call_args)
+    assert merged['temperature'] == 0.2
+    assert model_args['request_policy'] == {'retries': 2, 'read_timeout': 90}
 
 
 def get_handler_module():
diff --git a/tests/unit_tests/plugin/test_plugin_file_transfer_root.py b/tests/unit_tests/plugin/test_plugin_file_transfer_root.py
new file mode 100644
index 000000000..fad2751ba
--- /dev/null
+++ b/tests/unit_tests/plugin/test_plugin_file_transfer_root.py
@@ -0,0 +1,51 @@
+"""The host must share the Runtime's plugin file-transfer root in shared deployments.
+
+A mismatch leaves plugin icons and assets failing with
+"Invalid file transfer capability" because the SDK binds one directory inode per
+process (mode 0700) and picks the default from the runtime profile.
+"""
+
+from __future__ import annotations
+
+import os
+from importlib import import_module
+from types import SimpleNamespace
+
+from langbot_plugin.runtime.io.handler import SHARED_WORKER_FILE_STORAGE_DIR
+from langbot_plugin.runtime.security import PLUGIN_FILE_STORAGE_DIR_ENV
+
+
+def connector_module():
+    import_module('langbot.pkg.core.app')
+    return import_module('langbot.pkg.plugin.connector')
+
+
+def align(profile: str) -> None:
+    connector = connector_module()
+    connector.PluginRuntimeConnector._align_plugin_file_transfer_root(
+        SimpleNamespace(runtime_profile=profile)
+    )
+
+
+def test_shared_profile_pins_the_runtime_transfer_root(monkeypatch):
+    monkeypatch.delenv(PLUGIN_FILE_STORAGE_DIR_ENV, raising=False)
+
+    align('shared')
+
+    assert os.environ[PLUGIN_FILE_STORAGE_DIR_ENV] == SHARED_WORKER_FILE_STORAGE_DIR
+
+
+def test_explicit_operator_value_is_never_overridden(monkeypatch):
+    monkeypatch.setenv(PLUGIN_FILE_STORAGE_DIR_ENV, '/var/lib/custom-transfer')
+
+    align('shared')
+
+    assert os.environ[PLUGIN_FILE_STORAGE_DIR_ENV] == '/var/lib/custom-transfer'
+
+
+def test_oss_profile_leaves_the_default_alone(monkeypatch):
+    monkeypatch.delenv(PLUGIN_FILE_STORAGE_DIR_ENV, raising=False)
+
+    align('oss_dev')
+
+    assert PLUGIN_FILE_STORAGE_DIR_ENV not in os.environ
diff --git a/tests/unit_tests/plugin/test_runtime_ops_reporter.py b/tests/unit_tests/plugin/test_runtime_ops_reporter.py
new file mode 100644
index 000000000..c5fbd4b5b
--- /dev/null
+++ b/tests/unit_tests/plugin/test_runtime_ops_reporter.py
@@ -0,0 +1,327 @@
+"""Tests for the Cloud plugin runtime ops sample reporter.
+
+Covers the frozen Core -> Space sample contract (C2):
+- Runtime /healthz URL derivation from the control WebSocket URL
+- Disablement without a runtime WebSocket URL / outside the cloud edition
+- Sample assembly from connector internals (desired states, failures, health)
+- Best-effort upload semantics (non-2xx and transport failures never raise)
+- The pushed payload never carries the runtime control token
+"""
+
+from __future__ import annotations
+
+import json
+from types import SimpleNamespace
+from unittest.mock import Mock
+
+import httpx
+import pytest
+
+from langbot.pkg.plugin.runtime_ops import RuntimeOpsReporter
+from langbot_plugin.entities.io.context import PluginExecutionMode
+from langbot_plugin.runtime.security import PLUGIN_RUNTIME_CONTROL_TOKEN_ENV
+
+_CONTROL_PLANE_TOKEN_ENV = 'LANGBOT_SPACE_CONTROL_PLANE_TOKEN'
+
+_DEFAULT_WS_URL = 'ws://langbot_plugin_runtime:5400/control/ws'
+
+
+def make_policy() -> SimpleNamespace:
+    return SimpleNamespace(model_dump=lambda: {'max_workers': 32, 'max_cpus': 0.25})
+
+
+def make_connector(
+    *,
+    desired_states: dict | None = None,
+    failures: dict | None = None,
+    identities: dict | None = None,
+    connected: bool = True,
+) -> SimpleNamespace:
+    return SimpleNamespace(
+        runtime_profile='shared',
+        worker_policy=make_policy(),
+        _known_desired_states=desired_states if desired_states is not None else {},
+        _installation_failures=failures if failures is not None else {},
+        _reconcile_summary={
+            'last_started_at': '2026-10-01T07:11:00Z',
+            'last_duration_ms': 1234,
+            'last_ok': True,
+            'failed_installations': 0,
+            'missing_artifacts': 0,
+        },
+        handler=SimpleNamespace(_installation_bindings=identities if identities is not None else {}),
+        _runtime_available=lambda: connected,
+    )
+
+
+def make_app(
+    *,
+    runtime_ws_url: str | None = _DEFAULT_WS_URL,
+    runtime_ops: dict | None = None,
+    space_url: str = 'https://space.example',
+    deployment_mode: str = 'cloud',
+    edition: str = 'cloud',
+    connector: SimpleNamespace | None = None,
+) -> SimpleNamespace:
+    plugin: dict = {}
+    if runtime_ws_url is not None:
+        plugin['runtime_ws_url'] = runtime_ws_url
+    if runtime_ops is not None:
+        plugin['runtime_ops'] = runtime_ops
+    return SimpleNamespace(
+        logger=Mock(),
+        instance_config=SimpleNamespace(
+            data={
+                'plugin': plugin,
+                'space': {'url': space_url},
+                'system': {'instance_id': 'inst-1', 'edition': edition},
+            }
+        ),
+        deployment=SimpleNamespace(mode=deployment_mode),
+        plugin_connector=connector if connector is not None else make_connector(),
+    )
+
+
+def make_desired_state(workspace_uuid: str, mode: PluginExecutionMode, *, enabled: bool = True) -> SimpleNamespace:
+    return SimpleNamespace(
+        enabled=enabled,
+        execution_mode=mode,
+        binding=SimpleNamespace(workspace_uuid=workspace_uuid),
+    )
+
+
+class TestHealthUrlDerivation:
+    @pytest.mark.parametrize(
+        ('ws_url', 'expected'),
+        [
+            ('ws://langbot_plugin_runtime:5400/control/ws', 'http://langbot_plugin_runtime:5400/healthz'),
+            ('wss://runtime.example/control/ws', 'https://runtime.example/healthz'),
+            ('wss://runtime.example:8443/control/ws', 'https://runtime.example:8443/healthz'),
+            ('ws://127.0.0.1:5400', 'http://127.0.0.1:5400/healthz'),
+            ('  ws://runtime.example:5400/control/ws  ', 'http://runtime.example:5400/healthz'),
+        ],
+    )
+    def test_derives_health_url(self, ws_url: str, expected: str):
+        reporter = RuntimeOpsReporter(make_app(runtime_ws_url=ws_url))
+
+        assert reporter.health_url == expected
+
+    @pytest.mark.parametrize('ws_url', ['', None, 'http://runtime.example:5400', 'not a url'])
+    def test_rejects_non_websocket_url(self, ws_url):
+        reporter = RuntimeOpsReporter(make_app(runtime_ws_url=ws_url))
+
+        assert reporter.health_url is None
+
+
+class TestEnablement:
+    def test_disabled_without_runtime_websocket_url(self):
+        reporter = RuntimeOpsReporter(make_app(runtime_ws_url=None))
+
+        assert reporter.enabled is False
+
+    def test_disabled_on_non_cloud_deployment(self):
+        reporter = RuntimeOpsReporter(make_app(deployment_mode='oss'))
+
+        assert reporter.enabled is False
+
+    def test_explicit_disable_on_cloud(self):
+        reporter = RuntimeOpsReporter(make_app(runtime_ops={'enabled': False}))
+
+        assert reporter.enabled is False
+
+    def test_enabled_by_default_on_cloud(self):
+        reporter = RuntimeOpsReporter(make_app())
+
+        assert reporter.enabled is True
+        assert reporter.interval_seconds == 30.0
+        assert reporter.timeout_seconds == 10.0
+        assert reporter.max_failures == 5
+
+
+class TestSampleAssembly:
+    def test_build_payload_matches_contract(self):
+        desired = {
+            'inst-dedicated': make_desired_state('ws-1', PluginExecutionMode.DEDICATED),
+            'inst-shared-off': make_desired_state(
+                'ws-2', PluginExecutionMode.SHARED_CERTIFIED, enabled=False
+            ),
+            'inst-shared-on': make_desired_state('ws-3', PluginExecutionMode.SHARED_CERTIFIED),
+        }
+        failures = {
+            'inst-dedicated': {
+                'installation_uuid': 'inst-dedicated',
+                'error_code': 'worker_launch_failed',
+                'message': 'boom\nstack line',
+            },
+            'inst-shared-off': {
+                'installation_uuid': 'inst-shared-off',
+                'error_code': 'dependency_prepare_failed',
+                'message': 'prepare failed',
+            },
+        }
+        identities = {
+            'inst-dedicated': (
+                None,
+                SimpleNamespace(plugin_author='acme', plugin_name='dedicated-plugin'),
+            ),
+        }
+        connector = make_connector(desired_states=desired, failures=failures, identities=identities)
+        reporter = RuntimeOpsReporter(make_app(connector=connector))
+        health = {'live': True, 'runtime': {'version': '0.7.6'}, 'shared_pool': {'workers': 12}}
+
+        payload = reporter.build_payload(health)
+
+        assert set(payload) == {'event', 'kind', 'instance_uuid', 'generated_at', 'core', 'health', 'failures'}
+        assert payload['event'] == 'runtime_ops_sample'
+        assert payload['kind'] == 'aggregate'
+        assert payload['instance_uuid'] == 'inst-1'
+        assert payload['generated_at'].endswith('Z')
+        # The health blob is forwarded verbatim.
+        assert payload['health'] is health
+
+        core = payload['core']
+        assert core['edition'] == 'cloud'
+        assert core['connected'] is True
+        assert core['runtime_profile'] == 'shared'
+        assert core['worker_policy'] == {'max_workers': 32, 'max_cpus': 0.25}
+        assert core['desired_states'] == {
+            'total': 3,
+            'enabled': 2,
+            'by_mode': {'dedicated': 1, 'shared-runtime-v1': 2},
+        }
+        assert core['reconcile'] == {
+            'last_started_at': '2026-10-01T07:11:00Z',
+            'last_duration_ms': 1234,
+            'last_ok': True,
+            'failed_installations': 0,
+            'missing_artifacts': 0,
+        }
+
+        assert len(payload['failures']) == 2
+        first = payload['failures'][0]
+        assert first['workspace_uuid'] == 'ws-1'
+        assert first['installation_uuid'] == 'inst-dedicated'
+        assert first['plugin_author'] == 'acme'
+        assert first['plugin_name'] == 'dedicated-plugin'
+        assert first['execution_mode'] == 'dedicated'
+        assert first['error_code'] == 'worker_launch_failed'
+        assert first['message'] == 'boom stack line'
+        assert first['observed_at'].endswith('Z')
+        second = payload['failures'][1]
+        assert second['workspace_uuid'] == 'ws-2'
+        assert second['plugin_author'] == ''
+        assert second['plugin_name'] == ''
+        assert second['execution_mode'] == 'shared-runtime-v1'
+
+    def test_failures_capped_by_max_failures(self):
+        failures = {
+            f'inst-{index}': {
+                'installation_uuid': f'inst-{index}',
+                'error_code': 'worker_launch_failed',
+                'message': 'boom',
+            }
+            for index in range(8)
+        }
+        connector = make_connector(failures=failures)
+        reporter = RuntimeOpsReporter(make_app(runtime_ops={'max_failures': 3}, connector=connector))
+
+        payload = reporter.build_payload({'live': True})
+
+        assert len(payload['failures']) == 3
+
+    def test_disconnected_connector_reports_not_connected(self):
+        connector = make_connector(connected=False)
+        reporter = RuntimeOpsReporter(make_app(connector=connector))
+
+        payload = reporter.build_payload({'live': False})
+
+        assert payload['core']['connected'] is False
+
+
+class TestUpload:
+    @pytest.mark.asyncio
+    async def test_successful_post_increments_sent(self):
+        reporter = RuntimeOpsReporter(make_app())
+        reporter._client = httpx.AsyncClient(
+            transport=httpx.MockTransport(lambda request: httpx.Response(200, json={'code': 0}))
+        )
+
+        await reporter._post({'event': 'runtime_ops_sample'})
+
+        assert reporter.get_stats() == {'sent_total': 1, 'dropped_total': 0, 'last_error': None}
+        await reporter.stop()
+
+    @pytest.mark.asyncio
+    async def test_non_2xx_increments_dropped_without_raising(self):
+        reporter = RuntimeOpsReporter(make_app())
+        reporter._client = httpx.AsyncClient(
+            transport=httpx.MockTransport(lambda request: httpx.Response(503, text='unavailable'))
+        )
+
+        await reporter._post({'event': 'runtime_ops_sample'})
+
+        stats = reporter.get_stats()
+        assert stats['sent_total'] == 0
+        assert stats['dropped_total'] == 1
+        assert '503' in stats['last_error']
+        await reporter.stop()
+
+    @pytest.mark.asyncio
+    async def test_transport_exception_increments_dropped_without_raising(self):
+        def raise_connect_error(request: httpx.Request) -> httpx.Response:
+            raise httpx.ConnectError('connection refused')
+
+        reporter = RuntimeOpsReporter(make_app())
+        reporter._client = httpx.AsyncClient(transport=httpx.MockTransport(raise_connect_error))
+
+        await reporter._post({'event': 'runtime_ops_sample'})
+
+        stats = reporter.get_stats()
+        assert stats['dropped_total'] == 1
+        assert 'connection refused' in stats['last_error']
+        await reporter.stop()
+
+
+class TestFailureTimestamps:
+    def test_publishes_when_the_failure_happened_not_when_it_was_sampled(self):
+        failures = {
+            'inst-1': {
+                'installation_uuid': 'inst-1',
+                'error_code': 'dependency_prepare_failed',
+                'message': 'prepare failed',
+                'failed_at': '2026-10-01T07:11:00Z',
+            }
+        }
+        reporter = RuntimeOpsReporter(make_app(connector=make_connector(failures=failures)))
+
+        payload = reporter.build_payload({'live': True})
+
+        entry = payload['failures'][0]
+        assert entry['observed_at'] == '2026-10-01T07:11:00Z'
+        assert entry['observed_at'] != payload['generated_at']
+
+    def test_legacy_records_without_a_timestamp_fall_back_to_the_sample_time(self):
+        failures = {
+            'inst-2': {
+                'installation_uuid': 'inst-2',
+                'error_code': 'worker_launch_failed',
+                'message': 'boom',
+            }
+        }
+        reporter = RuntimeOpsReporter(make_app(connector=make_connector(failures=failures)))
+
+        payload = reporter.build_payload({'live': True})
+
+        assert payload['failures'][0]['observed_at'] == payload['generated_at']
+
+
+def test_payload_never_contains_a_control_token(monkeypatch):
+    monkeypatch.setenv(PLUGIN_RUNTIME_CONTROL_TOKEN_ENV, 'runtime-control-secret')
+    monkeypatch.setenv(_CONTROL_PLANE_TOKEN_ENV, 'space-control-plane-secret')
+    reporter = RuntimeOpsReporter(make_app())
+
+    payload = reporter.build_payload({'live': True})
+    serialized = json.dumps(payload)
+
+    assert 'runtime-control-secret' not in serialized
+    assert 'space-control-plane-secret' not in serialized
diff --git a/tests/unit_tests/plugin/test_worker_mode_quotas.py b/tests/unit_tests/plugin/test_worker_mode_quotas.py
new file mode 100644
index 000000000..6a11be777
--- /dev/null
+++ b/tests/unit_tests/plugin/test_worker_mode_quotas.py
@@ -0,0 +1,41 @@
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+
+from langbot.pkg.core.stages.load_config import _complete_runtime_policy_defaults, _apply_env_overrides_to_config
+from langbot.pkg.plugin.connector import PluginRuntimeConnector
+
+
+@pytest.mark.parametrize('mode', ['cloud', 'oss'])
+def test_mode_overrides_only_reach_cloud_runtime(mode, monkeypatch):
+    monkeypatch.setenv('PLUGIN__WORKER__DEDICATED__MAX_CPUS', '0.5')
+    monkeypatch.setenv('PLUGIN__WORKER__DEDICATED__MAX_MEMORY_MB', '256')
+    monkeypatch.setenv('PLUGIN__WORKER__SHARED__MAX_CPUS', '1.0')
+    monkeypatch.setenv('PLUGIN__WORKER__SHARED__MAX_MEMORY_MB', '512')
+    config = _apply_env_overrides_to_config(_complete_runtime_policy_defaults({'plugin': {'enable': True}}))
+    app = SimpleNamespace(instance_config=SimpleNamespace(data=config), deployment=SimpleNamespace(mode=mode))
+    connector = PluginRuntimeConnector(app, AsyncMock())
+    policy = connector._load_worker_policy()
+    if mode == 'cloud':
+        assert policy.dedicated.max_cpus == 0.5
+        assert policy.dedicated.max_memory_mb == 256
+        assert policy.shared.max_cpus == 1.0
+        assert policy.shared.max_memory_mb == 512
+    else:
+        assert policy.dedicated is None
+        assert policy.shared is None
+        assert policy.max_cpus == 1.0
+        assert policy.max_memory_mb == 512
+
+
+def test_cloud_unset_overrides_inherit_existing_limits():
+    config = _complete_runtime_policy_defaults(
+        {'plugin': {'enable': True, 'worker': {'max_cpus': 1.5, 'max_memory_mb': 768}}}
+    )
+    app = SimpleNamespace(instance_config=SimpleNamespace(data=config), deployment=SimpleNamespace(mode='cloud'))
+    policy = PluginRuntimeConnector(app, AsyncMock())._load_worker_policy()
+    for mode in ('dedicated', 'shared-runtime-v1'):
+        resolved = policy.for_execution_mode(mode, 'shared')
+        assert resolved.max_cpus == 1.5
+        assert resolved.max_memory_mb == 768
diff --git a/tests/unit_tests/provider/test_codex.py b/tests/unit_tests/provider/test_codex.py
index 14dbd6832..cb0e9e046 100644
--- a/tests/unit_tests/provider/test_codex.py
+++ b/tests/unit_tests/provider/test_codex.py
@@ -1,6 +1,7 @@
 """Replay synthetic HTTP/SSE traffic through the real Codex requester."""
 
 import json
+import asyncio
 from collections import OrderedDict
 from types import SimpleNamespace
 from unittest.mock import AsyncMock
@@ -16,6 +17,32 @@ TOKENS = {'access_token': 'access-secret', 'account_id': 'account', 'connection_
 MODEL = SimpleNamespace(model_entity=SimpleNamespace(name='codex-test', extra_args={}, reasoning_config=None))
 
 
+@pytest.mark.asyncio
+async def test_response_headers_obey_host_first_event_deadline(monkeypatch):
+    from langbot.pkg.provider.modelmgr.errors import RequesterError
+    from langbot.pkg.provider.modelmgr.request_policy import CallState, RequestPolicy, current_call
+
+    cancelled = asyncio.Event()
+
+    async def handler(request):
+        try:
+            await asyncio.sleep(10)
+        finally:
+            cancelled.set()
+
+    obj = requester(monkeypatch, handler)
+    state = CallState(RequestPolicy(first_event_timeout=0.02))
+    token = current_call.set(state)
+    try:
+        with pytest.raises(RequesterError) as caught:
+            await obj.invoke_llm(None, MODEL, [])
+    finally:
+        current_call.reset(token)
+    assert caught.value.code == 'timeout'
+    assert caught.value.phase == 'first_event'
+    assert cancelled.is_set()
+
+
 def requester(monkeypatch, handler):
     real_client = httpx.AsyncClient
     monkeypatch.setattr(
@@ -198,3 +225,51 @@ def test_advanced_parameters_cannot_override_transport(monkeypatch, key):
     obj = requester(monkeypatch, lambda request: httpx.Response(200))
     with pytest.raises(ValueError, match='Unsupported Codex advanced'):
         obj._body(None, MODEL, [], None, {key: 'secret'}, TOKENS)
+
+
+@pytest.mark.asyncio
+async def test_early_close_releases_nested_http_stream(monkeypatch):
+    closed = asyncio.Event()
+
+    class Body(httpx.AsyncByteStream):
+        async def __aiter__(self):
+            yield b'data: {"type":"response.output_text.delta","delta":"first"}\n\n'
+            await asyncio.Event().wait()
+
+        async def aclose(self):
+            closed.set()
+
+    obj = requester(monkeypatch, lambda request: httpx.Response(200, stream=Body()))
+    stream = obj.invoke_llm_stream(None, MODEL, [])
+    assert (await anext(stream)).content == 'first'
+    await stream.aclose()
+    assert closed.is_set()
+
+
+@pytest.mark.asyncio
+async def test_host_deadline_does_not_cancel_suspended_consumer(monkeypatch, runtime_provider, runtime_llm_model):
+    from langbot.pkg.provider.modelmgr.errors import RequesterError
+    from tests.unit_tests.provider.conftest import TEST_EXECUTION_CONTEXT
+
+    runtime_provider.requester = requester(
+        monkeypatch,
+        lambda request: stream(
+            [
+                {'type': 'response.output_text.delta', 'delta': 'first'},
+                {'type': 'response.completed', 'response': {'status': 'completed', 'output': []}},
+            ]
+        ),
+    )
+    runtime_llm_model.model_entity.extra_args = {}
+    result = runtime_provider.invoke_llm_stream(
+        None,
+        runtime_llm_model,
+        [],
+        extra_args={'request_policy': {'total_timeout': 0.05}},
+        execution_context=TEST_EXECUTION_CONTEXT,
+    )
+    assert (await anext(result)).content == 'first'
+    await asyncio.sleep(0.07)
+    with pytest.raises(RequesterError) as caught:
+        await anext(result)
+    assert caught.value.phase == 'total'
diff --git a/tests/unit_tests/provider/test_http_pool_streams.py b/tests/unit_tests/provider/test_http_pool_streams.py
new file mode 100644
index 000000000..b474e19d5
--- /dev/null
+++ b/tests/unit_tests/provider/test_http_pool_streams.py
@@ -0,0 +1,285 @@
+import asyncio
+import json
+import threading
+from types import SimpleNamespace
+from unittest.mock import Mock
+
+import httpx
+import pytest
+from aiohttp import web
+
+from langbot.pkg.provider.modelmgr.http_pool import _CompletedSSEStream, RequesterHTTPPool
+from langbot.pkg.provider.modelmgr.errors import RequesterError
+from langbot.pkg.provider.modelmgr.requesters.litellmchat import LiteLLMRequester
+from langbot.pkg.provider.modelmgr.request_policy import CallState, RequestPolicy, current_call
+from tests.unit_tests.provider.test_reasoning_control import _runtime_model
+from scripts.llm_latency_probe import SSEMeter, cache_tokens
+
+
+def test_request_timing_uses_high_resolution_monotonic_clock(monkeypatch):
+    from langbot.pkg.provider.modelmgr import request_policy
+
+    ticks = iter([100.0, 100.0004])
+    monkeypatch.setattr(request_policy.time, 'perf_counter', lambda: next(ticks))
+    state = CallState(RequestPolicy())
+    state.mark('request_prepared_ms')
+    assert state.metrics['request_prepared_ms'] == 0.4
+
+
+def test_meter_accepts_split_utf8_and_usage_without_content():
+    meter = SSEMeter()
+    event = {'choices': [{'delta': {'content': '你好'}}]}
+    data = ('data: ' + json.dumps(event, ensure_ascii=False) + '\r\n\r\n').encode()
+    for byte in data:
+        meter.feed(bytes([byte]), 1)
+    meter.feed(b'data: {"choices":[],"usage":{"prompt_tokens_details":{"cached_tokens":12}}}\n\n', 2)
+    meter.feed(b'data: [DONE]\n\n', 3)
+    assert meter.first_content == 1 and meter.content_chars == 2
+    assert cache_tokens(meter.usage) == 12 and meter.done
+    assert cache_tokens(None) is None
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('provider,name', [('openai', 'gpt-4o'), ('openrouter', 'openai/gpt-4o')])
+@pytest.mark.parametrize('drain', [False, True])
+async def test_real_stream_reuses_socket_and_preserves_late_usage(monkeypatch, provider, name, drain):
+    monkeypatch.setenv('NO_PROXY', '127.0.0.1,localhost')
+    sockets = []
+
+    async def handle(request):
+        await request.read()
+        sockets.append(request.transport)
+        response = web.StreamResponse(headers={'Content-Type': 'text/event-stream'})
+        await response.prepare(request)
+        events = [
+            {'choices': [{'index': 0, 'delta': {'role': 'assistant', 'content': 'hello'}, 'finish_reason': None}]},
+            {'choices': [{'index': 0, 'delta': {}, 'finish_reason': 'stop'}]},
+            {
+                'choices': [],
+                'usage': {
+                    'prompt_tokens': 14,
+                    'completion_tokens': 1,
+                    'total_tokens': 15,
+                    'prompt_tokens_details': {'cached_tokens': 12},
+                },
+            },
+        ]
+        for event in events:
+            await response.write(
+                (
+                    'data: '
+                    + json.dumps(
+                        {
+                            'id': 'r',
+                            'object': 'chat.completion.chunk',
+                            'created': 1,
+                            'model': name,
+                            **event,
+                        }
+                    )
+                    + '\n\n'
+                ).encode()
+            )
+            await asyncio.sleep(0.02)
+        await response.write(b'data: [DO')
+        await response.write(b'NE]\n\n')
+        await response.write_eof()
+        return response
+
+    app = web.Application()
+    app.router.add_post('/chat/completions', handle)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    adapter = LiteLLMRequester(
+        SimpleNamespace(logger=Mock()),
+        {
+            'custom_llm_provider': provider,
+            'base_url': f'http://127.0.0.1:{site._server.sockets[0].getsockname()[1]}',
+        },
+    )
+    await adapter.initialize()
+    adapter._http_pool.drain_sse = drain
+    model = _runtime_model(adapter, name=name)
+    model.provider.token_mgr.get_token = lambda: 'test'
+    try:
+        for _ in range(2):
+            state = CallState(RequestPolicy())
+            token = current_call.set(state)
+            try:
+                chunks = [chunk async for chunk in adapter.invoke_llm_stream(None, model, [])]
+                assert ''.join(chunk.content or '' for chunk in chunks) == 'hello'
+                assert state.usage['prompt_tokens_details']['cached_tokens'] == 12
+            finally:
+                current_call.reset(token)
+        # OpenRouter's generic HTTP adapter already consumes EOF; the OpenAI
+        # SDK path is the regression that needs terminal draining.
+        assert (sockets[0] is sockets[1]) == (drain or provider == 'openrouter')
+    finally:
+        await adapter.aclose()
+        await runner.cleanup()
+
+
+@pytest.mark.asyncio
+async def test_terminal_drain_is_bounded_but_early_close_does_not_drain():
+    class Stream(httpx.AsyncByteStream):
+        def __init__(self, terminal):
+            self.terminal = terminal
+            self.tail_read = False
+            self.closed = False
+
+        async def __aiter__(self):
+            yield b'data: [DONE]\n\n' if self.terminal else b'data: {"choices":[]}\n\n'
+            self.tail_read = True
+            await asyncio.Event().wait()
+
+        async def aclose(self):
+            self.closed = True
+
+    for terminal in (False, True):
+        raw = Stream(terminal)
+        wrapped = _CompletedSSEStream(raw)
+        iterator = wrapped.__aiter__()
+        async with asyncio.timeout(1):
+            await anext(iterator)
+        await wrapped.aclose()
+        await iterator.aclose()
+        assert raw.closed and raw.tail_read == terminal
+
+
+@pytest.mark.asyncio
+async def test_cancellation_during_terminal_drain_is_not_swallowed():
+    reading = asyncio.Event()
+
+    class Stream(httpx.AsyncByteStream):
+        async def __aiter__(self):
+            yield b'data: [DONE]\n\n'
+            reading.set()
+            await asyncio.Event().wait()
+
+        async def aclose(self):
+            pass
+
+    wrapped = _CompletedSSEStream(Stream())
+    iterator = wrapped.__aiter__()
+    pending = asyncio.create_task(anext(iterator))
+    await reading.wait()
+    pending.cancel()
+    with pytest.raises(asyncio.CancelledError):
+        await pending
+    await iterator.aclose()
+    await wrapped.aclose()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('retire', [False, True])
+async def test_concurrent_pool_initialization_loads_tls_once_off_loop(monkeypatch, retire):
+    context = httpx.create_ssl_context()
+    entered, release = threading.Event(), threading.Event()
+    main_thread = threading.get_ident()
+    calls = []
+
+    def load():
+        assert threading.get_ident() != main_thread
+        calls.append(True)
+        entered.set()
+        assert release.wait(2)
+        return context
+
+    monkeypatch.setattr(httpx, 'create_ssl_context', load)
+    pool = RequesterHTTPPool()
+
+    async def lease():
+        async with pool.lease() as client:
+            return client
+
+    first, second = asyncio.create_task(lease()), asyncio.create_task(lease())
+    await asyncio.to_thread(entered.wait, 2)
+    if retire:
+        await pool.aclose()
+    release.set()
+    results = await asyncio.gather(first, second, return_exceptions=True)
+    assert len(calls) == 1
+    if retire:
+        assert all(isinstance(result, RequesterError) for result in results)
+        assert pool.client is None
+    else:
+        assert results[0] is results[1] is pool.client
+        assert not pool.client.is_closed
+    await pool.aclose()
+
+
+@pytest.mark.asyncio
+async def test_cancelling_tls_initialization_leaks_no_http_client(monkeypatch):
+    context = httpx.create_ssl_context()
+    entered, release = threading.Event(), threading.Event()
+
+    def load():
+        entered.set()
+        assert release.wait(2)
+        return context
+
+    monkeypatch.setattr(httpx, 'create_ssl_context', load)
+    pool = RequesterHTTPPool()
+
+    async def lease():
+        async with pool.lease():
+            pytest.fail('Cancelled initializer must not publish a client')
+
+    pending = asyncio.create_task(lease())
+    await asyncio.to_thread(entered.wait, 2)
+    pending.cancel()
+    try:
+        with pytest.raises(asyncio.CancelledError):
+            await pending
+        await pool.aclose()
+        assert pool.client is None and pool.active == 0
+    finally:
+        release.set()
+
+
+@pytest.mark.asyncio
+async def test_real_openai_stream_close_disconnects_upstream(monkeypatch):
+    monkeypatch.setenv('NO_PROXY', '127.0.0.1,localhost')
+    disconnected = asyncio.Event()
+
+    async def handle(request):
+        await request.read()
+        response = web.StreamResponse(headers={'Content-Type': 'text/event-stream'})
+        await response.prepare(request)
+        await response.write(
+            b'data: {"id":"r","object":"chat.completion.chunk","created":1,"model":"gpt-4o","choices":[{"index":0,"delta":{"content":"hello"},"finish_reason":null}]}\n\n'
+        )
+        for _ in range(100):
+            if request.transport is None or request.transport.is_closing():
+                disconnected.set()
+                break
+            await asyncio.sleep(0.02)
+        return response
+
+    app = web.Application()
+    app.router.add_post('/chat/completions', handle)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    adapter = LiteLLMRequester(
+        SimpleNamespace(logger=Mock()),
+        {
+            'custom_llm_provider': 'openai',
+            'base_url': f'http://127.0.0.1:{site._server.sockets[0].getsockname()[1]}',
+        },
+    )
+    await adapter.initialize()
+    model = _runtime_model(adapter, name='gpt-4o')
+    model.provider.token_mgr.get_token = lambda: 'test'
+    stream = adapter.invoke_llm_stream(None, model, [])
+    try:
+        assert (await anext(stream)).content == 'hello'
+        await stream.aclose()
+        await asyncio.wait_for(disconnected.wait(), 3)
+    finally:
+        await stream.aclose()
+        await adapter.aclose()
+        await runner.cleanup()
diff --git a/tests/unit_tests/provider/test_litellmchat.py b/tests/unit_tests/provider/test_litellmchat.py
index 490f7aa12..311367d5f 100644
--- a/tests/unit_tests/provider/test_litellmchat.py
+++ b/tests/unit_tests/provider/test_litellmchat.py
@@ -53,6 +53,37 @@ class MockRuntimeRerankModel:
         self.provider.token_mgr.get_token = Mock(return_value=api_key)
 
 
+@pytest.mark.asyncio
+async def test_closing_consumer_closes_provider_http_stream():
+    import asyncio
+    from types import SimpleNamespace
+
+    app = Mock()
+    app.tool_mgr.generate_tools_for_openai = AsyncMock(return_value=None)
+    requester = litellmchat.LiteLLMRequester(ap=app, config={})
+    closed = asyncio.Event()
+
+    async def provider():
+        try:
+            yield SimpleNamespace(
+                choices=[
+                    SimpleNamespace(
+                        delta=SimpleNamespace(model_dump=lambda: {'role': 'assistant', 'content': 'first'}),
+                        finish_reason=None,
+                    )
+                ]
+            )
+            await asyncio.Event().wait()
+        finally:
+            closed.set()
+
+    with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: provider())):
+        output = requester.invoke_llm_stream(query=None, model=MockRuntimeModel(), messages=[])
+        assert (await anext(output)).content == 'first'
+        await output.aclose()
+    assert closed.is_set()
+
+
 class TestBuildLiteLLMModelName:
     """Test _build_litellm_model_name method"""
 
@@ -348,43 +379,33 @@ class TestInvokeLLMStreamUsage:
         assert ''.join(chunk.content or '' for chunk in collected) == 'Visible answer'
 
     @pytest.mark.asyncio
-    async def test_stream_removes_initial_orphan_think_close(self):
-        """Initial reasoning content without an open tag is removed until </think>."""
-        import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-        import langbot_plugin.api.entities.builtin.provider.message as provider_message
-
+    @pytest.mark.parametrize('separate_reasoning', [False, True])
+    async def test_remove_think_preserves_incremental_plain_text(self, separate_reasoning):
+        """Visible text must arrive before the provider produces the next delta."""
         mock_ap = Mock()
-        mock_ap.tool_mgr = Mock()
         mock_ap.tool_mgr.generate_tools_for_openai = AsyncMock(return_value=None)
         requester = litellmchat.LiteLLMRequester(ap=mock_ap, config={})
-        model = MockRuntimeModel('minimax-m3', 'test-api-key')
+        model = MockRuntimeModel('gpt-4o', 'test-api-key')
+        consumed = []
 
-        chunks = [
-            self._make_chunk(content='hidden reasoning'),
-            self._make_chunk(content=' still hidden</thi'),
-            self._make_chunk(content='nk>Visible answer', finish_reason='stop'),
-        ]
+        async def provider_stream():
+            if separate_reasoning:
+                reasoning_chunk = self._make_chunk()
+                reasoning_chunk.choices[0].delta.model_dump.return_value['reasoning_content'] = 'hidden reasoning'
+                yield reasoning_chunk
+            for text in ['Hello', ' world']:
+                consumed.append(text)
+                yield self._make_chunk(content=text)
+            yield self._make_chunk(content=None, finish_reason='stop')
 
-        async def _aiter(*args, **kwargs):
-            for c in chunks:
-                yield c
-
-        query = Mock(spec=pipeline_query.Query)
-        query.variables = {}
-        messages = [provider_message.Message(role='user', content='Hi')]
-
-        with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: _aiter())):
-            collected = [
-                chunk
-                async for chunk in requester.invoke_llm_stream(
-                    query=query,
-                    model=model,
-                    messages=messages,
-                    remove_think=True,
-                )
-            ]
-
-        assert ''.join(chunk.content or '' for chunk in collected) == 'Visible answer'
+        with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: provider_stream())):
+            received = []
+            async for chunk in requester.invoke_llm_stream(query=None, model=model, messages=[], remove_think=True):
+                if chunk.content:
+                    received.append(chunk.content)
+                    assert received == consumed
+                    assert not chunk.is_final
+            assert received == ['Hello', ' world']
 
     @pytest.mark.asyncio
     async def test_stream_removes_non_leading_think_content(self):
@@ -746,7 +767,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(error)
 
-        assert '请求参数错误' in str(exc_info.value)
+        assert 'Invalid request parameters' in str(exc_info.value)
 
     def test_authentication_error(self):
         """Test AuthenticationError translation"""
@@ -757,7 +778,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(error)
 
-        assert 'API key 无效' in str(exc_info.value)
+        assert 'Invalid API key' in str(exc_info.value)
 
     def test_rate_limit_error(self):
         """Test RateLimitError translation"""
@@ -768,7 +789,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(error)
 
-        assert '请求过于频繁' in str(exc_info.value)
+        assert 'Rate limit exceeded' in str(exc_info.value)
 
     def test_timeout_error(self):
         """Test Timeout translation"""
@@ -779,7 +800,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(error)
 
-        assert '请求超时' in str(exc_info.value)
+        assert 'Request timed out' in str(exc_info.value)
 
     def test_context_window_error(self):
         """Test ContextWindowExceededError translation"""
@@ -790,7 +811,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(error)
 
-        assert '上下文长度超限' in str(exc_info.value)
+        assert 'Context length exceeded' in str(exc_info.value)
 
     def test_unknown_error(self):
         """Test unknown error translation"""
@@ -799,7 +820,7 @@ class TestHandleLiteLLMError:
         with pytest.raises(errors.RequesterError) as exc_info:
             requester._handle_litellm_error(Exception('unknown'))
 
-        assert '未知错误' in str(exc_info.value)
+        assert 'Unknown error' in str(exc_info.value)
 
 
 class TestInvokeLLM:
@@ -955,7 +976,7 @@ class TestInvokeLLM:
                     messages=messages,
                 )
 
-            assert 'API key 无效' in str(exc_info.value)
+            assert 'Invalid API key' in str(exc_info.value)
 
 
 class TestInvokeEmbedding:
@@ -1509,3 +1530,33 @@ class TestScanModels:
 
 if __name__ == '__main__':
     pytest.main([__file__, '-v'])
+
+
+class TestStreamingThinkFilter:
+    @pytest.mark.parametrize(
+        'opening,closing',
+        [
+            ('<think>', '</think>'),
+            ('CRETIRE_REASONING_BEGINk', 'CRETIRE_REASONING_ENDk'),
+        ],
+    )
+    def test_split_tags_and_repeated_blocks(self, opening, closing):
+        state = litellmchat._ThinkStripState()
+        assert state.feed('Before ') == 'Before '
+        for char in opening + 'hidden' + closing:
+            assert state.feed(char) == ''
+        assert state.feed('Answer ') == 'Answer '
+        assert state.feed(opening + 'more hidden' + closing + 'end') == 'end'
+        assert state.flush() == ''
+
+    def test_unclosed_think_is_discarded(self):
+        state = litellmchat._ThinkStripState()
+        assert state.feed('<think>hidden') == ''
+        assert state.flush() == ''
+
+    def test_partial_non_tag_is_released(self):
+        state = litellmchat._ThinkStripState()
+        assert state.feed('Answer <thi') == 'Answer '
+        assert state.feed('ng>') == '<thing>'
+        assert state.feed(' <') == ' '
+        assert state.flush() == '<'
diff --git a/tests/unit_tests/provider/test_llm_preparation.py b/tests/unit_tests/provider/test_llm_preparation.py
new file mode 100644
index 000000000..626003982
--- /dev/null
+++ b/tests/unit_tests/provider/test_llm_preparation.py
@@ -0,0 +1,516 @@
+"""Connection ownership, preparation reuse and provider-specific request policy."""
+
+import asyncio
+import copy
+import threading
+import json
+from pathlib import Path
+from collections import OrderedDict
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, Mock, patch
+
+import httpx
+import pytest
+import yaml
+from aiohttp import web
+
+from langbot.pkg.provider.modelmgr import reasoning
+from langbot.pkg.provider.modelmgr.errors import RequesterError
+from langbot.pkg.provider.modelmgr.http_pool import RequesterHTTPPool
+from langbot.pkg.provider.modelmgr.prompt_caching import apply_prompt_caching
+from langbot.pkg.provider.modelmgr.request_policy import CallState, RequestPolicy, current_call
+from langbot.pkg.provider.modelmgr.requesters.litellmchat import LiteLLMRequester
+from langbot.pkg.provider.modelmgr.requesters.litellmchat import _LeasedHTTPHandler
+from langbot.pkg.provider.modelmgr.requesters import codex
+from langbot_plugin.api.entities.builtin.provider.message import Message
+from tests.unit_tests.provider.test_reasoning_control import _runtime_model, _requester
+
+
+@pytest.mark.asyncio
+async def test_pool_retirement_waits_for_stream_lease():
+    pool = RequesterHTTPPool()
+    async with pool.lease() as first:
+        async with pool.lease() as second:
+            assert first is second
+            await pool.aclose()
+            assert not first.is_closed
+            with pytest.raises(RequesterError, match='retired'):
+                async with pool.lease():
+                    pass
+        assert not first.is_closed
+    assert first.is_closed
+    await pool.aclose()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('provider,name', [('openai', 'gpt-4o'), ('openrouter', 'openai/gpt-4o')])
+async def test_real_http_reuses_socket_with_rotating_key_and_timeout(monkeypatch, provider, name):
+    monkeypatch.setenv('NO_PROXY', '127.0.0.1,localhost')
+    sockets, headers, bodies = [], [], []
+
+    async def handle(request):
+        sockets.append(request.transport)
+        headers.append(request.headers['Authorization'])
+        bodies.append(await request.json())
+        return web.json_response(
+            {
+                'id': 'r',
+                'object': 'chat.completion',
+                'created': 1,
+                'model': 'gpt-4o',
+                'choices': [{'index': 0, 'message': {'role': 'assistant', 'content': 'ok'}, 'finish_reason': 'stop'}],
+                'usage': {'prompt_tokens': 1, 'completion_tokens': 1, 'total_tokens': 2},
+            }
+        )
+
+    app = web.Application()
+    app.router.add_post('/v1/chat/completions', handle)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    port = site._server.sockets[0].getsockname()[1]
+    adapter = LiteLLMRequester(Mock(), {'custom_llm_provider': provider, 'base_url': f'http://127.0.0.1:{port}/v1'})
+    await adapter.initialize()
+    model = _runtime_model(adapter, name=name)
+    model.provider.token_mgr.get_token = Mock(side_effect=['first-key', 'second-key'])
+    try:
+        for seconds in (10, 20):
+            state = CallState(RequestPolicy(read_timeout=seconds), 'lb-scoped')
+            token = current_call.set(state)
+            try:
+                message, _ = await adapter.invoke_llm(
+                    None,
+                    model,
+                    [Message(role='user', content='Hi')],
+                    extra_args={'prompt_caching': {'mode': 'openai', 'retention': 'in_memory'}},
+                )
+                assert message.content == 'ok'
+            finally:
+                current_call.reset(token)
+        assert sockets[0] is sockets[1]
+        assert headers == ['Bearer first-key', 'Bearer second-key']
+        assert bodies[0]['prompt_cache_key'] == bodies[1]['prompt_cache_key']
+        assert bodies[0]['prompt_cache_retention'] == 'in_memory'
+        assert all('prompt_caching' not in body and 'reasoning_config' not in body for body in bodies)
+    finally:
+        await adapter.aclose()
+        await runner.cleanup()
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('cancel', [False, True])
+async def test_stream_holds_pool_lease_until_consumer_close(cancel):
+    adapter = _requester('openai')
+    await adapter.initialize()
+    model = _runtime_model(adapter, name='gpt-4o')
+    model.provider.token_mgr.get_token = Mock(return_value='test')
+    closed = asyncio.Event()
+
+    async def raw():
+        try:
+            yield SimpleNamespace(
+                choices=[
+                    SimpleNamespace(delta=SimpleNamespace(model_dump=lambda: {'content': 'hi'}), finish_reason=None)
+                ]
+            )
+            await asyncio.Event().wait()
+        finally:
+            closed.set()
+
+    with patch('langbot.pkg.provider.modelmgr.requesters.litellmchat.acompletion', new=AsyncMock(return_value=raw())):
+        stream = adapter.invoke_llm_stream(None, model, [])
+        await anext(stream)
+        client = adapter._http_pool.client
+        await adapter.aclose()
+        assert not client.is_closed
+        if cancel:
+            pending = asyncio.create_task(anext(stream))
+            await asyncio.sleep(0)
+            pending.cancel()
+            with pytest.raises(asyncio.CancelledError):
+                await pending
+        else:
+            await stream.aclose()
+        assert closed.is_set() and client.is_closed
+
+
+@pytest.mark.asyncio
+async def test_preparation_reused_only_within_call_and_is_not_mutated():
+    adapter = _requester('openai')
+    adapter.ap.tool_mgr = SimpleNamespace(
+        generate_tools_for_openai=AsyncMock(
+            return_value=[{'type': 'function', 'function': {'name': 'tool', 'parameters': {'type': 'object'}}}]
+        )
+    )
+    model = _runtime_model(adapter, name='gpt-4o')
+    model.provider.token_mgr.get_token = Mock(side_effect=['key-1', 'key-2', 'key-3'])
+    messages, funcs = [Message(role='user', content='original')], [object()]
+    state = CallState(RequestPolicy(retries=1))
+    with patch.object(adapter, '_convert_messages', wraps=adapter._convert_messages) as convert:
+        token = current_call.set(state)
+        try:
+            first = await adapter._build_completion_args(model, messages, funcs)
+            first['messages'][0]['content'] = 'mutated'
+            first['tools'][0]['function']['name'] = 'mutated'
+            second = await adapter._build_completion_args(model, messages, funcs)
+        finally:
+            current_call.reset(token)
+        assert convert.call_count == 1
+        assert adapter.ap.tool_mgr.generate_tools_for_openai.await_count == 1
+        assert second['messages'][0]['content'] == 'original'
+        assert second['tools'][0]['function']['name'] == 'tool'
+        assert second['api_key'] == 'key-2'
+        await adapter._build_completion_args(model, messages, funcs)
+        assert convert.call_count == 2
+    assert messages[0].content == 'original'
+
+
+@pytest.mark.asyncio
+async def test_token_counter_runs_off_event_loop():
+    adapter = _requester('openai')
+    model = _runtime_model(adapter, name='gpt-4o')
+    model.provider.token_mgr.get_token = Mock(return_value='test')
+    main_thread = threading.get_ident()
+
+    def count(**kwargs):
+        assert threading.get_ident() != main_thread
+        return 12
+
+    with patch('litellm.token_counter', new=count):
+        assert await adapter.count_tokens(model, []) == 12
+
+
+def test_metadata_cache_is_bounded_and_returns_copies():
+    adapter = _requester('openai')
+    compute = Mock(return_value={'levels': ['low']})
+    first = adapter._cached_metadata('model', compute)
+    first['levels'].append('bad')
+    assert adapter._cached_metadata('model', compute) == {'levels': ['low']}
+    assert compute.call_count == 1
+    for index in range(150):
+        adapter._cached_metadata(index, lambda: index)
+    assert len(adapter._metadata_cache) == 128
+
+
+@pytest.mark.parametrize(
+    'config',
+    [
+        {'budget_tokens': True},
+        {'budget_tokens': 0},
+        {'budget_tokens': -1},
+        {'budget_tokens': 1.5},
+        {'level': 'high', 'budget_tokens': 1024},
+        {'level': 'disabled', 'budget_tokens': 1024},
+    ],
+)
+def test_invalid_reasoning_budget(config):
+    with pytest.raises(ValueError):
+        reasoning.normalize_reasoning_config(config)
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    'provider,name,budget,expected',
+    [
+        ('anthropic', 'claude-sonnet-4-5', 2048, {'thinking': {'type': 'enabled', 'budget_tokens': 2048}}),
+        ('openai', 'qwen3-32b', 2048, {'extra_body': {'enable_thinking': True, 'thinking_budget': 2048}}),
+        ('gemini', 'gemini-2.5-flash', 2048, {'thinking': {'type': 'enabled', 'budget_tokens': 2048}}),
+        ('openrouter', 'anthropic/claude-sonnet-4.5', 2048, {'extra_body': {'reasoning': {'max_tokens': 2048}}}),
+    ],
+)
+async def test_budget_mapping_through_existing_extra_args(provider, name, budget, expected):
+    adapter = _requester(provider)
+    model = _runtime_model(adapter, name=name)
+    model.provider.token_mgr.get_token = Mock(return_value='test')
+    args = await adapter._build_completion_args(
+        model, [], extra_args={'reasoning_config': {'budget_tokens': budget}, 'max_tokens': 4096}
+    )
+    for key, value in expected.items():
+        assert args[key] == value
+    assert 'reasoning_config' not in args
+    assert model.reasoning_config_override is None
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    'provider,name,budget',
+    [
+        ('openai', 'gpt-5', 2048),
+        ('anthropic', 'claude-opus-4-7', 2048),
+        ('anthropic', 'claude-sonnet-4-5', 512),
+        ('gemini', 'gemini-2.5-flash', 30000),
+        ('gemini', 'gemini-2.5-flash-lite', 128),
+    ],
+)
+async def test_budget_not_silently_dropped_on_unsupported_model(provider, name, budget):
+    adapter = _requester(provider)
+    model = _runtime_model(adapter, name=name)
+    model.provider.token_mgr.get_token = Mock(return_value='test')
+    with pytest.raises(RequesterError, match='budget'):
+        await adapter._build_completion_args(model, [], extra_args={'reasoning_config': {'budget_tokens': budget}})
+
+
+def test_runner_level_replaces_call_and_model_budget():
+    adapter = _requester('anthropic')
+    model = _runtime_model(adapter, name='claude-sonnet-4-5')
+    model.model_entity.reasoning_config = {'budget_tokens': 2048}
+    call = {'reasoning_config': {'budget_tokens': 4096}}
+    assert reasoning.resolve_reasoning_config(model, call)['budget_tokens'] == 4096
+    model.reasoning_config_override = {'level': 'low'}
+    assert reasoning.resolve_reasoning_config(model, call) == {'level': 'low'}
+
+
+def test_anthropic_cache_boundary_stable_and_input_unchanged():
+    messages = [{'role': 'system', 'content': 'stable instructions'}, {'role': 'user', 'content': 'first'}]
+    original = copy.deepcopy(messages)
+    args = {'messages': messages, 'prompt_caching': {'mode': 'anthropic', 'ttl': '1h'}}
+    apply_prompt_caching(args, None, provider='anthropic', family='anthropic')
+    prefix = args['messages'][0]
+    assert prefix['content'][-1]['cache_control'] == {'type': 'ephemeral', 'ttl': '1h'}
+    assert messages == original
+    assert 'cache_control' not in args['messages'][1]
+    args['messages'].append({'role': 'user', 'content': 'second'})
+    args['prompt_caching'] = {'mode': 'anthropic'}
+    apply_prompt_caching(args, None, provider='anthropic', family='anthropic')
+    assert args['messages'][0] == prefix
+
+
+def test_explicit_cache_settings_win_and_default_does_nothing():
+    args = {
+        'model': 'openai/gpt-4o',
+        'extra_body': {'prompt_cache_key': 'explicit', 'prompt_cache_retention': '24h'},
+        'prompt_caching': {'mode': 'openai', 'retention': 'in_memory'},
+    }
+    apply_prompt_caching(args, CallState(RequestPolicy(), 'lb-scoped'), provider='openai', family='openai')
+    assert 'prompt_cache_key' not in args and 'prompt_cache_retention' not in args
+    args = {'messages': [{'role': 'system', 'content': 'same'}], 'prompt_caching': {'mode': 'provider_default'}}
+    apply_prompt_caching(args, None, provider='anthropic', family='anthropic')
+    assert args == {'messages': [{'role': 'system', 'content': 'same'}]}
+
+
+def test_native_cache_options_bypass_model_filter_without_legacy_hints():
+    args = {
+        'model': 'openai/gpt-test',
+        'prompt_cache_options': {'mode': 'implicit', 'ttl': '30m'},
+        'prompt_caching': {'mode': 'openai', 'retention': '24h'},
+    }
+    apply_prompt_caching(args, CallState(RequestPolicy(), 'lb-scoped'), provider='openai', family='openai')
+    assert args == {
+        'model': 'openai/gpt-test',
+        'extra_body': {'prompt_cache_options': {'mode': 'implicit', 'ttl': '30m'}},
+    }
+
+
+@pytest.mark.parametrize(
+    'config',
+    [
+        {'mode': 'typo'},
+        {'mode': 'anthropic', 'ttl': '24h'},
+        {'mode': 'openai', 'retention': 'forever'},
+        {'mode': 'provider_default', 'ttl': '1h'},
+    ],
+)
+def test_invalid_cache_config(config):
+    with pytest.raises(RequesterError):
+        apply_prompt_caching(
+            {'model': 'test', 'messages': [], 'prompt_caching': config}, None, provider='openai', family='openai'
+        )
+
+
+@pytest.mark.asyncio
+async def test_gateway_transport_never_retries_or_closes_borrowed_client():
+    calls = 0
+
+    async def fail(request):
+        nonlocal calls
+        calls += 1
+        raise httpx.ConnectError('unavailable', request=request)
+
+    async with httpx.AsyncClient(transport=httpx.MockTransport(fail)) as client:
+        handler = _LeasedHTTPHandler(client, 10)
+        with pytest.raises(httpx.ConnectError):
+            await handler.post('https://example.com', json={})
+        await handler.close()
+        assert calls == 1 and not client.is_closed
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('provider', ['anthropic', 'openrouter', 'openrouter-metadata'])
+async def test_anthropic_cache_and_budget_reach_upstream(provider):
+    bodies = []
+
+    async def handle(request):
+        bodies.append(json.loads(request.content))
+        if provider == 'anthropic':
+            payload = {
+                'id': 'msg_test',
+                'type': 'message',
+                'role': 'assistant',
+                'model': 'claude-sonnet-4-5',
+                'content': [{'type': 'text', 'text': 'ok'}],
+                'stop_reason': 'end_turn',
+                'usage': {'input_tokens': 1, 'output_tokens': 1, 'cache_read_input_tokens': 50},
+            }
+        else:
+            payload = {
+                'id': 'r',
+                'object': 'chat.completion',
+                'created': 1,
+                'model': 'anthropic/claude-sonnet-4.5',
+                'choices': [{'index': 0, 'message': {'role': 'assistant', 'content': 'ok'}, 'finish_reason': 'stop'}],
+                'usage': {
+                    'prompt_tokens': 51,
+                    'completion_tokens': 1,
+                    'total_tokens': 52,
+                    'prompt_tokens_details': {'cached_tokens': 50},
+                },
+            }
+        return httpx.Response(200, json=payload)
+
+    adapter = _requester('openai' if provider == 'openrouter-metadata' else provider)
+    if provider == 'openrouter-metadata':
+        metadata = yaml.safe_load(
+            (
+                Path(__file__).parents[3] / 'src/langbot/pkg/provider/modelmgr/requesters/openrouterchatcmpl.yaml'
+            ).read_text(encoding='utf-8')
+        )
+        adapter.requester_cfg.update(
+            requester_name=metadata['metadata']['name'], custom_llm_provider=metadata['spec']['litellm_provider']
+        )
+    name = 'claude-sonnet-4-5' if provider == 'anthropic' else 'anthropic/claude-sonnet-4.5'
+    model = _runtime_model(adapter, name=name)
+    model.provider.token_mgr.get_token = Mock(return_value='test')
+    async with httpx.AsyncClient(transport=httpx.MockTransport(handle)) as client:
+        result, usage = await adapter.invoke_llm(
+            None,
+            model,
+            [
+                Message(role='system', content='stable'),
+                Message(role='user', content='hello'),
+            ],
+            extra_args={
+                'client': _LeasedHTTPHandler(client, 10),
+                'reasoning_config': {'budget_tokens': 2048},
+                'max_tokens': 4096,
+                'prompt_caching': {'mode': 'anthropic', 'ttl': '1h'},
+            },
+        )
+    assert result.content == 'ok'
+    assert usage
+    body = bodies[0]
+    if provider == 'anthropic':
+        assert body['system'][-1]['cache_control'] == {'type': 'ephemeral', 'ttl': '1h'}
+        assert body['thinking']['budget_tokens'] == 2048
+    else:
+        assert body['messages'][0]['content'][-1]['cache_control'] == {'type': 'ephemeral', 'ttl': '1h'}
+        assert body['reasoning']['max_tokens'] == 2048
+    assert 'reasoning_config' not in body and 'prompt_caching' not in body
+
+
+@pytest.mark.asyncio
+async def test_codex_reuses_real_connection(monkeypatch):
+    monkeypatch.setenv('NO_PROXY', '127.0.0.1,localhost')
+    sockets = []
+
+    async def handle(request):
+        sockets.append(request.transport)
+        await request.read()
+        events = [
+            {'type': 'response.output_text.delta', 'delta': 'ok'},
+            {'type': 'response.completed', 'response': {'id': 'r', 'status': 'completed', 'output': []}},
+        ]
+        return web.Response(
+            text=''.join('data: ' + json.dumps(event) + '\n\n' for event in events), content_type='text/event-stream'
+        )
+
+    app = web.Application()
+    app.router.add_post('/responses', handle)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    monkeypatch.setattr(codex, 'BASE_URL', f'http://127.0.0.1:{site._server.sockets[0].getsockname()[1]}')
+    adapter = object.__new__(codex.CodexRequester)
+    adapter.workspace, adapter.provider = 'w', 'p'
+    adapter._replay = OrderedDict()
+    adapter._http_pool = RequesterHTTPPool()
+    adapter.auth = SimpleNamespace(
+        access=AsyncMock(
+            return_value={
+                'access_token': 'test',
+                'account_id': 'a',
+                'connection_id': 'c',
+            }
+        )
+    )
+    model = SimpleNamespace(model_entity=SimpleNamespace(name='test', extra_args={}, reasoning_config=None))
+    try:
+        for _ in range(2):
+            message, _ = await adapter.invoke_llm(None, model, [])
+            assert message.content == 'ok'
+        assert sockets[0] is sockets[1]
+    finally:
+        await adapter.aclose()
+        await runner.cleanup()
+
+
+@pytest.mark.asyncio
+async def test_explicit_client_and_redirected_provider_are_preserved():
+    adapter = _requester('openai')
+    await adapter.initialize()
+    explicit = object()
+    for args in ({'model': 'openai/gpt-4o', 'client': explicit}, {'model': 'anthropic/claude-sonnet-4-5'}):
+        async with adapter._client_args(args) as result:
+            assert result is args
+    assert adapter._http_pool.client is None
+    await adapter.aclose()
+
+
+@pytest.mark.asyncio
+async def test_codex_terminal_tail_is_bounded_and_releases_stream():
+    closed = asyncio.Event()
+
+    class HangingTail(httpx.AsyncByteStream):
+        async def __aiter__(self):
+            yield b'data: {"type":"response.completed"}\n\n'
+            await asyncio.Event().wait()
+
+        async def aclose(self):
+            closed.set()
+
+    async with httpx.AsyncClient(
+        transport=httpx.MockTransport(lambda request: httpx.Response(200, stream=HangingTail()))
+    ) as client:
+        async with client.stream('GET', 'https://example.com') as response:
+            events = [event async for event in codex.sse_events(response)]
+            assert events == [{'type': 'response.completed'}]
+    assert closed.is_set()
+
+
+@pytest.mark.asyncio
+async def test_custom_tls_keeps_litellm_transport(monkeypatch):
+    import litellm
+
+    adapter = _requester('openai')
+    await adapter.initialize()
+    monkeypatch.setattr(litellm, 'ssl_verify', '/custom/ca.pem')
+    args = {'model': 'openai/gpt-4o'}
+    async with adapter._client_args(args) as result:
+        assert result is args
+    assert adapter._http_pool.client is None
+    await adapter.aclose()
+
+
+@pytest.mark.asyncio
+async def test_pooled_sdk_preserves_environment_fallbacks(monkeypatch):
+    monkeypatch.setenv('OPENAI_BASE_URL', 'https://gateway.example/v1')
+    monkeypatch.setenv('OPENAI_API_KEY', 'env-test-key')
+    adapter = _requester('openai')
+    await adapter.initialize()
+    try:
+        async with adapter._client_args({'model': 'openai/gpt-4o', 'api_key': ''}) as args:
+            assert str(args['client'].base_url) == 'https://gateway.example/v1/'
+            assert args['client'].api_key == 'env-test-key'
+    finally:
+        await adapter.aclose()
diff --git a/tests/unit_tests/provider/test_mcp_resources.py b/tests/unit_tests/provider/test_mcp_resources.py
index 0ba164e9a..97b136d56 100644
--- a/tests/unit_tests/provider/test_mcp_resources.py
+++ b/tests/unit_tests/provider/test_mcp_resources.py
@@ -24,7 +24,6 @@ from langbot.pkg.provider.tools.loaders.mcp import (
     MCPToolCallTimeoutError,
     RuntimeMCPSession,
 )
-from langbot.pkg.telemetry import features as telemetry_features
 from langbot.pkg.workspace.errors import WorkspaceGenerationMismatchError
 
 
@@ -298,10 +297,6 @@ async def test_read_resource_envelope_truncates_caches_and_records_trace():
     traces = query.variables[MCP_RESOURCE_TRACE_QUERY_KEY]
     assert [trace['source'] for trace in traces] == ['ui_preview', 'agent_tool']
     assert traces[1]['cache_hit'] is True
-    assert query.variables[telemetry_features.FEATURES_KEY]['mcp_resource_reads'] == {
-        'ui_preview': 1,
-        'agent_tool': 1,
-    }
 
 
 @pytest.mark.asyncio
diff --git a/tests/unit_tests/provider/test_request_policy.py b/tests/unit_tests/provider/test_request_policy.py
new file mode 100644
index 000000000..c8b0eb448
--- /dev/null
+++ b/tests/unit_tests/provider/test_request_policy.py
@@ -0,0 +1,549 @@
+import asyncio
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, Mock, patch
+
+import httpx
+import litellm
+import pytest
+
+from langbot.pkg.provider.modelmgr.errors import RequesterError
+from langbot.pkg.provider.modelmgr.request_policy import (
+    CallState,
+    RequestPolicy,
+    current_call,
+    error_metadata,
+    opaque_conversation_id,
+)
+from langbot.pkg.provider.modelmgr.requesters.litellmchat import LiteLLMRequester
+from langbot_plugin.api.entities.builtin.provider.message import Message, MessageChunk
+from tests.unit_tests.provider.conftest import TEST_EXECUTION_CONTEXT
+from tests.unit_tests.provider.test_litellmchat import MockRuntimeModel
+
+
+def test_policy_precedence_zero_and_no_mutation():
+    model = {'num_retries': 3, 'request_policy': {'read_timeout': 42}}
+    policy = RequestPolicy.resolve({'request_policy': {'retries': 2}}, model, {'num_retries': 0})
+    assert policy.retries == 0
+    assert policy.read_timeout == 42
+    assert model['num_retries'] == 3
+
+
+@pytest.mark.parametrize(
+    'policy',
+    [
+        {'retries': True},
+        {'retries': -1},
+        {'retries': 6},
+        {'read_timeout': 0},
+        {'total_timeout': float('nan')},
+        {'connect_timeout': float('inf')},
+        {'attribution': 'false'},
+        {'conversation_header': 'authorization'},
+        {'typo': 1},
+    ],
+)
+def test_invalid_policy(policy):
+    with pytest.raises(RequesterError, match='Model request failed|模型请求失败'):
+        RequestPolicy.resolve({'request_policy': policy})
+
+
+def test_retry_metadata_does_not_include_upstream_secrets():
+    response = httpx.Response(429, headers={'retry-after': '2', 'x-request-id': 'req-123'})
+    meta = error_metadata(SimpleNamespace(status_code=429, response=response))
+    assert meta == {'status_code': 429, 'retry_after': 2, 'request_id': 'req-123'}
+    adapter = LiteLLMRequester(Mock(), {})
+    exc = litellm.RateLimitError('secret-key-and-prompt', llm_provider='openai', model='test', response=response)
+    with pytest.raises(RequesterError) as caught:
+        adapter._handle_litellm_error(exc)
+    assert caught.value.retryable
+    assert caught.value.request_id == 'req-123'
+    assert 'secret-key' not in str(caught.value)
+
+
+def test_quota_is_not_retryable():
+    adapter = LiteLLMRequester(Mock(), {})
+    exc = litellm.RateLimitError('quota', llm_provider='openai', model='test')
+    exc.body = {'error': {'code': 'insufficient_quota'}}
+    with pytest.raises(RequesterError) as caught:
+        adapter._handle_litellm_error(exc)
+    assert caught.value.code == 'quota_exceeded'
+    assert not caught.value.retryable
+
+
+@pytest.mark.asyncio
+async def test_headers_and_sdk_retry_ownership():
+    adapter = LiteLLMRequester(Mock(), {'base_url': 'https://openrouter.ai/api/v1'})
+    state = CallState(RequestPolicy(retries=2), 'lb-opaque')
+    token = current_call.set(state)
+    try:
+        args = await adapter._build_completion_args(
+            MockRuntimeModel(),
+            [],
+            extra_args={
+                'request_policy': {'retries': 2},
+                'num_retries': 2,
+                'extra_headers': {'x-openrouter-title': 'Custom app', 'x-other': 'keep'},
+            },
+        )
+    finally:
+        current_call.reset(token)
+    assert 'request_policy' not in args
+    assert args['num_retries'] == args['max_retries'] == 0
+    assert isinstance(args['timeout'], httpx.Timeout)
+    assert args['extra_headers'] == {
+        'x-openrouter-title': 'Custom app',
+        'x-other': 'keep',
+        'HTTP-Referer': 'https://langbot.app',
+        'x-session-id': 'lb-opaque',
+    }
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('url', ['https://gateway.example/v1', 'https://openrouter.ai.evil.example/v1'])
+async def test_no_vendor_headers_on_unrelated_hosts(url):
+    adapter = LiteLLMRequester(Mock(), {'base_url': url})
+    args = await adapter._build_completion_args(MockRuntimeModel(), [])
+    assert not args.get('extra_headers')
+    assert args['num_retries'] == 0
+
+
+def test_conversation_is_stable_and_scoped():
+    query = SimpleNamespace(session=None, launcher_type='person', launcher_id='123', bot_uuid='bot', pipeline_uuid='p')
+    first = opaque_conversation_id(query, TEST_EXECUTION_CONTEXT)
+    assert first == opaque_conversation_id(query, TEST_EXECUTION_CONTEXT)
+    query.bot_uuid = 'another-bot'
+    assert first != opaque_conversation_id(query, TEST_EXECUTION_CONTEXT)
+    assert opaque_conversation_id(None, TEST_EXECUTION_CONTEXT) is None
+
+
+@pytest.mark.asyncio
+async def test_runtime_retries_only_before_output(runtime_provider, runtime_llm_model):
+    provider = runtime_provider
+    attempts = 0
+    closed = []
+
+    async def stream(**kwargs):
+        nonlocal attempts
+        attempts += 1
+        try:
+            if attempts == 1:
+                raise RequesterError('transient', retryable=True)
+            yield MessageChunk(role='assistant', content='first')
+            raise RequesterError('after output', retryable=True)
+        finally:
+            closed.append(attempts)
+
+    provider.requester.invoke_llm_stream = stream
+    result = []
+    with pytest.raises(RequesterError, match='after output'):
+        async for chunk in provider.invoke_llm_stream(
+            None,
+            runtime_llm_model,
+            [],
+            extra_args={'request_policy': {'retries': 3, 'backoff_base': 0.001}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        ):
+            result.append(chunk.content)
+    assert result == ['first']
+    assert attempts == 2
+    assert closed == [1, 2]
+    assert current_call.get() is None
+
+
+@pytest.mark.asyncio
+async def test_runtime_deadline_and_close(runtime_provider, runtime_llm_model):
+    closed = asyncio.Event()
+
+    async def stream(**kwargs):
+        try:
+            await asyncio.Event().wait()
+            yield MessageChunk(role='assistant', content='never')
+        finally:
+            closed.set()
+
+    runtime_provider.requester.invoke_llm_stream = stream
+    with pytest.raises(RequesterError) as caught:
+        async for _ in runtime_provider.invoke_llm_stream(
+            None,
+            runtime_llm_model,
+            [],
+            extra_args={'request_policy': {'first_event_timeout': 0.02}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        ):
+            pass
+    assert caught.value.phase == 'first_event'
+    assert closed.is_set()
+
+
+@pytest.mark.asyncio
+async def test_retry_after_cannot_extend_total_budget():
+    state = CallState(RequestPolicy(retries=5, total_timeout=0.1))
+    state.metrics['attempts'] = 1
+    assert not await state.retry(RequesterError('busy', retryable=True, retry_after=30))
+
+
+@pytest.mark.asyncio
+async def test_cancellation_is_not_retried(runtime_provider, runtime_llm_model):
+    invoke = AsyncMock(side_effect=asyncio.CancelledError)
+    runtime_provider.requester.invoke_llm = invoke
+    with pytest.raises(asyncio.CancelledError):
+        await runtime_provider.invoke_llm(
+            None,
+            runtime_llm_model,
+            [],
+            extra_args={'request_policy': {'retries': 3}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        )
+    assert invoke.await_count == 1
+
+
+@pytest.mark.asyncio
+async def test_complete_usage_and_timings_persist(runtime_provider, runtime_llm_model):
+    from langbot_plugin.api.entities.builtin.pipeline.query import Query
+
+    query = Query.model_construct(
+        query_id=1, launcher_type='person', launcher_id=1, variables={}, bot_uuid='bot', pipeline_uuid='p', session=None
+    )
+    object.__setattr__(query, '_execution_context', TEST_EXECUTION_CONTEXT)
+    usage = {
+        'prompt_tokens': 20,
+        'completion_tokens': 10,
+        'total_tokens': 30,
+        'prompt_tokens_details': {'cached_tokens': 15},
+        'completion_tokens_details': {'reasoning_tokens': 8},
+        'cost': 0.01,
+    }
+    runtime_provider.requester.invoke_llm = AsyncMock(return_value=(Message(role='assistant', content='ok'), usage))
+    with patch(
+        'langbot.pkg.pipeline.monitoring_helper.MonitoringHelper.record_llm_call', new_callable=AsyncMock
+    ) as record:
+        result = await runtime_provider.invoke_llm(query, runtime_llm_model, [])
+    assert result.content == 'ok'
+    details = record.call_args.kwargs['details']
+    assert details['usage'] == usage
+    assert details['timing']['attempts'] == 1
+    assert details['timing']['total_ms'] >= details['timing']['response_completed_ms']
+    assert query.variables['_llm_usage'] == usage
+
+
+@pytest.mark.asyncio
+async def test_hidden_reasoning_marks_raw_event_and_usage():
+    state = CallState(RequestPolicy())
+    adapter = LiteLLMRequester(Mock(), {})
+
+    async def raw():
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(
+                    delta=SimpleNamespace(model_dump=lambda: {'reasoning_content': 'hidden'}), finish_reason=None
+                )
+            ]
+        )
+        yield SimpleNamespace(choices=[], usage={'prompt_tokens': 3, 'completion_tokens': 2, 'total_tokens': 5})
+
+    token = current_call.set(state)
+    try:
+        with patch(
+            'langbot.pkg.provider.modelmgr.requesters.litellmchat.acompletion', new=AsyncMock(return_value=raw())
+        ):
+            chunks = [c async for c in adapter.invoke_llm_stream(None, MockRuntimeModel(), [], remove_think=True)]
+    finally:
+        current_call.reset(token)
+    assert state.raw_started
+    assert 'first_event_ms' in state.metrics and 'first_reasoning_ms' in state.metrics
+    assert 'first_content_ms' not in state.metrics
+    assert state.usage['total_tokens'] == 5
+    assert not any(c.content for c in chunks)
+
+
+@pytest.mark.asyncio
+async def test_real_http_retry_count_and_request_body(runtime_provider, runtime_llm_model):
+    """Exercise the pinned LiteLLM/OpenAI SDK against a local HTTP server."""
+    from aiohttp import web
+
+    received = []
+
+    async def handler(request):
+        received.append(await request.json())
+        if len(received) == 1:
+            return web.json_response({'error': {'message': 'busy', 'type': 'server_error'}}, status=503)
+        return web.json_response(
+            {
+                'id': 'chatcmpl-local',
+                'object': 'chat.completion',
+                'created': 1,
+                'model': 'gpt-4o',
+                'choices': [{'index': 0, 'message': {'role': 'assistant', 'content': 'ok'}, 'finish_reason': 'stop'}],
+                'usage': {'prompt_tokens': 4, 'completion_tokens': 1, 'total_tokens': 5},
+            }
+        )
+
+    app = web.Application()
+    app.router.add_post('/v1/chat/completions', handler)
+    runner = web.AppRunner(app)
+    await runner.setup()
+    site = web.TCPSite(runner, '127.0.0.1', 0)
+    await site.start()
+    port = site._server.sockets[0].getsockname()[1]
+    runtime_provider.requester = LiteLLMRequester(
+        Mock(),
+        {
+            'base_url': f'http://127.0.0.1:{port}/v1',
+            'custom_llm_provider': 'openai',
+        },
+    )
+    runtime_llm_model.model_entity.name = 'gpt-4o'
+    runtime_llm_model.model_entity.extra_args = {}
+    try:
+        with pytest.raises(RequesterError):
+            await runtime_provider.invoke_llm(
+                None,
+                runtime_llm_model,
+                [Message(role='user', content='hello')],
+                extra_args={'num_retries': 0},
+                execution_context=TEST_EXECUTION_CONTEXT,
+            )
+        assert len(received) == 1
+        received.clear()
+        result = await runtime_provider.invoke_llm(
+            None,
+            runtime_llm_model,
+            [Message(role='user', content='hello')],
+            extra_args={'request_policy': {'retries': 1, 'backoff_base': 0.001}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        )
+        assert result.content == 'ok'
+        assert len(received) == 2
+        assert all('request_policy' not in body and 'num_retries' not in body for body in received)
+    finally:
+        await runner.cleanup()
+
+
+@pytest.mark.asyncio
+async def test_stream_idle_timeout_after_output(runtime_provider, runtime_llm_model):
+    async def stream(**kwargs):
+        yield MessageChunk(role='assistant', content='first')
+        await asyncio.Event().wait()
+
+    runtime_provider.requester.invoke_llm_stream = stream
+    with pytest.raises(RequesterError) as caught:
+        async for _ in runtime_provider.invoke_llm_stream(
+            None,
+            runtime_llm_model,
+            [],
+            extra_args={'request_policy': {'stream_idle_timeout': 0.02}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        ):
+            pass
+    assert caught.value.phase == 'stream_idle'
+
+
+def test_details_migration_preserves_existing_records_and_is_idempotent():
+    import importlib
+    import sqlalchemy as sa
+    from alembic.migration import MigrationContext
+    from alembic.operations import Operations
+
+    migration = importlib.import_module('langbot.pkg.persistence.alembic.versions.0038_llm_call_details')
+    engine = sa.create_engine('sqlite://')
+    try:
+        with engine.begin() as connection:
+            connection.exec_driver_sql('CREATE TABLE monitoring_llm_calls (id TEXT PRIMARY KEY)')
+            connection.exec_driver_sql("INSERT INTO monitoring_llm_calls (id) VALUES ('old')")
+            with Operations.context(MigrationContext.configure(connection)):
+                migration.upgrade()
+                migration.upgrade()
+            assert connection.exec_driver_sql('SELECT id, details FROM monitoring_llm_calls').one() == ('old', None)
+            table = sa.Table('monitoring_llm_calls', sa.MetaData(), autoload_with=connection)
+            payload = {'usage': {'completion_tokens_details': {'reasoning_tokens': 8}}, 'timing': {'first_event_ms': 1}}
+            connection.execute(table.insert().values(id='new', details=payload))
+            assert connection.execute(sa.select(table.c.details).where(table.c.id == 'new')).scalar() == payload
+    finally:
+        engine.dispose()
+
+
+@pytest.mark.asyncio
+async def test_hidden_raw_events_reset_idle_not_first_visible_deadline(runtime_provider, runtime_llm_model):
+    adapter = LiteLLMRequester(Mock(), {})
+    runtime_provider.requester = adapter
+    runtime_llm_model.model_entity.extra_args = {}
+
+    async def raw():
+        for _ in range(5):
+            await asyncio.sleep(0.3)
+            yield SimpleNamespace(
+                choices=[
+                    SimpleNamespace(delta=SimpleNamespace(model_dump=lambda: {'role': 'assistant'}), finish_reason=None)
+                ]
+            )
+        yield SimpleNamespace(
+            choices=[
+                SimpleNamespace(delta=SimpleNamespace(model_dump=lambda: {'content': 'answer'}), finish_reason='stop')
+            ]
+        )
+
+    with patch('langbot.pkg.provider.modelmgr.requesters.litellmchat.acompletion', new=AsyncMock(return_value=raw())):
+        chunks = [
+            chunk
+            async for chunk in runtime_provider.invoke_llm_stream(
+                None,
+                runtime_llm_model,
+                [],
+                extra_args={'request_policy': {'first_event_timeout': 1, 'stream_idle_timeout': 2}},
+                execution_context=TEST_EXECUTION_CONTEXT,
+            )
+        ]
+    assert ''.join(chunk.content or '' for chunk in chunks) == 'answer'
+
+
+@pytest.mark.asyncio
+async def test_monitoring_service_persists_full_details():
+    from langbot.pkg.api.http.service.monitoring import MonitoringService
+
+    persistence = SimpleNamespace(execute_async=AsyncMock())
+    service = MonitoringService(SimpleNamespace(persistence_mgr=persistence))
+    details = {
+        'usage': {
+            'prompt_tokens': 4,
+            'completion_tokens': 3,
+            'total_tokens': 9,
+            'cache_read_input_tokens': 2,
+            'provider_extra': {'cost': 0.01},
+        },
+        'timing': {'attempts': 2},
+        'error': None,
+    }
+    await service.record_llm_call(
+        TEST_EXECUTION_CONTEXT,
+        bot_id='b',
+        bot_name='b',
+        pipeline_id='p',
+        pipeline_name='p',
+        session_id='s',
+        model_name='m',
+        input_tokens=4,
+        output_tokens=3,
+        duration=10,
+        details=details,
+    )
+    values = persistence.execute_async.call_args.args[0].compile().params
+    assert values['details'] == details
+    assert values['total_tokens'] == 9
+
+
+def test_codex_usage_preserves_provider_fields():
+    from langbot.pkg.provider.modelmgr.requesters.codex import _usage
+
+    raw = {
+        'input_tokens': 4,
+        'output_tokens': 2,
+        'cost': 0.02,
+        'output_tokens_details': {'reasoning_tokens': 1},
+        'future_counter': 7,
+    }
+    normalized = _usage({'usage': raw})
+    assert all(normalized[key] == value for key, value in raw.items())
+    assert normalized['total_tokens'] == 6
+
+
+@pytest.mark.asyncio
+async def test_nonstream_uses_response_budget_not_first_event():
+    adapter = LiteLLMRequester(Mock(), {})
+    state = CallState(RequestPolicy(first_event_timeout=0.001, total_timeout=1))
+
+    async def completion(**kwargs):
+        await asyncio.sleep(0.02)
+        return 'complete'
+
+    token = current_call.set(state)
+    try:
+        with patch('langbot.pkg.provider.modelmgr.requesters.litellmchat.acompletion', new=completion):
+            assert await adapter._completion({'stream': False}) == 'complete'
+    finally:
+        current_call.reset(token)
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize('cancel_read', [False, True])
+async def test_litellm_stream_close_and_cancel_release_upstream(runtime_provider, runtime_llm_model, cancel_read):
+    adapter = LiteLLMRequester(Mock(), {})
+    runtime_provider.requester = adapter
+    runtime_llm_model.model_entity.extra_args = {}
+    closed = asyncio.Event()
+    reading = asyncio.Event()
+
+    async def raw():
+        try:
+            yield SimpleNamespace(
+                choices=[
+                    SimpleNamespace(delta=SimpleNamespace(model_dump=lambda: {'content': 'first'}), finish_reason=None)
+                ]
+            )
+            reading.set()
+            await asyncio.Event().wait()
+        finally:
+            closed.set()
+
+    completion = AsyncMock(return_value=raw())
+    with patch('langbot.pkg.provider.modelmgr.requesters.litellmchat.acompletion', new=completion):
+        stream = runtime_provider.invoke_llm_stream(
+            None,
+            runtime_llm_model,
+            [],
+            extra_args={'request_policy': {'retries': 3}},
+            execution_context=TEST_EXECUTION_CONTEXT,
+        )
+        assert (await anext(stream)).content == 'first'
+        if cancel_read:
+            task = asyncio.create_task(anext(stream))
+            await reading.wait()
+            task.cancel()
+            with pytest.raises(asyncio.CancelledError):
+                await task
+        else:
+            await stream.aclose()
+        assert closed.is_set()
+        assert completion.await_count == 1
+        assert current_call.get() is None
+
+
+@pytest.mark.asyncio
+async def test_cancel_during_backoff_does_not_start_another_attempt(runtime_provider, runtime_llm_model):
+    runtime_provider.requester.invoke_llm = AsyncMock(side_effect=RequesterError('busy', retryable=True))
+    sleeping = asyncio.Event()
+
+    async def backoff(delay):
+        sleeping.set()
+        await asyncio.Event().wait()
+
+    with patch('langbot.pkg.provider.modelmgr.request_policy.asyncio.sleep', new=backoff):
+        task = asyncio.create_task(
+            runtime_provider.invoke_llm(
+                None,
+                runtime_llm_model,
+                [],
+                extra_args={'request_policy': {'retries': 3}},
+                execution_context=TEST_EXECUTION_CONTEXT,
+            )
+        )
+        await sleeping.wait()
+        task.cancel()
+        with pytest.raises(asyncio.CancelledError):
+            await task
+    assert runtime_provider.requester.invoke_llm.await_count == 1
+
+
+@pytest.mark.asyncio
+async def test_legacy_none_chunk_is_ignored(runtime_provider, runtime_llm_model):
+    async def stream(**kwargs):
+        yield None
+        yield MessageChunk(role='assistant', content='ok')
+
+    runtime_provider.requester.invoke_llm_stream = stream
+    chunks = [
+        chunk
+        async for chunk in runtime_provider.invoke_llm_stream(
+            None, runtime_llm_model, [], execution_context=TEST_EXECUTION_CONTEXT
+        )
+    ]
+    assert [chunk.content for chunk in chunks] == ['ok']
diff --git a/tests/unit_tests/provider/test_requester_base.py b/tests/unit_tests/provider/test_requester_base.py
index 4ed1335fb..017c9e705 100644
--- a/tests/unit_tests/provider/test_requester_base.py
+++ b/tests/unit_tests/provider/test_requester_base.py
@@ -563,7 +563,7 @@ def test_requester_error_message_format():
     """Test RequesterError message format."""
     error = RequesterError('API returned 500')
 
-    assert '模型请求失败' in str(error)
+    assert 'Model request failed' in str(error)
     assert 'API returned 500' in str(error)
 
 
diff --git a/tests/unit_tests/provider/test_session_manager.py b/tests/unit_tests/provider/test_session_manager.py
index 469743c5b..456a3a9c9 100644
--- a/tests/unit_tests/provider/test_session_manager.py
+++ b/tests/unit_tests/provider/test_session_manager.py
@@ -11,7 +11,7 @@ from __future__ import annotations
 import pytest
 import asyncio
 from types import SimpleNamespace
-from unittest.mock import Mock
+from unittest.mock import Mock, AsyncMock
 from importlib import import_module
 
 import langbot_plugin.api.entities.builtin.provider.session as provider_session
@@ -26,6 +26,13 @@ from langbot.pkg.pipeline.pool import (
 )
 
 
+@pytest.fixture(autouse=True)
+def reset_generation(monkeypatch):
+    lookup = AsyncMock(return_value=None)
+    monkeypatch.setattr('langbot.pkg.agent.runner.context_reset.get_reset_generation', lookup)
+    return lookup
+
+
 TEST_CONTEXT = ExecutionContext(
     instance_uuid='instance-test',
     workspace_uuid='workspace-test',
@@ -447,7 +454,7 @@ class TestSessionManagerWorkspaceIsolation:
             sessions.append(await manager.get_session(query))
 
         assert len(manager.session_list) == 2
-        assert sessions[0] not in manager.session_list
+        assert all(session is not sessions[0] for session in manager.session_list)
         assert sessions[1:] == manager.session_list
         assert await manager.get_session(queries[-1]) is manager.session_list[-1]
 
@@ -556,3 +563,18 @@ class TestSessionManagerWorkspaceIsolation:
         content = conversation.messages[0].content
         assert content[1].image_base64 is None
         assert content[2].file_base64 is None
+
+
+@pytest.mark.asyncio
+async def test_reset_generation_replaces_cached_session(reset_generation):
+    app = Mock()
+    app.instance_config.data = {'concurrency': {'session': 5}}
+    manager = get_session_module().SessionManager(app)
+    query = scoped_query()
+    previous = await manager.get_session(query)
+    assert await manager.get_session(query) is previous
+    reset_generation.return_value = 'new-context'
+    fresh = await manager.get_session(query)
+    assert fresh is not previous
+    assert not fresh.conversations
+    assert await manager.get_session(query) is fresh
diff --git a/tests/unit_tests/provider/test_skill_tools.py b/tests/unit_tests/provider/test_skill_tools.py
index 7396c00ed..2c2f49c24 100644
--- a/tests/unit_tests/provider/test_skill_tools.py
+++ b/tests/unit_tests/provider/test_skill_tools.py
@@ -135,49 +135,16 @@ class TestSkillPathHelpers:
 
         assert list(result.keys()) == ['visible']
 
-    @pytest.mark.asyncio
-    async def test_restore_activated_skills_uses_exact_revision_bindings(self):
-        from langbot.pkg.provider.tools.loaders.skill import (
-            ACTIVATED_SKILLS_KEY,
-            PIPELINE_BOUND_SKILLS_KEY,
-            get_activated_skill_bindings,
-            get_activated_skill_names,
-            restore_activated_skills,
-        )
+    def test_activation_keeps_first_revision_and_rejects_unversioned_data(self):
+        from langbot.pkg.provider.tools.loaders.skill import register_activated_skill
 
-        ap = _make_ap()
-        pinned = _make_skill_data(name='visible', revision='sha256:' + '2' * 64)
-        ap.skill_repository = SimpleNamespace(
-            get_skill=AsyncMock(return_value=pinned),
-        )
-        query = _make_query(variables={PIPELINE_BOUND_SKILLS_KEY: ['visible']})
-
-        restored = await restore_activated_skills(
-            ap,
-            query,
-            [{'name': 'visible', 'revision': pinned['revision']}],
-        )
-
-        assert restored == ['visible']
-        assert list(query.variables[ACTIVATED_SKILLS_KEY].keys()) == ['visible']
-        assert get_activated_skill_names(query) == ['visible']
-        assert get_activated_skill_bindings(query) == [{'name': 'visible', 'revision': pinned['revision']}]
-        ap.skill_repository.get_skill.assert_awaited_once_with(
-            _CONTEXT,
-            'visible',
-            snapshot=True,
-            revision=pinned['revision'],
-        )
-
-    @pytest.mark.asyncio
-    async def test_restore_rejects_name_only_state(self):
-        from langbot.pkg.provider.tools.loaders.skill import restore_activated_skills
-
-        ap = _make_ap()
-        ap.skill_repository = SimpleNamespace(get_skill=AsyncMock())
-
-        with pytest.raises(ValueError, match='cannot use names without revisions'):
-            await restore_activated_skills(ap, _make_query(), ['visible'])
+        query = _make_query()
+        first = _make_skill_data(name='visible', revision='sha256:' + '1' * 64)
+        second = _make_skill_data(name='visible', revision='sha256:' + '2' * 64)
+        assert register_activated_skill(query, first)['revision'] == first['revision']
+        assert register_activated_skill(query, second)['revision'] == first['revision']
+        with pytest.raises(ValueError, match='immutable revision'):
+            register_activated_skill(query, {'name': 'unversioned'})
 
     def test_resolve_virtual_skill_path_allows_visible_skill_reads(self):
         from langbot.pkg.provider.tools.loaders.skill import (
@@ -785,8 +752,8 @@ class TestNativeToolLoaderSkillPaths:
 
             assert result['ok'] is True
             tool_parameters = ap.box_service.execute_tool.await_args.args[0]
-            assert tool_parameters['command'] == 'python /workspace/.skills/demo/scripts/run.py'
-            assert tool_parameters['workdir'] == '/workspace/.skills/demo'
+            assert tool_parameters['command'] == f'python /opt/langbot-artifacts/{"1" * 64}/scripts/run.py'
+            assert tool_parameters['workdir'] == '/opt/langbot-artifacts/' + '1' * 64
             assert 'skill_name' not in ap.box_service.execute_tool.await_args.kwargs
             ap.skill_mgr.refresh_skill_from_disk.assert_not_called()
 
@@ -832,7 +799,7 @@ class TestNativeToolLoaderSkillPaths:
         tool_parameters = ap.box_service.execute_tool.await_args.args[0]
         wrapped = tool_parameters['command']
         assert f'_LB_VENV_DIR="/workspace/.skill-envs/demo/{"1" * 64}/.venv"' in wrapped
-        assert 'root = "/workspace/.skills/demo"' in wrapped
+        assert f'root = "/opt/langbot-artifacts/{"1" * 64}"' in wrapped
         assert tmpdir not in wrapped
         assert 'skill_name' not in ap.box_service.execute_tool.await_args.kwargs
 
@@ -898,3 +865,52 @@ async def test_native_skill_tools_require_runner_box_binding():
     with pytest.raises(BoxValidationError, match='Runner must bind a Box'):
         await loader.invoke_tool('exec', {'command': 'true'}, _make_query())
     ap.box_service.execute_tool.assert_not_awaited()
+
+
+@pytest.mark.asyncio
+async def test_register_skill_hidden_when_cloud_scanning_is_disabled():
+    from langbot.pkg.provider.tools.loaders.skill_authoring import SkillToolLoader
+
+    ap = _make_ap()
+    ap.skill_repository = SimpleNamespace()
+    ap.skill_mgr = _make_skill_manager({'demo': _make_skill_data(name='demo')})
+    ap.box_service = SimpleNamespace(
+        available=True,
+        managed_admission_required=True,
+        get_backend_status=AsyncMock(return_value={'backend': {'available': True}}),
+    )
+
+    loader = SkillToolLoader(ap)
+    await loader.initialize()
+
+    assert {tool.name for tool in await loader.get_tools(sandbox_available=True)} == {
+        'activate',
+        'list_skill_resources',
+        'read_skill_resource',
+    }
+    assert await loader.has_tool('register_skill', sandbox_available=True) is False
+    assert await loader.has_tool('activate') is True
+
+
+@pytest.mark.asyncio
+async def test_catalog_refresh_does_not_hide_skills_from_another_run():
+    import asyncio
+    from langbot.pkg.skill.manager import SkillManager
+
+    app = _make_ap()
+    entered, release = asyncio.Event(), asyncio.Event()
+
+    async def read_catalog(context):
+        entered.set()
+        await release.wait()
+        return [_make_skill_data(name='new')]
+
+    app.skill_repository = SimpleNamespace(list_skills=read_catalog)
+    manager = SkillManager(app)
+    manager._skills_by_scope[manager._scope_key(_CONTEXT)] = {'old': _make_skill_data(name='old')}
+    refresh = asyncio.create_task(manager.reload_skills(_CONTEXT))
+    await entered.wait()
+    assert set(manager.get_skills(_CONTEXT)) == {'old'}
+    release.set()
+    await refresh
+    assert set(manager.get_skills(_CONTEXT)) == {'new'}
diff --git a/tests/unit_tests/provider/test_tool_manager.py b/tests/unit_tests/provider/test_tool_manager.py
index d404400a4..0e81ae5e2 100644
--- a/tests/unit_tests/provider/test_tool_manager.py
+++ b/tests/unit_tests/provider/test_tool_manager.py
@@ -7,6 +7,8 @@ Tests cover:
 
 from __future__ import annotations
 
+import asyncio
+import contextvars
 from importlib import import_module
 from types import SimpleNamespace
 from unittest.mock import AsyncMock, Mock
@@ -380,6 +382,127 @@ class TestToolManagerExecuteFuncCall:
         assert 'MCP resource agent reads are disabled' in record.await_args.kwargs['error_message']
 
 
+class _RecordingTelemetrySender:
+    """Telemetry sender stand-in that keeps every payload it was handed."""
+
+    def __init__(self):
+        self.telemetry_config = {'url': 'https://space.example.test', 'execution_trace': 'all'}
+        self.sent: list[dict] = []
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+
+class TestToolCallTracing:
+    """Every tool call a Run makes is its own step of the execution chain."""
+
+    @pytest.fixture
+    def traced_app(self):
+        execution = import_module('langbot.pkg.telemetry.execution')
+        counters = execution.ExecutionCounters(_RecordingTelemetrySender())
+        app = Mock()
+        app.logger = Mock()
+        app.telemetry = SimpleNamespace(execution=counters)
+        # Tool-call monitoring is asserted by the tests above; this path only
+        # needs the execution chain.
+        app.monitoring_service = None
+
+        def _inert_loader():
+            loader = Mock()
+            loader.has_tool = AsyncMock(return_value=False)
+            loader.invoke_tool = AsyncMock(return_value=None)
+            return loader
+
+        app._native_loader = _inert_loader()
+        app._plugin_loader = _inert_loader()
+        app._mcp_loader = _inert_loader()
+        app._skill_loader = _inert_loader()
+        return app, counters
+
+    @pytest.fixture
+    def sample_query(self):
+        query = Mock(spec=pipeline_query.Query)
+        query._execution_context = _CONTEXT
+        return query
+
+    @staticmethod
+    def _manager(app):
+        """ToolManager wired to app-local inert loaders."""
+        manager = get_toolmgr_module().ToolManager(app)
+        manager.native_tool_loader = app._native_loader
+        manager.plugin_tool_loader = app._plugin_loader
+        manager.mcp_tool_loader = app._mcp_loader
+        manager.skill_tool_loader = app._skill_loader
+        return manager
+
+    @staticmethod
+    def _tool_nodes(counters):
+        return [record for record in counters.records if record['event_type'] == 'execution_node']
+
+    @pytest.mark.asyncio
+    async def test_a_tool_call_becomes_a_node_under_the_step_that_issued_it(self, traced_app, sample_query):
+        trace = import_module('langbot.pkg.telemetry.trace')
+        execution = import_module('langbot.pkg.telemetry.execution')
+        app, counters = traced_app
+        manager = self._manager(app)
+
+        binding = execution.bind_trace(app, 'exec-tool')
+        # The Runtime's RPC task executes the call with no inherited trace
+        # context; the Run id on the query is what resolves the owning chain.
+        sample_query._agent_run_session = {'run_id': 'exec-tool'}
+        try:
+            with trace.stage_scope() as lane:
+                foreign = contextvars.Context()
+                task = asyncio.get_running_loop().create_task(
+                    manager._invoke_tool_with_monitoring(
+                        source='native',
+                        name='exec',
+                        parameters={'command': 'ls'},
+                        query=sample_query,
+                        invoke=AsyncMock(return_value='sandbox output'),
+                    ),
+                    context=foreign,
+                )
+                assert await task == 'sandbox output'
+        finally:
+            trace.unbind_root(binding)
+
+        rows = self._tool_nodes(counters)
+        assert len(rows) == 1
+        assert rows[0]['family'] == 'tool'
+        assert rows[0]['operation'] == 'exec'
+        assert rows[0]['adapter'] == 'native'
+        assert rows[0]['outcome'] == 'success'
+        assert rows[0]['parent_node_id'] == lane
+
+    @pytest.mark.asyncio
+    async def test_a_failed_tool_call_carries_its_reason(self, traced_app, sample_query):
+        execution = import_module('langbot.pkg.telemetry.execution')
+        trace = import_module('langbot.pkg.telemetry.trace')
+        app, counters = traced_app
+        manager = self._manager(app)
+
+        binding = execution.bind_trace(app, 'exec-tool-failed')
+        try:
+            with pytest.raises(RuntimeError):
+                await manager._invoke_tool_with_monitoring(
+                    source='mcp',
+                    name='read_resource',
+                    parameters={},
+                    query=sample_query,
+                    invoke=AsyncMock(side_effect=RuntimeError('server unreachable')),
+                )
+        finally:
+            trace.unbind_root(binding)
+
+        rows = self._tool_nodes(counters)
+        assert len(rows) == 1
+        assert rows[0]['family'] == 'tool'
+        assert rows[0]['outcome'] == 'failed'
+        assert rows[0]['error'] == 'server unreachable'
+
+
 class TestToolManagerSourceResolution:
     @pytest.mark.asyncio
     @pytest.mark.parametrize('source', ['mcp', 'plugin'])
diff --git a/tests/unit_tests/provider/test_tool_manager_native.py b/tests/unit_tests/provider/test_tool_manager_native.py
index b3b06dd08..767e1d7b6 100644
--- a/tests/unit_tests/provider/test_tool_manager_native.py
+++ b/tests/unit_tests/provider/test_tool_manager_native.py
@@ -862,3 +862,35 @@ async def test_box_grep_script_serializes_optional_include_as_python(monkeypatch
             values[statement.targets[0].id] = ast.literal_eval(statement.value)
     assert values['include'] == include
     assert values['path'] == '/workspace'
+
+
+@pytest.mark.asyncio
+async def test_workspace_file_script_resolves_the_sandbox_interpreter():
+    """Sandboxes built from a host rootfs may only ship `python3`."""
+
+    captured: dict = {}
+
+    async def execute_tool(payload, query):
+        captured.update(payload)
+        return {'ok': True, 'stdout': '{"ok": true}'}
+
+    box_service = SimpleNamespace(
+        available=True,
+        execute_tool=execute_tool,
+        require_workspace_sandbox=AsyncMock(),
+    )
+    loader = NativeToolLoader(SimpleNamespace(box_service=box_service, logger=Mock()))
+    query = SimpleNamespace(
+        _execution_context=_CONTEXT,
+        bot_uuid=None,
+        pipeline_uuid=None,
+        query_uuid=None,
+    )
+    query._box_binding = RunBoxBinding('run', 'box', {}, 'run')
+
+    result = await loader._run_workspace_file_script('print("x")', query)
+
+    assert result == {'ok': True}
+    command = captured['command']
+    assert command.startswith('PYTHON_BIN=$(command -v python3 || command -v python)')
+    assert command.endswith(' - <<\'PY\'\nprint("x")\nPY')
diff --git a/tests/unit_tests/telemetry/test_adapter_acceptance.py b/tests/unit_tests/telemetry/test_adapter_acceptance.py
deleted file mode 100644
index ceffe2288..000000000
--- a/tests/unit_tests/telemetry/test_adapter_acceptance.py
+++ /dev/null
@@ -1,208 +0,0 @@
-"""Adapter acceptance boundaries and strict Beta-only production."""
-
-import json
-from types import SimpleNamespace
-from uuid import uuid4
-
-import pytest
-from langbot_plugin.api.entities.builtin.platform.events import MessageReceivedEvent
-from langbot_plugin.api.entities.builtin.platform.message import MessageChain, Plain, Image
-
-from langbot.pkg.telemetry import diagnostics as d
-from langbot.pkg.telemetry import adapter_diagnostics as adapter
-from langbot.pkg.telemetry.diagnostic_catalog import snapshot_bot, catalog
-
-
-def make_manager(version='4.11.0b2', **config):
-    ap = SimpleNamespace(instance_config=SimpleNamespace(data={'space': {'url': 'https://example.invalid', **config}}))
-    ap.diagnostics = d.DiagnosticsManager(ap, version=version, instance_id='instance-test', capacity=2048)
-    return ap.diagnostics
-
-
-def boundary(fn, operation='send_message', kind='api'):
-    fn.__module__ = 'langbot.pkg.platform.adapters.telegram.adapter'
-    return d.observe(kind, operation, source='platform', stage='convert' if kind == 'event' else 'accepted')(fn)
-
-
-def evidence(m):
-    return [e for e in m.pending if e['attributes'].get('adapter_evidence') and e['outcome'] != 'started']
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'version,config',
-    [
-        ('4.11.0', {}),
-        ('4.11.0a1', {}),
-        ('4.11.0rc1', {}),
-        ('4.11.0.dev1', {}),
-        ('4.11.0b2+local', {}),
-        ('invalid', {}),
-        ('4.11.0b2', {'disable_telemetry': True}),
-        ('4.11.0b2', {'disable_beta_diagnostics': True}),
-    ],
-)
-async def test_disabled_adapter_produces_nothing(version, config, monkeypatch):
-    m = make_manager(version, **config)
-
-    def forbidden(*args, **kwargs):
-        raise AssertionError('disabled producer projected adapter metadata')
-
-    monkeypatch.setattr(adapter, 'boundary_fields', forbidden)
-
-    async def call(self):
-        return 42
-
-    assert await boundary(call)(SimpleNamespace(ap=m.ap)) == 42
-    snapshot_bot(SimpleNamespace(ap=m.ap), listener_registered=True)
-    d.adapter_event_received(SimpleNamespace(ap=m.ap), MessageReceivedEvent())
-    assert not m.pending
-    assert m.counters['generated'] == 0
-    await m.flush_once()
-    assert m.counters['acked'] == 0
-
-
-@pytest.mark.asyncio
-async def test_nested_api_counted_once_and_only_finite_scenario():
-    m = make_manager()
-    owner = SimpleNamespace(ap=m.ap)
-
-    async def leaf(self, message, target_type):
-        return {'ok': True, 'private_result': 'SECRET_CANARY'}
-
-    third = boundary(leaf)
-
-    async def middle(self, message, target_type):
-        return await third(self, message, target_type)
-
-    second = boundary(middle)
-
-    async def outer(self, message, target_type):
-        return await second(self, message, target_type)
-
-    await boundary(outer)(
-        owner, MessageChain([Plain(text='SECRET_CANARY'), Image(url='https://secret.invalid')]), 'person'
-    )
-    rows = evidence(m)
-    assert len(rows) == 1
-    assert rows[0]['attributes'] == {'adapter_evidence': True, 'chat_type': 'person', 'content_type': 'mixed'}
-    assert len(m.pending) == 6
-    assert 'SECRET_CANARY' not in json.dumps(list(m.pending))
-    assert '_adapter_api_active' not in json.dumps(list(m.pending))
-
-
-@pytest.mark.asyncio
-async def test_api_specific_interaction_unknown_and_failure():
-    m = make_manager()
-
-    async def call(self, action, params):
-        return {'ok': False, 'description': 'SECRET_CANARY'}
-
-    call = boundary(call, 'call_platform_api')
-    actions = catalog()['telegram']['specific_apis']
-    assert actions
-    action = actions[0].removeprefix('platform_api.')
-    for name in (action, 'interaction.request', 'SECRET_CANARY'):
-        await call(SimpleNamespace(ap=m.ap), name, {'token': 'SECRET_CANARY'})
-    rows = evidence(m)
-    assert [r['operation'] for r in rows] == [actions[0], 'interaction.request']
-    assert all(r['outcome'] == 'failed' for r in rows)
-    assert 'SECRET_CANARY' not in json.dumps(list(m.pending))
-
-
-@pytest.mark.asyncio
-async def test_conversion_count_and_private_message_scenario():
-    m = make_manager()
-
-    async def convert(self, event):
-        return event
-
-    call = boundary(convert, 'platform.target2yiri', 'event')
-    event = await call(
-        SimpleNamespace(ap=m.ap), MessageReceivedEvent(message_chain=MessageChain([Plain(text='SECRET_CANARY')]))
-    )
-    assert not evidence(m)
-    d.adapter_event_received(SimpleNamespace(ap=m.ap), event)
-    await call(SimpleNamespace(ap=m.ap), None)
-    await call(SimpleNamespace(ap=m.ap), {'legacy': 'SECRET_CANARY'})
-    rows = evidence(m)
-    assert len(rows) == 1
-    assert rows[0]['platform_event_type'] == 'message.received'
-    assert rows[0]['attributes']['chat_type'] == 'person'
-    assert rows[0]['attributes']['content_type'] == 'text'
-    assert 'SECRET_CANARY' not in json.dumps(list(m.pending))
-
-
-def test_registered_snapshot_includes_specific_apis_without_claiming_connection():
-    m = make_manager()
-    entry = catalog()['telegram']
-    cls = type(
-        'Adapter',
-        (),
-        {
-            '__module__': 'langbot.pkg.platform.adapters.telegram.adapter',
-            'get_supported_events': lambda self: entry['events'],
-            'get_supported_apis': lambda self: entry['apis'],
-        },
-    )
-    snapshot_bot(
-        SimpleNamespace(ap=m.ap, adapter=cls(), execution_context=SimpleNamespace(workspace_uuid=str(uuid4()))),
-        listener_registered=True,
-    )
-    rows = list(m.pending)
-    assert rows
-    assert all('available' not in r['attributes'] for r in rows)
-    assert all(r['attributes']['listener_registered'] for r in rows if r['attributes']['capability_type'] == 'event')
-    names = {r['attributes']['capability_name'] for r in rows}
-    assert set(entry['specific_apis']) <= names
-
-
-@pytest.mark.asyncio
-async def test_real_dispatch_success_precedes_listener_failure():
-    from unittest.mock import AsyncMock
-    from langbot.pkg.platform.adapters.telegram.adapter import TelegramAdapter
-    from langbot_plugin.api.entities.builtin.platform.events import EBAEvent
-
-    m = make_manager()
-    owner = SimpleNamespace(ap=m.ap, listeners={EBAEvent: AsyncMock(side_effect=ValueError('SECRET_CANARY'))})
-    with pytest.raises(ValueError):
-        await TelegramAdapter._dispatch_eba_event(owner, MessageReceivedEvent())
-    rows = evidence(m)
-    assert len(rows) == 1 and rows[0]['outcome'] == 'succeeded'
-    assert rows[0]['operation'] == 'platform.adapter_event'
-
-
-def test_event_dispatch_does_not_borrow_another_workspace_trace():
-    from langbot.pkg.api.http.context import ExecutionContext
-
-    m = make_manager()
-    workspace = str(uuid4())
-    owner = SimpleNamespace(
-        ap=m.ap,
-        execution_context=ExecutionContext(
-            instance_uuid=m.instance_id, workspace_uuid=workspace, placement_generation=1
-        ),
-    )
-    with d.Span(m, 'event', 'platform.receive', {'workspace_uuid': str(uuid4()), 'source': 'webui_debug'}).activate():
-        d.adapter_event_received(owner, MessageReceivedEvent())
-    row = evidence(m)[0]
-    assert row['workspace_uuid'] == workspace
-    assert not row.get('parent_span_id')
-    assert row['source'] == 'platform'
-    assert not row['attributes'].get('synthetic')
-
-
-def test_every_adapter_records_native_and_interaction_dispatch():
-    import ast
-    from pathlib import Path
-
-    base = Path(__file__).resolve().parents[3] / 'src/langbot/pkg/platform/adapters'
-    for directory in catalog():
-        tree = ast.parse((base / directory / 'adapter.py').read_text(encoding='utf-8'))
-        dispatch = next(
-            n for n in ast.walk(tree) if isinstance(n, ast.AsyncFunctionDef) and n.name == '_dispatch_eba_event'
-        )
-        assert any(
-            isinstance(n, ast.Call) and ast.unparse(n.func) == 'diagnostics.adapter_event_received'
-            for n in ast.walk(dispatch)
-        ), directory
diff --git a/tests/unit_tests/telemetry/test_adapter_confirmation.py b/tests/unit_tests/telemetry/test_adapter_confirmation.py
deleted file mode 100644
index 48307db67..000000000
--- a/tests/unit_tests/telemetry/test_adapter_confirmation.py
+++ /dev/null
@@ -1,318 +0,0 @@
-"""Real adapter acceptance must reflect confirmed operations, not normal returns."""
-
-import base64
-import json
-from types import SimpleNamespace
-from unittest.mock import AsyncMock
-from uuid import uuid4
-
-import httpx
-import pytest
-from langbot_plugin.api.entities.builtin.platform.message import File, MessageChain, Plain
-
-from langbot.pkg.api.http.context import ExecutionContext
-from langbot.pkg.platform.adapters.telegram.adapter import TelegramAdapter
-from langbot.pkg.telemetry import diagnostics as d
-
-
-def make_adapter(version='4.11.0b3', **config):
-    ap = SimpleNamespace(instance_config=SimpleNamespace(data={'space': {'url': 'https://example.invalid', **config}}))
-    manager = d.DiagnosticsManager(ap, version=version, instance_id='instance-test')
-    ap.diagnostics = manager
-    context = ExecutionContext(instance_uuid=manager.instance_id, workspace_uuid=str(uuid4()), placement_generation=1)
-    sdk = SimpleNamespace(edit_message_text=AsyncMock(return_value=True), send_message=AsyncMock(return_value=True))
-    adapter = TelegramAdapter.model_construct(
-        bot=sdk, config={'markdown_card': False}, logger=SimpleNamespace(ap=ap, execution_context=context), listeners={}
-    )
-    return adapter, manager, sdk
-
-
-async def wire_events(manager):
-    batches = []
-
-    async def sender(request):
-        batch = json.loads(request.content)
-        batches.append(batch)
-        return httpx.Response(
-            200,
-            json={
-                'code': 200,
-                'data': {'accepted_event_ids': [e['event_id'] for e in batch['events']], 'rejected': []},
-            },
-        )
-
-    manager.client = httpx.AsyncClient(transport=httpx.MockTransport(sender))
-    await manager.flush_once()
-    assert not manager.pending
-    await manager.shutdown(drain_timeout=0)
-    assert 'PRIVATE_' not in json.dumps(batches)
-    return [event for batch in batches for event in batch['events']]
-
-
-def successes(events):
-    return [e for e in events if e['attributes'].get('adapter_evidence') and e['outcome'] == 'succeeded']
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_file_only_edit_is_not_acceptance_through_sender():
-    adapter, manager, sdk = make_adapter()
-    for _ in range(2):
-        result = await adapter.edit_message(
-            'person',
-            'PRIVATE_CHAT',
-            'PRIVATE_MESSAGE',
-            MessageChain([File(name='PRIVATE_FILE', base64=base64.b64encode(b'PRIVATE_BYTES').decode())]),
-        )
-        assert result is None
-    sdk.edit_message_text.assert_not_awaited()
-    events = await wire_events(manager)
-    assert len([e for e in events if e['operation'] == 'edit_message']) == 4
-    assert not successes(events)
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_acknowledged_void_edit_is_acceptance_through_sender():
-    adapter, manager, sdk = make_adapter()
-    assert (
-        await adapter.edit_message(
-            'person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE', MessageChain([Plain(text='PRIVATE_TEXT')])
-        )
-        is None
-    )
-    sdk.edit_message_text.assert_awaited_once_with(
-        chat_id='PRIVATE_CHAT', message_id='PRIVATE_MESSAGE', text='PRIVATE_TEXT'
-    )
-    events = await wire_events(manager)
-    assert len(successes(events)) == 1
-    assert successes(events)[0]['attributes']['content_type'] == 'text'
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_empty_send_is_not_acceptance():
-    adapter, manager, sdk = make_adapter()
-    assert await adapter.send_message('person', 'PRIVATE_CHAT', MessageChain([])) is None
-    sdk.send_message.assert_not_awaited()
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_edit_exception_is_unchanged():
-    adapter, manager, sdk = make_adapter()
-    error = ValueError('PRIVATE_ERROR')
-    sdk.edit_message_text.side_effect = error
-    with pytest.raises(ValueError) as raised:
-        await adapter.edit_message(
-            'person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE', MessageChain([Plain(text='PRIVATE_TEXT')])
-        )
-    assert raised.value is error
-    events = await wire_events(manager)
-    assert not successes(events)
-    assert any(e['outcome'] == 'failed' and e['operation'] == 'edit_message' for e in events)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'version,config',
-    [('4.11.0', {}), ('4.11.0b3', {'disable_telemetry': True}), ('4.11.0b3', {'disable_beta_diagnostics': True})],
-)
-async def test_real_adapter_disabled_gates_keep_business_behavior(version, config):
-    adapter, manager, sdk = make_adapter(version, **config)
-    assert (
-        await adapter.edit_message(
-            'person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE', MessageChain([Plain(text='PRIVATE_TEXT')])
-        )
-        is None
-    )
-    sdk.edit_message_text.assert_awaited_once()
-    assert await wire_events(manager) == []
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('result', [None, False, {}, {'queued': True}, {'stream': False}])
-async def test_unconfirmed_normal_return_is_not_acceptance(result):
-    adapter, manager, _ = make_adapter()
-
-    async def call(self):
-        return result
-
-    call.__module__ = 'langbot.pkg.platform.adapters.telegram.adapter'
-    observed = d.observe('api', 'send_message', source='platform', stage='accepted')(call)
-    assert await observed(adapter) is result
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('raw', [{}, {'results': []}, {'result': None}, {'queued': True}])
-async def test_message_result_source_id_is_not_confirmation(raw):
-    from langbot_plugin.api.entities.builtin.platform.events import MessageResult
-
-    adapter, manager, _ = make_adapter()
-    result = MessageResult(message_id='PRIVATE_SOURCE_ID', raw=raw)
-
-    async def call(self):
-        return result
-
-    call.__module__ = 'langbot.pkg.platform.adapters.wecom.adapter'
-    observed = d.observe('api', 'reply_message', source='platform', stage='accepted')(call)
-    assert await observed(adapter) is result
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_discord_void_delete_is_confirmed():
-    from langbot.pkg.platform.adapters.discord.adapter import DiscordAdapter
-
-    adapter, manager, _ = make_adapter()
-    message = SimpleNamespace(delete=AsyncMock(return_value=None))
-    channel = SimpleNamespace(fetch_message=AsyncMock(return_value=message))
-    discord = DiscordAdapter.model_construct(
-        bot=SimpleNamespace(get_channel=lambda _: channel), logger=adapter.logger, config={}, listeners={}
-    )
-    assert await discord.delete_message('group', '123', '456') is None
-    message.delete.assert_awaited_once()
-    assert len(successes(await wire_events(manager))) == 1
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('result', [False, None])
-async def test_real_telegram_unacknowledged_edit_keeps_void_return(result):
-    adapter, manager, sdk = make_adapter()
-    sdk.edit_message_text.return_value = result
-    assert (
-        await adapter.edit_message(
-            'person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE', MessageChain([Plain(text='PRIVATE_TEXT')])
-        )
-        is None
-    )
-    sdk.edit_message_text.assert_awaited_once()
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_acknowledged_void_delete_is_acceptance():
-    adapter, manager, sdk = make_adapter()
-    sdk.delete_message = AsyncMock(return_value=True)
-    assert await adapter.delete_message('person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE') is None
-    sdk.delete_message.assert_awaited_once()
-    assert len(successes(await wire_events(manager))) == 1
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_sdk_message_send_is_acceptance():
-    import datetime
-    import telegram
-
-    adapter, manager, sdk = make_adapter()
-    sdk.send_message.return_value = telegram.Message(
-        message_id=123, date=datetime.datetime.now(datetime.timezone.utc), chat=telegram.Chat(id=456, type='private')
-    )
-    assert await adapter.send_message('person', 'PRIVATE_CHAT', MessageChain([Plain(text='PRIVATE_TEXT')])) is None
-    sdk.send_message.assert_awaited_once()
-    assert len(successes(await wire_events(manager))) == 1
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'raw', [{'ok': True, 'raw': None}, {'ok': True, 'raw': {'errcode': 42}}, {'results': [{'ok': True}, None]}]
-)
-async def test_wrapped_ack_does_not_hide_missing_or_failed_response(raw):
-    adapter, manager, _ = make_adapter()
-
-    async def call(self):
-        return raw
-
-    call.__module__ = 'langbot.pkg.platform.adapters.wecombot.adapter'
-    observed = d.observe('api', 'reply_message', source='platform', stage='accepted')(call)
-    assert await observed(adapter) is raw
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize(
-    'action', ['check_tenant_access_token', 'refresh_app_access_token', 'refresh_tenant_access_token']
-)
-async def test_real_lark_self_app_token_noop_is_not_acceptance(action):
-    from langbot.pkg.platform.adapters.lark.adapter import LarkAdapter
-
-    adapter, manager, _ = make_adapter()
-    lark = LarkAdapter.model_construct(config={'app_type': 'self'}, logger=adapter.logger, listeners={})
-    assert await lark.call_platform_api(action, {}) == {'ok': True}
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_mixed_edit_cannot_confirm_ignored_file():
-    adapter, manager, sdk = make_adapter()
-    message = MessageChain([Plain(text='PRIVATE_TEXT'), File(name='PRIVATE_FILE', base64='eA==')])
-    assert await adapter.edit_message('person', 'PRIVATE_CHAT', 'PRIVATE_MESSAGE', message) is None
-    sdk.edit_message_text.assert_awaited_once()
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_discord_sent_message_result_is_confirmed():
-    from langbot.pkg.platform.adapters.discord.adapter import DiscordAdapter
-
-    adapter, manager, _ = make_adapter()
-    channel = SimpleNamespace(send=AsyncMock(return_value=SimpleNamespace(id=123)))
-    discord = DiscordAdapter.model_construct(
-        bot=SimpleNamespace(get_channel=lambda _: channel), logger=adapter.logger, config={}, listeners={}
-    )
-    result = await discord.send_message('group', '456', MessageChain([Plain(text='PRIVATE_TEXT')]))
-    assert result.message_id == 123
-    channel.send.assert_awaited_once_with(content='PRIVATE_TEXT')
-    assert len(successes(await wire_events(manager))) == 1
-
-
-@pytest.mark.asyncio
-async def test_real_aiocqhttp_empty_forward_is_not_acceptance():
-    from langbot.pkg.platform.adapters.aiocqhttp.adapter import AiocqhttpAdapter
-    from langbot_plugin.api.entities.builtin.platform.message import Forward
-
-    adapter, manager, _ = make_adapter()
-    sdk = SimpleNamespace(call_action=AsyncMock())
-    onebot = AiocqhttpAdapter.model_construct(bot=sdk, logger=adapter.logger, config={}, listeners={})
-    result = await onebot.send_message('group', '123', MessageChain([Forward(node_list=[])]))
-    assert result.message_id is None and result.raw == {}
-    sdk.call_action.assert_not_awaited()
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_unsupported_exception_is_unchanged():
-    from langbot_plugin.api.entities.builtin.platform.errors import NotSupportedError
-
-    adapter, manager, _ = make_adapter()
-    with pytest.raises(NotSupportedError):
-        await adapter.upload_file(b'PRIVATE_BYTES', 'PRIVATE_FILE')
-    assert not successes(await wire_events(manager))
-
-
-@pytest.mark.asyncio
-async def test_real_lark_sent_message_result_is_confirmed():
-    from langbot.pkg.platform.adapters.lark.adapter import LarkAdapter
-
-    adapter, manager, _ = make_adapter()
-    create = AsyncMock(
-        return_value=SimpleNamespace(success=lambda: True, data=SimpleNamespace(message_id='PRIVATE_SENT_ID'))
-    )
-    lark = LarkAdapter.model_construct(
-        config={'app_type': 'self'},
-        logger=adapter.logger,
-        listeners={},
-        api_client=SimpleNamespace(im=SimpleNamespace(v1=SimpleNamespace(message=SimpleNamespace(acreate=create)))),
-    )
-    result = await lark.send_message('group', 'PRIVATE_CHAT', MessageChain([Plain(text='PRIVATE_TEXT')]))
-    assert result.message_id == 'PRIVATE_SENT_ID'
-    create.assert_awaited_once()
-    assert len(successes(await wire_events(manager))) == 1
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_returned_user_info_is_confirmed():
-    adapter, manager, sdk = make_adapter()
-    sdk.get_chat = AsyncMock(return_value=SimpleNamespace(id=123, first_name='PRIVATE_NAME', username='PRIVATE_USER'))
-    result = await adapter.get_user_info('PRIVATE_USER')
-    assert result.id == 123
-    sdk.get_chat.assert_awaited_once_with(chat_id='PRIVATE_USER')
-    assert len(successes(await wire_events(manager))) == 1
diff --git a/tests/unit_tests/telemetry/test_batch_ack.py b/tests/unit_tests/telemetry/test_batch_ack.py
new file mode 100644
index 000000000..0aeff5448
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_batch_ack.py
@@ -0,0 +1,29 @@
+from unittest.mock import Mock
+
+import httpx
+import pytest
+
+from langbot.pkg.telemetry.telemetry import TelemetryManager
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    'ack,expected',
+    [
+        ({'stored': 1, 'rejected': 0}, True),
+        ({'stored': 0, 'rejected': 1}, False),
+        ({'stored': 1, 'rejected': 1}, False),
+        ({'stored': 0, 'rejected': 0}, False),
+        ({}, False),
+        ({'stored': True, 'rejected': 0}, False),
+    ],
+)
+async def test_batch_requires_complete_ack(ack, expected):
+    app = Mock()
+    manager = TelemetryManager(app)
+    manager.telemetry_config = {'url': 'https://telemetry.test'}
+    async with httpx.AsyncClient(
+        transport=httpx.MockTransport(lambda request: httpx.Response(200, json={'code': 0, 'data': ack}))
+    ) as client:
+        manager._client = client
+        assert await manager.send({'records': [{'event_id': 'a' * 163}]}) is expected
diff --git a/tests/unit_tests/telemetry/test_beta_diagnostic_hooks.py b/tests/unit_tests/telemetry/test_beta_diagnostic_hooks.py
deleted file mode 100644
index e4ce1e93e..000000000
--- a/tests/unit_tests/telemetry/test_beta_diagnostic_hooks.py
+++ /dev/null
@@ -1,143 +0,0 @@
-"""Execute real Core boundaries with content canaries and early failures."""
-
-import json
-from types import SimpleNamespace
-from unittest.mock import AsyncMock
-from uuid import uuid4
-
-import pytest
-
-from langbot.pkg.telemetry import diagnostics as d
-from langbot.pkg.api.http.context import ExecutionContext
-from langbot.pkg.agent.runner.orchestrator import AgentRunOrchestrator
-from langbot.pkg.agent.runner.reply_stream import ReplyStreamSession, ReplyStreamRequest
-
-
-def make_ap():
-    ap = SimpleNamespace(instance_config=SimpleNamespace(data={'space': {'url': 'https://example.invalid'}}))
-    ap.persistence_mgr = SimpleNamespace(get_db_engine=lambda: None)
-    ap.diagnostics = d.DiagnosticsManager(ap, version='4.11.0b2', instance_id='instance-test')
-    return ap
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('processor', ['pipeline', 'agent', 'event_processor'])
-async def test_real_orchestrator_prepare_failure(processor):
-    ap = make_ap()
-    registry = SimpleNamespace(get=AsyncMock(side_effect=ValueError('CANARY private runner URL')))
-    orchestrator = AgentRunOrchestrator(ap, registry)
-    context = ExecutionContext(instance_uuid='instance-test', workspace_uuid=str(uuid4()), placement_generation=1)
-    event = SimpleNamespace(workspace_id=context.workspace_uuid, event_type='message.received')
-    binding = SimpleNamespace(runner_id='CANARY', processor_type=processor)
-    with pytest.raises(ValueError, match='CANARY'):
-        await anext(orchestrator.run(event, binding, adapter_context={'_execution_context': context}))
-    records = ap.diagnostics.pending
-    assert [r['outcome'] for r in records] == ['started', 'failed']
-    assert records[-1]['stage'] == 'prepare'
-    assert records[-1]['workspace_uuid'] == context.workspace_uuid
-    assert records[-1]['processor_type'] == processor
-    assert 'CANARY' not in json.dumps(records)
-
-
-@pytest.mark.asyncio
-async def test_real_reply_stream_mock_is_not_platform_success():
-    ap = make_ap()
-    event = SimpleNamespace(
-        delivery=SimpleNamespace(reply_target={}, surface='webui', platform_capabilities={'debug_mock': True})
-    )
-    session = ReplyStreamSession(event)
-    # Real runtime supplies the manager at construction from the orchestrator.
-    session.diagnostics = ap.diagnostics
-    result = await session.apply(ReplyStreamRequest(stream_id=uuid4(), operation='finish', text='CANARY user reply'))
-    assert result['mock'] is True and result['text'] == 'CANARY user reply'
-    assert ap.diagnostics.pending[-1]['source'] == 'webui_debug'
-    assert ap.diagnostics.pending[-1]['attributes']['synthetic'] is True
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
-@pytest.mark.asyncio
-async def test_real_bot_route_projects_status_not_reason():
-    from langbot.pkg.platform.botmgr import RuntimeBot
-
-    ap = make_ap()
-    bot = object.__new__(RuntimeBot)
-    bot.ap = ap
-    bot.logger = SimpleNamespace(info=AsyncMock())
-    await bot._record_event_route_trace(
-        event_type='message.received',
-        status='not_matched',
-        text='CANARY secret',
-        reason='CANARY',
-        failure_code='route_not_found',
-    )
-    records = ap.diagnostics.pending
-    assert records[-1]['outcome'] == 'skipped'
-    assert records[-1]['reason_code'] == 'route_not_found'
-    assert 'CANARY' not in json.dumps(records)
-
-
-@pytest.mark.asyncio
-async def test_real_telegram_conversion_failure_before_bot_manager(monkeypatch):
-    from langbot.pkg.platform.adapters.telegram.adapter import TelegramAdapter
-    from langbot.pkg.platform.adapters.telegram.event_converter import TelegramEventConverter
-
-    ap = make_ap()
-    context = ExecutionContext(instance_uuid='instance-test', workspace_uuid=str(uuid4()), placement_generation=1)
-    from langbot.pkg.platform.logger import EventLogger
-
-    logger = EventLogger('test', ap, context, 'test')
-    logger.error = AsyncMock()
-    logger.warning = AsyncMock()
-    adapter = TelegramAdapter({'token': '123456:ABCDEFGHIJKLMNOPQRSTUVWXYZ_123456789'}, logger)
-    adapter.listeners = {}
-    monkeypatch.setattr(
-        TelegramEventConverter, '_convert_message', AsyncMock(side_effect=ValueError('CANARY conversion'))
-    )
-    update = SimpleNamespace(
-        message=SimpleNamespace(from_user=SimpleNamespace(is_bot=False), text='CANARY text'),
-        edited_message=None,
-        chat_member=None,
-        my_chat_member=None,
-        callback_query=None,
-        message_reaction=None,
-    )
-    callback = adapter.application.handlers[0][0].callback
-    await callback(update, None)
-    records = [e for e in ap.diagnostics.pending if e['stage'] == 'convert' and e['outcome'] == 'failed']
-    assert records and records[-1]['adapter'] == 'telegram-omni'
-    assert records[-1]['workspace_uuid'] == context.workspace_uuid
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
-@pytest.mark.asyncio
-async def test_real_interaction_ack_skip_and_failure():
-    from langbot.pkg.agent.runner.interaction_manager import InteractionManager
-
-    ap = make_ap()
-    interactions = InteractionManager(ap, store=SimpleNamespace(record_delivery_success=AsyncMock()))
-    await interactions.acknowledge_submission({}, SimpleNamespace(get_supported_apis=lambda: []))
-    assert ap.diagnostics.pending[-1]['outcome'] == 'skipped'
-    ap.diagnostics.pending.clear()
-    ap.logger = SimpleNamespace(warning=lambda *a: None)
-    adapter = SimpleNamespace(
-        get_supported_apis=lambda: ['interaction.acknowledge'],
-        call_platform_api=AsyncMock(side_effect=ValueError('CANARY ack')),
-    )
-    await interactions.acknowledge_submission({'delivery_result': {'secret': 'CANARY'}}, adapter)
-    assert ap.diagnostics.pending[-1]['outcome'] == 'failed'
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
-def test_actual_adapter_capability_snapshot():
-    from langbot.pkg.platform.adapters.telegram.adapter import TelegramAdapter
-    from langbot.pkg.telemetry.diagnostic_catalog import snapshot_bot
-
-    ap = make_ap()
-    context = ExecutionContext(instance_uuid='instance-test', workspace_uuid=str(uuid4()), placement_generation=1)
-    adapter = TelegramAdapter.model_construct(config={}, listeners={})
-    snapshot_bot(SimpleNamespace(ap=ap, adapter=adapter, execution_context=context))
-    records = ap.diagnostics.pending
-    assert records and all(e['kind'] == 'capability' for e in records)
-    api_rows = [e for e in records if e['attributes']['capability_type'] == 'api']
-    assert any(e['operation'] == 'send_message' and e['attributes']['supported'] for e in api_rows)
-    assert all(e['workspace_uuid'] == context.workspace_uuid for e in records)
diff --git a/tests/unit_tests/telemetry/test_beta_diagnostic_review_regressions.py b/tests/unit_tests/telemetry/test_beta_diagnostic_review_regressions.py
deleted file mode 100644
index 67deb287b..000000000
--- a/tests/unit_tests/telemetry/test_beta_diagnostic_review_regressions.py
+++ /dev/null
@@ -1,405 +0,0 @@
-"""Independent-review regressions through real ownership and lifecycle boundaries."""
-
-import asyncio
-import importlib.metadata
-import json
-from collections.abc import AsyncGenerator
-from types import SimpleNamespace as NS
-from unittest.mock import AsyncMock, Mock
-from uuid import uuid4
-
-import pytest
-
-from langbot.pkg.api.http.context import ExecutionContext
-from langbot.pkg.provider.modelmgr.requester import RuntimeProvider
-from langbot.pkg.telemetry import diagnostics as d
-
-
-def make_ap(identity='instance-test', disabled=False):
-    ap = NS(instance_config=NS(data={'space': {'url': 'https://example.invalid', 'disable_telemetry': disabled}}))
-    ap.diagnostics = d.DiagnosticsManager(ap, version='4.11.0b2', instance_id=identity)
-    return ap
-
-
-def context(identity='instance-test'):
-    return ExecutionContext(instance_uuid=identity, workspace_uuid=str(uuid4()), placement_generation=1)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [True, False])
-async def test_real_runtime_provider_cannot_use_other_app_manager(disabled):
-    a, b = make_ap('instance-A'), make_ap('instance-B', disabled)
-    ca, cb = context('instance-A'), context('instance-B')
-    requester = NS(ap=b, invoke_llm=AsyncMock(return_value='business result'))
-    provider = RuntimeProvider(cb, NS(workspace_uuid=cb.workspace_uuid), None, requester)
-    model = NS(execution_context=cb, provider=provider)
-    parent = d.Span(a.diagnostics, 'api', 'review.parent', {'workspace_uuid': ca.workspace_uuid})
-    with parent.activate():
-        assert await provider.invoke_llm(None, model, [], execution_context=cb) == 'business result'
-    assert not [e for e in a.diagnostics.pending if e['operation'] == 'model.invoke_llm']
-    assert len(b.diagnostics.pending) == (0 if disabled else 2)
-    for event in b.diagnostics.pending:
-        assert event['instance_id'] == 'instance-B'
-        assert event['workspace_uuid'] == cb.workspace_uuid
-        assert event['trace_id'] != parent.fields['trace_id']
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('shape', ['app', 'ap', 'logger', 'requester', 'adapter', 'direct', 'absent'])
-async def test_explicit_disabled_owner_blocks_parent_even_in_ownerless_children(shape):
-    a, b = make_ap('instance-A'), make_ap('instance-B', True)
-    shapes = {
-        'app': b,
-        'ap': NS(ap=b),
-        'logger': NS(logger=NS(ap=b)),
-        'requester': NS(requester=NS(ap=b)),
-        'adapter': NS(adapter=NS(logger=NS(ap=b))),
-        'direct': NS(diagnostics=b.diagnostics),
-        'absent': NS(ap=None),
-    }
-
-    @d.observe('api', 'review.child')
-    async def child(data):
-        d.event(data, 'api', 'review.point', 'succeeded')
-        return 42
-
-    @d.observe('api', 'review.owner')
-    async def call(owner):
-        return await child({})
-
-    @d.observe('run', 'review.stream')
-    async def stream(owner):
-        yield await child({})
-
-    parent = d.Span(a.diagnostics, 'api', 'review.parent', {})
-    with parent.activate():
-        assert await call(shapes[shape]) == 42
-        assert [x async for x in stream(shapes[shape])] == [42]
-        d.event(shapes[shape], 'api', 'review.point', 'succeeded')
-    assert len(a.diagnostics.pending) == 1
-    assert not b.diagnostics.pending
-
-
-@pytest.mark.asyncio
-async def test_context_cannot_select_other_instance_or_owned_workspace():
-    ap = make_ap()
-    ca, cb = context(), context('instance-other')
-
-    @d.observe('api', 'review.context')
-    async def call(owner, execution_context):
-        return 42
-
-    owner = NS(ap=ap, execution_context=ca)
-    with d.Span(ap.diagnostics, 'api', 'review.parent', {'workspace_uuid': ca.workspace_uuid}).activate():
-        assert await call(owner, cb) == 42
-        assert await call(owner, context()) == 42
-    assert not [e for e in ap.diagnostics.pending if e['operation'] == 'review.context']
-    # A misplaced manager attachment is not an authoritative manager for B.
-    other = make_ap('instance-other')
-    other.diagnostics = ap.diagnostics
-    assert await call(NS(ap=other), cb) == 42
-    assert not [e for e in ap.diagnostics.pending if e['operation'] == 'review.context']
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [False, True])
-@pytest.mark.parametrize('action', ['aclose', 'athrow_exit', 'athrow_value', 'athrow_cancel'])
-async def test_cleanup_exception_identity_matches_native(disabled, action):
-    ap = make_ap(disabled=disabled)
-    cleanup_error = ValueError('CANARY cleanup')
-
-    async def original(owner):
-        try:
-            yield 1
-        finally:
-            raise cleanup_error
-
-    for fn in (original, d.observe('run', 'review.cleanup')(original)):
-        gen = fn(ap)
-        assert isinstance(gen, AsyncGenerator)
-        assert await anext(gen) == 1
-        with pytest.raises(ValueError) as caught:
-            if action == 'aclose':
-                await gen.aclose()
-            else:
-                error = {
-                    'athrow_exit': GeneratorExit(),
-                    'athrow_value': KeyError('business'),
-                    'athrow_cancel': asyncio.CancelledError('cancel'),
-                }[action]
-                await gen.athrow(error)
-        assert caught.value is cleanup_error
-        assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [False, True])
-async def test_athrow_generator_exit_can_yield_and_close_remains_native(disabled):
-    ap = make_ap(disabled=disabled)
-
-    async def original(owner):
-        try:
-            yield 1
-        except GeneratorExit:
-            yield 2
-        yield 3
-
-    for fn in (original, d.observe('run', 'review.exit')(original)):
-        gen = fn(ap)
-        assert await anext(gen) == 1
-        assert await gen.athrow(GeneratorExit()) == 2
-        assert await anext(gen) == 3
-        await gen.aclose()
-        gen = fn(ap)
-        assert await anext(gen) == 1
-        with pytest.raises(RuntimeError, match='ignored GeneratorExit'):
-            await gen.aclose()
-        # The native generator is still suspended after the refused close.
-        assert await anext(gen) == 3
-        with pytest.raises(StopAsyncIteration):
-            await anext(gen)
-        await gen.aclose()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [False, True])
-async def test_generator_send_throw_cancellation_and_primary_exception(disabled):
-    ap = make_ap(disabled=disabled)
-    entered = asyncio.Event()
-    primary = ValueError('CANARY business')
-
-    async def original(owner):
-        value = yield 1
-        try:
-            yield value
-        except KeyError:
-            yield 3
-        entered.set()
-        await asyncio.Event().wait()
-
-    for fn in (original, d.observe('run', 'review.protocol')(original)):
-        gen = fn(ap)
-        assert await anext(gen) == 1
-        assert await gen.asend(7) == 7
-        assert await gen.athrow(KeyError('throw')) == 3
-        entered.clear()
-        task = asyncio.create_task(anext(gen))
-        await entered.wait()
-        task.cancel('native cancellation')
-        with pytest.raises(asyncio.CancelledError, match='native cancellation'):
-            await task
-        await gen.aclose()
-
-    async def failing(owner):
-        yield 1
-        raise primary
-
-    gen = d.observe('run', 'review.primary')(failing)(ap)
-    assert await anext(gen) == 1
-    with pytest.raises(ValueError) as caught:
-        await anext(gen)
-    assert caught.value is primary
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_real_kook_native_entry_conversion_failure_has_trusted_workspace():
-    from langbot.pkg.platform.adapters.kook.adapter import KookAdapter
-    from langbot.pkg.platform.logger import EventLogger
-
-    ap, ctx = make_ap(), context()
-    logger = EventLogger('test', ap, ctx, 'test')
-    logger.error = AsyncMock()
-    adapter = KookAdapter({'token': 'test-placeholder'}, logger)
-    # Invalid native timestamp fails inside the real static converter.
-    await adapter._handle_event({'type': 255, 'msg_timestamp': 'CANARY invalid', 'workspace_uuid': str(uuid4())}, 1)
-    logger.error.assert_awaited_once()
-    failures = [e for e in ap.diagnostics.pending if e['outcome'] == 'failed']
-    assert any(e['stage'] == 'convert' for e in failures)
-    assert any(e['operation'] == 'platform.receive' for e in failures)
-    assert all(e['workspace_uuid'] == ctx.workspace_uuid for e in ap.diagnostics.pending)
-    assert all(e['adapter'] == 'kook-omni' for e in ap.diagnostics.pending)
-    assert 'CANARY' not in json.dumps(ap.diagnostics.pending)
-
-
-@pytest.fixture
-def boot_stages(monkeypatch):
-    from langbot.pkg.core import boot
-
-    stages_run = []
-
-    class Stage:
-        async def run(self, app):
-            stages_run.append(app)
-
-    # Earlier registry tests clear this shared dictionary; cached imports do not
-    # re-register stages. Own the registry per test and restore it on teardown.
-    monkeypatch.setattr(boot.stage, 'preregistered_stages', {'LoadConfigStage': Stage, 'GenKeysStage': Stage})
-    return stages_run
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('failure', ['constructor', 'metadata', 'session', 'start'])
-@pytest.mark.parametrize('disabled', [False, True])
-async def test_real_make_app_optional_initialization_fail_open(monkeypatch, boot_stages, failure, disabled):
-    from langbot.pkg.core import boot
-
-    ap = NS(
-        instance_config=NS(data={'space': {'url': 'https://example.invalid', 'disable_telemetry': disabled}}),
-        initialize=AsyncMock(),
-        shutdown=AsyncMock(),
-    )
-    stages_run = boot_stages
-    monkeypatch.setattr(boot.app, 'Application', lambda: ap)
-    monkeypatch.setattr(boot, 'stage_order', ['LoadConfigStage', 'GenKeysStage'])
-    error = RuntimeError('optional diagnostics')
-    if failure == 'constructor':
-        monkeypatch.setattr(boot.diagnostics, 'DiagnosticsManager', Mock(side_effect=error))
-    elif failure == 'metadata':
-        monkeypatch.setattr(
-            importlib.metadata, 'version', Mock(side_effect=importlib.metadata.PackageNotFoundError('langbot'))
-        )
-    else:
-        manager = NS(
-            start_session=AsyncMock(), start=Mock(), shutdown=AsyncMock(side_effect=RuntimeError('optional cleanup'))
-        )
-        getattr(manager, 'start_session' if failure == 'session' else 'start').side_effect = error
-        monkeypatch.setattr(boot.diagnostics, 'DiagnosticsManager', Mock(return_value=manager))
-    assert await boot.make_app(asyncio.get_running_loop()) is ap
-    assert len(stages_run) == 2
-    ap.initialize.assert_awaited_once()
-    ap.shutdown.assert_not_awaited()
-    assert getattr(ap, 'diagnostics', None) is None
-
-
-@pytest.mark.asyncio
-async def test_real_make_app_session_cancellation_and_shutdown_failure_preserve_primary(monkeypatch, boot_stages):
-    from langbot.pkg.core import boot
-
-    cancelled = asyncio.CancelledError('genuine cancellation')
-    ap = NS(
-        instance_config=NS(data={'space': {'url': 'https://example.invalid'}}),
-        initialize=AsyncMock(),
-        shutdown=AsyncMock(side_effect=RuntimeError('shutdown error')),
-    )
-    manager = NS(start_session=AsyncMock(side_effect=cancelled), start=Mock(), shutdown=AsyncMock())
-    monkeypatch.setattr(boot.app, 'Application', lambda: ap)
-    monkeypatch.setattr(boot, 'stage_order', ['GenKeysStage'])
-    monkeypatch.setattr(boot.diagnostics, 'DiagnosticsManager', Mock(return_value=manager))
-    with pytest.raises(asyncio.CancelledError) as caught:
-        await boot.make_app(asyncio.get_running_loop())
-    assert caught.value is cancelled
-    ap.initialize.assert_not_awaited()
-    ap.shutdown.assert_awaited_once()
-
-
-@pytest.mark.asyncio
-async def test_real_make_app_business_failure_not_masked_by_optional_shutdown(monkeypatch, boot_stages):
-    from langbot.pkg.core import boot
-
-    primary = ValueError('business startup')
-    ap = boot.app.Application()
-    ap.instance_config = NS(data={'space': {'url': 'https://example.invalid'}})
-    ap.initialize = AsyncMock(side_effect=primary)
-    manager = make_ap().diagnostics
-    manager.start_session = AsyncMock()
-    manager.start = Mock()
-    manager.shutdown = AsyncMock(side_effect=RuntimeError('optional shutdown'))
-    monkeypatch.setattr(boot.app, 'Application', lambda: ap)
-    monkeypatch.setattr(boot.diagnostics, 'DiagnosticsManager', lambda *a, **kw: manager)
-    monkeypatch.setattr(boot, 'stage_order', ['GenKeysStage'])
-
-    with pytest.raises(ValueError) as caught:
-        await boot.make_app(asyncio.get_running_loop())
-    assert caught.value is primary
-    manager.shutdown.assert_awaited_once()
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [False, True])
-async def test_protocol_rejected_calls_and_unstarted_throw_match_native(disabled):
-    ap = make_ap(disabled=disabled)
-
-    async def original(owner):
-        try:
-            yield 1
-        except GeneratorExit:
-            yield 2
-        yield 3
-
-    async def record(fn, actions):
-        gen = fn(ap)
-        results = []
-        for name, values in actions:
-            try:
-                results.append(('value', await getattr(gen, name)(*values)))
-            except BaseException as exc:
-                results.append((type(exc).__name__, str(exc)))
-        # Fully exhaust any generator left suspended by a refused close.
-        try:
-            while True:
-                await anext(gen)
-        except (StopAsyncIteration, GeneratorExit):
-            pass
-        return results
-
-    wrapped = d.observe('run', 'review.protocol_matrix')(original)
-    for actions in (
-        [('aclose', ())],
-        [('athrow', (GeneratorExit(),))],
-        [('athrow', (ValueError('unstarted'),))],
-        [('asend', (7,)), ('__anext__', ()), ('athrow', (ValueError('primary'),))],
-        [('__anext__', ()), ('aclose', ()), ('__anext__', ()), ('aclose', ())],
-        [('__anext__', ()), ('athrow', (GeneratorExit(),)), ('__anext__', ()), ('aclose', ())],
-    ):
-        assert await record(wrapped, actions) == await record(original, actions)
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize('disabled', [False, True])
-async def test_native_task_cancel_cleanup_error_wins_without_extra_close(disabled):
-    ap = make_ap(disabled=disabled)
-    error = ValueError('native cleanup')
-    entered = asyncio.Event()
-
-    async def original(owner):
-        yield 1
-        try:
-            entered.set()
-            await asyncio.Event().wait()
-        finally:
-            raise error
-
-    for fn in (original, d.observe('run', 'review.cancel_cleanup')(original)):
-        gen = fn(ap)
-        await anext(gen)
-        entered.clear()
-        task = asyncio.create_task(anext(gen))
-        await entered.wait()
-        task.cancel()
-        with pytest.raises(ValueError) as caught:
-            await task
-        assert caught.value is error
-        await gen.aclose()
-
-
-@pytest.mark.asyncio
-async def test_sdk_b2_handler_consumes_observed_async_generator(tmp_path):
-    from langbot_plugin.runtime.io.handler import Handler, ActionResponse
-    from langbot_plugin.entities.io.resp import ChunkStatus
-
-    ap = make_ap()
-    handler = Handler(NS(), file_storage_dir=str(tmp_path))
-    handler._send_message = AsyncMock()
-
-    @d.observe('api', 'host.review_stream', ap=ap)
-    async def stream(data):
-        yield ActionResponse.success({'business': 'unchanged'})
-
-    handler.actions['review_stream'] = stream
-    await handler._handle_action({'seq_id': 1, 'action': 'review_stream', 'data': {}})
-    responses = [call.args[0] for call in handler._send_message.await_args_list]
-    assert len(responses) == 2
-    assert responses[0].data == {'business': 'unchanged'}
-    assert [response.chunk_status for response in responses] == [ChunkStatus.CONTINUE, ChunkStatus.END]
-    assert [e['outcome'] for e in ap.diagnostics.pending] == ['started', 'succeeded']
diff --git a/tests/unit_tests/telemetry/test_beta_diagnostics.py b/tests/unit_tests/telemetry/test_beta_diagnostics.py
deleted file mode 100644
index a73d8d8a9..000000000
--- a/tests/unit_tests/telemetry/test_beta_diagnostics.py
+++ /dev/null
@@ -1,350 +0,0 @@
-"""Content-free diagnostics contract and lifecycle regression tests."""
-
-import asyncio
-import json
-from types import SimpleNamespace
-from uuid import uuid4
-
-import httpx
-import pytest
-
-from langbot.pkg.telemetry import diagnostics as d
-
-
-def manager(version='4.11.0-beta.2', **config):
-    ap = SimpleNamespace(instance_config=SimpleNamespace(data={'space': {'url': 'https://example.invalid', **config}}))
-    ap.diagnostics = d.DiagnosticsManager(ap, version=version, instance_id='instance-test', capacity=4)
-    return ap.diagnostics
-
-
-def test_release_gate_and_privacy():
-    assert not manager('4.11.0').enabled
-    assert manager().enabled
-    assert manager('4.11.0b2').enabled
-    assert not manager(disable_telemetry=True).enabled
-    assert not manager(disable_beta_diagnostics=True).enabled
-    m = manager()
-    m.emit(
-        'api',
-        'test.operation',
-        'failed',
-        attributes={'prompt': 'CANARY', 'plugin_id': 'CANARY', 'attempts': 1},
-        error=ValueError('CANARY https://secret/token'),
-        workspace_uuid=str(uuid4()),
-    )
-    payload = m.pending[0]
-    assert 'CANARY' not in json.dumps(payload)
-    assert payload['attributes'] == {'attempts': 1}
-    assert payload['error_type'] == 'ValueError'
-    assert payload['instance_id'] != payload['workspace_uuid']
-    assert payload['sample_rate'] == 1
-
-
-@pytest.mark.asyncio
-async def test_disabled_boundary_skips_all_projection(monkeypatch):
-    m = manager('4.11.0')
-
-    def broken(*args, **kwargs):
-        raise AssertionError('diagnostic machinery ran while disabled')
-
-    monkeypatch.setattr(d, 'Span', broken)
-    monkeypatch.setattr(d, 'result_outcome', broken)
-
-    class Service:
-        ap = m.ap
-
-        @d.observe('api', 'test.disabled', fields=broken)
-        async def call(self):
-            return 42
-
-        @d.observe('run', 'test.disabled', fields=broken)
-        async def stream(self):
-            yield 42
-
-    assert await Service().call() == 42
-    assert [v async for v in Service().stream()] == [42]
-    assert not m.pending
-
-
-def test_bounds_and_disable_clear():
-    m = manager()
-    for _ in range(8):
-        m.emit('api', 'test.operation', 'succeeded')
-    assert len(m.pending) == 4
-    assert m.counters['dropped'] == 4
-    m.ap.instance_config.data['space']['disable_telemetry'] = True
-    m.emit('api', 'test.operation', 'succeeded')
-    assert not m.pending
-
-
-@pytest.mark.asyncio
-async def test_partial_ack_retries_same_identity_and_drops_rejected():
-    m = manager()
-    for _ in range(3):
-        m.emit('api', 'test.operation', 'succeeded')
-    ids = [e['event_id'] for e in m.pending]
-    requests = []
-
-    async def handler(req):
-        requests.append(json.loads(req.content))
-        return httpx.Response(
-            200,
-            json={
-                'code': 200,
-                'data': {'accepted_event_ids': [ids[0]], 'rejected': [{'event_id': ids[1], 'code': 'invalid_event'}]},
-            },
-        )
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    await m.flush_once()
-    assert [e['event_id'] for e in m.pending] == [ids[2]]
-    assert m.counters['acked'] == 1
-    assert m.counters['dropped'] == 1
-    assert requests[0]['schema_version'] == 1
-    await m.shutdown(drain_timeout=0)
-
-
-@pytest.mark.asyncio
-async def test_outage_finite_retry_and_slow_credentials():
-    m = manager()
-    m.max_attempts = 2
-    m.emit('run', 'test.operation', 'started')
-    calls = []
-
-    async def handler(req):
-        calls.append(req)
-        return httpx.Response(503)
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    await m.flush_once()
-    await m.flush_once()
-    assert not m.pending
-    assert m.counters['dropped'] == 1
-    assert len(calls) == 2
-
-    async def credentials(workspace):
-        await asyncio.sleep(60)
-
-    m.credentials = credentials
-    m.request_timeout = 0.01
-    m.emit('api', 'test.operation', 'succeeded')
-    await asyncio.wait_for(m.flush_once(), 0.2)
-    await m.shutdown(drain_timeout=0)
-
-
-@pytest.mark.asyncio
-async def test_inflight_disable_cancels_and_clears():
-    m = manager()
-    entered = asyncio.Event()
-
-    async def handler(req):
-        entered.set()
-        await asyncio.sleep(60)
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    m.start()
-    m.emit('api', 'test.operation', 'succeeded')
-    await asyncio.wait_for(entered.wait(), 1)
-    m.ap.instance_config.data['space']['disable_beta_diagnostics'] = True
-    await asyncio.sleep(0.3)
-    assert not m.pending
-    await asyncio.wait_for(m.shutdown(drain_timeout=0), 0.5)
-
-
-@pytest.mark.asyncio
-async def test_observe_returns_errors_and_cancellation():
-    m = manager()
-
-    class Service:
-        ap = m.ap
-
-        @d.observe('api', 'test.operation')
-        async def call(self, error=None):
-            if error:
-                raise error
-            return {'secret': 'CANARY'}
-
-    s = Service()
-    assert await s.call() == {'secret': 'CANARY'}
-    with pytest.raises(ValueError):
-        await s.call(ValueError('CANARY'))
-    m.pending.clear()
-    with pytest.raises(asyncio.CancelledError):
-        await s.call(asyncio.CancelledError())
-    assert m.pending[-1]['outcome'] == 'cancelled'
-    assert 'CANARY' not in json.dumps(m.pending)
-
-
-@pytest.mark.asyncio
-async def test_generator_send_throw_close_and_context_isolation():
-    m = manager()
-    m.capacity = 30
-    closed = []
-
-    class Service:
-        ap = m.ap
-
-        @d.observe('run', 'test.operation')
-        async def stream(self):
-            try:
-                value = yield 1
-                try:
-                    yield value
-                except ValueError:
-                    yield 3
-            finally:
-                closed.append(True)
-
-    gen = Service().stream()
-    assert await anext(gen) == 1
-    assert d.current_span() is None
-    assert await gen.asend(7) == 7
-    assert await gen.athrow(ValueError('CANARY')) == 3
-    await gen.aclose()
-    assert closed == [True]
-    assert m.pending[-1]['outcome'] == 'cancelled'
-    assert d.current_span() is None
-
-
-@pytest.mark.asyncio
-async def test_generator_early_failure_and_explicit_terminal():
-    m = manager()
-
-    class Service:
-        ap = m.ap
-
-        @d.observe('run', 'test.operation')
-        async def stream(self, fail):
-            if fail:
-                raise ValueError('prepare CANARY')
-            d.set_outcome('failed', reason_code='runner_failed')
-            yield 1
-
-    with pytest.raises(ValueError):
-        await anext(Service().stream(True))
-    assert m.pending[-1]['outcome'] == 'failed'
-    m.pending.clear()
-    assert [v async for v in Service().stream(False)] == [1]
-    assert m.pending[-1]['outcome'] == 'failed'
-
-
-@pytest.mark.asyncio
-async def test_projection_fault_does_not_replace_business_return(monkeypatch):
-    m = manager()
-
-    def broken(*args, **kwargs):
-        raise RuntimeError('CANARY projection')
-
-    monkeypatch.setattr(d, 'result_outcome', broken)
-
-    class Service:
-        ap = m.ap
-
-        @d.observe('api', 'test.projection_fault')
-        async def call(self):
-            return 42
-
-    assert await Service().call() == 42
-    assert m.pending[-1]['outcome'] == 'succeeded'
-
-
-@pytest.mark.asyncio
-async def test_workspace_batches_and_credential_failure_are_anonymous():
-    m = manager()
-    workspaces = [str(uuid4()), str(uuid4())]
-    for workspace in workspaces:
-        m.emit('api', 'test.operation', 'succeeded', workspace_uuid=workspace)
-    requests = []
-
-    async def broken(workspace):
-        raise RuntimeError('CANARY credentials')
-
-    m.credentials = broken
-
-    async def handler(request):
-        data = json.loads(request.content)
-        requests.append((data, dict(request.headers)))
-        return httpx.Response(
-            200,
-            json={'code': 200, 'data': {'accepted_event_ids': [e['event_id'] for e in data['events']], 'rejected': []}},
-        )
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    await m.flush_once()
-    await m.flush_once()
-    assert len(requests) == 2
-    assert [request[0]['events'][0]['workspace_uuid'] for request in requests] == workspaces
-    assert all('authorization' not in headers for _, headers in requests)
-    assert not m.pending
-    await m.shutdown(drain_timeout=0)
-
-
-@pytest.mark.asyncio
-async def test_summary_is_interval_delta_and_ack_replay_is_idempotent():
-    m = manager()
-    m.capacity = 20
-    m.emit('api', 'test.operation', 'succeeded')
-    m.report_transport()
-    first = m.pending[-1]
-    m.report_transport()
-    second = m.pending[-1]
-    assert first['attributes']['generated'] == 1
-    assert second['attributes']['generated'] == 1  # Only first summary itself.
-    assert first['event_id'] != second['event_id']
-    ids = [e['event_id'] for e in m.pending]
-    seen = []
-
-    async def handler(request):
-        body = json.loads(request.content)
-        seen.append([e['event_id'] for e in body['events']])
-        return (
-            httpx.Response(503)
-            if len(seen) == 1
-            else httpx.Response(200, json={'code': 200, 'data': {'accepted_event_ids': ids + ids, 'rejected': []}})
-        )
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    await m.flush_once()
-    await m.flush_once()
-    assert seen == [ids, ids]
-    assert m.counters['acked'] == len(ids)
-    assert not m.pending
-    await m.shutdown(drain_timeout=0)
-
-
-@pytest.mark.asyncio
-async def test_retention_and_bounded_shutdown():
-    m = manager()
-    m.emit('api', 'test.operation', 'succeeded')
-    m.retention_seconds = -1
-    await m.flush_once()
-    assert not m.pending and m.counters['dropped'] == 1
-    m.retention_seconds = 900
-
-    async def handler(request):
-        await asyncio.sleep(60)
-        return httpx.Response(503)
-
-    m.client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
-    m.emit('api', 'test.operation', 'succeeded')
-    await asyncio.wait_for(m.shutdown(drain_timeout=0.01), 0.5)
-    assert not m.pending and m.client.is_closed
-
-
-@pytest.mark.asyncio
-async def test_session_marker_restart_and_disable(tmp_path):
-    marker = tmp_path / 'session.json'
-    m = manager()
-    m.marker_path = marker
-    await m.start_session()
-    assert marker.exists()
-    recovered = manager()
-    recovered.marker_path = marker
-    await recovered.start_session()
-    assert recovered.pending[-1]['attributes']['previous_session_unclean'] is True
-    recovered.ap.instance_config.data['space']['disable_beta_diagnostics'] = True
-    recovered.start()
-    await asyncio.sleep(0.2)
-    assert not marker.exists()
-    await recovered.shutdown(drain_timeout=0)
-    await m.shutdown(drain_timeout=0)
diff --git a/tests/unit_tests/telemetry/test_chat_emitter.py b/tests/unit_tests/telemetry/test_chat_emitter.py
new file mode 100644
index 000000000..0c66d3f83
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_chat_emitter.py
@@ -0,0 +1,110 @@
+"""The Chat handler must not emit the legacy per-query telemetry payload.
+
+Per-execution telemetry is produced by the trace that owns the Pipeline lane;
+the handler only annotates that trace with execution-scoped record fields.
+"""
+
+from __future__ import annotations
+
+import types
+from importlib import import_module
+from unittest.mock import AsyncMock, Mock
+
+import langbot_plugin.api.entities.builtin.provider.session as provider_session
+
+
+def get_modules():
+    # Import the application package first: the handler package is wired into
+    # it during initialization, so importing the handler directly would re-enter
+    # the partially initialized pipeline package.
+    import_module('langbot.pkg.core.app')
+    return (
+        import_module('langbot.pkg.pipeline.process.handlers.chat'),
+        import_module('langbot.pkg.telemetry.trace'),
+    )
+
+
+class FakeEvent:
+    """Stand-in for the plugin event models, which need a real Query."""
+
+    def __init__(self, **kwargs):
+        self.__dict__.update(kwargs)
+        self.event_name = 'person_normal_message_received'
+
+
+class FakeEventContext:
+    event = types.SimpleNamespace(user_message_alter=None)
+
+    def is_prevented_default(self) -> bool:
+        return False
+
+
+async def empty_run_from_query(query):
+    if False:  # pragma: no cover - async generator with no results
+        yield
+
+
+def build_ap():
+    ap = Mock()
+    ap.logger = Mock()
+    ap.plugin_connector = Mock()
+    ap.plugin_connector.emit_event = AsyncMock(return_value=FakeEventContext())
+    ap.agent_run_orchestrator = Mock()
+    ap.agent_run_orchestrator.try_claim_steering_from_query = AsyncMock(return_value=False)
+    ap.agent_run_orchestrator.run_from_query = empty_run_from_query
+    ap.agent_run_orchestrator.resolve_runner_id_for_telemetry = Mock(return_value='runner-1')
+    ap.survey = Mock()
+    ap.survey.trigger_event = AsyncMock()
+    ap.survey.record_bot_response_success = AsyncMock()
+    ap.telemetry = Mock()
+    ap.telemetry.start_send_task = AsyncMock()
+    ap.telemetry.send = AsyncMock()
+    return ap
+
+
+def build_query():
+    query = Mock()
+    query.query_id = 7
+    query.query_uuid = 'exec-uuid'
+    query.launcher_type = provider_session.LauncherTypes.PERSON
+    query.launcher_id = 'user-1'
+    query.sender_id = 'user-1'
+    query.message_chain = 'hello'
+    query.message_event = object()
+    query.variables = {'_pipeline_bound_plugins': ['plugin:a']}
+    query.pipeline_config = {'ai': {}, 'output': {'misc': {}}}
+    query.session = Mock(using_conversation=object())
+    query.use_llm_model_uuid = None
+    query.adapter = Mock()
+    query.adapter.is_stream_output_supported = AsyncMock(return_value=False)
+    query.resp_messages = []
+    return query
+
+
+async def test_normal_query_sends_no_legacy_payload_and_annotates_the_trace(monkeypatch):
+    chat, trace_mod = get_modules()
+    monkeypatch.setattr(chat.events, 'PersonNormalMessageReceived', FakeEvent)
+    monkeypatch.setattr(chat.events, 'GroupNormalMessageReceived', FakeEvent)
+
+    ap = build_ap()
+    handler = chat.ChatMessageHandler(ap)
+    query = build_query()
+
+    binding = trace_mod.bind('exec-uuid')
+    try:
+        async for _ in handler.handle(query):
+            pass
+    finally:
+        trace_mod.unbind_root(binding)
+
+    # The legacy `event_type='query'` emitter is gone.
+    ap.telemetry.start_send_task.assert_not_called()
+    ap.telemetry.send.assert_not_called()
+    # The survey trigger and its milestone bookkeeping still fire.
+    ap.survey.trigger_event.assert_awaited_once_with('first_bot_response_success')
+    ap.survey.record_bot_response_success.assert_awaited_once()
+    # Execution-scoped fields are attached to the owning trace instead.
+    assert binding.state.runner == 'runner-1'
+    assert binding.state.runner_category == 'unknown'
+    assert binding.state.pipeline_plugins == ['plugin:a']
+    assert binding.state.model_name == ''
diff --git a/tests/unit_tests/telemetry/test_cloud_sampling.py b/tests/unit_tests/telemetry/test_cloud_sampling.py
new file mode 100644
index 000000000..ef4de5e2e
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_cloud_sampling.py
@@ -0,0 +1,37 @@
+from types import SimpleNamespace
+
+from langbot.pkg.telemetry.execution import ExecutionCounters
+from langbot.pkg.telemetry.trace import TraceState
+from langbot.pkg.utils import constants
+
+
+def test_cloud_counts_are_not_sampled(monkeypatch):
+    monkeypatch.setattr(constants, 'edition', 'cloud')
+    counters = ExecutionCounters(SimpleNamespace(telemetry_config={'execution_trace': 'sampled'}))
+    assert counters.trace_mode() == 'all'
+    state = TraceState('00000001')
+    counters.configure(state)
+    assert state.emit_nodes is True
+    assert counters._chain_emitted(state) is True
+
+
+def test_community_retains_sampling(monkeypatch):
+    monkeypatch.setattr(constants, 'edition', 'community')
+    counters = ExecutionCounters(SimpleNamespace(telemetry_config={'execution_trace': 'sampled'}))
+    assert counters.trace_mode() == 'sampled'
+
+
+def test_dns_survives_saturated_executor():
+    import socket
+    import threading
+    from langbot.pkg.utils.bounded_executor import BoundedThreadPoolExecutor
+
+    gate = threading.Event()
+    executor = BoundedThreadPoolExecutor(max_workers=1, max_pending=0)
+    try:
+        blocked = executor.submit(gate.wait)
+        assert executor.submit(socket.getaddrinfo, '127.0.0.1', 443).result(timeout=2)
+    finally:
+        gate.set()
+        blocked.result(timeout=2)
+        executor.shutdown()
diff --git a/tests/unit_tests/telemetry/test_features.py b/tests/unit_tests/telemetry/test_features.py
deleted file mode 100644
index 6e19d0828..000000000
--- a/tests/unit_tests/telemetry/test_features.py
+++ /dev/null
@@ -1,92 +0,0 @@
-"""Unit tests for telemetry feature counters (pkg/telemetry/features.py)."""
-
-from __future__ import annotations
-
-from importlib import import_module
-
-
-def get_features_module():
-    return import_module('langbot.pkg.telemetry.features')
-
-
-class FakeQuery:
-    def __init__(self):
-        self.variables = {}
-
-
-class TestIncrement:
-    def test_increment_nested_counter(self):
-        features = get_features_module()
-        q = FakeQuery()
-        features.increment(q, 'tool_calls', 'native')
-        features.increment(q, 'tool_calls', 'native')
-        features.increment(q, 'tool_calls', 'mcp')
-        assert q.variables[features.FEATURES_KEY]['tool_calls'] == {'native': 2, 'mcp': 1}
-
-    def test_increment_flat_counter(self):
-        features = get_features_module()
-        q = FakeQuery()
-        features.increment(q, 'something')
-        features.increment(q, 'something', amount=2)
-        assert q.variables[features.FEATURES_KEY]['something'] == 3
-
-    def test_increment_never_raises_on_broken_query(self):
-        features = get_features_module()
-
-        class Broken:
-            @property
-            def variables(self):
-                raise RuntimeError('boom')
-
-        # Must not raise
-        features.increment(Broken(), 'tool_calls', 'native')
-
-    def test_set_value(self):
-        features = get_features_module()
-        q = FakeQuery()
-        features.set_value(q, 'tool_call_rounds', 5)
-        assert q.variables[features.FEATURES_KEY]['tool_call_rounds'] == 5
-
-
-class TestCollectFeatures:
-    def test_collect_empty(self):
-        features = get_features_module()
-        q = FakeQuery()
-        assert features.collect_features(q) == {}
-
-    def test_collect_combines_counters_and_snapshots(self):
-        features = get_features_module()
-        q = FakeQuery()
-        features.increment(q, 'sandbox', 'execs')
-        features.set_value(q, 'kb', {'kb_count': 2, 'engine_plugins': ['builtin'], 'retrieved_entries': 7})
-        q.variables['_activated_skills'] = {'pdf-tools': {}, 'a-skill': {}}
-        q.variables['_pipeline_bound_mcp_servers'] = ['srv1', 'srv2']
-
-        result = features.collect_features(q)
-        assert result['sandbox'] == {'execs': 1}
-        assert result['kb']['kb_count'] == 2
-        assert result['activated_skills'] == ['a-skill', 'pdf-tools']  # sorted
-        assert result['mcp_servers'] == ['srv1', 'srv2']
-
-    def test_collect_omits_mcp_when_all_enabled(self):
-        """None means 'all enabled' and is not reported."""
-        features = get_features_module()
-        q = FakeQuery()
-        q.variables['_pipeline_bound_mcp_servers'] = None
-        assert 'mcp_servers' not in features.collect_features(q)
-
-    def test_collect_drops_non_json_serializable(self):
-        features = get_features_module()
-        q = FakeQuery()
-        features.set_value(q, 'good', 1)
-        features.set_value(q, 'bad', object())
-        result = features.collect_features(q)
-        assert result == {'good': 1}
-
-    def test_collect_is_json_serializable(self):
-        import json
-
-        features = get_features_module()
-        q = FakeQuery()
-        features.increment(q, 'tool_calls', 'skill')
-        json.dumps(features.collect_features(q))
diff --git a/tests/unit_tests/telemetry/test_heartbeat.py b/tests/unit_tests/telemetry/test_heartbeat.py
index 86b70094b..f559f1679 100644
--- a/tests/unit_tests/telemetry/test_heartbeat.py
+++ b/tests/unit_tests/telemetry/test_heartbeat.py
@@ -183,6 +183,7 @@ class TestBuildHeartbeatPayload:
         assert by_workspace['workspace-b']['execution_generation'] == 9
         assert by_workspace['workspace-b']['adapters'] == ['WorkspaceBAdapter']
         assert 'workspace_resources' not in by_workspace['workspace-a']
+        ap.plugin_connector.list_plugins.assert_not_awaited()
         ap.persistence_mgr.execute_async.assert_not_awaited()
         ap.workspace_service.list_active_execution_bindings.assert_awaited_once()
 
diff --git a/tests/unit_tests/telemetry/test_pipeline_lane.py b/tests/unit_tests/telemetry/test_pipeline_lane.py
new file mode 100644
index 000000000..80e635436
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_pipeline_lane.py
@@ -0,0 +1,316 @@
+"""The Pipeline lane must land its nodes on one v2 execution chain."""
+
+from __future__ import annotations
+
+import asyncio
+import types
+from importlib import import_module
+
+import pytest
+
+
+def get_modules():
+    # Import the application package first; the pipeline package is wired into it.
+    import_module('langbot.pkg.core.app')
+    return (
+        import_module('langbot.pkg.pipeline.pipelinemgr'),
+        import_module('langbot.pkg.telemetry.execution'),
+    )
+
+
+def get_trace():
+    import_module('langbot.pkg.core.app')
+    return import_module('langbot.pkg.telemetry.trace')
+
+
+class FakeManager:
+    def __init__(self):
+        self.telemetry_config = {'url': 'https://space.example.test', 'execution_trace': 'all'}
+        self.sent: list[dict] = []
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+
+CONTEXT = types.SimpleNamespace(instance_uuid='instance-1', workspace_uuid='workspace-1')
+
+
+def build_pipeline(pipelinemgr, ap):
+    pipeline = object.__new__(pipelinemgr.RuntimePipeline)
+    pipeline.ap = ap
+    pipeline.execution_context = CONTEXT
+    return pipeline
+
+
+def test_lane_failure_marks_the_owned_chain_without_any_prior_stage():
+    pipelinemgr, execution = get_modules()
+    manager = FakeManager()
+    counters = execution.ExecutionCounters(manager)
+    ap = types.SimpleNamespace(telemetry=types.SimpleNamespace(execution=counters))
+    pipeline = build_pipeline(pipelinemgr, ap)
+    query = types.SimpleNamespace(pipeline_config={})
+
+    with execution.ingress(ap, 'pipeline_done', CONTEXT, execution_id='query-7'):
+        pipeline._record_lane_failure(query, 'sandbox exited 127')
+
+    rows = [record for record in counters.records if record['event_type'] == 'execution_node']
+    assert len(rows) == 1
+    node = rows[0]
+    assert node['event_id'] == 'query-7'
+    assert node['error'] == 'sandbox exited 127'
+    assert (node['family'], node['operation'], node['outcome']) == ('runner', 'execute', 'failed')
+    chain = [record for record in counters.records if record['event_type'] == 'execution_chain'][0]
+    assert chain['event_id'] == 'query-7'
+    assert chain['outcome'] == 'failed'
+    assert chain['error'] == 'sandbox exited 127'
+
+
+def test_lane_failure_does_not_duplicate_the_runner_node_the_orchestrator_recorded():
+    """A break inside a runner is already a node; the lane only carries the reason over."""
+    pipelinemgr, execution = get_modules()
+    manager = FakeManager()
+    counters = execution.ExecutionCounters(manager)
+    ap = types.SimpleNamespace(telemetry=types.SimpleNamespace(execution=counters))
+    pipeline = build_pipeline(pipelinemgr, ap)
+    query = types.SimpleNamespace(pipeline_config={})
+
+    with execution.ingress(ap, 'pipeline_done', CONTEXT, execution_id='query-8'):
+        execution.record(
+            ap,
+            CONTEXT,
+            family='runner',
+            operation='execute',
+            mode='pipeline',
+            runner='plugin:langbot-team/LocalAgent/default',
+            outcome='failed',
+            error='No authorized model for local-agent',
+        )
+        pipeline._record_lane_failure(query, 'No authorized model for local-agent')
+
+    rows = [record for record in counters.records if record['event_type'] == 'execution_node']
+    assert [(record['family'], record['operation'], record['outcome']) for record in rows] == [
+        ('runner', 'execute', 'failed')
+    ]
+    chain = [record for record in counters.records if record['event_type'] == 'execution_chain'][-1]
+    assert chain['outcome'] == 'failed'
+    assert chain['error'] == 'No authorized model for local-agent'
+
+
+def build_counters():
+    manager = FakeManager()
+    counters = import_module('langbot.pkg.telemetry.execution').ExecutionCounters(manager)
+    ap = types.SimpleNamespace(telemetry=types.SimpleNamespace(execution=counters))
+    return counters, ap
+
+
+def build_lane(pipelinemgr, ap):
+    # ``process_query`` lives on the RuntimePipeline lane, not on PipelineManager.
+    lane = object.__new__(pipelinemgr.RuntimePipeline)
+    lane.ap = ap
+    lane.execution_context = CONTEXT
+    return lane
+
+
+def build_query(**overrides):
+    query = types.SimpleNamespace(
+        query_uuid='exec-lane',
+        query_id=1,
+        pipeline_config={},
+        variables={},
+        adapter=None,
+        launcher_type=types.SimpleNamespace(value='private'),
+        launcher_id='user',
+    )
+    for key, value in overrides.items():
+        setattr(query, key, value)
+    return query
+
+
+def lane_nodes(counters):
+    return [record for record in counters.records if record['event_type'] == 'execution_node']
+
+
+def lane_shapes(counters):
+    return [
+        (record['family'], record['operation'], record['node_id'], record['parent_node_id'])
+        for record in lane_nodes(counters)
+    ]
+
+
+async def stop_counters(counters):
+    """Cancel the telemetry flush task so no loop is left with pending work."""
+    task = counters.task
+    if task is not None and not task.done():
+        task.cancel()
+        try:
+            await task
+        except asyncio.CancelledError:
+            pass
+    counters.task = None
+
+
+async def test_lane_step_owns_the_apis_and_runner_it_calls():
+    pipelinemgr, execution = get_modules()
+    trace = get_trace()
+    counters, ap = build_counters()
+    lane = build_lane(pipelinemgr, ap)
+    query = build_query()
+    seen = {}
+
+    async def fake_process(_query):
+        execution.record(
+            ap,
+            CONTEXT,
+            family='platform_api',
+            operation='send_message',
+            adapter='AiocqhttpAdapter',
+            mode='pipeline',
+            outcome='success',
+        )
+        with trace.stage_scope() as run_node:
+            seen['run'] = run_node
+            execution.record(
+                ap, CONTEXT, family='runner', operation='execute', mode='pipeline', outcome='success', node=run_node
+            )
+        # A reply issued after the runner still belongs to the lane.
+        execution.record(
+            ap,
+            CONTEXT,
+            family='platform_api',
+            operation='reply_message',
+            adapter='AiocqhttpAdapter',
+            mode='pipeline',
+            outcome='success',
+        )
+
+    lane._process_query = fake_process
+
+    async def scenario():
+        await lane.process_query(query)
+        await stop_counters(counters)
+
+    await scenario()
+
+    shapes = lane_shapes(counters)
+    # The synthetic inbound event roots the chain; the lane step hangs under it.
+    assert [shape[0:2] for shape in shapes] == [
+        ('platform_event', 'message.received'),
+        ('platform_api', 'send_message'),
+        ('runner', 'execute'),
+        ('platform_api', 'reply_message'),
+        ('pipeline', 'run'),
+    ]
+    event_node = shapes[0][2]
+    lane_node = shapes[-1][2]
+    assert shapes[0][3] == ''
+    assert shapes[1][3] == lane_node
+    assert shapes[2] == ('runner', 'execute', seen['run'], lane_node)
+    assert shapes[3][3] == lane_node
+    # The lane step is recorded last and nests under the synthetic event node.
+    assert shapes[4] == ('pipeline', 'run', lane_node, event_node)
+
+
+async def test_lane_node_nests_under_the_platform_route_that_called_it():
+    pipelinemgr, execution = get_modules()
+    trace = get_trace()
+    counters, ap = build_counters()
+    lane = build_lane(pipelinemgr, ap)
+    query = build_query(query_uuid='exec-routed')
+
+    async def fake_process(_query):
+        execution.record(
+            ap, CONTEXT, family='platform_api', operation='send_message', mode='pipeline', outcome='success'
+        )
+
+    lane._process_query = fake_process
+
+    async def scenario():
+        # A platform event owns the chain; the route step wraps the Pipeline lane.
+        with execution.ingress(ap, 'event_done', CONTEXT, execution_id='exec-routed'):
+            with trace.stage_scope() as event_node:
+                execution.record(
+                    ap,
+                    CONTEXT,
+                    family='platform_event',
+                    operation='message.received',
+                    adapter='AiocqhttpAdapter',
+                    outcome='success',
+                    node=event_node,
+                )
+                with trace.stage_scope() as route_node:
+                    execution.record(
+                        ap,
+                        CONTEXT,
+                        family='event_route',
+                        operation='message.received',
+                        adapter='AiocqhttpAdapter',
+                        outcome='success',
+                        node=route_node,
+                    )
+                    await lane.process_query(query)
+        await stop_counters(counters)
+        return event_node, route_node
+
+    event_node, route_node = await scenario()
+
+    assert event_node != route_node
+    shapes = lane_shapes(counters)
+    assert shapes[0] == ('platform_event', 'message.received', event_node, '')
+    assert shapes[1] == ('event_route', 'message.received', route_node, event_node)
+    lane_node = shapes[-1][2]
+    # The API is called by the lane, so it hangs on the lane node.
+    assert shapes[2][0:2] == ('platform_api', 'send_message')
+    assert shapes[2][3] == lane_node and shapes[2][2] != lane_node
+    # The lane nests under the route step that called it.
+    assert shapes[3] == ('pipeline', 'run', lane_node, route_node)
+
+
+async def test_broken_lane_records_a_failed_step_with_its_reason():
+    pipelinemgr, execution = get_modules()
+    counters, ap = build_counters()
+    lane = build_lane(pipelinemgr, ap)
+
+    async def fake_process(_query):
+        raise RuntimeError('lane exploded')
+
+    lane._process_query = fake_process
+
+    async def scenario():
+        try:
+            await lane.process_query(build_query())
+        finally:
+            await stop_counters(counters)
+
+    with pytest.raises(RuntimeError, match='lane exploded'):
+        await scenario()
+
+    shapes = lane_shapes(counters)
+    family, operation, node, parent = shapes[-1]
+    row = lane_nodes(counters)[-1]
+    # The lane step nests under the synthetic inbound event that roots the chain.
+    assert (family, operation) == ('pipeline', 'run')
+    assert shapes[0] == ('platform_event', 'message.received', parent, '')
+    assert row['outcome'] == 'failed'
+    assert row['error'] == 'lane exploded'
+
+
+async def test_lane_reported_failure_marks_the_step_failed():
+    pipelinemgr, execution = get_modules()
+    counters, ap = build_counters()
+    lane = build_lane(pipelinemgr, ap)
+    query = build_query(variables={'_monitoring_has_error': True})
+
+    async def fake_process(_query):
+        return None
+
+    lane._process_query = fake_process
+
+    async def scenario():
+        await lane.process_query(query)
+        await stop_counters(counters)
+
+    await scenario()
+
+    row = lane_nodes(counters)[-1]
+    assert (row['family'], row['operation'], row['outcome']) == ('pipeline', 'run', 'failed')
diff --git a/tests/unit_tests/telemetry/test_platform_model.py b/tests/unit_tests/telemetry/test_platform_model.py
new file mode 100644
index 000000000..8bca383ea
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_platform_model.py
@@ -0,0 +1,27 @@
+from types import SimpleNamespace
+
+import pytest
+from pydantic import BaseModel
+
+from langbot.pkg.telemetry.platform import observe_adapter
+
+
+class ModelAdapter(BaseModel):
+    def get_supported_apis(self):
+        return ['send_message']
+
+    async def send_message(self):
+        return {'message_id': 'ok'}
+
+
+@pytest.mark.asyncio
+async def test_model_adapter_instrumentation_is_instance_local():
+    first, second = ModelAdapter(), ModelAdapter()
+    ap = SimpleNamespace(telemetry=None)
+    observe_adapter(ap, None, first)
+    wrapped = first.send_message
+    observe_adapter(ap, None, first)
+    assert first.send_message is wrapped
+    assert not getattr(second, '_execution_observed', False)
+    assert await first.send_message() == {'message_id': 'ok'}
+    assert await second.send_message() == {'message_id': 'ok'}
diff --git a/tests/unit_tests/telemetry/test_telemetry.py b/tests/unit_tests/telemetry/test_telemetry.py
index 291264f61..87699a18d 100644
--- a/tests/unit_tests/telemetry/test_telemetry.py
+++ b/tests/unit_tests/telemetry/test_telemetry.py
@@ -6,7 +6,6 @@ Tests cover:
 - Early return conditions (disabled, empty config, no server)
 - URL construction (with actual URL verification)
 - HTTP request success/failure scenarios
-- Source code bug: send_tasks should be instance variable
 """
 
 from __future__ import annotations
@@ -431,6 +430,7 @@ class TestHTTPScenarios:
         with patch.object(httpx, 'AsyncClient', return_value=mock_client):
             await manager.send({'query_id': 'test'})
 
+        mock_app.logger.warning.assert_not_called()
         mock_app.logger.debug.assert_called()
         # Verify debug message contains URL and status
         debug_call_args = mock_app.logger.debug.call_args[0][0]
@@ -459,8 +459,9 @@ class TestHTTPScenarios:
         with patch.object(httpx, 'AsyncClient', return_value=mock_client):
             await manager.send({'query_id': 'test'})
 
-        mock_app.logger.warning.assert_called()
-        warning_call_args = mock_app.logger.warning.call_args[0][0]
+        mock_app.logger.warning.assert_not_called()
+        mock_app.logger.debug.assert_called()
+        warning_call_args = mock_app.logger.debug.call_args[0][0]
         assert 'status 500' in warning_call_args
 
     @pytest.mark.asyncio
@@ -488,9 +489,9 @@ class TestHTTPScenarios:
             await manager.send({'query_id': 'test'})
 
         # Source code calls warning twice for application errors
-        assert mock_app.logger.warning.call_count >= 1
+        assert mock_app.logger.debug.call_count >= 1
         # Check that one of the calls contains application error info
-        all_warnings = [call[0][0] for call in mock_app.logger.warning.call_args_list]
+        all_warnings = [call[0][0] for call in mock_app.logger.debug.call_args_list]
         assert any('400' in w for w in all_warnings), f'No warning contained error code 400: {all_warnings}'
 
     @pytest.mark.asyncio
@@ -516,8 +517,9 @@ class TestHTTPScenarios:
         with patch.object(httpx, 'AsyncClient', return_value=mock_client):
             await manager.send({'query_id': 'test'})
 
-        mock_app.logger.warning.assert_called()
-        warning_call_args = mock_app.logger.warning.call_args[0][0]
+        mock_app.logger.warning.assert_not_called()
+        mock_app.logger.debug.assert_called()
+        warning_call_args = mock_app.logger.debug.call_args[0][0]
         assert 'timed out' in warning_call_args
 
     @pytest.mark.asyncio
@@ -542,7 +544,8 @@ class TestHTTPScenarios:
             # Should not raise exception
             await manager.send({'query_id': 'test'})
 
-        mock_app.logger.warning.assert_called()
+        mock_app.logger.warning.assert_not_called()
+        mock_app.logger.debug.assert_called()
 
     @pytest.mark.asyncio
     async def test_send_never_raises_exception(self):
@@ -551,7 +554,7 @@ class TestHTTPScenarios:
         mock_app = Mock()
         # Even logger may fail
         mock_app.logger = Mock()
-        mock_app.logger.warning = Mock(side_effect=Exception('Logger failed'))
+        mock_app.logger.debug = Mock(side_effect=Exception('Logger failed'))
 
         manager = telemetry.TelemetryManager(mock_app)
         manager.telemetry_config = {'url': 'https://example.com'}
@@ -625,53 +628,3 @@ class TestAuthenticatedWorkspaceReporter:
             'Authorization': 'Bearer refreshed-workspace-owner-token'
         }
 
-
-class TestStartSendTask:
-    """Tests for start_send_task() method."""
-
-    @pytest.mark.asyncio
-    async def test_start_send_task_creates_task(self):
-        """Test that start_send_task creates an asyncio task."""
-        telemetry = get_telemetry_module()
-        mock_app = Mock()
-        mock_app.logger = Mock()
-        mock_app.instance_config = Mock()
-        mock_app.instance_config.data = {}
-
-        manager = telemetry.TelemetryManager(mock_app)
-        manager.telemetry_config = {}
-
-        await manager.start_send_task({'query_id': 'test'})
-
-        # Task should be added to send_tasks list
-        assert len(manager.send_tasks) >= 1
-
-        # Clean up the task
-        for task in manager.send_tasks:
-            if not task.done():
-                task.cancel()
-        manager.send_tasks.clear()
-
-    @pytest.mark.asyncio
-    async def test_start_send_task_multiple_tasks(self):
-        """Test that multiple tasks are tracked."""
-        telemetry = get_telemetry_module()
-        mock_app = Mock()
-        mock_app.logger = Mock()
-        mock_app.instance_config = Mock()
-        mock_app.instance_config.data = {}
-
-        manager = telemetry.TelemetryManager(mock_app)
-        manager.telemetry_config = {}
-
-        await manager.start_send_task({'query_id': 'test1'})
-        await manager.start_send_task({'query_id': 'test2'})
-        await manager.start_send_task({'query_id': 'test3'})
-
-        assert len(manager.send_tasks) >= 3
-
-        # Clean up
-        for task in manager.send_tasks:
-            if not task.done():
-                task.cancel()
-        manager.send_tasks.clear()
diff --git a/tests/unit_tests/telemetry/test_trace.py b/tests/unit_tests/telemetry/test_trace.py
new file mode 100644
index 000000000..cf7e63701
--- /dev/null
+++ b/tests/unit_tests/telemetry/test_trace.py
@@ -0,0 +1,1014 @@
+"""Unit tests for per-execution execution chains (pkg/telemetry/trace.py + execution.py).
+
+One execution is one chain of independent records: one ``execution_node`` per
+workflow node, emitted the moment the node completes, plus exactly one
+``execution_chain`` when the execution closes. Every record of a chain carries
+the same ``event_id``.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import contextvars
+import json
+import time
+import types
+from datetime import datetime, timedelta, timezone
+from importlib import import_module
+
+
+def get_modules():
+    return (
+        import_module('langbot.pkg.telemetry.trace'),
+        import_module('langbot.pkg.telemetry.execution'),
+    )
+
+
+class FakeManager:
+    """Stand-in for TelemetryManager: records payloads instead of posting them."""
+
+    def __init__(self, config=None):
+        self.telemetry_config = (
+            {'url': 'https://space.example.test', 'disable_telemetry': False} if config is None else config
+        )
+        self.sent: list[dict] = []
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+
+class RejectingManager(FakeManager):
+    """Stand-in whose sender refuses every batch until told to accept them."""
+
+    def __init__(self, config=None):
+        super().__init__(config)
+        self.rejecting = True
+        self.attempts = 0
+
+    async def send(self, payload: dict) -> bool:
+        self.attempts += 1
+        if self.rejecting:
+            return False
+        self.sent.append(payload)
+        return True
+
+
+def make_counters(config=None):
+    _, execution = get_modules()
+    manager = FakeManager(config)
+    return manager, execution.ExecutionCounters(manager)
+
+
+def make_ap(counters):
+    return types.SimpleNamespace(telemetry=types.SimpleNamespace(execution=counters))
+
+
+def trace_config(**overrides):
+    config = {'url': 'https://space.example.test', 'execution_trace': 'all'}
+    config.update(overrides)
+    return config
+
+
+def records_sent(manager):
+    """Flatten the buffered-record batches a FakeManager received."""
+    return [record for payload in manager.sent for record in payload.get('records', [])]
+
+
+def nodes(counters):
+    return [record for record in counters.records if record.get('event_type') == 'execution_node']
+
+
+def chains(counters):
+    return [record for record in counters.records if record.get('event_type') == 'execution_chain']
+
+
+CONTEXT = types.SimpleNamespace(instance_uuid='instance-1', workspace_uuid='workspace-1')
+
+STAGE = {
+    'family': 'platform_event',
+    'operation': 'message.received',
+    'mode': 'none',
+    'adapter': 'AiocqhttpAdapter',
+    'outcome': 'success',
+}
+
+NODE_KEYS = {
+    'event_type',
+    'schema',
+    'event_id',
+    'instance_id',
+    'workspace_uuid',
+    'version',
+    'edition',
+    'timestamp',
+    'debug',
+    'sample',
+    'node_id',
+    'parent_node_id',
+    'root',
+    'seq',
+    'family',
+    'operation',
+    'mode',
+    'adapter',
+    'runner',
+    'outcome',
+    'error',
+    'route_ref',
+    'run_id',
+    'synthetic',
+    'started_at',
+    'ended_at',
+}
+
+
+class TestTraceIdentity:
+    def test_nested_bind_reuses_the_in_flight_trace(self):
+        trace, _ = get_modules()
+        outer = trace.bind()
+        inner = trace.bind()
+        try:
+            assert outer.created is True
+            assert inner.created is False
+            assert inner.state is outer.state
+            assert trace.current() is outer.state
+        finally:
+            trace.unbind_root(inner)
+            trace.unbind_root(outer)
+        assert trace.current() is None
+
+    def test_unbind_root_leaves_a_nested_trace_attached(self):
+        trace, _ = get_modules()
+        outer = trace.bind()
+        inner = trace.bind()
+        assert trace.unbind_root(inner) is False
+        assert trace.current() is outer.state
+        assert trace.unbind_root(outer) is True
+        assert trace.current() is None
+        # A repeated unbind of the same binding must never raise.
+        assert trace.unbind_root(outer) is True
+
+    def test_event_id_is_the_execution_id(self):
+        trace, _ = get_modules()
+        binding = trace.bind('exec-1234')
+        try:
+            assert binding.created is True
+            assert binding.state.event_id == 'exec-1234'
+        finally:
+            trace.unbind_root(binding)
+
+    def test_nested_boundary_keeps_the_enclosing_execution_id(self):
+        trace, _ = get_modules()
+        outer = trace.bind('event-1')
+        inner = trace.bind('query-9')
+        try:
+            assert inner.created is False
+            assert inner.state.event_id == 'event-1'
+        finally:
+            trace.unbind_root(inner)
+            trace.unbind_root(outer)
+
+    def test_route_and_run_are_only_pinned_inside_the_block(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind()
+        token = trace.set_run('run-1')
+        try:
+            with trace.scope(route_ref='agent:agent-1'):
+                counters.record(
+                    CONTEXT,
+                    family='runner',
+                    operation='execute',
+                    mode='agent',
+                    adapter='',
+                    runner='langbot/runner',
+                    outcome='success',
+                    synthetic=False,
+                )
+            counters.record(
+                CONTEXT,
+                family='platform_api',
+                operation='send_message',
+                mode='agent',
+                adapter='AiocqhttpAdapter',
+                runner='',
+                outcome='success',
+                synthetic=False,
+            )
+        finally:
+            trace.reset_run(token)
+            trace.unbind_root(binding)
+        assert [(row['route_ref'], row['run_id']) for row in nodes(counters)] == [
+            ('agent:agent-1', 'run-1'),
+            ('', 'run-1'),
+        ]
+
+
+class TestTraceRecording:
+    def test_stage_carries_sequence_identity_and_timestamps(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind()
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.record(CONTEXT, **{**STAGE, 'family': 'runner', 'operation': 'execute', 'mode': 'pipeline'})
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        rows = nodes(counters)
+        assert [row['seq'] for row in rows] == [0, 1]
+        # A plain observation starts and ends at the same instant.
+        assert rows[0]['started_at'] == rows[0]['ended_at']
+        assert {(row['instance_id'], row['workspace_uuid']) for row in rows} == {('instance-1', 'workspace-1')}
+        chain = chains(counters)[0]
+        assert chain['outcome'] == 'success'
+        assert chain['stage_count'] == 2
+
+    def test_oversized_identifier_never_becomes_a_stage(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind()
+        try:
+            counters.record(CONTEXT, **{**STAGE, 'operation': 'x' * 513})
+            counters.record(CONTEXT, **{**STAGE, 'family': 'not-a-family'})
+        finally:
+            trace.unbind_root(binding)
+        assert nodes(counters) == []
+
+    def test_observation_without_an_owning_execution_is_dropped(self):
+        _, _ = get_modules()
+        _, counters = make_counters()
+        # An observation that belongs to no execution is discarded instead of
+        # being aggregated into a synthetic bucket.
+        counters.record(CONTEXT, **STAGE)
+        assert counters.records == []
+        assert counters.registry == {}
+
+
+class TestTraceNodeScope:
+    """Workflow step nodes: a step's own record and the observations it calls."""
+
+    @staticmethod
+    def _step(**overrides):
+        return {
+            'family': 'runner',
+            'operation': 'execute',
+            'mode': 'agent',
+            'runner': 'runner-1',
+            'outcome': 'success',
+            **overrides,
+        }
+
+    def test_nested_observation_carries_the_step_node_as_parent(self):
+        trace, _ = get_modules()
+        _, counters = make_counters()
+        binding = trace.bind()
+        try:
+            with trace.stage_scope() as step:
+                counters.record(CONTEXT, node=step, **self._step())
+                counters.record(CONTEXT, **{**STAGE, 'family': 'platform_api', 'operation': 'send_message'})
+        finally:
+            trace.unbind_root(binding)
+
+        step_row, api_row = nodes(counters)
+        assert step_row['node_id'] == step
+        assert step_row['parent_node_id'] == ''  # a root step carries no parent
+        assert step_row['root'] is True
+        assert api_row['parent_node_id'] == step
+        assert api_row['root'] is False
+        assert api_row['node_id'] != step  # observations are not step nodes
+
+    def test_nested_step_records_its_enclosing_step_as_parent(self):
+        trace, _ = get_modules()
+        _, counters = make_counters()
+        binding = trace.bind()
+        try:
+            with trace.stage_scope() as outer:
+                counters.record(CONTEXT, node=outer, **self._step(mode='pipeline'))
+                with trace.stage_scope() as inner:
+                    counters.record(CONTEXT, node=inner, **self._step())
+                    counters.record(CONTEXT, **{**STAGE, 'family': 'platform_api', 'operation': 'send_message'})
+        finally:
+            trace.unbind_root(binding)
+
+        outer_row, inner_row, api_row = nodes(counters)
+        assert (outer_row['node_id'], outer_row['parent_node_id']) == (outer, '')
+        # The step's own record sits under its enclosing step, never under itself.
+        assert (inner_row['node_id'], inner_row['parent_node_id']) == (inner, outer)
+        assert api_row['parent_node_id'] == inner and api_row['node_id'] != inner
+
+    def test_sequential_steps_get_distinct_ids_and_roots_have_no_parent(self):
+        trace, _ = get_modules()
+        _, counters = make_counters()
+        binding = trace.bind()
+        try:
+            with trace.stage_scope() as first:
+                counters.record(CONTEXT, node=first, **self._step())
+            with trace.stage_scope() as second:
+                counters.record(CONTEXT, node=second, **self._step())
+        finally:
+            trace.unbind_root(binding)
+
+        assert first and second and first != second
+        rows = nodes(counters)
+        assert [row['node_id'] for row in rows] == [first, second]
+        assert [row['seq'] for row in rows] == [0, 1]
+        assert rows[0]['root'] is True
+        assert all(row['parent_node_id'] == '' for row in rows)
+
+    def test_cross_task_observation_joins_the_owning_execution(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = execution.bind_trace(make_ap(counters), 'exec-cross')
+        try:
+            with trace.stage_scope() as lane:
+                counters.record(
+                    CONTEXT, family='pipeline', operation='run', mode='pipeline', outcome='success', node=lane
+                )
+                # A plugin/RPC observation arrives on a task that never had the
+                # trace context; it is resolved through the execution registry and
+                # still lands on the owning chain.
+                foreign = contextvars.Context()
+                foreign.run(execution.set_execution_id, 'exec-cross')
+                foreign.run(
+                    counters.record,
+                    CONTEXT,
+                    family='platform_api',
+                    operation='send_message',
+                    mode='pipeline',
+                    outcome='success',
+                )
+        finally:
+            trace.unbind_root(binding)
+
+        lane_row, foreign_row = nodes(counters)
+        assert lane_row['node_id'] == lane and lane_row['parent_node_id'] == '' and lane_row['root'] is True
+        # The cross-task node joins the same execution and hangs off the step the
+        # chain had open - the step that caused the work - instead of mirroring
+        # any task's own stack.
+        assert foreign_row['event_id'] == 'exec-cross'
+        assert foreign_row['parent_node_id'] == lane
+        assert foreign_row['root'] is False
+
+    def test_cross_task_observation_attaches_to_the_innermost_open_step(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = execution.bind_trace(make_ap(counters), 'exec-nested')
+        try:
+            with trace.stage_scope() as outer:
+                counters.record(CONTEXT, **{**STAGE, 'node': outer})
+                with trace.stage_scope() as inner:
+                    counters.record(CONTEXT, **{**STAGE, 'node': inner})
+                    foreign = contextvars.Context()
+                    foreign.run(execution.set_execution_id, 'exec-nested')
+                    foreign.run(counters.record, CONTEXT, **{**STAGE, 'family': 'tool', 'operation': 'exec'})
+        finally:
+            trace.unbind_root(binding)
+
+        outer_row, inner_row, tool_row = nodes(counters)
+        assert outer_row['parent_node_id'] == ''
+        assert inner_row['parent_node_id'] == outer
+        # The lane that caused the cross-task work is the innermost one open.
+        assert tool_row['parent_node_id'] == inner
+
+    def test_cross_task_observation_after_the_lane_closed_is_a_root(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = execution.bind_trace(make_ap(counters), 'exec-late')
+        try:
+            with trace.stage_scope() as lane:
+                counters.record(CONTEXT, **{**STAGE, 'node': lane})
+            # A late callback arrives after the owning step returned: nothing is
+            # open anymore, so the node is a root of the chain rather than a
+            # child of a finished step.
+            foreign = contextvars.Context()
+            foreign.run(execution.set_execution_id, 'exec-late')
+            foreign.run(counters.record, CONTEXT, family='platform_api', operation='reply', mode='pipeline')
+        finally:
+            trace.unbind_root(binding)
+
+        lane_row, late_row = nodes(counters)
+        assert late_row['parent_node_id'] == '' and late_row['root'] is False
+        assert lane_row['root'] is True
+
+    def test_cross_task_observation_without_an_open_step_has_no_parent(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = execution.bind_trace(make_ap(counters), 'exec-bare')
+        try:
+            foreign = contextvars.Context()
+            foreign.run(execution.set_execution_id, 'exec-bare')
+            foreign.run(
+                counters.record,
+                CONTEXT,
+                family='platform_api',
+                operation='send_message',
+                mode='pipeline',
+                outcome='success',
+            )
+        finally:
+            trace.unbind_root(binding)
+
+        row = nodes(counters)[0]
+        assert row['parent_node_id'] == '' and row['root'] is True
+
+    def test_scope_outside_any_trace_is_inert(self):
+        trace, _ = get_modules()
+        with trace.stage_scope() as node:
+            assert node == ''
+            assert trace.current_parent() == ''
+
+    def test_records_outside_any_scope_are_root_nodes(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-legacy')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+
+        node = nodes(counters)[0]
+        assert set(node) == NODE_KEYS
+        assert node['node_id'] and node['parent_node_id'] == '' and node['root'] is True
+        assert node['seq'] == 0
+        # Node placement lives on node records only; the chain carries none.
+        assert 'node_id' not in json.dumps(chains(counters)[0], sort_keys=True)
+
+
+class TestTraceDelivery:
+    def test_close_emits_one_terminal_chain_record_per_execution(self):
+        trace, _ = get_modules()
+        manager, counters = make_counters(trace_config())
+        binding = trace.bind('exec-1')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.record(
+                CONTEXT, **{**STAGE, 'family': 'runner', 'operation': 'execute', 'mode': 'agent', 'runner': 'runner-1'}
+            )
+            counters.close_trace(binding.state, 'event_done')
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+
+        # Closing buffers; nothing is sent before the flush cadence.
+        assert manager.sent == []
+        assert len(nodes(counters)) == 2
+        chain_rows = chains(counters)
+        assert len(chain_rows) == 1
+        record = chain_rows[0]
+        assert record['schema'] == 2
+        assert record['event_id'] == 'exec-1'
+        assert record['instance_id'] == 'instance-1'
+        assert record['workspace_uuid'] == 'workspace-1'
+        assert record['closed_by'] == 'event_done'
+        assert record['outcome'] == 'success'
+        assert record['stage_count'] == 2
+        assert isinstance(record['duration_ms'], int) and record['duration_ms'] >= 0
+        assert 'query_id' not in record and 'features' not in record
+
+    def test_sampling_decision_is_deterministic_and_keeps_failures(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config(execution_trace='sampled', execution_trace_sample=2))
+        in_id = next(i for i in (f'sample-{n}' for n in range(64)) if counters._sampling_decision(i)[2])
+        out_id = next(i for i in (f'sample-{n}' for n in range(64)) if not counters._sampling_decision(i)[2])
+        # Same id, same answer: the decision is a pure function of the identity.
+        assert counters._sampling_decision(in_id) == counters._sampling_decision(in_id)
+
+        excluded = trace.bind(out_id)
+        counters.record(CONTEXT, **STAGE)
+        assert excluded.state.emit_nodes is False
+        assert counters._chain_emitted(excluded.state) is False
+        counters.close_trace(excluded.state, 'event_done')
+        trace.unbind_root(excluded)
+        assert counters.records == []
+
+        included = trace.bind(in_id)
+        counters.record(CONTEXT, **STAGE)
+        assert included.state.emit_nodes is True
+        counters.close_trace(included.state, 'event_done')
+        trace.unbind_root(included)
+        assert [record['event_type'] for record in counters.records] == ['execution_node', 'execution_chain']
+
+        counters.records.clear()
+        broken = trace.bind(out_id)
+        counters.record(CONTEXT, **{**STAGE, 'outcome': 'timeout'})
+        assert counters._chain_emitted(broken.state) is True
+        counters.close_trace(broken.state, 'event_done')
+        trace.unbind_root(broken)
+        assert [record['event_type'] for record in counters.records] == ['execution_chain']
+        assert counters.records[0]['outcome'] == 'timeout'
+
+    def test_failures_mode_keeps_failed_and_debug_chains_only(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config(execution_trace='failures'))
+
+        failed = trace.bind('exec-failed')
+        counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'})
+        counters.close_trace(failed.state, 'event_done')
+        trace.unbind_root(failed)
+
+        debug = trace.bind('exec-debug')
+        counters.configure(debug.state, debug=True)
+        counters.record(CONTEXT, **STAGE)
+        counters.close_trace(debug.state, 'event_done')
+        trace.unbind_root(debug)
+
+        sampled_out = trace.bind('exec-sampled')
+        counters.record(CONTEXT, **STAGE)
+        counters.close_trace(sampled_out.state, 'event_done')
+        trace.unbind_root(sampled_out)
+
+        assert [record['event_id'] for record in chains(counters)] == ['exec-failed', 'exec-debug']
+        assert chains(counters)[0]['outcome'] == 'failed'
+        # Its node was sampled out, but the broken chain still explains itself.
+        assert chains(counters)[0]['dropped_nodes'] == 1
+        assert chains(counters)[1]['debug'] is True
+        assert [record['event_id'] for record in nodes(counters)] == ['exec-debug']
+
+    def test_off_mode_and_opt_out_buffer_nothing(self):
+        trace, _ = get_modules()
+        manager, counters = make_counters(trace_config(execution_trace='off'))
+        binding = trace.bind()
+        try:
+            counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'})
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        assert manager.sent == []
+        assert counters.records == []
+
+        disabled_manager, disabled = make_counters(
+            {'url': 'https://space.example.test', 'disable_telemetry': True, 'execution_trace': 'all'}
+        )
+        binding = trace.bind()
+        try:
+            disabled.record(CONTEXT, **STAGE)
+            disabled.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        assert disabled_manager.sent == []
+        assert disabled.records == []
+        assert disabled.registry == {}
+
+    async def test_shutdown_does_not_claim_an_unfinished_execution(self):
+        trace, _ = get_modules()
+        manager, counters = make_counters(trace_config())
+        binding = trace.bind('exec-open')
+        try:
+            counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'})
+            await counters.shutdown()
+        finally:
+            trace.unbind_root(binding)
+        # The already-emitted node is flushed, but shutdown never fabricates a
+        # terminal chain for an execution that never completed.
+        assert [record['event_type'] for record in records_sent(manager)] == ['execution_node']
+        assert counters.records == []
+        assert chains(counters) == []
+        assert counters.registry == {}
+
+    async def test_flush_sends_one_batch_of_complete_records(self):
+        _, execution = get_modules()
+        manager, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        for index in range(2):
+            with execution.ingress(ap, 'event_done', CONTEXT, execution_id=f'exec-{index}'):
+                counters.record(CONTEXT, **STAGE)
+        assert len(counters.records) == 4
+
+        await counters.flush()
+        assert len(manager.sent) == 1
+        payload = manager.sent[0]
+        assert set(payload) == {'records'}
+        assert [record['event_type'] for record in payload['records']] == [
+            'execution_node',
+            'execution_chain',
+            'execution_node',
+            'execution_chain',
+        ]
+        assert [record['event_id'] for record in payload['records']] == ['exec-0', 'exec-0', 'exec-1', 'exec-1']
+        await counters.shutdown()
+
+    async def test_flush_interval_is_honoured(self):
+        _, execution = get_modules()
+        manager, counters = make_counters(trace_config(telemetry_flush_seconds=1))
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', CONTEXT, execution_id='exec-timed'):
+            counters.record(CONTEXT, **STAGE)
+        assert counters.flush_seconds() == 1
+        assert manager.sent == []
+        await asyncio.sleep(1.3)
+        sent = records_sent(manager)
+        assert [record['event_type'] for record in sent] == ['execution_node', 'execution_chain']
+        assert {record['event_id'] for record in sent} == {'exec-timed'}
+        await counters.shutdown()
+
+    def test_flush_interval_config_is_bounded(self):
+        _, counters = make_counters(trace_config())
+        assert counters.flush_seconds() == 180
+        counters.manager.telemetry_config = trace_config(telemetry_flush_seconds='nope')
+        assert counters.flush_seconds() == 180
+        counters.manager.telemetry_config = trace_config(telemetry_flush_seconds=0)
+        assert counters.flush_seconds() == 180
+        counters.manager.telemetry_config = trace_config(telemetry_flush_seconds=300)
+        assert counters.flush_seconds() == 300
+
+    def test_buffered_record_cap_drops_oldest(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        original = execution.MAX_BUFFERED_RECORDS
+        execution.MAX_BUFFERED_RECORDS = 3
+        try:
+            for index in range(5):
+                binding = trace.bind(f'exec-{index}')
+                counters.record(CONTEXT, **STAGE)
+                trace.unbind_root(binding)
+        finally:
+            execution.MAX_BUFFERED_RECORDS = original
+        assert [record['event_id'] for record in nodes(counters)] == ['exec-2', 'exec-3', 'exec-4']
+        assert counters.dropped == 2
+
+
+class TestIngress:
+    def test_ingress_closes_only_the_trace_it_started(self):
+        _, execution = get_modules()
+        manager, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', execution_id='exec-outer'):
+            with execution.ingress(ap, 'pipeline_done', execution_id='exec-inner'):
+                counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'})
+        assert [record['event_type'] for record in counters.records] == ['execution_node', 'execution_chain']
+        assert all(record['event_id'] == 'exec-outer' for record in counters.records)
+        assert chains(counters)[0]['closed_by'] == 'event_done'
+        assert manager.sent == []
+
+    def test_ingress_without_telemetry_is_a_no_op(self):
+        _, execution = get_modules()
+        with execution.ingress(types.SimpleNamespace(), 'event_done', execution_id='exec-x'):
+            pass
+
+    def test_ingress_registers_every_execution_alias(self):
+        _, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', execution_id='event-1') as outer:
+            with execution.ingress(ap, 'pipeline_done', execution_id='query-1') as inner:
+                assert inner.state is outer.state
+        assert counters.registry == {}
+        # Closing the owning trace unregisters every alias it held.
+        assert counters.registry_deadlines == {}
+
+
+class TestCrossTaskAttribution:
+    def test_registry_joins_nodes_from_a_task_without_a_trace(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        binding = execution.bind_trace(ap, 'run-1')
+        trace.unbind_root(binding)
+        # The plugin/RPC task has no inherited trace context.
+        assert trace.current() is None
+
+        counters.record(CONTEXT, **{**STAGE, 'family': 'platform_api', 'operation': 'send_message'})
+        counters.record(
+            CONTEXT, **{**STAGE, 'family': 'platform_api', 'operation': 'send_message'}, execution_id='run-1'
+        )
+        rows = nodes(counters)
+        assert len(rows) == 1
+        assert rows[0]['operation'] == 'send_message' and rows[0]['event_id'] == 'run-1'
+
+        counters.close_trace(binding.state, 'event_done')
+        assert chains(counters)[0]['event_id'] == 'run-1'
+
+    def test_registry_aliases_a_nested_execution(self):
+        _, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', CONTEXT, execution_id='event-1'):
+            with execution.ingress(ap, 'pipeline_done', CONTEXT, execution_id='query-1'):
+                pass
+            assert set(counters.registry) == {'event-1', 'query-1'}
+            counters.record(CONTEXT, **STAGE, execution_id='query-1')
+        rows = nodes(counters)
+        assert len(rows) == 1 and rows[0]['event_id'] == 'event-1'
+
+    def test_contextvar_execution_id_is_used_when_no_kwarg(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        binding = execution.bind_trace(ap, 'run-2')
+        trace.unbind_root(binding)
+        token = execution.set_execution_id('run-2')
+        try:
+            counters.record(CONTEXT, **STAGE)
+        finally:
+            execution.reset_execution_id(token)
+        rows = nodes(counters)
+        assert len(rows) == 1 and rows[0]['event_id'] == 'run-2'
+
+    def test_registry_alias_is_swept_when_never_closed(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        binding = execution.bind_trace(ap, 'run-3')
+        trace.unbind_root(binding)
+        counters.registry_deadlines['run-3'] = 0
+        counters._sweep_registry(time.monotonic())
+        assert counters.registry == {}
+        assert counters.registry_deadlines == {}
+
+
+class TestTraceDispatch:
+    async def test_flush_awaits_the_manager_send(self):
+        _, execution = get_modules()
+        manager = AsyncSendManager(trace_config())
+        counters = execution.ExecutionCounters(manager)
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', CONTEXT, execution_id='exec-async'):
+            counters.record(CONTEXT, **STAGE)
+        await counters.flush()
+        sent = records_sent(manager)
+        assert [record['event_type'] for record in sent] == ['execution_node', 'execution_chain']
+        assert all(record['event_id'] == 'exec-async' for record in sent)
+        await counters.shutdown()
+
+
+class AsyncSendManager(FakeManager):
+    """Stand-in whose send is a coroutine, like TelemetryManager."""
+
+    async def send(self, payload: dict) -> bool:
+        self.sent.append(payload)
+        return True
+
+
+class TestTraceFailureDetail:
+    def test_failed_stage_carries_the_reason(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-fail')
+        try:
+            counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'}, error='boom: sandbox exited 127')
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        assert nodes(counters)[0]['error'] == 'boom: sandbox exited 127'
+        # The detail lives on the node; the chain carries the terminal outcome.
+        assert chains(counters)[0]['outcome'] == 'failed'
+
+    def test_successful_stage_reports_no_error(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-ok')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        assert nodes(counters)[0]['error'] == ''
+        assert chains(counters)[0]['error'] == ''
+
+    def test_error_detail_is_bounded_and_single_line(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-noisy')
+        try:
+            counters.record(CONTEXT, **{**STAGE, 'outcome': 'failed'}, error='x' * 1000 + '\n\tboom')
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        detail = chains(counters)[0]['error']
+        assert len(detail) <= 200
+        assert '\n' not in detail and '\t' not in detail
+
+
+class TestIngressFailureDetail:
+    def test_ingress_records_the_failure_reason(self):
+        _, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        raised = False
+        try:
+            with execution.ingress(ap, 'event_done', execution_id='exec-boom'):
+                counters.record(CONTEXT, **STAGE)
+                raise RuntimeError('dispatch exploded')
+        except RuntimeError:
+            raised = True
+        assert raised is True
+        chain = chains(counters)[0]
+        assert chain['error'].startswith('RuntimeError: dispatch exploded')
+        assert chain['outcome'] == 'failed'
+        assert chain['closed_by'] == 'event_done'
+
+    def test_successful_ingress_still_reports_success(self):
+        _, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        with execution.ingress(ap, 'event_done', execution_id='exec-fine'):
+            counters.record(CONTEXT, **STAGE)
+        chain = chains(counters)[0]
+        assert chain['error'] == ''
+        assert chain['outcome'] == 'success'
+
+
+class TestFailureOnlyTrace:
+    def test_failure_without_stages_still_uploads(self):
+        _, execution = get_modules()
+        # failures mode: nodes are not reported, but a broken chain always is.
+        _, counters = make_counters({'url': 'https://space.example.test', 'execution_trace': 'failures'})
+        ap = make_ap(counters)
+        try:
+            with execution.ingress(ap, 'pipeline_done', CONTEXT, execution_id='exec-empty'):
+                raise RuntimeError('pipeline exploded before any stage')
+        except RuntimeError:
+            pass
+        rows = chains(counters)
+        assert len(rows) == 1
+        record = rows[0]
+        assert record['error'].startswith('RuntimeError: pipeline exploded')
+        assert record['outcome'] == 'failed'
+        assert record['stage_count'] == 0
+        assert nodes(counters) == []
+
+    def test_cancelled_and_timeout_chains_report_their_outcome(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        for outcome, reason in (('cancelled', 'run cancelled by operator'), ('timeout', 'deadline exhausted')):
+            binding = trace.bind(f'exec-{outcome}')
+            try:
+                counters.record(CONTEXT, **{**STAGE, 'outcome': outcome}, error=reason)
+                counters.close_trace(binding.state, 'runner_done')
+            finally:
+                trace.unbind_root(binding)
+            record = chains(counters)[-1]
+            assert record['event_id'] == f'exec-{outcome}'
+            assert record['outcome'] == outcome
+            assert nodes(counters)[-1]['error'] == reason
+
+
+class TestLegacyRecordFields:
+    def test_chain_carries_execution_scoped_columns(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-legacy')
+        binding.state.started_at = (datetime.now(timezone.utc) - timedelta(seconds=2)).isoformat()
+        binding.state.model_name = 'gpt-4o'
+        binding.state.pipeline_plugins = ['plugin:a']
+        binding.state.runner_category = 'cloud'
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.record(
+                CONTEXT, **{**STAGE, 'family': 'runner', 'operation': 'execute', 'mode': 'agent', 'runner': 'runner-1'}
+            )
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        record = chains(counters)[0]
+        assert record['model_name'] == 'gpt-4o'
+        assert record['pipeline_plugins'] == ['plugin:a']
+        assert record['runner_category'] == 'cloud'
+        assert record['adapter'] == 'AiocqhttpAdapter'
+        assert record['runner'] == 'runner-1'
+        assert record['duration_ms'] >= 1000
+
+    def test_legacy_string_fields_are_never_null(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-blank')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        record = chains(counters)[0]
+        assert record['model_name'] == ''
+        assert record['runner'] == ''
+        assert record['runner_category'] == ''
+        assert isinstance(record['duration_ms'], int) and record['duration_ms'] >= 0
+        assert record['pipeline_plugins'] is None
+
+
+class TestNoTraceTtl:
+    def test_nodes_more_than_two_minutes_apart_still_emit(self):
+        trace, _ = get_modules()
+        _, counters = make_counters(trace_config())
+        binding = trace.bind('exec-long')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            # A gap far beyond the old 120s TTL must not sweep the chain: there is
+            # no local trace deadline any more, so both nodes and the terminal
+            # chain are still emitted.
+            binding.state.started_at = (datetime.now(timezone.utc) - timedelta(seconds=200)).isoformat()
+            counters.record(CONTEXT, **{**STAGE, 'operation': 'still.alive'})
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        rows = nodes(counters)
+        assert [record['seq'] for record in rows] == [0, 1]
+        assert [record['operation'] for record in rows] == ['message.received', 'still.alive']
+        chain = chains(counters)[0]
+        assert chain['stage_count'] == 2
+        assert chain['duration_ms'] >= 120_000
+
+
+class TestFlushDelivery:
+    async def test_rejected_batch_stays_buffered_and_is_retried(self):
+        trace, execution = get_modules()
+        manager = RejectingManager(trace_config())
+        counters = execution.ExecutionCounters(manager)
+        binding = trace.bind('exec-retry')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            assert await counters.flush() is False
+            assert manager.attempts == execution.MAX_FLUSH_ATTEMPTS
+            # The record was never lost: it is still buffered and unsent.
+            assert len(counters.records) == 1 and manager.sent == []
+            manager.rejecting = False
+            assert await counters.flush() is True
+        finally:
+            trace.unbind_root(binding)
+        assert counters.records == []
+        assert [record['event_id'] for record in records_sent(manager)] == ['exec-retry']
+        await counters.shutdown()
+
+    async def test_health_reports_buffered_sent_and_dropped(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        binding = execution.bind_trace(ap, 'exec-health')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            health = counters.health()
+            assert health['records_buffered'] == 1
+            assert health['records_sent'] == 0
+            assert health['records_dropped'] == 0
+            assert health['executions_in_flight'] == 1
+            await counters.flush()
+            health = counters.health()
+            assert health['records_buffered'] == 0
+            assert health['records_sent'] == 1
+        finally:
+            trace.unbind_root(binding)
+
+        # Overflowing the outbound buffer is surfaced as dropped records.
+        original = execution.MAX_BUFFERED_RECORDS
+        execution.MAX_BUFFERED_RECORDS = 1
+        try:
+            for index in range(3):
+                inner = trace.bind(f'exec-drop-{index}')
+                counters.record(CONTEXT, **STAGE)
+                trace.unbind_root(inner)
+        finally:
+            execution.MAX_BUFFERED_RECORDS = original
+        health = counters.health()
+        assert health['records_dropped'] == 2
+        assert health['records_buffered'] == 1
+        await counters.shutdown()
+
+
+class TestSiblingTaskIsolation:
+    async def test_sibling_stage_scopes_do_not_corrupt_parent_links(self):
+        trace, execution = get_modules()
+        _, counters = make_counters(trace_config())
+        ap = make_ap(counters)
+        opened = {}
+
+        async def work(name):
+            with trace.stage_scope() as node:
+                opened[name] = node
+                counters.record(CONTEXT, **{**STAGE, 'operation': name, 'node': node})
+
+        async def scenario():
+            with execution.ingress(ap, 'event_done', CONTEXT, execution_id='exec-siblings'):
+                await asyncio.gather(work('alpha'), work('beta'))
+            snapshot = list(counters.records)
+            await counters.shutdown()
+            return snapshot
+
+        records = await scenario()
+        rows = {record['operation']: record for record in records if record['event_type'] == 'execution_node'}
+        assert opened['alpha'] != opened['beta']
+        # Each sibling's step is its own root: neither becomes the other's parent.
+        assert rows['alpha']['node_id'] == opened['alpha']
+        assert rows['beta']['node_id'] == opened['beta']
+        assert rows['alpha']['parent_node_id'] == ''
+        assert rows['beta']['parent_node_id'] == ''
+        assert [record['event_type'] for record in records].count('execution_chain') == 1
+
+
+class TestSamplingShape:
+    def test_platform_prefixed_event_id_is_emitted_at_default_config(self):
+        trace, _ = get_modules()
+        _, counters = make_counters()  # no execution_trace override -> default mode
+        assert counters.trace_mode() == 'all'
+        # The identity is hashed, not parsed: a platform-prefixed id samples the
+        # same way any other id shape does.
+        assert counters._sampling_decision('platform:bot-1:message-1') == ('all', 1, True)
+        binding = trace.bind('platform:bot-1:message-1')
+        try:
+            counters.record(CONTEXT, **STAGE)
+            counters.close_trace(binding.state, 'event_done')
+        finally:
+            trace.unbind_root(binding)
+        assert [record['event_type'] for record in counters.records] == ['execution_node', 'execution_chain']
+        assert all(record['event_id'] == 'platform:bot-1:message-1' for record in counters.records)
diff --git a/tests/unit_tests/test_backend_i18n.py b/tests/unit_tests/test_backend_i18n.py
new file mode 100644
index 000000000..a43f4113d
--- /dev/null
+++ b/tests/unit_tests/test_backend_i18n.py
@@ -0,0 +1,107 @@
+"""Runtime language isolation, catalog completeness, and Pipeline overrides."""
+
+import asyncio
+import json
+import string
+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+from langbot.pkg import i18n
+from langbot.pkg.i18n.adapters import adapter_t, localized_adapter
+from langbot.pkg.provider.modelmgr.errors import RequesterError
+
+
+def test_catalogs_have_matching_keys_and_placeholders():
+    folder = Path(i18n.__file__).parent / 'locales'
+    catalogs = {
+        locale: json.loads((folder / f'{locale}.json').read_text(encoding='utf-8'))
+        for locale in i18n.SUPPORTED_LANGUAGES
+    }
+    english = catalogs['en_US']
+    formatter = string.Formatter()
+
+    def fields(text):
+        return {name for _, name, _, _ in formatter.parse(text) if name is not None}
+
+    for locale, catalog in catalogs.items():
+        assert catalog.keys() == english.keys(), locale
+        for key, text in catalog.items():
+            assert text.strip(), (locale, key)
+            assert fields(text) == fields(key), (locale, key)
+
+
+@pytest.mark.asyncio
+async def test_concurrent_requests_and_children_keep_their_language():
+    async def render(locale):
+        with i18n.language_scope(locale):
+
+            async def child():
+                await asyncio.sleep(0)
+                return str(RequesterError(i18n.t('Request timed out')))
+
+            return await asyncio.create_task(child())
+
+    assert await asyncio.gather(render('zh_Hans'), render('ja_JP'), render('en_US')) == [
+        '模型请求失败: 请求超时',
+        'モデルへのリクエストに失敗しました: リクエストがタイムアウトしました',
+        'Model request failed: Request timed out',
+    ]
+    assert i18n.language.get() == 'en_US'
+    with pytest.raises(RuntimeError), i18n.language_scope('zh_Hans'):
+        raise RuntimeError('interrupted')
+    assert i18n.language.get() == 'en_US'
+
+
+def test_pipeline_overrides_inherit_language_and_preserve_literal_braces():
+    query = SimpleNamespace(pipeline_config={'output': {'misc': {}}})
+    with i18n.language_scope('zh_Hans'):
+        assert i18n.pipeline_notice(query, 'failure-hint', 'Request failed.') == '请求失败。'
+        query.pipeline_config['output']['misc']['failure-hint'] = 'Request failed.'
+        assert i18n.pipeline_notice(query, 'failure-hint', 'Request failed.') == '请求失败。'
+        query.pipeline_config['output']['misc']['failure-hint'] = 'My custom notice {literal}'
+        assert i18n.pipeline_notice(query, 'failure-hint', 'Request failed.') == 'My custom notice {literal}'
+        query.pipeline_config['output']['misc']['tool-call-hint'] = 'Using {tools}; keep {other}'
+        assert i18n.pipeline_notice(query, 'tool-call-hint', 'Calling {tools}...', tools='search') == (
+            'Using search; keep {other}'
+        )
+
+
+def test_diagnostics_are_never_parsed_or_rewritten():
+    diagnostic = 'upstream: {"error": "some provider detail"}'
+    with i18n.language_scope('zh_Hans'):
+        assert i18n.t(diagnostic) == diagnostic
+        assert i18n.error_text('permission_denied', 'Missing Workspace permission: workspace.update') == (
+            '没有执行此操作的权限'
+        )
+        assert i18n.t('Please do not replace model not found in user content') == (
+            'Please do not replace model not found in user content'
+        )
+
+
+@pytest.mark.asyncio
+async def test_adapter_callbacks_use_their_owner_not_the_callers_language():
+    app = SimpleNamespace(_message_language_cache={'bot-workspace': ('ja_JP', float('inf'))})
+
+    class Adapter:
+        logger = SimpleNamespace(ap=app, execution_context=SimpleNamespace(workspace_uuid='bot-workspace'))
+
+        @localized_adapter
+        async def reply(self):
+            return i18n.t('Submitted')
+
+    with i18n.language_scope('zh_Hans'):
+        assert await Adapter().reply() == '送信済み'
+        assert adapter_t(Adapter(), 'Submitted') == '送信済み'
+        assert i18n.language.get() == 'zh_Hans'
+
+
+def test_lark_card_uses_selected_language():
+    from langbot.pkg.platform.adapters.lark.stream_card import build_card
+
+    with i18n.language_scope('ja_JP'):
+        card = json.dumps(build_card('<think>reason</think>answer'), ensure_ascii=False)
+        assert '思考過程' in card
+        assert 'AI による生成コンテンツ' in card
+        assert 'Thinking' not in card
diff --git a/tests/unit_tests/test_execution_lease.py b/tests/unit_tests/test_execution_lease.py
new file mode 100644
index 000000000..2a851eec7
--- /dev/null
+++ b/tests/unit_tests/test_execution_lease.py
@@ -0,0 +1,54 @@
+import datetime
+
+import pytest
+import sqlalchemy as sa
+from sqlalchemy.ext.asyncio import create_async_engine
+
+from langbot.pkg.entity.persistence.base import Base
+from langbot.pkg.entity.persistence.agent_run import AgentRun
+from langbot.pkg.entity.persistence.monitoring import MonitoringMessage, MonitoringError
+from langbot.pkg.entity.persistence import workspace, user  # noqa: F401
+from langbot.pkg.utils.execution_lease import reconcile_execution_leases
+
+
+@pytest.mark.asyncio
+async def test_host_lease_reaping_preserves_live_and_external_work():
+    engine = create_async_engine('sqlite+aiosqlite:///:memory:')
+    now = datetime.datetime(2026, 10, 10)
+    async with engine.begin() as conn:
+        await conn.run_sync(Base.metadata.create_all)
+        for name, owner, expiry, status, queue in (
+            ('dead', 'dead-host', -1, 'running', None),
+            ('other-tenant', 'dead-host', -1, 'running', None),
+            ('ours', 'live-host', 30, 'running', None),
+            ('other', 'other-host', 30, 'running', None),
+            ('done', 'dead-host', -1, 'completed', None),
+            ('legacy', None, None, 'running', None),
+            ('remote', 'dead-host', -1, 'running', 'remote'),
+        ):
+            await conn.execute(sa.insert(AgentRun).values(
+                run_id=name, runner_id='runner', workspace_id='other-workspace' if name == 'other-tenant' else 'workspace', status=status,
+                queue_name=queue, execution_owner_id=owner,
+                execution_lease_expires_at=now+datetime.timedelta(seconds=expiry) if expiry else None,
+            ))
+        await conn.execute(sa.insert(MonitoringMessage).values(
+            id='pipeline', workspace_uuid='workspace', timestamp=now,
+            bot_id='bot', bot_name='bot', pipeline_id='pipeline', pipeline_name='pipeline',
+            message_content='hello', session_id='session', status='pending', level='info',
+            execution_owner_id='dead-host', execution_lease_expires_at=now-datetime.timedelta(seconds=1),
+        ))
+    assert await reconcile_execution_leases(engine, 'workspace', now=now, owner_id='live-host') == 2
+    assert await reconcile_execution_leases(engine, 'workspace', now=now, owner_id='live-host') == 0
+    async with engine.connect() as conn:
+        rows = {r['run_id']: r for r in (await conn.execute(sa.select(AgentRun))).mappings()}
+        assert rows['dead']['status'] == 'failed'
+        assert rows['dead']['status_reason']
+        assert rows['dead']['finished_at'] == now
+        assert rows['ours']['execution_lease_expires_at'] == now+datetime.timedelta(seconds=120)
+        for key in ('ours', 'other', 'legacy', 'remote', 'other-tenant'):
+            assert rows[key]['status'] == 'running'
+        assert rows['done']['status'] == 'completed'
+        assert (await conn.execute(sa.select(MonitoringMessage.status))).scalar_one() == 'error'
+        assert (await conn.execute(sa.select(sa.func.count()).select_from(MonitoringError))).scalar_one() == 1
+    await engine.dispose()
+
diff --git a/tests/unit_tests/test_operation_trace_tenant_scope.py b/tests/unit_tests/test_operation_trace_tenant_scope.py
new file mode 100644
index 000000000..2ed132e7e
--- /dev/null
+++ b/tests/unit_tests/test_operation_trace_tenant_scope.py
@@ -0,0 +1,303 @@
+"""Regression tests for the operation-trace tenant scope.
+
+Cloud runtime refuses every database call made without an explicit scope, and the
+audit table is row-level-secured on the bound Workspace. The writer drains its
+queue outside any request lifetime, so these tests drive the real service through
+persistence doubles that enforce exactly that contract: without a bound scope the
+call fails. A writer (or a retention pass) that forgets to bind a scope records
+nothing, which is the defect pinned here.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import contextlib
+import contextvars
+import datetime
+import logging
+import typing
+from types import SimpleNamespace
+
+import pytest
+import sqlalchemy
+from sqlalchemy.ext.asyncio import create_async_engine
+
+from langbot.pkg.api.http.authz import WorkspaceRole
+from langbot.pkg.api.http.context import PrincipalType
+from langbot.pkg.entity.persistence.base import Base
+from langbot.pkg.entity.persistence import metadata as metadata_module
+from langbot.pkg.entity.persistence import operation_log as operation_log_module
+from langbot.pkg.operation_trace import service as service_module
+from langbot.pkg.persistence.tenant_uow import CrossScopeTransactionError, TenantScopeRequiredError
+
+pytestmark = pytest.mark.asyncio
+
+
+@pytest.fixture(autouse=True)
+def _closed_gate(monkeypatch):
+    """The gate latch is process-wide: start every test from a disabled instance."""
+
+    monkeypatch.setattr(service_module, '_global_enabled', False)
+
+
+TRACED_WORKSPACE = '00000000-0000-0000-0000-00000000000a'
+IDLE_WORKSPACE = '00000000-0000-0000-0000-00000000000b'
+MODEL = operation_log_module.WorkspaceOperationLog
+LEVEL_KEY = service_module.OPERATION_LEVEL_KEY
+
+
+class _CloudRuntimePersistence:
+    """Persistence double mirroring the cloud-runtime scope contract."""
+
+    def __init__(self, engine) -> None:
+        self.engine = engine
+        self._bound: contextvars.ContextVar[tuple[str, typing.Any] | None] = contextvars.ContextVar(
+            'bound_workspace', default=None
+        )
+
+    def current_scope(self) -> str | None:
+        bound = self._bound.get()
+        if bound is None:
+            return None
+        workspace_uuid, owner_task = bound
+        if owner_task is not asyncio.current_task():
+            # The real manager refuses a boundary a child task inherited.
+            raise CrossScopeTransactionError('scoped boundaries cannot be inherited by child tasks')
+        return workspace_uuid
+
+    @contextlib.asynccontextmanager
+    async def tenant_scope(self, workspace_uuid: str):
+        token = self._bound.set((workspace_uuid, asyncio.current_task()))
+        try:
+            yield self
+        finally:
+            self._bound.reset(token)
+
+    async def execute_async(self, *args, **kwargs):
+        if self.current_scope() is None:
+            raise TenantScopeRequiredError('cloud persistence access requires an explicit Workspace scope')
+        # The database carries the scope as an RLS setting, so a statement that
+        # writes another Workspace's rows would be rejected as well.
+        async with self.engine.connect() as connection:
+            result = await connection.execute(*args, **kwargs)
+            await connection.commit()
+            return result
+
+
+class _SingleDatabasePersistence:
+    """Persistence double for the unscoped single-database deployment."""
+
+    def __init__(self, engine) -> None:
+        self.engine = engine
+
+    async def execute_async(self, *args, **kwargs):
+        async with self.engine.connect() as connection:
+            result = await connection.execute(*args, **kwargs)
+            await connection.commit()
+            return result
+
+
+@pytest.fixture
+async def trace_env(tmp_path):
+    """A real service against SQLite, with the enforcement double installed."""
+
+    engine = create_async_engine(f'sqlite+aiosqlite:///{tmp_path / "trace-scope.db"}')
+    async with engine.begin() as connection:
+        await connection.run_sync(Base.metadata.create_all)
+    persistence = _CloudRuntimePersistence(engine)
+    application = SimpleNamespace(
+        persistence_mgr=persistence,
+        logger=logging.getLogger('operation-trace-scope-test'),
+        user_service=None,
+        operation_trace_active=False,
+    )
+    try:
+        yield service_module.WorkspaceSettingsService(application), application, engine
+    finally:
+        await engine.dispose()
+
+
+async def _set_level(engine, workspace_uuid: str, level: int) -> None:
+    async with engine.begin() as connection:
+        await connection.execute(
+            sqlalchemy.insert(metadata_module.WorkspaceMetadata).values(
+                workspace_uuid=workspace_uuid,
+                key=LEVEL_KEY,
+                value=str(level),
+            )
+        )
+
+
+async def _rows(engine) -> list:
+    async with engine.begin() as connection:
+        result = await connection.execute(sqlalchemy.select(MODEL))
+    return list(result.all())
+
+
+async def test_writer_binds_workspace_scope_when_recording(trace_env):
+    """A queued trace is written by the background writer holding its own scope."""
+
+    service, application, engine = trace_env
+    rule = service_module.classify('POST', '/api/v1/pipelines')
+
+    queued = await service.record(
+        TRACED_WORKSPACE,
+        rule=rule,
+        level=service_module.OPERATION_LEVEL_MUTATION,
+        http_method='POST',
+        route='/api/v1/pipelines',
+    )
+    assert queued is True
+
+    await service.flush_pending(timeout=5.0)
+
+    rows = await _rows(engine)
+    assert len(rows) == 1
+    assert rows[0].workspace_uuid == TRACED_WORKSPACE
+    assert service._write_failures == 0
+
+
+async def test_retention_prune_binds_workspace_scope(trace_env):
+    """The maintenance loop prunes without a request, so prune binds its own scope."""
+
+    service, application, engine = trace_env
+    stale = datetime.datetime.now(datetime.timezone.utc).replace(tzinfo=None) - datetime.timedelta(days=90)
+    async with engine.begin() as connection:
+        await connection.execute(
+            sqlalchemy.insert(MODEL).values(
+                workspace_uuid=TRACED_WORKSPACE,
+                level=service_module.OPERATION_LEVEL_MUTATION,
+                outcome='ok',
+                created_at=stale,
+            )
+        )
+    await _set_level(engine, TRACED_WORKSPACE, service_module.OPERATION_LEVEL_MUTATION)
+
+    report = await service.prune(TRACED_WORKSPACE, retention_days=30)
+
+    assert report['expired'] == 1
+    assert await _rows(engine) == []
+
+
+async def test_request_wrapper_records_after_the_handler_released_its_scope(trace_env):
+    """The route wrapper traces outside the request scope and must still record.
+
+    Level resolution and the queued write both happen after the handler released
+    its tenant scope. Each of them has to bind the Workspace itself: an unscoped
+    read reported level 0, so every request was dropped without a trace.
+    """
+
+    service, application, engine = trace_env
+    await _set_level(engine, TRACED_WORKSPACE, service_module.OPERATION_LEVEL_MUTATION)
+    context = SimpleNamespace(
+        workspace_uuid=TRACED_WORKSPACE,
+        workspace=SimpleNamespace(role=WorkspaceRole.OWNER.value),
+        principal=SimpleNamespace(
+            principal_type=PrincipalType.ACCOUNT,
+            account_uuid=None,
+            actor_account_uuid=None,
+            api_key_uuid=None,
+        ),
+        auth_type='user_token',
+        request_id='req-scope-test',
+    )
+
+    recorded = await service.record_request(
+        context,
+        method='POST',
+        route=f'/api/v1/workspaces/{TRACED_WORKSPACE}/invitations',
+        status_code=200,
+    )
+
+    assert recorded is True
+    await service.flush_pending(timeout=5.0)
+    rows = await _rows(engine)
+    assert len(rows) == 1
+    assert rows[0].action == 'member_invite'
+
+
+async def test_writer_started_inside_a_request_scope_still_records(trace_env):
+    """The writer outlives the request that first enqueued a trace.
+
+    A task created while its caller held a Workspace boundary inherits it, and a
+    child task may not use an inherited boundary. Starting the writer with the
+    caller's context made it reject every later write instead of recording it.
+    """
+
+    service, application, engine = trace_env
+    await _set_level(engine, TRACED_WORKSPACE, service_module.OPERATION_LEVEL_MUTATION)
+    rule = service_module.classify('POST', '/api/v1/pipelines')
+
+    async with application.persistence_mgr.tenant_scope(TRACED_WORKSPACE):
+        queued = await service.record(
+            TRACED_WORKSPACE,
+            rule=rule,
+            level=service_module.OPERATION_LEVEL_MUTATION,
+            http_method='POST',
+            route='/api/v1/pipelines',
+        )
+        assert queued is True
+        await service.flush_pending(timeout=5.0)
+
+    rows = await _rows(engine)
+    assert len(rows) == 1
+    assert service._write_failures == 0
+
+
+async def test_prime_opens_gate_from_discovered_workspace_bindings(trace_env):
+    """A restart resumes tracing: the gate opens from the Workspaces this instance owns."""
+
+    service, application, engine = trace_env
+    await _set_level(engine, IDLE_WORKSPACE, service_module.OPERATION_LEVEL_NONE)
+    await _set_level(engine, TRACED_WORKSPACE, service_module.OPERATION_LEVEL_MUTATION)
+
+    async def list_active_execution_bindings():
+        return [
+            SimpleNamespace(workspace_uuid=IDLE_WORKSPACE),
+            SimpleNamespace(workspace_uuid=TRACED_WORKSPACE),
+        ]
+
+    application.workspace_service = SimpleNamespace(list_active_execution_bindings=list_active_execution_bindings)
+
+    await service.prime_global_flag()
+
+    assert application.operation_trace_active is True
+
+
+async def test_prime_keeps_gate_closed_when_no_workspace_traces(trace_env):
+    """A disabled instance still pays nothing on the request path."""
+
+    service, application, engine = trace_env
+    await _set_level(engine, IDLE_WORKSPACE, service_module.OPERATION_LEVEL_NONE)
+
+    async def list_active_execution_bindings():
+        return [SimpleNamespace(workspace_uuid=IDLE_WORKSPACE)]
+
+    application.workspace_service = SimpleNamespace(list_active_execution_bindings=list_active_execution_bindings)
+
+    await service.prime_global_flag()
+
+    assert application.operation_trace_active is False
+
+
+async def test_prime_falls_back_to_metadata_scan_without_discovery(tmp_path):
+    """A single-database deployment without a Workspace directory still resumes."""
+
+    engine = create_async_engine(f'sqlite+aiosqlite:///{tmp_path / "trace-single.db"}')
+    async with engine.begin() as connection:
+        await connection.run_sync(Base.metadata.create_all)
+    application = SimpleNamespace(
+        persistence_mgr=_SingleDatabasePersistence(engine),
+        logger=logging.getLogger('operation-trace-single-test'),
+        user_service=None,
+        operation_trace_active=False,
+    )
+    try:
+        await _set_level(engine, TRACED_WORKSPACE, service_module.OPERATION_LEVEL_MUTATION)
+        service = service_module.WorkspaceSettingsService(application)
+
+        await service.prime_global_flag()
+
+        assert application.operation_trace_active is True
+    finally:
+        await engine.dispose()
diff --git a/tests/unit_tests/test_skill_repository.py b/tests/unit_tests/test_skill_repository.py
index 765ee6a92..053303af4 100644
--- a/tests/unit_tests/test_skill_repository.py
+++ b/tests/unit_tests/test_skill_repository.py
@@ -288,30 +288,12 @@ async def test_existing_run_keeps_v1_while_new_run_resolves_v2(tmp_path):
 
 @pytest.mark.asyncio
 async def test_deleted_skill_can_restore_exact_recoverable_run_revision(tmp_path):
-    from langbot.pkg.provider.tools.loaders import skill as skill_loader
-
     repository = _repository(tmp_path)
     published = await repository.create_skill(
         _CONTEXT,
         {'name': 'recoverable', 'instructions': 'Pinned instructions'},
     )
     await repository.delete_skill(_CONTEXT, 'recoverable')
-    app = SimpleNamespace(skill_repository=repository)
-    query = SimpleNamespace(
-        variables={skill_loader.PIPELINE_BOUND_SKILLS_KEY: ['recoverable']},
-        instance_uuid=_CONTEXT.instance_uuid,
-        workspace_uuid=_CONTEXT.workspace_uuid,
-        placement_generation=_CONTEXT.placement_generation,
-        bot_uuid=None,
-        pipeline_uuid=None,
-        query_uuid='recovered-query',
-    )
-
-    restored = await skill_loader.restore_activated_skills(
-        app,
-        query,
-        [{'name': 'recoverable', 'revision': published['revision']}],
-    )
-
-    assert restored == ['recoverable']
-    assert skill_loader.get_activated_skill(query, 'recoverable')['revision'] == published['revision']
+    restored = await repository.get_skill(_CONTEXT, 'recoverable', snapshot=True, revision=published['revision'])
+    assert restored['revision'] == published['revision']
+    assert restored['instructions'] == 'Pinned instructions'
diff --git a/tests/unit_tests/test_social_preview.py b/tests/unit_tests/test_social_preview.py
new file mode 100644
index 000000000..c71ec6341
--- /dev/null
+++ b/tests/unit_tests/test_social_preview.py
@@ -0,0 +1,28 @@
+import importlib.util
+from pathlib import Path
+import unittest
+
+ROOT = Path(__file__).resolve().parents[2]
+spec = importlib.util.spec_from_file_location('social_preview', ROOT / 'src/langbot/pkg/api/http/controller/social_preview.py')
+assert spec is not None and spec.loader is not None
+module = importlib.util.module_from_spec(spec)
+spec.loader.exec_module(module)
+
+
+class SocialPreviewTest(unittest.TestCase):
+    def test_cloud_has_crawler_visible_image(self):
+        html = (ROOT / 'web/index.html').read_text()
+        result = module.cloud_preview_html(html, 'cloud.langbot.app')
+        self.assertIn(module.CLOUD_IMAGE, result)
+        self.assertIn('summary_large_image', result)
+        self.assertEqual(result.count('property="og:image"'), 1)
+        self.assertTrue((ROOT / 'web/public/social/cloud-v3.png').is_file())
+
+    def test_self_hosted_is_unchanged(self):
+        html = '<head><title>LangBot</title></head>'
+        for host in ['localhost:5300', 'example.com', 'cloud.langbot.app.evil.com']:
+            self.assertEqual(module.cloud_preview_html(html, host), html)
+
+
+if __name__ == '__main__':
+    unittest.main()
diff --git a/tests/unit_tests/test_telemetry.py b/tests/unit_tests/test_telemetry.py
deleted file mode 100644
index ed3994923..000000000
--- a/tests/unit_tests/test_telemetry.py
+++ /dev/null
@@ -1,29 +0,0 @@
-from __future__ import annotations
-
-from types import SimpleNamespace
-import asyncio
-
-import pytest
-
-from langbot.pkg.telemetry.telemetry import TelemetryManager
-
-
-@pytest.mark.asyncio
-async def test_send_tasks_are_scoped_to_manager_instance(monkeypatch):
-    async def fake_send(self, payload):
-        return payload
-
-    monkeypatch.setattr(TelemetryManager, 'send', fake_send)
-
-    first = TelemetryManager(SimpleNamespace())
-    second = TelemetryManager(SimpleNamespace())
-
-    assert first.send_tasks is not second.send_tasks
-
-    await first.start_send_task({'event': 'first'})
-    task = first.send_tasks[0]
-    await task
-    await asyncio.sleep(0)
-
-    assert first.send_tasks == []
-    assert second.send_tasks == []
diff --git a/tests/unit_tests/test_workspace_subscription_limits.py b/tests/unit_tests/test_workspace_subscription_limits.py
new file mode 100644
index 000000000..c16604ed9
--- /dev/null
+++ b/tests/unit_tests/test_workspace_subscription_limits.py
@@ -0,0 +1,41 @@
+import asyncio
+from types import SimpleNamespace
+
+import pytest
+
+from langbot.pkg.cloud.entitlements import EntitlementSnapshot, EntitlementUnavailableError
+from langbot.pkg.api.http.service.workspace_limits import workspace_limits
+
+
+def test_oss_retains_instance_limits():
+    ap = SimpleNamespace(instance_config=SimpleNamespace(data={'system': {'limitation': {'max_bots': 7}}}))
+    assert asyncio.run(workspace_limits(ap, 'one')) == {'max_bots': 7}
+
+
+def test_cloud_uses_signed_workspace_limits():
+    class Resolver:
+        async def resolve(self, workspace_uuid, **kwargs):
+            assert workspace_uuid == 'one'
+            return EntitlementSnapshot(
+                instance_uuid='i',
+                workspace_uuid='one',
+                entitlement_revision=1,
+                status='active',
+                not_before=0,
+                expires_at=9999999999,
+                limits={'bots.max': 2, 'pipelines.max': 3, 'plugins.max': 4, 'knowledge_bases.max': 5},
+            )
+
+    ap = SimpleNamespace(deployment=SimpleNamespace(mode='cloud'), entitlement_resolver=Resolver())
+    assert asyncio.run(workspace_limits(ap, 'one')) == {
+        'max_bots': 2,
+        'max_pipelines': 3,
+        'max_extensions': 4,
+        'max_knowledge_bases': 5,
+    }
+
+
+def test_cloud_missing_resolver_fails_closed():
+    ap = SimpleNamespace(deployment=SimpleNamespace(mode='cloud'))
+    with pytest.raises(EntitlementUnavailableError):
+        asyncio.run(workspace_limits(ap, 'one'))
diff --git a/uv.lock b/uv.lock
index 2788ba272..61b3495ef 100644
--- a/uv.lock
+++ b/uv.lock
@@ -2059,7 +2059,7 @@ wheels = [
 
 [[package]]
 name = "langbot"
-version = "4.11.0b3"
+version = "4.11.0"
 source = { editable = "." }
 dependencies = [
     { name = "aiocqhttp" },
@@ -2185,7 +2185,7 @@ requires-dist = [
     { name = "gewechat-client", specifier = ">=0.1.5" },
     { name = "html2text", specifier = ">=2024.2.26" },
     { name = "httpx", extras = ["socks"], specifier = ">=0.28.1" },
-    { name = "langbot-plugin", git = "https://github.com/langbot-app/langbot-plugin-sdk.git?rev=661b1475091000c94595016d3a4914ff3a1ad81d" },
+    { name = "langbot-plugin", git = "https://github.com/langbot-app/langbot-plugin-sdk.git?rev=dc3d98efd4f02565a4d56999f91152393542fd73" },
     { name = "langchain", specifier = ">=1.3.9" },
     { name = "langchain-core", specifier = ">=1.3.3" },
     { name = "langchain-text-splitters", specifier = ">=1.1.2" },
@@ -2255,8 +2255,8 @@ dev = [
 
 [[package]]
 name = "langbot-plugin"
-version = "0.6.21"
-source = { git = "https://github.com/langbot-app/langbot-plugin-sdk.git?rev=661b1475091000c94595016d3a4914ff3a1ad81d#661b1475091000c94595016d3a4914ff3a1ad81d" }
+version = "0.7.12"
+source = { git = "https://github.com/langbot-app/langbot-plugin-sdk.git?rev=dc3d98efd4f02565a4d56999f91152393542fd73#dc3d98efd4f02565a4d56999f91152393542fd73" }
 dependencies = [
     { name = "aiofiles" },
     { name = "aiohttp" },
diff --git a/web/package-lock.json b/web/package-lock.json
index 044c59c31..194f6d2b0 100644
--- a/web/package-lock.json
+++ b/web/package-lock.json
@@ -11,6 +11,7 @@
         "@dnd-kit/core": "^6.3.1",
         "@dnd-kit/sortable": "^10.0.0",
         "@dnd-kit/utilities": "^3.2.2",
+        "@fingerprintjs/fingerprintjs": "^4.6.2",
         "@hookform/resolvers": "^5.0.1",
         "@radix-ui/react-alert-dialog": "^1.1.15",
         "@radix-ui/react-avatar": "^1.1.11",
@@ -26,18 +27,21 @@
         "@radix-ui/react-scroll-area": "^1.2.9",
         "@radix-ui/react-select": "^2.2.4",
         "@radix-ui/react-separator": "^1.1.8",
+        "@radix-ui/react-slider": "^1.4.7",
         "@radix-ui/react-slot": "^1.2.3",
         "@radix-ui/react-switch": "^1.2.4",
         "@radix-ui/react-tabs": "^1.1.11",
         "@radix-ui/react-toggle": "^1.1.8",
         "@radix-ui/react-toggle-group": "^1.1.9",
         "@radix-ui/react-tooltip": "^1.2.7",
+        "@simplewebauthn/browser": "^14.0.0",
         "@tailwindcss/postcss": "^4.1.5",
         "@tanstack/react-table": "^8.21.3",
         "@vitejs/plugin-react": "^6.0.1",
         "axios": "^1.16.0",
         "class-variance-authority": "^0.7.1",
         "clsx": "^2.1.1",
+        "cmdk": "^1.1.1",
         "highlight.js": "^11.11.1",
         "i18next": "^25.1.2",
         "i18next-browser-languagedetector": "^8.1.0",
@@ -46,6 +50,7 @@
         "lucide-react": "^0.507.0",
         "postcss": "^8.5.10",
         "qrcode": "^1.5.4",
+        "radix-ui": "^1.6.0",
         "react": "19.2.1",
         "react-dom": "19.2.1",
         "react-hook-form": "^7.56.3",
@@ -344,6 +349,15 @@
         "node": "^18.18.0 || ^20.9.0 || >=21.1.0"
       }
     },
+    "node_modules/@fingerprintjs/fingerprintjs": {
+      "version": "4.6.2",
+      "resolved": "https://registry.npmjs.org/@fingerprintjs/fingerprintjs/-/fingerprintjs-4.6.2.tgz",
+      "integrity": "sha512-g8mXuqcFKbgH2CZKwPfVtsUJDHyvcgIABQI7Y0tzWEFXpGxJaXuAuzlifT2oTakjDBLTK4Gaa9/5PERDhqUjtw==",
+      "license": "BUSL-1.1",
+      "dependencies": {
+        "tslib": "^2.4.1"
+      }
+    },
     "node_modules/@floating-ui/core": {
       "version": "1.7.4",
       "resolved": "https://registry.npmjs.org/@floating-ui/core/-/core-1.7.4.tgz",
@@ -548,29 +562,24 @@
       }
     },
     "node_modules/@radix-ui/number": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/number/-/number-1.1.1.tgz",
-      "integrity": "sha512-MkKCwxlXTgz6CFoJx3pCwn07GKp36+aZyu/u2Ln2VrA5DcdyCZkASEDBTd8x5whTQQL5CiYf4prXKLcgQdv29g==",
+      "version": "1.1.3",
+      "resolved": "https://registry.npmjs.org/@radix-ui/number/-/number-1.1.3.tgz",
+      "integrity": "sha512-Road2bidD0uu/1BGDOWNdPI06g0lIRy6IF9GZcIrDK2KGItfor8IQwQa+yM2ERgHM1MmHxaxpTzk0/Jp42lNfA==",
       "license": "MIT"
     },
     "node_modules/@radix-ui/primitive": {
-      "version": "1.1.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/primitive/-/primitive-1.1.3.tgz",
-      "integrity": "sha512-JTF99U/6XIjCBo0wqkU5sK10glYe27MRRsfwoiq5zzOEZLHU3A3KCMa5X/azekYRCJ0HlwI0crAXS/5dEHTzDg==",
+      "version": "1.1.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/primitive/-/primitive-1.1.7.tgz",
+      "integrity": "sha512-rqWnm76nYT8HoNNqEjpgJ7Pw/DrBj5iBTrmEPo6HTX5+VJyBNOqTdv4g89G63HuR5g0AaENoAcH7Is5fF2kZ8Q==",
       "license": "MIT"
     },
-    "node_modules/@radix-ui/react-alert-dialog": {
+    "node_modules/@radix-ui/react-accessible-icon": {
       "version": "1.1.15",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-alert-dialog/-/react-alert-dialog-1.1.15.tgz",
-      "integrity": "sha512-oTVLkEw5GpdRe29BqJ0LSDFWI3qu0vR1M0mUkOQWDIUnY/QIkLpgDMWuKxP94c2NAC2LGcgVhG1ImF3jkZ5wXw==",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-accessible-icon/-/react-accessible-icon-1.1.15.tgz",
+      "integrity": "sha512-WTQwcAvQf5sOcuUyi90lKPbhwcvQ+j55cjrSmeaN+L2vKU3DooOvlKw2MDeiJ5IkV5N905KW0/fGojKOBhD11A==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-dialog": "1.1.15",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3"
+        "@radix-ui/react-visually-hidden": "1.2.11"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -587,14 +596,42 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-alert-dialog/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-accordion": {
+      "version": "1.2.20",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-accordion/-/react-accordion-1.2.20.tgz",
+      "integrity": "sha512-jDhG9FvAEnlhnjrsINbNXcUa4G+L1KqSkJSunkbKEzFRcAb52jvM0PjPxPRvhe1HNc5F5yc0yzzWeeqlH4yBIg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collapsible": "1.1.20",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
       },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-accordion/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -605,13 +642,170 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-arrow": {
-      "version": "1.1.7",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-arrow/-/react-arrow-1.1.7.tgz",
-      "integrity": "sha512-F+M1tLhO+mlQaOWspE8Wstg+z6PwxwRd8oQ8IXceWz92kfAmalTRf0EjrouQeo7QssEPfCn05B4Ihs1K9WQ/7w==",
+    "node_modules/@radix-ui/react-accordion/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-primitive": "2.1.3"
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-alert-dialog": {
+      "version": "1.1.23",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-alert-dialog/-/react-alert-dialog-1.1.23.tgz",
+      "integrity": "sha512-VAYOiQRqj3GPpYJE0I9J+X8Ip05cyVlNdKOFeiGS2Ou1HHGfpl0BxOyZm6nmVDyU+W+NF3/XLzmjHmVGydhwgA==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dialog": "1.1.23",
+        "@radix-ui/react-primitive": "2.1.10"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-alert-dialog/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-alert-dialog/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-arrow": {
+      "version": "1.1.15",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-arrow/-/react-arrow-1.1.15.tgz",
+      "integrity": "sha512-v4zggRcjadnI+ClKDuijlQEW4tw3NoaeHc/PwpKnLoLLKNUG4InLegkstooLcRIUWCs+8L22dGURCVuFfOKfnA==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-primitive": "2.1.10"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-arrow/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-aspect-ratio": {
+      "version": "1.1.15",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-aspect-ratio/-/react-aspect-ratio-1.1.15.tgz",
+      "integrity": "sha512-fy+dyVR+90nelK8rqIznFlxzx7uPcGbhxH8Nfr2bHb4UfSe+e3hklOC0luK0hDwVwnRX7xTRySpsrQVeW+/oNQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-primitive": "2.1.10"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-aspect-ratio/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -629,16 +823,17 @@
       }
     },
     "node_modules/@radix-ui/react-avatar": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-avatar/-/react-avatar-1.1.11.tgz",
-      "integrity": "sha512-0Qk603AHGV28BOBO34p7IgD5m+V5Sg/YovfayABkoDDBM5d3NCx0Mp4gGrjzLGes1jV5eNOE1r3itqOR33VC6Q==",
+      "version": "1.2.6",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-avatar/-/react-avatar-1.2.6.tgz",
+      "integrity": "sha512-4ULOTJ/mqy2hT9GlWa/MFHxHSvH3nJzHnZM1waNsc5Bonv7i70aNenghXmD97S6OJ81ekXONGGt4nT1r0PfEdA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-context": "1.1.3",
-        "@radix-ui/react-primitive": "2.1.4",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-is-hydrated": "0.1.0",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-is-hydrated": "0.1.3",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -655,10 +850,33 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-avatar/node_modules/@radix-ui/react-context": {
-      "version": "1.1.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-context/-/react-context-1.1.3.tgz",
-      "integrity": "sha512-ieIFACdMpYfMEjF0rEf5KLvfVyIkOz6PDGyNnP+u+4xQ6jny3VCgA4OgXOwNx2aUkxn8zx9fiVcM8CfFYv9Lxw==",
+    "node_modules/@radix-ui/react-avatar/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-avatar/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -670,13 +888,19 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-avatar/node_modules/@radix-ui/react-primitive": {
-      "version": "2.1.4",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.4.tgz",
-      "integrity": "sha512-9hQc4+GNVtJAIEPEqlYqW5RiYdrr8ea5XQ0ZOnD6fgru+83kqT15mq2OCcbe8KnjRZl5vF3ks69AKz3kh1jrhg==",
+    "node_modules/@radix-ui/react-checkbox": {
+      "version": "1.3.11",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-checkbox/-/react-checkbox-1.3.11.tgz",
+      "integrity": "sha512-Gnptr9pDDQxD3hgq2dtPbtrp/c2qH1mBwIzw3X/ivrMb2e1t0jMTi606fVEqFPaQR1ggXIVQWKj3P2WW9v7zGQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-slot": "1.2.4"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-size": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -693,20 +917,28 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-checkbox": {
-      "version": "1.3.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-checkbox/-/react-checkbox-1.3.3.tgz",
-      "integrity": "sha512-wBbpv+NQftHDdG86Qc0pIyXk5IR3tM8Vd0nWLKDcX8nNn4nXFOFwsKuqw2okA/1D/mpaAkmuyndrPJTYDNZtFw==",
+    "node_modules/@radix-ui/react-checkbox/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-checkbox/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
-        "@radix-ui/react-use-previous": "1.1.1",
-        "@radix-ui/react-use-size": "1.1.1"
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -724,19 +956,57 @@
       }
     },
     "node_modules/@radix-ui/react-collapsible": {
-      "version": "1.1.12",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-collapsible/-/react-collapsible-1.1.12.tgz",
-      "integrity": "sha512-Uu+mSh4agx2ib1uIGPP4/CKNULyajb3p92LsVXmH2EHVMTfZWpll88XJ0j4W0z3f8NK1eYl1+Mf/szHPmcHzyA==",
+      "version": "1.1.20",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-collapsible/-/react-collapsible-1.1.20.tgz",
+      "integrity": "sha512-mcGesGplBnzN2sbvJETzpCNfSMyPnb29q1GRLU+Ib7bJrpIG2ywmRoh2V5VbA2uNvKikKUlVbAPks7JDjz4A8Q==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-collapsible/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-collapsible/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -754,15 +1024,15 @@
       }
     },
     "node_modules/@radix-ui/react-collection": {
-      "version": "1.1.7",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-collection/-/react-collection-1.1.7.tgz",
-      "integrity": "sha512-Fh9rGN0MoI4ZFUNyfFVNU4y9LUz93u9/0K+yLgA2bwRojxM8JU1DyvvMBabnZPBgMWREAJvU2jjVzq+LrFUglw==",
+      "version": "1.1.15",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-collection/-/react-collection-1.1.15.tgz",
+      "integrity": "sha512-9W+B9NPF0NaaPh/1NJd3+KqsnlLqU9H7T2rvww+fp+T/evVXdNAyYcnfRQZFOjkR1ajQp3yORlqnI8soawLvNA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3"
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -779,14 +1049,11 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-collection/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-collection/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -797,6 +1064,29 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-collection/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-compose-refs": {
       "version": "1.1.2",
       "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.2.tgz",
@@ -813,9 +1103,9 @@
       }
     },
     "node_modules/@radix-ui/react-context": {
-      "version": "1.1.2",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-context/-/react-context-1.1.2.tgz",
-      "integrity": "sha512-jCi/QKUM2r1Ju5a3J64TH2A5SpKAgh0LpknyqdQ4m6DCV0xJ2HG1xARRwNGPQfi1SLdLWZ1OJz6F4OMBBNiGJA==",
+      "version": "1.2.2",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-context/-/react-context-1.2.2.tgz",
+      "integrity": "sha512-RHCUGwKHDr0hDGg4X7ma4JG4/+12qxw8rkh5QKdDldlCvtja6nUx1Ef/8HVrJze81lEsgLQlqjzjGNHantgnQA==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -828,17 +1118,39 @@
       }
     },
     "node_modules/@radix-ui/react-context-menu": {
-      "version": "2.2.16",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-context-menu/-/react-context-menu-2.2.16.tgz",
-      "integrity": "sha512-O8morBEW+HsVG28gYDZPTrT9UUovQUlJue5YO836tiTJhuIWBm/zQHc7j388sHWtdH/xUZurK9olD2+pcqx5ww==",
+      "version": "2.3.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-context-menu/-/react-context-menu-2.3.7.tgz",
+      "integrity": "sha512-CtXP35dxaB5T3zXSd+E3uHe/QpXcpYnZmxp6OaIbfthtfW4wyb77M23BG+bwIJDtsMwEP/YssdsmNyZu7jhWew==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-menu": "2.1.16",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-menu": "2.1.24",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-context-menu/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -856,25 +1168,26 @@
       }
     },
     "node_modules/@radix-ui/react-dialog": {
-      "version": "1.1.15",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-dialog/-/react-dialog-1.1.15.tgz",
-      "integrity": "sha512-TCglVRtzlffRNxRMEyR36DGBLJpeusFcgMVD9PZEzAKnUs1lKCgX5u9BmC2Yg+LL9MgZDugFFs1Vl+Jp4t/PGw==",
+      "version": "1.1.23",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-dialog/-/react-dialog-1.1.23.tgz",
+      "integrity": "sha512-Ksw4WeROkO4rC9k/onilX/Ao2Cr1ku1unMNH+XSCcP4jSXYu7HDsg9n4ojMjVb22XpYjAQ9qfrFlVbru1vXDUA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-focus-guards": "1.1.3",
-        "@radix-ui/react-focus-scope": "1.1.7",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-focus-guards": "1.1.6",
+        "@radix-ui/react-focus-scope": "1.1.16",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
         "aria-hidden": "^1.2.4",
-        "react-remove-scroll": "^2.6.3"
+        "react-remove-scroll": "^2.7.2"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -891,14 +1204,11 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-dialog/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-dialog/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -909,10 +1219,33 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-dialog/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-direction": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-direction/-/react-direction-1.1.1.tgz",
-      "integrity": "sha512-1UEWRX6jnOA2y4H5WczZ44gOOjTEmlqv1uNW4GAJEO5+bauCBhv8snY65Iw5/VOS/ghKN9gr2KjnLKxrsvoMVw==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-direction/-/react-direction-1.1.4.tgz",
+      "integrity": "sha512-5pzg4FGQNpExhnhT2zlrP1wZFaYCd1K0nYWoFAdcYoYK868IEigqMX3B3f8yIoRlAhAeDWciLI6ZdCKHF9P4Vg==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -925,16 +1258,16 @@
       }
     },
     "node_modules/@radix-ui/react-dismissable-layer": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-dismissable-layer/-/react-dismissable-layer-1.1.11.tgz",
-      "integrity": "sha512-Nqcp+t5cTB8BinFkZgXiMJniQH0PsUt2k51FUhbdfeKvc4ACcG2uQniY/8+h1Yv6Kza4Q7lD7PQV0z0oicE0Mg==",
+      "version": "1.1.19",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-dismissable-layer/-/react-dismissable-layer-1.1.19.tgz",
+      "integrity": "sha512-8g4pfOL9HoKKLWGiypT+dphVqjFfmcXO5GBnhsG6zI+lxAx/8feQpr+1LSN8Re3hiZ+XkLNS4O9ztK11/LzQ6w==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-escape-keydown": "1.1.1"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-effect-event": "0.0.5"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -951,19 +1284,110 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-dropdown-menu": {
-      "version": "2.1.16",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-dropdown-menu/-/react-dropdown-menu-2.1.16.tgz",
-      "integrity": "sha512-1PLGQEynI/3OX/ftV54COn+3Sud/Mn8vALg2rWnBLnRaGtJDduNW/22XjlGgPdpcIbiQxjKtb7BkcjP00nqfJw==",
+    "node_modules/@radix-ui/react-dismissable-layer/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-dismissable-layer/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-menu": "2.1.16",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-dismissable-layer/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-dropdown-menu": {
+      "version": "2.1.24",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-dropdown-menu/-/react-dropdown-menu-2.1.24.tgz",
+      "integrity": "sha512-geq8l2rJkxvkXsT9RMgtUE3P8pITFpTsvYpbySi1IH4fZEABD/Gp85myayFgxk0ktljGMJnCbeFkyTusvSvv7g==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-menu": "2.1.24",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-dropdown-menu/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-dropdown-menu/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -981,9 +1405,9 @@
       }
     },
     "node_modules/@radix-ui/react-focus-guards": {
-      "version": "1.1.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-focus-guards/-/react-focus-guards-1.1.3.tgz",
-      "integrity": "sha512-0rFg/Rj2Q62NCm62jZw0QX7a3sz6QCQU0LpZdNrJX8byRGaGVTqbrW9jAoIAHyMQqsNpeZ81YgSizOt5WXq0Pw==",
+      "version": "1.1.6",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-focus-guards/-/react-focus-guards-1.1.6.tgz",
+      "integrity": "sha512-RNOJjfZMTyBM6xYmV3IVGXkPjIhcBAuv48POevAXwrGJhkWZ9p1rFoIS1JFooPuT193AZmRsCPhpoVJxx6OPoQ==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -1020,21 +1444,125 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-hover-card": {
-      "version": "1.1.15",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-hover-card/-/react-hover-card-1.1.15.tgz",
-      "integrity": "sha512-qgTkjNT1CfKMoP0rcasmlH2r1DAiYicWsDsufxl940sT2wHNEWWv6FMWIQXWhVdmC1d/HYfbhQx60KYyAtKxjg==",
+    "node_modules/@radix-ui/react-form": {
+      "version": "0.1.16",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-form/-/react-form-0.1.16.tgz",
+      "integrity": "sha512-Q4TLEn2A7TAypxwmd6R9EwrlXDvkfYSDMrq9/887AXAGh+G1rH+kYJKSTv+Si9Y0JPKTwKYv6PviAJosysNimA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-popper": "1.2.8",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-label": "2.1.15",
+        "@radix-ui/react-primitive": "2.1.10"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-form/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-form/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-hover-card": {
+      "version": "1.1.23",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-hover-card/-/react-hover-card-1.1.23.tgz",
+      "integrity": "sha512-H8qONfZd3ltrU3+jHCIgITbWo6e1iTKvP9DHdrvYbX48ooRM5FjEDTn16AMwdfuOGkWdZEhpl3PLL/Wk/AnHDQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-hover-card/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-hover-card/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1052,12 +1580,12 @@
       }
     },
     "node_modules/@radix-ui/react-id": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-id/-/react-id-1.1.1.tgz",
-      "integrity": "sha512-kGkGegYIdQsOb4XjsfM97rXsiHaBwco+hFI66oO4s9LU+PLAC5oJ7khdOVFxkhsmlbpUqDAvXw11CluXP+jkHg==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-id/-/react-id-1.1.4.tgz",
+      "integrity": "sha512-TMQp2llA+RYn7JcjnrMnz7wN4pcVttPZnRZo52PLQsoLVKzNlVwUeHmfePgTgRluXFvlD3GD5g5MOVVTJCO0qA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1070,12 +1598,12 @@
       }
     },
     "node_modules/@radix-ui/react-label": {
-      "version": "2.1.8",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-label/-/react-label-2.1.8.tgz",
-      "integrity": "sha512-FmXs37I6hSBVDlO4y764TNz1rLgKwjJMQ0EGte6F3Cb3f4bIuHB/iLa/8I9VKkmOy+gNHq8rql3j686ACVV21A==",
+      "version": "2.1.15",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-label/-/react-label-2.1.15.tgz",
+      "integrity": "sha512-o/rdYEwZTTo5tjknnPeyQFU45kUC4i/XyeDPP+HGyi6XqpOP6Zf5Ya5vh/Yfe9Id5JiuWnnAx2XqIeD3UYZt0g==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-primitive": "2.1.4"
+        "@radix-ui/react-primitive": "2.1.10"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1093,12 +1621,12 @@
       }
     },
     "node_modules/@radix-ui/react-label/node_modules/@radix-ui/react-primitive": {
-      "version": "2.1.4",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.4.tgz",
-      "integrity": "sha512-9hQc4+GNVtJAIEPEqlYqW5RiYdrr8ea5XQ0ZOnD6fgru+83kqT15mq2OCcbe8KnjRZl5vF3ks69AKz3kh1jrhg==",
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-slot": "1.2.4"
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1116,29 +1644,29 @@
       }
     },
     "node_modules/@radix-ui/react-menu": {
-      "version": "2.1.16",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-menu/-/react-menu-2.1.16.tgz",
-      "integrity": "sha512-72F2T+PLlphrqLcAotYPp0uJMr5SjP5SL01wfEspJbru5Zs5vQaSHb4VB3ZMJPimgHHCHG7gMOeOB9H3Hdmtxg==",
+      "version": "2.1.24",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-menu/-/react-menu-2.1.24.tgz",
+      "integrity": "sha512-uW7RVuU6Lp/ZtfeY4b3kL32zccgEWvPv1+cf17ubYzHa9cL8AHokmk36cG/XEiH/smbQvumnieXX9j/e9RqJWA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-collection": "1.1.7",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-focus-guards": "1.1.3",
-        "@radix-ui/react-focus-scope": "1.1.7",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-popper": "1.2.8",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-roving-focus": "1.1.11",
-        "@radix-ui/react-slot": "1.2.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-focus-guards": "1.1.6",
+        "@radix-ui/react-focus-scope": "1.1.16",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
         "aria-hidden": "^1.2.4",
-        "react-remove-scroll": "^2.6.3"
+        "react-remove-scroll": "^2.7.2"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1155,14 +1683,11 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-menu/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-menu/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -1173,27 +1698,364 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-menu/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-menu/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-menubar": {
+      "version": "1.1.24",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-menubar/-/react-menubar-1.1.24.tgz",
+      "integrity": "sha512-eeVs0vf7cuqXaM0qLQCPcufImiJNVBXdJDLu7ZGYl2732UH23Qat/foNGrr6vYV3/DdTsBqASoggUFgH14OcZA==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-menu": "2.1.24",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-menubar/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-menubar/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-navigation-menu": {
+      "version": "1.2.22",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-navigation-menu/-/react-navigation-menu-1.2.22.tgz",
+      "integrity": "sha512-ou7iLEJ+yrhQndkkA4U21XIdS/CS45F4iXIkTZcb6/Ne9EMsOuDudVmCwmDnfFZZ+y1FZqXRNSIgBy+YMvZVZg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-use-previous": "1.1.4",
+        "@radix-ui/react-visually-hidden": "1.2.11"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-navigation-menu/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-navigation-menu/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-navigation-menu/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-one-time-password-field": {
+      "version": "0.1.16",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-one-time-password-field/-/react-one-time-password-field-0.1.16.tgz",
+      "integrity": "sha512-Tj9P6ntAJEw52oq/F0AGknXR4XncxEt7XU47O3xJQOiWfLzEy3d9gtgKfvjSzGxzHkfL+VzvxGu2KTFsloJqXw==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/number": "1.1.3",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-effect-event": "0.0.5",
+        "@radix-ui/react-use-is-hydrated": "0.1.3",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-one-time-password-field/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-one-time-password-field/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-password-toggle-field": {
+      "version": "0.1.11",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-password-toggle-field/-/react-password-toggle-field-0.1.11.tgz",
+      "integrity": "sha512-4gvFnmDXu3dgj21CqsufzIameRvlRd4SBqaWhcrlrNhRo0Y5i/49AmRJYe1fdAM3G2VNBbmin4b0D6cdQocwgw==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-effect-event": "0.0.5",
+        "@radix-ui/react-use-is-hydrated": "0.1.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-password-toggle-field/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-password-toggle-field/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-popover": {
-      "version": "1.1.15",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-popover/-/react-popover-1.1.15.tgz",
-      "integrity": "sha512-kr0X2+6Yy/vJzLYJUPCZEc8SfQcf+1COFoAqauJm74umQhta9M7lNJHP7QQS3vkvcGLQUbWpMzwrXYwrYztHKA==",
+      "version": "1.1.23",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-popover/-/react-popover-1.1.23.tgz",
+      "integrity": "sha512-mw58MrBlyHWFisTOYignD0vf/3gdcgAR+9of1s9G/38CbFiUwH1nCDkc0AUM9IrXFgN5Ue8n45j9WCgyM1sbiQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-focus-guards": "1.1.3",
-        "@radix-ui/react-focus-scope": "1.1.7",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-popper": "1.2.8",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-focus-guards": "1.1.6",
+        "@radix-ui/react-focus-scope": "1.1.16",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
         "aria-hidden": "^1.2.4",
-        "react-remove-scroll": "^2.6.3"
+        "react-remove-scroll": "^2.7.2"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1210,14 +2072,11 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-popover/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-popover/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -1228,22 +2087,13 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-popper": {
-      "version": "1.2.8",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-popper/-/react-popper-1.2.8.tgz",
-      "integrity": "sha512-0NJQ4LFFUuWkE7Oxf0htBKS6zLkkjBH+hM1uk7Ng705ReR8m/uelduy1DBo0PyBXPKVnBA6YBlU94MBGXrSBCw==",
+    "node_modules/@radix-ui/react-popover/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@floating-ui/react-dom": "^2.0.0",
-        "@radix-ui/react-arrow": "1.1.7",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-layout-effect": "1.1.1",
-        "@radix-ui/react-use-rect": "1.1.1",
-        "@radix-ui/react-use-size": "1.1.1",
-        "@radix-ui/rect": "1.1.1"
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1260,14 +2110,122 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-portal": {
-      "version": "1.1.9",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-portal/-/react-portal-1.1.9.tgz",
-      "integrity": "sha512-bpIxvq03if6UNwXZ+HTK71JLh4APvnXntDc6XOX8UVq4XQOVl7lwok0AvIl+b8zgCw3fSaVTZMpAPPagXbKmHQ==",
+    "node_modules/@radix-ui/react-popper": {
+      "version": "1.3.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-popper/-/react-popper-1.3.7.tgz",
+      "integrity": "sha512-UsJrrd7w4wuKKTdvd/DNERVlwSlUcyXzjhyDwBk+3aPOsCjOY6ZSbxuw8E6lZTjjfP8Cpd0J8VVkrYUWyGYXyg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@floating-ui/react-dom": "^2.0.0",
+        "@radix-ui/react-arrow": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-use-rect": "1.1.4",
+        "@radix-ui/react-use-size": "1.1.4",
+        "@radix-ui/rect": "1.1.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-popper/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-popper/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-popper/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-portal": {
+      "version": "1.1.17",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-portal/-/react-portal-1.1.17.tgz",
+      "integrity": "sha512-vKQLcWypUnwZVvfV7UkGahH2g6ySe8M8R+zYBwPrv5byZ9QAW6cQVvNKo7GgmD+p8aYb6D9JBuvy8/WhOno2wQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-portal/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1285,13 +2243,12 @@
       }
     },
     "node_modules/@radix-ui/react-presence": {
-      "version": "1.1.5",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-presence/-/react-presence-1.1.5.tgz",
-      "integrity": "sha512-/jfEwNDdQVBCNvjkGit4h6pMOzq8bHkopq458dPt2lMjx+eBQUohZNG9A7DtO/O5ukSbxuaNGXMjHicgwy6rQQ==",
+      "version": "1.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-presence/-/react-presence-1.1.10.tgz",
+      "integrity": "sha512-3wyzCQ6+ubRA+D4uv9m95JYLXxmOHp05qjrkjeA7uKHHtjpPggQzc6DAb0URl7j67oR0K2foO4ip27TiX037Bw==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1350,13 +2307,13 @@
       }
     },
     "node_modules/@radix-ui/react-progress": {
-      "version": "1.1.8",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-progress/-/react-progress-1.1.8.tgz",
-      "integrity": "sha512-+gISHcSPUJ7ktBy9RnTqbdKW78bcGke3t6taawyZ71pio1JewwGSJizycs7rLhGTvMJYCQB1DBK4KQsxs7U8dA==",
+      "version": "1.1.16",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-progress/-/react-progress-1.1.16.tgz",
+      "integrity": "sha512-5XnomAsoZZCY+KNTxbIghpGqPruZvKFNlvcAljVAOdDRDsH4/OZQxhtwo5wdtoDM5R6MhJBb2sPnDuRFep3lzg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-context": "1.1.3",
-        "@radix-ui/react-primitive": "2.1.4"
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-primitive": "2.1.10"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1373,10 +2330,64 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-progress/node_modules/@radix-ui/react-context": {
-      "version": "1.1.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-context/-/react-context-1.1.3.tgz",
-      "integrity": "sha512-ieIFACdMpYfMEjF0rEf5KLvfVyIkOz6PDGyNnP+u+4xQ6jny3VCgA4OgXOwNx2aUkxn8zx9fiVcM8CfFYv9Lxw==",
+    "node_modules/@radix-ui/react-progress/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-radio-group": {
+      "version": "1.4.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-radio-group/-/react-radio-group-1.4.7.tgz",
+      "integrity": "sha512-cgYFEkntCxppHZgtSZ+7vh0wbZQ+IC7PPMw8DSnRG27B6kDd32/Zw0OJt7dGDigCoprMuWHjg2PvUn3PYvPFoQ==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-size": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-radio-group/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -1388,13 +2399,13 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-progress/node_modules/@radix-ui/react-primitive": {
-      "version": "2.1.4",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.4.tgz",
-      "integrity": "sha512-9hQc4+GNVtJAIEPEqlYqW5RiYdrr8ea5XQ0ZOnD6fgru+83kqT15mq2OCcbe8KnjRZl5vF3ks69AKz3kh1jrhg==",
+    "node_modules/@radix-ui/react-radio-group/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-slot": "1.2.4"
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1412,20 +2423,22 @@
       }
     },
     "node_modules/@radix-ui/react-roving-focus": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-roving-focus/-/react-roving-focus-1.1.11.tgz",
-      "integrity": "sha512-7A6S9jSgm/S+7MdtNDSb+IU859vQqJ/QAtcYQcfFC6W8RS4IxIZDldLR0xqCFZ6DCyrQLjLPsxtTNch5jVA4lA==",
+      "version": "1.1.19",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-roving-focus/-/react-roving-focus-1.1.19.tgz",
+      "integrity": "sha512-V9jI6hDjT7l3jsCQD9bLNvDLM3tH/gdbOTp7Tefp3hbbgCGQoK7tUvrWiRlcoBHIZ809ElXwNQwVo0B98LuTXQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-collection": "1.1.7",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-is-hydrated": "0.1.3",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1442,21 +2455,74 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-roving-focus/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-roving-focus/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-roving-focus/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-scroll-area": {
-      "version": "1.2.10",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-scroll-area/-/react-scroll-area-1.2.10.tgz",
-      "integrity": "sha512-tAXIa1g3sM5CGpVT0uIbUx/U3Gs5N8T52IICuCtObaos1S8fzsrPXG5WObkQN3S6NVl6wKgPhAIiBGbWnvc97A==",
+      "version": "1.2.18",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-scroll-area/-/react-scroll-area-1.2.18.tgz",
+      "integrity": "sha512-Zn5Cd171wxsO3Dfg8HaW6RifTb9CYTKQJHs/G4+LN1GfmJpaQMZQyQxMprVPHpaz7QY4l9BxK2JwQuzHsXC8nA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/number": "1.1.1",
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/number": "1.1.3",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1473,33 +2539,87 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-scroll-area/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-scroll-area/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-scroll-area/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-select": {
-      "version": "2.2.6",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-select/-/react-select-2.2.6.tgz",
-      "integrity": "sha512-I30RydO+bnn2PQztvo25tswPH+wFBjehVGtmagkU78yMdwTwVf12wnAOF+AeP8S2N8xD+5UPbGhkUfPyvT+mwQ==",
+      "version": "2.3.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-select/-/react-select-2.3.7.tgz",
+      "integrity": "sha512-WFGImkmbzcfxeIwq/+4HvRN0pizBwbwQUED4I13ezQsDdfl38ZntN6TmR8XaSzPBqoCToe8rF75j6NPNDSzhbg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/number": "1.1.1",
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-collection": "1.1.7",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-focus-guards": "1.1.3",
-        "@radix-ui/react-focus-scope": "1.1.7",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-popper": "1.2.8",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3",
-        "@radix-ui/react-use-callback-ref": "1.1.1",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
-        "@radix-ui/react-use-layout-effect": "1.1.1",
-        "@radix-ui/react-use-previous": "1.1.1",
-        "@radix-ui/react-visually-hidden": "1.2.3",
+        "@radix-ui/number": "1.1.3",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-focus-guards": "1.1.6",
+        "@radix-ui/react-focus-scope": "1.1.16",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-use-previous": "1.1.4",
+        "@radix-ui/react-visually-hidden": "1.2.11",
         "aria-hidden": "^1.2.4",
-        "react-remove-scroll": "^2.6.3"
+        "react-remove-scroll": "^2.7.2"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1516,14 +2636,49 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-select/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-select/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-select/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
+        "@radix-ui/react-slot": "1.3.3"
       },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-select/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -1535,12 +2690,12 @@
       }
     },
     "node_modules/@radix-ui/react-separator": {
-      "version": "1.1.8",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-separator/-/react-separator-1.1.8.tgz",
-      "integrity": "sha512-sDvqVY4itsKwwSMEe0jtKgfTh+72Sy3gPmQpjqcQneqQ4PFmr/1I0YA+2/puilhggCe2gJcx5EBAYFkWkdpa5g==",
+      "version": "1.1.15",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-separator/-/react-separator-1.1.15.tgz",
+      "integrity": "sha512-jOLO4lssEzWpoDu7G+Ze4VjwMRUBt291pnZD0gmalREZipnTX3wadQo7Fy48GCTfe14/YRN6rw/rOJqrE85Wxw==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-primitive": "2.1.4"
+        "@radix-ui/react-primitive": "2.1.10"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1558,12 +2713,83 @@
       }
     },
     "node_modules/@radix-ui/react-separator/node_modules/@radix-ui/react-primitive": {
-      "version": "2.1.4",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.4.tgz",
-      "integrity": "sha512-9hQc4+GNVtJAIEPEqlYqW5RiYdrr8ea5XQ0ZOnD6fgru+83kqT15mq2OCcbe8KnjRZl5vF3ks69AKz3kh1jrhg==",
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-slot": "1.2.4"
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-slider": {
+      "version": "1.4.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-slider/-/react-slider-1.4.7.tgz",
+      "integrity": "sha512-mTSLf1GC/C0moWjTbvCM6Qn/gBjvlFt1azuWF2v7MN5C3Zq2U2J2lN3ZEYkpujuOU5Ro7A28wkviSxaKnG0BYg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/number": "1.1.3",
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-use-previous": "1.1.4",
+        "@radix-ui/react-use-size": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-slider/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1581,12 +2807,12 @@
       }
     },
     "node_modules/@radix-ui/react-slot": {
-      "version": "1.2.4",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.4.tgz",
-      "integrity": "sha512-Jl+bCv8HxKnlTLVrcDE8zTMJ09R9/ukw4qBs/oZClOfoQk/cOTbDn+NceXfV7j09YPVQUryJPHurafcSg6EVKA==",
+      "version": "1.3.3",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.3.3.tgz",
+      "integrity": "sha512-qx7oqnYbxnK9kYI9m317qmFmEgo6ywqWvbTogdj7cL9p3/yx4M48p7Rnw5z3H890cL/ow/EeWJsuTykeZVXP5Q==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
+        "@radix-ui/react-compose-refs": "1.1.5"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1598,19 +2824,71 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-slot/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-switch": {
-      "version": "1.2.6",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-switch/-/react-switch-1.2.6.tgz",
-      "integrity": "sha512-bByzr1+ep1zk4VubeEVViV592vu2lHE2BZY5OnzehZqOOgogN80+mNtCqPkhn2gklJqOpxWgPoYTSnhBCqpOXQ==",
+      "version": "1.3.7",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-switch/-/react-switch-1.3.7.tgz",
+      "integrity": "sha512-48tB/4dn2UVLBCYhTu9AuR63IHl73l/qLbLgxd86noTUor4/K4LFDAcYjK+isP5313qxaFpjPVogE7+Y0/V3Kw==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
-        "@radix-ui/react-use-previous": "1.1.1",
-        "@radix-ui/react-use-size": "1.1.1"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-size": "1.1.4"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-switch/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-switch/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1628,19 +2906,19 @@
       }
     },
     "node_modules/@radix-ui/react-tabs": {
-      "version": "1.1.13",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-tabs/-/react-tabs-1.1.13.tgz",
-      "integrity": "sha512-7xdcatg7/U+7+Udyoj2zodtI9H/IIopqo+YOIcZOq1nJwXWBZ9p8xiu5llXlekDbZkca79a/fozEYQXIA4sW6A==",
+      "version": "1.1.21",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-tabs/-/react-tabs-1.1.21.tgz",
+      "integrity": "sha512-UKxJlZid7FVtsk/WTxj4i4uSEgj2Au+KBbS7SQyTlzMhhn+86Cz3tISZdTa87bfEfcuvZezf2ZsxD4xuEKtkog==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-roving-focus": "1.1.11",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1657,15 +2935,125 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-toggle": {
-      "version": "1.1.10",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-toggle/-/react-toggle-1.1.10.tgz",
-      "integrity": "sha512-lS1odchhFTeZv3xwHH31YPObmJn8gOg7Lq12inrr0+BH/l3Tsq32VfjqH1oh80ARM3mlkfMic15n0kg4sD1poQ==",
+    "node_modules/@radix-ui/react-tabs/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toast": {
+      "version": "1.2.23",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-toast/-/react-toast-1.2.23.tgz",
+      "integrity": "sha512-ofhyAsYaocRGOs/n0XWdUOSVzEAG6BfrMVM8z0c0kLEWY38w/0WuMFPTJP/HVaZPYkMvHZoKIIhNcjbTCBILPg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-visually-hidden": "1.2.11"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toast/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toast/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toast/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toggle": {
+      "version": "1.1.18",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-toggle/-/react-toggle-1.1.18.tgz",
+      "integrity": "sha512-7lonPlKfSacd20GlOBx2ltuVKz9oqWYZz+oMQyOltw6t1y2nyftj2ZmwwUHYn49kqfDWcp8dNZm5NgV+5Z+mug==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1683,18 +3071,116 @@
       }
     },
     "node_modules/@radix-ui/react-toggle-group": {
-      "version": "1.1.11",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-toggle-group/-/react-toggle-group-1.1.11.tgz",
-      "integrity": "sha512-5umnS0T8JQzQT6HbPyO7Hh9dgd82NmS36DQr+X/YJ9ctFNCiiQd6IJAYYZ33LUwm8M+taCz5t2ui29fHZc4Y6Q==",
+      "version": "1.1.19",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-toggle-group/-/react-toggle-group-1.1.19.tgz",
+      "integrity": "sha512-OtnwuSVjd1Ofi+AdnvhsjQdyuhCDwYs1w9RyB5BN/OavXOVQo42SYqQjwUnbPnaiPFBpQ9aX70dWeee+v2oBLA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-direction": "1.1.1",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-roving-focus": "1.1.11",
-        "@radix-ui/react-toggle": "1.1.10",
-        "@radix-ui/react-use-controllable-state": "1.2.2"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-toggle": "1.1.18",
+        "@radix-ui/react-use-controllable-state": "1.2.6"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toggle-group/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toggle/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toolbar": {
+      "version": "1.1.19",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-toolbar/-/react-toolbar-1.1.19.tgz",
+      "integrity": "sha512-Ph0IvtYw4VB12ZnZg+YtrGs8yJQsnizwo/zu0R4Y/nWugtJzA7Pg1eWeuDR9+LSqn+xjamss+UOSOJJJ4gx8jw==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-separator": "1.1.15",
+        "@radix-ui/react-toggle-group": "1.1.19"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-toolbar/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1712,23 +3198,24 @@
       }
     },
     "node_modules/@radix-ui/react-tooltip": {
-      "version": "1.2.8",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-tooltip/-/react-tooltip-1.2.8.tgz",
-      "integrity": "sha512-tY7sVt1yL9ozIxvmbtN5qtmH2krXcBCfjEiCgKGLqunJHvgvZG2Pcl2oQ3kbcZARb1BGEHdkLzcYGO8ynVlieg==",
+      "version": "1.2.16",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-tooltip/-/react-tooltip-1.2.16.tgz",
+      "integrity": "sha512-6EamKFRRnlpdadndbZ6LMwycfwkwPte1B42hs6QA0gYhjaOKqW4PZ4pjaW9UrlDX5eVt/OjncE7BFTPL5nmZhg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/primitive": "1.1.3",
-        "@radix-ui/react-compose-refs": "1.1.2",
-        "@radix-ui/react-context": "1.1.2",
-        "@radix-ui/react-dismissable-layer": "1.1.11",
-        "@radix-ui/react-id": "1.1.1",
-        "@radix-ui/react-popper": "1.2.8",
-        "@radix-ui/react-portal": "1.1.9",
-        "@radix-ui/react-presence": "1.1.5",
-        "@radix-ui/react-primitive": "2.1.3",
-        "@radix-ui/react-slot": "1.2.3",
-        "@radix-ui/react-use-controllable-state": "1.2.2",
-        "@radix-ui/react-visually-hidden": "1.2.3"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-id": "1.1.4",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-visually-hidden": "1.2.11"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1745,14 +3232,11 @@
         }
       }
     },
-    "node_modules/@radix-ui/react-tooltip/node_modules/@radix-ui/react-slot": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-slot/-/react-slot-1.2.3.tgz",
-      "integrity": "sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==",
+    "node_modules/@radix-ui/react-tooltip/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
       "license": "MIT",
-      "dependencies": {
-        "@radix-ui/react-compose-refs": "1.1.2"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -1763,6 +3247,29 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-tooltip/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-use-callback-ref": {
       "version": "1.1.1",
       "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.1.tgz",
@@ -1779,13 +3286,14 @@
       }
     },
     "node_modules/@radix-ui/react-use-controllable-state": {
-      "version": "1.2.2",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-controllable-state/-/react-use-controllable-state-1.2.2.tgz",
-      "integrity": "sha512-BjasUjixPFdS+NKkypcyyN5Pmg83Olst0+c6vGov0diwTEo6mgdqVR6hxcEgFuh4QrAs7Rc+9KuGJ9TVCj0Zzg==",
+      "version": "1.2.6",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-controllable-state/-/react-use-controllable-state-1.2.6.tgz",
+      "integrity": "sha512-uEQJGT97ZA/TgP/Hydw47lHu+/vQj6z/0jA+WeTbK1o9Rx45GImjpD0tc3W5ad3D6XTSR6e1yEO0FvGq6WQfVQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-use-effect-event": "0.0.2",
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-use-effect-event": "0.0.5",
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1798,12 +3306,12 @@
       }
     },
     "node_modules/@radix-ui/react-use-effect-event": {
-      "version": "0.0.2",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-effect-event/-/react-use-effect-event-0.0.2.tgz",
-      "integrity": "sha512-Qp8WbZOBe+blgpuUT+lw2xheLP8q0oatc9UpmiemEICxGvFLYmHm9QowVZGHtJlGbS6A6yJ3iViad/2cVjnOiA==",
+      "version": "0.0.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-effect-event/-/react-use-effect-event-0.0.5.tgz",
+      "integrity": "sha512-7cshFL8HGS/7HEiHH+9kL9HBwp2sa9yX18Knwek6KYWmXwM7pegMgta2AXMQKI+rq3JnfSj9x8wYqFMTdG1Jgg==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1816,12 +3324,12 @@
       }
     },
     "node_modules/@radix-ui/react-use-escape-keydown": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-escape-keydown/-/react-use-escape-keydown-1.1.1.tgz",
-      "integrity": "sha512-Il0+boE7w/XebUHyBjroE+DbByORGR9KKmITzbR7MyQ4akpORYP/ZmbhAr0DG7RmmBqoOnZdy2QlvajJ2QA59g==",
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-escape-keydown/-/react-use-escape-keydown-1.1.5.tgz",
+      "integrity": "sha512-ge3ipobwSXTj4JyVtswQ7qZj0ZHdtbGuOno/LrgAAeSxtsJ6Vs4Gz5IkPH2bmqpjcLUFoqGhA/mueuIf63UXlA==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-use-callback-ref": "1.1.1"
+        "@radix-ui/react-use-callback-ref": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1833,14 +3341,26 @@
         }
       }
     },
+    "node_modules/@radix-ui/react-use-escape-keydown/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/@radix-ui/react-use-is-hydrated": {
-      "version": "0.1.0",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-is-hydrated/-/react-use-is-hydrated-0.1.0.tgz",
-      "integrity": "sha512-U+UORVEq+cTnRIaostJv9AGdV3G6Y+zbVd+12e18jQ5A3c0xL03IhnHuiU4UV69wolOQp5GfR58NW/EgdQhwOA==",
+      "version": "0.1.3",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-is-hydrated/-/react-use-is-hydrated-0.1.3.tgz",
+      "integrity": "sha512-umO/aJ+82CpOnhDZUTbILCQf7kU/g0iv+oGs/Q8jw7IkhWBzaEP4sA268PhFAJTFetbwp3ICc6ktpI4TqtxcIw==",
       "license": "MIT",
-      "dependencies": {
-        "use-sync-external-store": "^1.5.0"
-      },
       "peerDependencies": {
         "@types/react": "*",
         "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
@@ -1852,9 +3372,9 @@
       }
     },
     "node_modules/@radix-ui/react-use-layout-effect": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-layout-effect/-/react-use-layout-effect-1.1.1.tgz",
-      "integrity": "sha512-RbJRS4UWQFkzHTTwVymMTUv8EqYhOp8dOOviLj2ugtTiXRaRQS7GLGxZTLL1jWhMeoSCf5zmcZkqTl9IiYfXcQ==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-layout-effect/-/react-use-layout-effect-1.1.4.tgz",
+      "integrity": "sha512-K20DkRkUwDnxEYMBPcg3Y6voLkEy5p5QQmszZgLngKKiC7dzBR/aEuK3w1qlx2JWDUNH6FluahYdgR3BP+QbYw==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -1867,9 +3387,9 @@
       }
     },
     "node_modules/@radix-ui/react-use-previous": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-previous/-/react-use-previous-1.1.1.tgz",
-      "integrity": "sha512-2dHfToCj/pzca2Ck724OZ5L0EVrr3eHRNsG/b3xQJLA2hZpVCS99bLAX+hm1IHXDEnzU6by5z/5MIY794/a8NQ==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-previous/-/react-use-previous-1.1.4.tgz",
+      "integrity": "sha512-XoSLhbRbqxFtgJoi2fNHA3C6pDlY34x508vUpUGoFZfvePfHXHbE1lC4FYFMnJWgiCRroSTw6fOsXQoVS9RwZg==",
       "license": "MIT",
       "peerDependencies": {
         "@types/react": "*",
@@ -1882,12 +3402,12 @@
       }
     },
     "node_modules/@radix-ui/react-use-rect": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-rect/-/react-use-rect-1.1.1.tgz",
-      "integrity": "sha512-QTYuDesS0VtuHNNvMh+CjlKJ4LJickCMUAqjlE3+j8w+RlRpwyX3apEQKGFzbZGdo7XNG1tXa+bQqIE7HIXT2w==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-rect/-/react-use-rect-1.1.4.tgz",
+      "integrity": "sha512-cSOCh6JlkmfjLyNcLiu2nB4v+nm+dkZ+Q5KHWk/soo4U7ZLiEQFKHK9/YmtBHjfCEaU43IBKQOc4/uJmCaiCTQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/rect": "1.1.1"
+        "@radix-ui/rect": "1.1.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1900,12 +3420,12 @@
       }
     },
     "node_modules/@radix-ui/react-use-size": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-size/-/react-use-size-1.1.1.tgz",
-      "integrity": "sha512-ewrXRDTAqAXlkl6t/fkXWNAhFX9I+CkKlw6zjEwk86RSPKwZr3xpBRso655aqYafwtnbpHLj6toFzmd6xdVptQ==",
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-size/-/react-use-size-1.1.4.tgz",
+      "integrity": "sha512-D3anSY15EJoxrihpsXI6SMrmmonnQtR2ni7arO+Lfdg3O95b9hNXxONk8jA5C8ANdF/h5HMAxejgs8PWJ6rlhw==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-use-layout-effect": "1.1.1"
+        "@radix-ui/react-use-layout-effect": "1.1.4"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1918,12 +3438,35 @@
       }
     },
     "node_modules/@radix-ui/react-visually-hidden": {
-      "version": "1.2.3",
-      "resolved": "https://registry.npmjs.org/@radix-ui/react-visually-hidden/-/react-visually-hidden-1.2.3.tgz",
-      "integrity": "sha512-pzJq12tEaaIhqjbzpCuv/OypJY/BPavOofm+dbab+MHLajy277+1lLm6JFcGgF5eskJ6mquGirhXY2GD/8u8Ug==",
+      "version": "1.2.11",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-visually-hidden/-/react-visually-hidden-1.2.11.tgz",
+      "integrity": "sha512-NFS86RYYZb4/exihaESBGOpMJFz8MGLAfu3mOBSGByVnVPC9JPASfYubxd/8KbkQK0sYAv8lVQDEQukDX/qXvQ==",
       "license": "MIT",
       "dependencies": {
-        "@radix-ui/react-primitive": "2.1.3"
+        "@radix-ui/react-primitive": "2.1.10"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/@radix-ui/react-visually-hidden/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
       },
       "peerDependencies": {
         "@types/react": "*",
@@ -1941,9 +3484,9 @@
       }
     },
     "node_modules/@radix-ui/rect": {
-      "version": "1.1.1",
-      "resolved": "https://registry.npmjs.org/@radix-ui/rect/-/rect-1.1.1.tgz",
-      "integrity": "sha512-HPwpGIzkl28mWyZqG52jiqDJ12waP11Pa1lGoiyUkIEuMLBP0oeK/C89esbXrxsky5we7dfd8U58nm0SgAWpVw==",
+      "version": "1.1.3",
+      "resolved": "https://registry.npmjs.org/@radix-ui/rect/-/rect-1.1.3.tgz",
+      "integrity": "sha512-JtyZR+mqgBibTo8xea3B6ZRmzZiM/YeVBtUkas6zMuXjAlfIFIW2FgqeM9eLyvEaYX66vr6DJMK+4U6LV0KhNw==",
       "license": "MIT"
     },
     "node_modules/@rolldown/binding-android-arm64": {
@@ -2194,6 +3737,12 @@
       "integrity": "sha512-qujRfC8sFVInYSPPMLQByRh7zhwkGFS4+tyMQ83srV1qrxL4g8E2tyxVVyxd0+8QeBM1mIk9KbWxkegRr76XzA==",
       "license": "MIT"
     },
+    "node_modules/@simplewebauthn/browser": {
+      "version": "14.0.0",
+      "resolved": "https://registry.npmjs.org/@simplewebauthn/browser/-/browser-14.0.0.tgz",
+      "integrity": "sha512-1odWVqeEBTl7lJ9zMKLEsmTlnyrDO5iRcTvfMKKk1WThUnp/i8JJdffdj2icP+tty159s4PgwE3BiMoEW9NFow==",
+      "license": "MIT"
+    },
     "node_modules/@standard-schema/utils": {
       "version": "0.3.0",
       "resolved": "https://registry.npmjs.org/@standard-schema/utils/-/utils-0.3.0.tgz",
@@ -3641,6 +5190,22 @@
         "node": ">=6"
       }
     },
+    "node_modules/cmdk": {
+      "version": "1.1.1",
+      "resolved": "https://registry.npmjs.org/cmdk/-/cmdk-1.1.1.tgz",
+      "integrity": "sha512-Vsv7kFaXm+ptHDMZ7izaRsP70GgrW9NBNGswt9OZaVBLlE0SNpDq8eu/VGXyF9r7M0azK3Wy7OlYXsuyYLFzHg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-compose-refs": "^1.1.1",
+        "@radix-ui/react-dialog": "^1.1.6",
+        "@radix-ui/react-id": "^1.1.0",
+        "@radix-ui/react-primitive": "^2.0.2"
+      },
+      "peerDependencies": {
+        "react": "^18 || ^19 || ^19.0.0-rc",
+        "react-dom": "^18 || ^19 || ^19.0.0-rc"
+      }
+    },
     "node_modules/color-convert": {
       "version": "2.0.1",
       "resolved": "https://registry.npmjs.org/color-convert/-/color-convert-2.0.1.tgz",
@@ -8127,6 +9692,136 @@
         "node": ">=10.13.0"
       }
     },
+    "node_modules/radix-ui": {
+      "version": "1.6.7",
+      "resolved": "https://registry.npmjs.org/radix-ui/-/radix-ui-1.6.7.tgz",
+      "integrity": "sha512-QBdhh1arIEUvPC0dQ5+nwWAxt7+N+oP/9jPwjJkGFoSk/sqxg32gJtSXGtFh8frAIcS6oC9cx2Q+7KYCQLOAeA==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/primitive": "1.1.7",
+        "@radix-ui/react-accessible-icon": "1.1.15",
+        "@radix-ui/react-accordion": "1.2.20",
+        "@radix-ui/react-alert-dialog": "1.1.23",
+        "@radix-ui/react-arrow": "1.1.15",
+        "@radix-ui/react-aspect-ratio": "1.1.15",
+        "@radix-ui/react-avatar": "1.2.6",
+        "@radix-ui/react-checkbox": "1.3.11",
+        "@radix-ui/react-collapsible": "1.1.20",
+        "@radix-ui/react-collection": "1.1.15",
+        "@radix-ui/react-compose-refs": "1.1.5",
+        "@radix-ui/react-context": "1.2.2",
+        "@radix-ui/react-context-menu": "2.3.7",
+        "@radix-ui/react-dialog": "1.1.23",
+        "@radix-ui/react-direction": "1.1.4",
+        "@radix-ui/react-dismissable-layer": "1.1.19",
+        "@radix-ui/react-dropdown-menu": "2.1.24",
+        "@radix-ui/react-focus-guards": "1.1.6",
+        "@radix-ui/react-focus-scope": "1.1.16",
+        "@radix-ui/react-form": "0.1.16",
+        "@radix-ui/react-hover-card": "1.1.23",
+        "@radix-ui/react-label": "2.1.15",
+        "@radix-ui/react-menu": "2.1.24",
+        "@radix-ui/react-menubar": "1.1.24",
+        "@radix-ui/react-navigation-menu": "1.2.22",
+        "@radix-ui/react-one-time-password-field": "0.1.16",
+        "@radix-ui/react-password-toggle-field": "0.1.11",
+        "@radix-ui/react-popover": "1.1.23",
+        "@radix-ui/react-popper": "1.3.7",
+        "@radix-ui/react-portal": "1.1.17",
+        "@radix-ui/react-presence": "1.1.10",
+        "@radix-ui/react-primitive": "2.1.10",
+        "@radix-ui/react-progress": "1.1.16",
+        "@radix-ui/react-radio-group": "1.4.7",
+        "@radix-ui/react-roving-focus": "1.1.19",
+        "@radix-ui/react-scroll-area": "1.2.18",
+        "@radix-ui/react-select": "2.3.7",
+        "@radix-ui/react-separator": "1.1.15",
+        "@radix-ui/react-slider": "1.4.7",
+        "@radix-ui/react-slot": "1.3.3",
+        "@radix-ui/react-switch": "1.3.7",
+        "@radix-ui/react-tabs": "1.1.21",
+        "@radix-ui/react-toast": "1.2.23",
+        "@radix-ui/react-toggle": "1.1.18",
+        "@radix-ui/react-toggle-group": "1.1.19",
+        "@radix-ui/react-toolbar": "1.1.19",
+        "@radix-ui/react-tooltip": "1.2.16",
+        "@radix-ui/react-use-callback-ref": "1.1.4",
+        "@radix-ui/react-use-controllable-state": "1.2.6",
+        "@radix-ui/react-use-effect-event": "0.0.5",
+        "@radix-ui/react-use-escape-keydown": "1.1.5",
+        "@radix-ui/react-use-is-hydrated": "0.1.3",
+        "@radix-ui/react-use-layout-effect": "1.1.4",
+        "@radix-ui/react-use-size": "1.1.4",
+        "@radix-ui/react-visually-hidden": "1.2.11"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/radix-ui/node_modules/@radix-ui/react-compose-refs": {
+      "version": "1.1.5",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-compose-refs/-/react-compose-refs-1.1.5.tgz",
+      "integrity": "sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/radix-ui/node_modules/@radix-ui/react-primitive": {
+      "version": "2.1.10",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-primitive/-/react-primitive-2.1.10.tgz",
+      "integrity": "sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==",
+      "license": "MIT",
+      "dependencies": {
+        "@radix-ui/react-slot": "1.3.3"
+      },
+      "peerDependencies": {
+        "@types/react": "*",
+        "@types/react-dom": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc",
+        "react-dom": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        },
+        "@types/react-dom": {
+          "optional": true
+        }
+      }
+    },
+    "node_modules/radix-ui/node_modules/@radix-ui/react-use-callback-ref": {
+      "version": "1.1.4",
+      "resolved": "https://registry.npmjs.org/@radix-ui/react-use-callback-ref/-/react-use-callback-ref-1.1.4.tgz",
+      "integrity": "sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==",
+      "license": "MIT",
+      "peerDependencies": {
+        "@types/react": "*",
+        "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc"
+      },
+      "peerDependenciesMeta": {
+        "@types/react": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/react": {
       "version": "19.2.1",
       "resolved": "https://registry.npmjs.org/react/-/react-19.2.1.tgz",
@@ -9815,15 +11510,6 @@
         }
       }
     },
-    "node_modules/use-sync-external-store": {
-      "version": "1.6.0",
-      "resolved": "https://registry.npmjs.org/use-sync-external-store/-/use-sync-external-store-1.6.0.tgz",
-      "integrity": "sha512-Pp6GSwGP/NrPIrxVFAIkOQeyw8lFenOHijQWkUTrDvrF4ALqylP2C/KCkeS9dpUM3KvYRQhna5vt7IL95+ZQ9w==",
-      "license": "MIT",
-      "peerDependencies": {
-        "react": "^16.8.0 || ^17.0.0 || ^18.0.0 || ^19.0.0"
-      }
-    },
     "node_modules/uuidjs": {
       "version": "5.1.0",
       "resolved": "https://registry.npmjs.org/uuidjs/-/uuidjs-5.1.0.tgz",
diff --git a/web/package.json b/web/package.json
index 554594214..358f146c6 100644
--- a/web/package.json
+++ b/web/package.json
@@ -55,6 +55,7 @@
     "@radix-ui/react-toggle": "^1.1.8",
     "@radix-ui/react-toggle-group": "^1.1.9",
     "@radix-ui/react-tooltip": "^1.2.7",
+    "@remixicon/react": "^4.9.0",
     "@simplewebauthn/browser": "^14.0.0",
     "@tailwindcss/postcss": "^4.1.5",
     "@tanstack/react-table": "^8.21.3",
diff --git a/web/pnpm-lock.yaml b/web/pnpm-lock.yaml
index fd0c6c5f6..4838d2a6f 100644
--- a/web/pnpm-lock.yaml
+++ b/web/pnpm-lock.yaml
@@ -29,85 +29,88 @@ dependencies:
     version: 4.6.2
   '@hookform/resolvers':
     specifier: ^5.0.1
-    version: 5.2.2(react-hook-form@7.71.1)
+    version: 5.9.1(react-hook-form@7.90.0)(zod@3.25.76)
   '@radix-ui/react-alert-dialog':
     specifier: ^1.1.15
-    version: 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-avatar':
     specifier: ^1.1.11
-    version: 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.2.7(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-checkbox':
     specifier: ^1.3.1
-    version: 1.3.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.3.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-collapsible':
     specifier: ^1.1.12
-    version: 1.1.12(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-context-menu':
     specifier: ^2.2.15
-    version: 2.2.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-dialog':
     specifier: ^1.1.14
-    version: 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-dropdown-menu':
     specifier: ^2.1.16
-    version: 2.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-hover-card':
     specifier: ^1.1.13
-    version: 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-label':
     specifier: ^2.1.6
-    version: 2.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 2.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-popover':
     specifier: ^1.1.14
-    version: 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-progress':
     specifier: ^1.1.8
-    version: 1.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.17(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-scroll-area':
     specifier: ^1.2.9
-    version: 1.2.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-select':
     specifier: ^2.2.4
-    version: 2.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-separator':
     specifier: ^1.1.8
-    version: 1.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-slider':
     specifier: ^1.4.7
-    version: 1.4.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.5.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-slot':
     specifier: ^1.2.3
-    version: 1.2.4(@types/react@19.2.10)(react@19.2.1)
+    version: 1.4.0(@types/react@19.2.18)(react@19.2.1)
   '@radix-ui/react-switch':
     specifier: ^1.2.4
-    version: 1.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-tabs':
     specifier: ^1.1.11
-    version: 1.1.13(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.22(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-toggle':
     specifier: ^1.1.8
-    version: 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.19(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-toggle-group':
     specifier: ^1.1.9
-    version: 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   '@radix-ui/react-tooltip':
     specifier: ^1.2.7
-    version: 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+  '@remixicon/react':
+    specifier: ^4.9.0
+    version: 4.9.0(react@19.2.1)
   '@simplewebauthn/browser':
     specifier: ^14.0.0
     version: 14.0.0
   '@tailwindcss/postcss':
     specifier: ^4.1.5
-    version: 4.1.18
+    version: 4.3.3
   '@tanstack/react-table':
     specifier: ^8.21.3
     version: 8.21.3(react-dom@19.2.1)(react@19.2.1)
   '@vitejs/plugin-react':
     specifier: ^6.0.1
-    version: 6.0.1(vite@8.0.16)
+    version: 6.1.2(vite@8.3.4)
   axios:
     specifier: ^1.16.0
-    version: 1.17.0
+    version: 1.20.0
   class-variance-authority:
     specifier: ^0.7.1
     version: 0.7.1
@@ -116,19 +119,19 @@ dependencies:
     version: 2.1.1
   cmdk:
     specifier: ^1.1.1
-    version: 1.1.1(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.1.1(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   highlight.js:
     specifier: ^11.11.1
-    version: 11.11.1
+    version: 11.12.0
   i18next:
     specifier: ^25.1.2
-    version: 25.8.0(typescript@5.9.3)
+    version: 25.10.10(typescript@5.9.3)
   i18next-browser-languagedetector:
     specifier: ^8.1.0
-    version: 8.2.0
+    version: 8.2.1
   input-otp:
     specifier: ^1.4.2
-    version: 1.4.2(react-dom@19.2.1)(react@19.2.1)
+    version: 1.5.0(react-dom@19.2.1)(react@19.2.1)
   lodash:
     specifier: ^4.18.0
     version: 4.18.1
@@ -137,13 +140,13 @@ dependencies:
     version: 0.507.0(react@19.2.1)
   postcss:
     specifier: ^8.5.10
-    version: 8.5.15
+    version: 8.5.29
   qrcode:
     specifier: ^1.5.4
     version: 1.5.4
   radix-ui:
     specifier: ^1.6.0
-    version: 1.6.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+    version: 1.7.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   react:
     specifier: 19.2.1
     version: 19.2.1
@@ -152,22 +155,22 @@ dependencies:
     version: 19.2.1(react@19.2.1)
   react-hook-form:
     specifier: ^7.56.3
-    version: 7.71.1(react@19.2.1)
+    version: 7.90.0(@types/react@19.2.18)(react@19.2.1)
   react-i18next:
     specifier: ^15.5.1
-    version: 15.7.4(i18next@25.8.0)(react-dom@19.2.1)(react@19.2.1)(typescript@5.9.3)
+    version: 15.7.4(i18next@25.10.10)(react-dom@19.2.1)(react@19.2.1)(typescript@5.9.3)
   react-markdown:
     specifier: ^10.1.0
-    version: 10.1.0(@types/react@19.2.10)(react@19.2.1)
+    version: 10.1.0(@types/react@19.2.18)(react@19.2.1)
   react-photo-view:
     specifier: ^1.2.7
     version: 1.2.7(react-dom@19.2.1)(react@19.2.1)
   react-router-dom:
     specifier: ^7.15.0
-    version: 7.17.0(react-dom@19.2.1)(react@19.2.1)
+    version: 7.18.4(react-dom@19.2.1)(react@19.2.1)
   react-syntax-highlighter:
     specifier: ^16.1.0
-    version: 16.1.0(react@19.2.1)
+    version: 16.1.2(react@19.2.1)
   recharts:
     specifier: 2.15.4
     version: 2.15.4(react-dom@19.2.1)(react@19.2.1)
@@ -191,19 +194,19 @@ dependencies:
     version: 4.0.1
   sonner:
     specifier: ^2.0.3
-    version: 2.0.7(react-dom@19.2.1)(react@19.2.1)
+    version: 2.0.8(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
   tailwind-merge:
     specifier: ^3.2.0
-    version: 3.4.0
+    version: 3.7.0
   tailwindcss:
     specifier: ^4.1.5
-    version: 4.1.18
+    version: 4.3.3
   uuidjs:
     specifier: ^5.1.0
     version: 5.1.0
   vite:
     specifier: ^8.0.16
-    version: 8.0.16(@types/node@20.19.30)
+    version: 8.3.4(@types/node@20.19.43)
   zod:
     specifier: ^3.24.4
     version: 3.25.76
@@ -211,22 +214,22 @@ dependencies:
 devDependencies:
   '@playwright/test':
     specifier: ^1.61.0
-    version: 1.61.0
+    version: 1.64.0
   '@types/debug':
     specifier: ^4.1.12
-    version: 4.1.12
+    version: 4.1.13
   '@types/estree':
     specifier: ^1.0.8
-    version: 1.0.8
+    version: 1.0.9
   '@types/estree-jsx':
     specifier: ^1.0.5
     version: 1.0.5
   '@types/hast':
     specifier: ^3.0.4
-    version: 3.0.4
+    version: 3.0.5
   '@types/lodash':
     specifier: ^4.17.16
-    version: 4.17.23
+    version: 4.17.26
   '@types/mdast':
     specifier: ^4.0.4
     version: 4.0.4
@@ -235,16 +238,16 @@ devDependencies:
     version: 2.1.0
   '@types/node':
     specifier: ^20
-    version: 20.19.30
+    version: 20.19.43
   '@types/qrcode':
     specifier: ^1.5.6
     version: 1.5.6
   '@types/react':
     specifier: ~19.2.7
-    version: 19.2.10
+    version: 19.2.18
   '@types/react-dom':
     specifier: ~19.2.3
-    version: 19.2.3(@types/react@19.2.10)
+    version: 19.2.7(@types/react@19.2.18)
   '@types/react-syntax-highlighter':
     specifier: ^15.5.13
     version: 15.5.13
@@ -253,25 +256,25 @@ devDependencies:
     version: 3.0.3
   eslint:
     specifier: ^9
-    version: 9.39.2
+    version: 9.39.5
   eslint-config-prettier:
     specifier: ^10.1.2
-    version: 10.1.8(eslint@9.39.2)
+    version: 10.1.8(eslint@9.39.5)
   eslint-plugin-prettier:
     specifier: ^5.2.6
-    version: 5.5.5(eslint-config-prettier@10.1.8)(eslint@9.39.2)(prettier@3.8.1)
+    version: 5.5.6(eslint-config-prettier@10.1.8)(eslint@9.39.5)(prettier@3.9.10)
   eslint-plugin-react:
     specifier: ^7.37.5
-    version: 7.37.5(eslint@9.39.2)
+    version: 7.37.5(eslint@9.39.5)
   eslint-plugin-react-hooks:
     specifier: ^5.2.0
-    version: 5.2.0(eslint@9.39.2)
+    version: 5.2.0(eslint@9.39.5)
   lint-staged:
     specifier: ^15.5.1
     version: 15.5.2
   prettier:
     specifier: ^3.5.3
-    version: 3.8.1
+    version: 3.9.10
   tw-animate-css:
     specifier: ^1.2.9
     version: 1.4.0
@@ -280,17 +283,17 @@ devDependencies:
     version: 5.9.3
   typescript-eslint:
     specifier: ^8.31.1
-    version: 8.54.0(eslint@9.39.2)(typescript@5.9.3)
+    version: 8.71.1(eslint@9.39.5)(typescript@5.9.3)
 
 packages:
 
-  /@alloc/quick-lru@5.2.0:
-    resolution: {integrity: sha512-UrcABB+4bUrFABwbluTIBErXwvbsU/V7TZWfmbgJfbkwiBuziS9gxdODUyuiecfdGQ85jglMW6juS3+z5TsKLw==}
+  /@alloc/quick-lru@5.3.0:
+    resolution: {integrity: sha512-U4+70Pc5ZS9osnCBCE5Jha/ciHM+Yp+CNMNC/7HvYbNRk1Ldd+f7qO65W5qfhu/TCv+/ozljlXXe9Nj8419DMA==}
     engines: {node: '>=10'}
     dev: false
 
-  /@babel/runtime@7.28.6:
-    resolution: {integrity: sha512-05WQkdpL9COIMz4LjTxGpPNCdlpyimKppYNoJ5Di5EUObifl8t4tuLuUBBZEpoLYOmfvIWrsp9fCl0HoPRVTdA==}
+  /@babel/runtime@7.29.10:
+    resolution: {integrity: sha512-NpugTzbKgGrNe45Ui32cN8vCFFMxVtP/9rtfwv0nP5FRDAjvYBxJcLVpYA9mr1jxP6cnIT7x9LUzuDTbWYYNRA==}
     engines: {node: '>=6.9.0'}
     dev: false
 
@@ -337,38 +340,13 @@ packages:
       tslib: 2.8.1
     dev: false
 
-  /@emnapi/core@1.10.0:
-    resolution: {integrity: sha512-yq6OkJ4p82CAfPl0u9mQebQHKPJkY7WrIuk205cTYnYe+k2Z8YBh11FrbRG/H6ihirqcacOgl2BIO8oyMQLeXw==}
-    requiresBuild: true
-    dependencies:
-      '@emnapi/wasi-threads': 1.2.1
-      tslib: 2.8.1
-    dev: false
-    optional: true
-
-  /@emnapi/runtime@1.10.0:
-    resolution: {integrity: sha512-ewvYlk86xUoGI0zQRNq/mC+16R1QeDlKQy21Ki3oSYXNgLb45GV1P6A0M+/s6nyCuNDqe5VpaY84BzXGwVbwFA==}
-    requiresBuild: true
-    dependencies:
-      tslib: 2.8.1
-    dev: false
-    optional: true
-
-  /@emnapi/wasi-threads@1.2.1:
-    resolution: {integrity: sha512-uTII7OYF+/Mes/MrcIOYp5yOtSMLBWSIoLPpcgwipoiKbli6k322tcoFsxoIIxPDqW01SQGAgko4EzZi2BNv2w==}
-    requiresBuild: true
-    dependencies:
-      tslib: 2.8.1
-    dev: false
-    optional: true
-
-  /@eslint-community/eslint-utils@4.9.1(eslint@9.39.2):
-    resolution: {integrity: sha512-phrYmNiYppR7znFEdqgfWHXR6NCkZEK7hwWDHZUjit/2/U0r6XvkDl0SYnoM51Hq7FhCGdLDT6zxCCOY1hexsQ==}
+  /@eslint-community/eslint-utils@4.10.1(eslint@9.39.5):
+    resolution: {integrity: sha512-cuadcxVFE8sDK6iWJbs8Sn0av2Nrh2QSGQhVlBW9AaAHqHwjWsZHT8LJ4hFGPh7ASBV2deFdM7H/DPjulmh8rg==}
     engines: {node: ^12.22.0 || ^14.17.0 || >=16.0.0}
     peerDependencies:
       eslint: ^6.0.0 || ^7.0.0 || >=8.0.0
     dependencies:
-      eslint: 9.39.2
+      eslint: 9.39.5
       eslint-visitor-keys: 3.4.3
     dev: true
 
@@ -377,13 +355,13 @@ packages:
     engines: {node: ^12.0.0 || ^14.0.0 || >=16.0.0}
     dev: true
 
-  /@eslint/config-array@0.21.1:
-    resolution: {integrity: sha512-aw1gNayWpdI/jSYVgzN5pL0cfzU02GT3NBpeT/DXbx1/1x7ZKxFPd9bwrzygx/qiwIQiJ1sw/zD8qY/kRvlGHA==}
+  /@eslint/config-array@0.21.2:
+    resolution: {integrity: sha512-nJl2KGTlrf9GjLimgIru+V/mzgSK0ABCDQRvxw5BjURL7WfH5uoWmizbH7QB6MmnMBd8cIC9uceWnezL1VZWWw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dependencies:
       '@eslint/object-schema': 2.1.7
       debug: 4.4.3
-      minimatch: 3.1.3
+      minimatch: 3.1.5
     transitivePeerDependencies:
       - supports-color
     dev: true
@@ -402,25 +380,25 @@ packages:
       '@types/json-schema': 7.0.15
     dev: true
 
-  /@eslint/eslintrc@3.3.3:
-    resolution: {integrity: sha512-Kr+LPIUVKz2qkx1HAMH8q1q6azbqBAsXJUxBl/ODDuVPX45Z9DfwB8tPjTi6nNZ8BuM3nbJxC5zCAg5elnBUTQ==}
+  /@eslint/eslintrc@3.3.7:
+    resolution: {integrity: sha512-F42g89Qd5oAWtp0k0nnSrjziAKza7w8SVT4mStc18LZMaRb4J1HQAHLCalEtDCxrTuksx7NU9qsmeLwpOfPqWw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dependencies:
-      ajv: 6.12.6
+      ajv: 6.15.0
       debug: 4.4.3
       espree: 10.4.0
       globals: 14.0.0
       ignore: 5.3.2
       import-fresh: 3.3.1
-      js-yaml: 4.2.0
-      minimatch: 3.1.3
+      js-yaml: 4.3.2
+      minimatch: 3.1.5
       strip-json-comments: 3.1.1
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@eslint/js@9.39.2:
-    resolution: {integrity: sha512-q1mjIoW1VX4IvSocvM/vbTiveKC4k9eLrajNEuSsmjymSDEbpGddtpfOoN7YGAqBK3NG+uqo8ia4PDTt8buCYA==}
+  /@eslint/js@9.39.5:
+    resolution: {integrity: sha512-QywQuszQh77pIXCsq998c8hbhSTI/azTty1Z6N53dmAudKHhy573j3yvRLsX2BSp8YpLtoCEG8E9DJe+8zUh4A==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dev: true
 
@@ -443,56 +421,138 @@ packages:
       tslib: 2.8.1
     dev: false
 
-  /@floating-ui/core@1.7.4:
-    resolution: {integrity: sha512-C3HlIdsBxszvm5McXlB8PeOEWfBhcGBTZGkGlWc2U0KFY5IwG5OQEuQ8rq52DZmcHDlPLd+YFBK+cZcytwIFWg==}
+  /@floating-ui/core@1.8.0:
+    resolution: {integrity: sha512-0CIZ5itps/8x7BG8dEIhs53BvCUH2PCoogtakwRTut+Arm58sJooJ0AuZhLw2HJYIR5cMLNPBSS728sPho2khQ==}
     dependencies:
-      '@floating-ui/utils': 0.2.10
+      '@floating-ui/utils': 0.2.12
     dev: false
 
-  /@floating-ui/dom@1.7.5:
-    resolution: {integrity: sha512-N0bD2kIPInNHUHehXhMke1rBGs1dwqvC9O9KYMyyjK7iXt7GAhnro7UlcuYcGdS/yYOlq0MAVgrow8IbWJwyqg==}
+  /@floating-ui/dom@1.8.0:
+    resolution: {integrity: sha512-yXSrzeHZBTZadLOlfyhCkJHNeLJnHRnRInwdZ40L7ZiaAtrBwoYlsDrX3v5zB1Utk7CLfzcOVnVVWoXEky7Ceg==}
     dependencies:
-      '@floating-ui/core': 1.7.4
-      '@floating-ui/utils': 0.2.10
+      '@floating-ui/core': 1.8.0
+      '@floating-ui/utils': 0.2.12
     dev: false
 
-  /@floating-ui/react-dom@2.1.7(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-0tLRojf/1Go2JgEVm+3Frg9A3IW8bJgKgdO0BN5RkF//ufuz2joZM63Npau2ff3J6lUVYgDSNzNkR+aH3IVfjg==}
+  /@floating-ui/react-dom@2.1.9(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-JDjEFGCpImxDCA7JJKviA0M9+RtmJdj0m/NVU5IMgBK+AmZouAQQ7/+2GLH0GXXY0YMw9oXPB8hKdbPYg5QLYg==}
     peerDependencies:
       react: '>=16.8.0'
       react-dom: '>=16.8.0'
     dependencies:
-      '@floating-ui/dom': 1.7.5
+      '@floating-ui/dom': 1.8.0
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@floating-ui/utils@0.2.10:
-    resolution: {integrity: sha512-aGTxbpbg8/b5JfU1HXSrbH3wXZuLPJcNEcZQFMxLs3oSzgtVu6nFPkbbGGUvBcUjKV2YyB9Wxxabo+HEH9tcRQ==}
+  /@floating-ui/utils@0.2.12:
+    resolution: {integrity: sha512-HpCo8tmWzLVad5s2d19EhAz5zqrrQ6s69qd6moPMQvkOuSwDT1YgRfWSVuc4ennqrgv3OHppiOGMQ7oC13yIww==}
     dev: false
 
-  /@hookform/resolvers@5.2.2(react-hook-form@7.71.1):
-    resolution: {integrity: sha512-A/IxlMLShx3KjV/HeTcTfaMxdwy690+L/ZADoeaTltLx+CVuzkeVIPuybK3jrRfw7YZnmdKsVVHAlEPIAEUNlA==}
+  /@hookform/resolvers@5.9.1(react-hook-form@7.90.0)(zod@3.25.76):
+    resolution: {integrity: sha512-7b7vsbraJxKgjVSA1Nur9tLwj539WGJUBLA7QNvXnFoT2pM5Z7G+6rlukk4B2/QrTZy6huRtH6wKeESPKuIr6w==}
     peerDependencies:
+      '@sinclair/typebox': '>=0.25.24'
+      '@standard-schema/spec': ^1.0.0
+      '@typeschema/main': '>=0.13.7'
+      '@vinejs/vine': ^2.0.0 || ^3.0.0 || ^4.0.0
+      ajv: ^8.12.0
+      ajv-errors: ^3.0.0
+      ajv-formats: ^2.1.1
+      arktype: ^2.0.0
+      ata-validator: ^1.2.0
+      class-transformer: '>=0.4.0'
+      class-validator: '>=0.12.0'
+      computed-types: ^1.0.0
+      effect: ^3.10.3
+      fluentvalidation-ts: ^3.0.0
+      fp-ts: ^2.7.0
+      io-ts: ^2.0.0
+      joi: ^17.0.0 || ^18.0.0
+      nope-validator: '>=0.12.0'
       react-hook-form: ^7.55.0
+      superstruct: '>=0.12.0'
+      typanion: ^3.3.2
+      valibot: '>=0.31.0 || ^1.0.0-beta.4 || ^1.0.0-rc'
+      vest: '>=6.0.0'
+      yup: ^1.0.0
+      zod: ^3.25.0 || ^4.0.0
+    peerDependenciesMeta:
+      '@sinclair/typebox':
+        optional: true
+      '@standard-schema/spec':
+        optional: true
+      '@typeschema/main':
+        optional: true
+      '@vinejs/vine':
+        optional: true
+      ajv:
+        optional: true
+      ajv-errors:
+        optional: true
+      ajv-formats:
+        optional: true
+      arktype:
+        optional: true
+      ata-validator:
+        optional: true
+      class-transformer:
+        optional: true
+      class-validator:
+        optional: true
+      computed-types:
+        optional: true
+      effect:
+        optional: true
+      fluentvalidation-ts:
+        optional: true
+      fp-ts:
+        optional: true
+      io-ts:
+        optional: true
+      joi:
+        optional: true
+      nope-validator:
+        optional: true
+      superstruct:
+        optional: true
+      typanion:
+        optional: true
+      valibot:
+        optional: true
+      vest:
+        optional: true
+      yup:
+        optional: true
+      zod:
+        optional: true
     dependencies:
       '@standard-schema/utils': 0.3.0
-      react-hook-form: 7.71.1(react@19.2.1)
+      react-hook-form: 7.90.0(@types/react@19.2.18)(react@19.2.1)
+      zod: 3.25.76
     dev: false
 
-  /@humanfs/core@0.19.1:
-    resolution: {integrity: sha512-5DyQ4+1JEUzejeK1JGICcideyfUbGixgS9jNgex5nqkW+cY7WZhxBigmieN5Qnw9ZosSNVC9KQKyb+GUaGyKUA==}
-    engines: {node: '>=18.18.0'}
-    dev: true
-
-  /@humanfs/node@0.16.7:
-    resolution: {integrity: sha512-/zUx+yOsIrG4Y43Eh2peDeKCxlRt/gET6aHfaKpuq267qXdYDFViVHfMaLyygZOnl0kGWxFIgsBy8QFuTLUXEQ==}
+  /@humanfs/core@0.19.2:
+    resolution: {integrity: sha512-UhXNm+CFMWcbChXywFwkmhqjs3PRCmcSa/hfBgLIb7oQ5HNb1wS0icWsGtSAUNgefHeI+eBrA8I1fxmbHsGdvA==}
     engines: {node: '>=18.18.0'}
     dependencies:
-      '@humanfs/core': 0.19.1
+      '@humanfs/types': 0.15.0
+    dev: true
+
+  /@humanfs/node@0.16.8:
+    resolution: {integrity: sha512-gE1eQNZ3R++kTzFUpdGlpmy8kDZD/MLyHqDwqjkVQI0JMdI1D51sy1H958PNXYkM2rAac7e5/CnIKZrHtPh3BQ==}
+    engines: {node: '>=18.18.0'}
+    dependencies:
+      '@humanfs/core': 0.19.2
+      '@humanfs/types': 0.15.0
       '@humanwhocodes/retry': 0.4.3
     dev: true
 
+  /@humanfs/types@0.15.0:
+    resolution: {integrity: sha512-ZZ1w0aoQkwuUuC7Yf+7sdeaNfqQiiLcSRbfI08oAxqLtpXQr9AIVX7Ay7HLDuiLYAaFPu8oBYNq/QIi9URHJ3Q==}
+    engines: {node: '>=18.18.0'}
+    dev: true
+
   /@humanwhocodes/module-importer@1.0.1:
     resolution: {integrity: sha512-bxveV4V8v5Yb4ncFTT3rPSgZBOpCkjfK0y4oVVVJwIuDVBRMDXrPyXRL988i5ap9m9bnyEEjWfm5WkBmtffLfA==}
     engines: {node: '>=12.22'}
@@ -506,7 +566,7 @@ packages:
   /@jridgewell/gen-mapping@0.3.13:
     resolution: {integrity: sha512-2kkt/7niJ6MgEPxF0bYdQ6etZaA+fQvDcLKckhy1yIQOzaoKjBBjSj63/aLVjYE3qhRt5dvM+uUyfCg6UKCBbA==}
     dependencies:
-      '@jridgewell/sourcemap-codec': 1.5.5
+      '@jridgewell/sourcemap-codec': 1.6.0
       '@jridgewell/trace-mapping': 0.3.31
     dev: false
 
@@ -522,65 +582,44 @@ packages:
     engines: {node: '>=6.0.0'}
     dev: false
 
-  /@jridgewell/sourcemap-codec@1.5.5:
-    resolution: {integrity: sha512-cYQ9310grqxueWbl+WuIUIaiUaDcj7WOq5fVhEljNVgRfOUhY9fy2zTvfoqWsnebh8Sl70VScFbICvJnLKB0Og==}
+  /@jridgewell/sourcemap-codec@1.6.0:
+    resolution: {integrity: sha512-T7jf+5zgsZHwNJ4lvQ7/aezbyk0nNX+zJVWpmHA7VYsEx7a7qr5Rg5IbtJFqkgze5Y2sruq1RUY8Q837Od7iFw==}
     dev: false
 
   /@jridgewell/trace-mapping@0.3.31:
     resolution: {integrity: sha512-zzNR+SdQSDJzc8joaeP8QQoCQr8NuYx2dIIytl1QeBEZHJ9uW6hebsrYgbz8hJwUQao3TWCMtmfV8Nu1twOLAw==}
     dependencies:
       '@jridgewell/resolve-uri': 3.1.2
-      '@jridgewell/sourcemap-codec': 1.5.5
+      '@jridgewell/sourcemap-codec': 1.6.0
     dev: false
 
-  /@napi-rs/wasm-runtime@1.1.4(@emnapi/core@1.10.0)(@emnapi/runtime@1.10.0):
-    resolution: {integrity: sha512-3NQNNgA1YSlJb/kMH1ildASP9HW7/7kYnRI2szWJaofaS1hWmbGI4H+d3+22aGzXXN9IJ+n+GiFVcGipJP18ow==}
-    requiresBuild: true
-    peerDependencies:
-      '@emnapi/core': ^1.7.1
-      '@emnapi/runtime': ^1.7.1
-    dependencies:
-      '@emnapi/core': 1.10.0
-      '@emnapi/runtime': 1.10.0
-      '@tybys/wasm-util': 0.10.1
-    dev: false
-    optional: true
-
-  /@oxc-project/types@0.133.0:
-    resolution: {integrity: sha512-KzkdCd6Uxqnf6l3HOw1xfatAlUURA0g14cvBYFyJ5SaNOQbOUvBr9PKArcPcrNIeRsBdgcUzOGrhKveVpvOIGA==}
+  /@oxc-project/types@0.153.0:
+    resolution: {integrity: sha512-jSNrRIyvPVLL5E1z/tS5zLAprosGz/FuI1PWwEOz59j3E4TTTzSpqKb/RDGqJhhW0QHYISGXxq582wGx6T11Tw==}
     dev: false
 
-  /@pkgr/core@0.2.9:
-    resolution: {integrity: sha512-QNqXyfVS2wm9hweSYD2O7F0G06uurj9kZ96TRQE5Y9hU7+tgdZwIkbAKc5Ocy1HxEY2kuDQa6cQ1WRs/O5LFKA==}
-    engines: {node: ^12.20.0 || ^14.18.0 || >=16.0.0}
+  /@pkgr/core@0.3.6:
+    resolution: {integrity: sha512-SEeaJLb3qBNF/OaXnaR1NmmBbFYk1zC0ZH/52fATcRPLFg/p791YrcyFFy44Bo9sLaGuSuLp5Q6axbb/O+v/RA==}
+    engines: {node: ^14.18.0 || >=16.0.0}
     dev: true
 
-  /@playwright/test@1.61.0:
-    resolution: {integrity: sha512-cKA5B6lpFEMyMGjxF54QihfYpB4FkEGH+qZhtArDEG+wezQAJY8Pq6C7T1SjWz+FFzt3TbyoXBQYk/0292TdJA==}
-    engines: {node: '>=18'}
+  /@playwright/test@1.64.0:
+    resolution: {integrity: sha512-oTN9FNYuN6bqQMwiho6jJKLCG2/QOYCixGXuCjRmzOq23eODeukuY2XiAVoVpAECbuezMt1MHQjAiQxK8F+Y+g==}
+    engines: {node: '>=20'}
     hasBin: true
     dependencies:
-      playwright: 1.61.0
+      playwright: 1.64.0
     dev: true
 
-  /@radix-ui/number@1.1.1:
-    resolution: {integrity: sha512-MkKCwxlXTgz6CFoJx3pCwn07GKp36+aZyu/u2Ln2VrA5DcdyCZkASEDBTd8x5whTQQL5CiYf4prXKLcgQdv29g==}
-    dev: false
-
   /@radix-ui/number@1.1.3:
     resolution: {integrity: sha512-Road2bidD0uu/1BGDOWNdPI06g0lIRy6IF9GZcIrDK2KGItfor8IQwQa+yM2ERgHM1MmHxaxpTzk0/Jp42lNfA==}
     dev: false
 
-  /@radix-ui/primitive@1.1.3:
-    resolution: {integrity: sha512-JTF99U/6XIjCBo0wqkU5sK10glYe27MRRsfwoiq5zzOEZLHU3A3KCMa5X/azekYRCJ0HlwI0crAXS/5dEHTzDg==}
-    dev: false
-
   /@radix-ui/primitive@1.1.7:
     resolution: {integrity: sha512-rqWnm76nYT8HoNNqEjpgJ7Pw/DrBj5iBTrmEPo6HTX5+VJyBNOqTdv4g89G63HuR5g0AaENoAcH7Is5fF2kZ8Q==}
     dev: false
 
-  /@radix-ui/react-accessible-icon@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-WTQwcAvQf5sOcuUyi90lKPbhwcvQ+j55cjrSmeaN+L2vKU3DooOvlKw2MDeiJ5IkV5N905KW0/fGojKOBhD11A==}
+  /@radix-ui/react-accessible-icon@1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-I+T/Hc6SY3WbFVZ9Ne1o0DJbpjQpTpLIBZcVzYiHCQ9z1yGUd/G88u5SKyM6eJ223txAOS7bp7bb9k0Ocfkjwg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -592,15 +631,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-accordion@1.2.20(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-jDhG9FvAEnlhnjrsINbNXcUa4G+L1KqSkJSunkbKEzFRcAb52jvM0PjPxPRvhe1HNc5F5yc0yzzWeeqlH4yBIg==}
+  /@radix-ui/react-accordion@1.2.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-2fmc6COuKLq2k3P7JZrOt9Zf8HT9nTQhHN5V+A3ih7ftqoyDcoEUrPoLq8ngr7QwTd7hpQ2xoUKWtvy1Xw9RAQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -613,47 +652,22 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collapsible': 1.1.20(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collapsible': 1.1.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-alert-dialog@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-oTVLkEw5GpdRe29BqJ0LSDFWI3qu0vR1M0mUkOQWDIUnY/QIkLpgDMWuKxP94c2NAC2LGcgVhG1ImF3jkZ5wXw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dialog': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-alert-dialog@1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-VAYOiQRqj3GPpYJE0I9J+X8Ip05cyVlNdKOFeiGS2Ou1HHGfpl0BxOyZm6nmVDyU+W+NF3/XLzmjHmVGydhwgA==}
+  /@radix-ui/react-alert-dialog@1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-dRvB59HhcjIWcCPKzK5Px1yQ8pIULQQrxqgyepCqin7tNXi/xvmC86WkyMOKhuCbjz3+v9Q6oE7/tUN5NQ9dCg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -666,18 +680,18 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dialog': 1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dialog': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-arrow@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-v4zggRcjadnI+ClKDuijlQEW4tw3NoaeHc/PwpKnLoLLKNUG4InLegkstooLcRIUWCs+8L22dGURCVuFfOKfnA==}
+  /@radix-ui/react-arrow@1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-nXPI1tzJsnQw8oXvB2r49rO9TIbbJ8VFeaWDiBsqY8+iwO5xb7WjUsrXDagjAgPKTdhW8WODQnAltUhBx8I+sg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -689,15 +703,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-arrow@1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-F+M1tLhO+mlQaOWspE8Wstg+z6PwxwRd8oQ8IXceWz92kfAmalTRf0EjrouQeo7QssEPfCn05B4Ihs1K9WQ/7w==}
+  /@radix-ui/react-aspect-ratio@1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-w2u6oiTzb7WkWwV3ZClX/llYYlTbilUOkHIKkwb+PWWytQ/IdEVRNLIChjbcXy/mkqcKKM3wEhrOilJW+uPHDg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -709,59 +723,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-aspect-ratio@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-fy+dyVR+90nelK8rqIznFlxzx7uPcGbhxH8Nfr2bHb4UfSe+e3hklOC0luK0hDwVwnRX7xTRySpsrQVeW+/oNQ==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-avatar@1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-0Qk603AHGV28BOBO34p7IgD5m+V5Sg/YovfayABkoDDBM5d3NCx0Mp4gGrjzLGes1jV5eNOE1r3itqOR33VC6Q==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-context': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.4(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.0(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-avatar@1.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-4ULOTJ/mqy2hT9GlWa/MFHxHSvH3nJzHnZM1waNsc5Bonv7i70aNenghXmD97S6OJ81ekXONGGt4nT1r0PfEdA==}
+  /@radix-ui/react-avatar@1.2.7(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-6AizCpbU3Zngojp9vQak8Lt4AXwdKJ2lkRYAfpoETraT+qKVNeRqytiADzneHetFEB0Wu4+qeSfmZEq0PiVDmQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -774,19 +744,19 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-checkbox@1.3.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Gnptr9pDDQxD3hgq2dtPbtrp/c2qH1mBwIzw3X/ivrMb2e1t0jMTi606fVEqFPaQR1ggXIVQWKj3P2WW9v7zGQ==}
+  /@radix-ui/react-checkbox@1.3.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-gNOS2lLrpQxNLWHhCX0sayfVRiyB5gQp7nQIzpl+PM1Hkab52utsKtD27alAr78uqhpvX51kwqDp71hM5hNSwg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -799,74 +769,20 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-checkbox@1.3.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-wBbpv+NQftHDdG86Qc0pIyXk5IR3tM8Vd0nWLKDcX8nNn4nXFOFwsKuqw2okA/1D/mpaAkmuyndrPJTYDNZtFw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-collapsible@1.1.12(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Uu+mSh4agx2ib1uIGPP4/CKNULyajb3p92LsVXmH2EHVMTfZWpll88XJ0j4W0z3f8NK1eYl1+Mf/szHPmcHzyA==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-collapsible@1.1.20(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-mcGesGplBnzN2sbvJETzpCNfSMyPnb29q1GRLU+Ib7bJrpIG2ywmRoh2V5VbA2uNvKikKUlVbAPks7JDjz4A8Q==}
+  /@radix-ui/react-collapsible@1.1.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-IzkusZHCZBnED5skW+b+syl8d+e3+zelvxZ2Ed9lfawqEBfa5osPj53sTJ+q/uNmGAj9e7wMYgyzwPu/luWJ4Q==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -879,21 +795,21 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-collection@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-9W+B9NPF0NaaPh/1NJd3+KqsnlLqU9H7T2rvww+fp+T/evVXdNAyYcnfRQZFOjkR1ajQp3yORlqnI8soawLvNA==}
+  /@radix-ui/react-collection@1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-yEQrjfI5eBbeYEJXOASpm+gs9eNHY8+UjkZ81GzcuwdLrSMOlEM8jnGoGdeGo92oiBMkuwZUhHTsnsrDiJLcyA==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -905,53 +821,17 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-collection@1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Fh9rGN0MoI4ZFUNyfFVNU4y9LUz93u9/0K+yLgA2bwRojxM8JU1DyvvMBabnZPBgMWREAJvU2jjVzq+LrFUglw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-compose-refs@1.1.2(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-z4eqJvfiNnFMHIIvXP3CY57y2WJs5g2v3X0zm9mEJkrkNv4rDxu+sg9Jh8EkXyeqBkB7SOcboo9dMVqhyrACIg==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-compose-refs@1.1.5(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-compose-refs@1.1.5(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-+48PbAAbq3didjJxa+OaWY2ZwgAKsNiRGyeHKszblZMQ+kcpd9pAaT11cMkGEie0vsOi3QdeTE6d5Fe3Gn61kA==}
     peerDependencies:
       '@types/react': '*'
@@ -960,37 +840,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-context-menu@2.2.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-O8morBEW+HsVG28gYDZPTrT9UUovQUlJue5YO836tiTJhuIWBm/zQHc7j388sHWtdH/xUZurK9olD2+pcqx5ww==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-context-menu@2.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-CtXP35dxaB5T3zXSd+E3uHe/QpXcpYnZmxp6OaIbfthtfW4wyb77M23BG+bwIJDtsMwEP/YssdsmNyZu7jhWew==}
+  /@radix-ui/react-context-menu@2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-s8ABBkEuAOyhDub0Y99f8lVRlVsd28mUnOHtuqGg8LuHnefkmPN5rYzA2g8MAPLjAZMisnDHhN4dBHUM7ZF7Rg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1003,43 +858,17 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-menu': 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-context@1.1.2(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-jCi/QKUM2r1Ju5a3J64TH2A5SpKAgh0LpknyqdQ4m6DCV0xJ2HG1xARRwNGPQfi1SLdLWZ1OJz6F4OMBBNiGJA==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-context@1.1.3(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-ieIFACdMpYfMEjF0rEf5KLvfVyIkOz6PDGyNnP+u+4xQ6jny3VCgA4OgXOwNx2aUkxn8zx9fiVcM8CfFYv9Lxw==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-context@1.2.2(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-context@1.2.2(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-RHCUGwKHDr0hDGg4X7ma4JG4/+12qxw8rkh5QKdDldlCvtja6nUx1Ef/8HVrJze81lEsgLQlqjzjGNHantgnQA==}
     peerDependencies:
       '@types/react': '*'
@@ -1048,12 +877,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-dialog@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-TCglVRtzlffRNxRMEyR36DGBLJpeusFcgMVD9PZEzAKnUs1lKCgX5u9BmC2Yg+LL9MgZDugFFs1Vl+Jp4t/PGw==}
+  /@radix-ui/react-dialog@1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-l5CX8HdNL+9cc4s7HkNyplumSA3FIx68sg9dttBHKr286fRBwq1MBniDVTeI/tkof92/APntMQvC1YG1Khb0Cg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1065,28 +894,42 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/primitive': 1.1.7
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       aria-hidden: 1.2.6
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
+      react-remove-scroll: 2.7.2(@types/react@19.2.18)(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-dialog@1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Ksw4WeROkO4rC9k/onilX/Ao2Cr1ku1unMNH+XSCcP4jSXYu7HDsg9n4ojMjVb22XpYjAQ9qfrFlVbru1vXDUA==}
+  /@radix-ui/react-direction@1.1.5(@types/react@19.2.18)(react@19.2.1):
+    resolution: {integrity: sha512-/NlO/rOqQabmg+GGr86vCmbJVhqTv85NfM9g6Gknf/JVygXeZ6e6HM/RW1omerW9HsFtecwNqgEAKV6UqDsvpA==}
+    peerDependencies:
+      '@types/react': '*'
+      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
+    peerDependenciesMeta:
+      '@types/react':
+        optional: true
+    dependencies:
+      '@types/react': 19.2.18
+      react: 19.2.1
+    dev: false
+
+  /@radix-ui/react-dismissable-layer@1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-35WHRrycfnj2m4LvPiKMKHGLM3Y2WvUA744CkRdxjs1BGDX1c8PVMfmiSGTUwnHRbqdtIoxveYTXvcJSt7VfgQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1099,78 +942,18 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      aria-hidden: 1.2.6
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-direction@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-1UEWRX6jnOA2y4H5WczZ44gOOjTEmlqv1uNW4GAJEO5+bauCBhv8snY65Iw5/VOS/ghKN9gr2KjnLKxrsvoMVw==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-direction@1.1.4(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-5pzg4FGQNpExhnhT2zlrP1wZFaYCd1K0nYWoFAdcYoYK868IEigqMX3B3f8yIoRlAhAeDWciLI6ZdCKHF9P4Vg==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-dismissable-layer@1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Nqcp+t5cTB8BinFkZgXiMJniQH0PsUt2k51FUhbdfeKvc4ACcG2uQniY/8+h1Yv6Kza4Q7lD7PQV0z0oicE0Mg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-escape-keydown': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-dismissable-layer@1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-8g4pfOL9HoKKLWGiypT+dphVqjFfmcXO5GBnhsG6zI+lxAx/8feQpr+1LSN8Re3hiZ+XkLNS4O9ztK11/LzQ6w==}
+  /@radix-ui/react-dropdown-menu@2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-45QO76jn/a+cRl6OPgZKCKnIMdfYZaOa6SuAJeKHFiM3orX6bnscuFQJDx3qS4Agp/RE+eKv+9gmUL8t3Ma7bw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1183,82 +966,19 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-menu': 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-dropdown-menu@2.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-1PLGQEynI/3OX/ftV54COn+3Sud/Mn8vALg2rWnBLnRaGtJDduNW/22XjlGgPdpcIbiQxjKtb7BkcjP00nqfJw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-dropdown-menu@2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-geq8l2rJkxvkXsT9RMgtUE3P8pITFpTsvYpbySi1IH4fZEABD/Gp85myayFgxk0ktljGMJnCbeFkyTusvSvv7g==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-focus-guards@1.1.3(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-0rFg/Rj2Q62NCm62jZw0QX7a3sz6QCQU0LpZdNrJX8byRGaGVTqbrW9jAoIAHyMQqsNpeZ81YgSizOt5WXq0Pw==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-focus-guards@1.1.6(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-focus-guards@1.1.6(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-RNOJjfZMTyBM6xYmV3IVGXkPjIhcBAuv48POevAXwrGJhkWZ9p1rFoIS1JFooPuT193AZmRsCPhpoVJxx6OPoQ==}
     peerDependencies:
       '@types/react': '*'
@@ -1267,12 +987,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-focus-scope@1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-wmRZ2WWLvmt6KHy2rNPOdPUjwq5xOHY02+m+udwJTn0aNIox/rkskAvJTyTLGhPK6KgrUjlJUJpgmx/+wFiFIQ==}
+  /@radix-ui/react-focus-scope@1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-I5S7vsAOozx2lWSBAyuXjWIJ/A8qx4fjbv6115WoaRYQSb1yZDaeNHonUAQzXyE8oD+0whJHbQRl4HIE2gQ0Xw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1284,39 +1004,17 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-focus-scope@1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-t2ODlkXBQyn7jkl6TNaw/MtVEVvIGelJDCG41Okq/KwUsJBwQ4XVZsHAVUkK4mBv3ewiAS3PGuUWuY2BoK4ZUw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-form@0.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Q4TLEn2A7TAypxwmd6R9EwrlXDvkfYSDMrq9/887AXAGh+G1rH+kYJKSTv+Si9Y0JPKTwKYv6PviAJosysNimA==}
+  /@radix-ui/react-form@0.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-5ELeC+bQUHXEjU51svw11DynjGFlaUjo5QOUCr5hptvLLtBiqWtva2O3IWdjrFWQwitw6k6KpHHEyI61/kCVeg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1329,47 +1027,19 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-label': 2.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-label': 2.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-hover-card@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-qgTkjNT1CfKMoP0rcasmlH2r1DAiYicWsDsufxl940sT2wHNEWWv6FMWIQXWhVdmC1d/HYfbhQx60KYyAtKxjg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-popper': 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-hover-card@1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-H8qONfZd3ltrU3+jHCIgITbWo6e1iTKvP9DHdrvYbX48ooRM5FjEDTn16AMwdfuOGkWdZEhpl3PLL/Wk/AnHDQ==}
+  /@radix-ui/react-hover-card@1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-0w/tZHRo1fv4WsfcyaOdi5eDftFA8wQkr8WCOOuxw7fx9S2ERfshQioM269aFZsmyAHd0X1LWoZ2BQTlA/JYUw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1382,35 +1052,22 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-id@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-kGkGegYIdQsOb4XjsfM97rXsiHaBwco+hFI66oO4s9LU+PLAC5oJ7khdOVFxkhsmlbpUqDAvXw11CluXP+jkHg==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-id@1.1.4(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-id@1.1.4(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-TMQp2llA+RYn7JcjnrMnz7wN4pcVttPZnRZo52PLQsoLVKzNlVwUeHmfePgTgRluXFvlD3GD5g5MOVVTJCO0qA==}
     peerDependencies:
       '@types/react': '*'
@@ -1419,13 +1076,13 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-label@2.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-o/rdYEwZTTo5tjknnPeyQFU45kUC4i/XyeDPP+HGyi6XqpOP6Zf5Ya5vh/Yfe9Id5JiuWnnAx2XqIeD3UYZt0g==}
+  /@radix-ui/react-label@2.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-2vTQj72YFmeoTxWKJEMbO/9uQt6zzLgbzi0+nWBffq49oo6KhIDP/D+fA2im0wsxBCAX7QfaF3xgsYBVdDbNFQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1437,15 +1094,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-label@2.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-FmXs37I6hSBVDlO4y764TNz1rLgKwjJMQ0EGte6F3Cb3f4bIuHB/iLa/8I9VKkmOy+gNHq8rql3j686ACVV21A==}
+  /@radix-ui/react-menu@2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-e8JaQc0ijBbTx37YVI09Nfq8ThUWJ4L1rvZ3Hvaufehco9/ug9+U3PHz9ue6w3gKd91UcKf9fgDWejeSO+VSKQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1457,52 +1114,32 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.4(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-menu@2.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-72F2T+PLlphrqLcAotYPp0uJMr5SjP5SL01wfEspJbru5Zs5vQaSHb4VB3ZMJPimgHHCHG7gMOeOB9H3Hdmtxg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-collection': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/primitive': 1.1.7
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       aria-hidden: 1.2.6
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
+      react-remove-scroll: 2.7.2(@types/react@19.2.18)(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-menu@2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-uW7RVuU6Lp/ZtfeY4b3kL32zccgEWvPv1+cf17ubYzHa9cL8AHokmk36cG/XEiH/smbQvumnieXX9j/e9RqJWA==}
+  /@radix-ui/react-menubar@1.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-JIQiKcXsgl6BtwcMe7BMLvCGH2p9Czgcmxl6Ie9Zoi+MW1mmitkcm5We01m/ndLYVI8PpRJnM/RcLV0eOCxaJQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1515,31 +1152,23 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      aria-hidden: 1.2.6
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-menu': 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-menubar@1.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-eeVs0vf7cuqXaM0qLQCPcufImiJNVBXdJDLu7ZGYl2732UH23Qat/foNGrr6vYV3/DdTsBqASoggUFgH14OcZA==}
+  /@radix-ui/react-navigation-menu@1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-hZqTqDwc8Z5aI4rGIyPqCSJ2JoSNorqX/F1/gqNHkZzWJO6JuYzL7hn5sWOZJjpkfT+cbxWWRbKVU/IeiOM9ag==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1552,56 +1181,28 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-navigation-menu@1.2.22(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-ou7iLEJ+yrhQndkkA4U21XIdS/CS45F4iXIkTZcb6/Ne9EMsOuDudVmCwmDnfFZZ+y1FZqXRNSIgBy+YMvZVZg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-one-time-password-field@0.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Tj9P6ntAJEw52oq/F0AGknXR4XncxEt7XU47O3xJQOiWfLzEy3d9gtgKfvjSzGxzHkfL+VzvxGu2KTFsloJqXw==}
+  /@radix-ui/react-one-time-password-field@0.1.17(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-+6jmOqjSNBr7F1uSc7HSDqD3XL54A2qA99I9rIH4JoOD/jDScw8tUmDz6HCAkjytbrE/cdL66i+QkeDyZ0WNyw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1615,24 +1216,24 @@ packages:
     dependencies:
       '@radix-ui/number': 1.1.3
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-password-toggle-field@0.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-4gvFnmDXu3dgj21CqsufzIameRvlRd4SBqaWhcrlrNhRo0Y5i/49AmRJYe1fdAM3G2VNBbmin4b0D6cdQocwgw==}
+  /@radix-ui/react-password-toggle-field@0.1.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-+SWSgf3HRpE4zUira+p6t5KUR473PSqewtQgj9qzl3znZPRWPCGycwXNTfHwkiRTD3tEXwNIXJg0i1NTteiULw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1645,55 +1246,21 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-popover@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-kr0X2+6Yy/vJzLYJUPCZEc8SfQcf+1COFoAqauJm74umQhta9M7lNJHP7QQS3vkvcGLQUbWpMzwrXYwrYztHKA==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      aria-hidden: 1.2.6
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-popover@1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-mw58MrBlyHWFisTOYignD0vf/3gdcgAR+9of1s9G/38CbFiUwH1nCDkc0AUM9IrXFgN5Ue8n45j9WCgyM1sbiQ==}
+  /@radix-ui/react-popover@1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-e70Nh+0iQ+hRAIM2nAGib5BhaDL2Rjlzxw9kIzsU2nz+t3/t82Z5qtPhY5nTnfj8FbmrU6RRIKH8RF3ynCe+vw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1706,28 +1273,29 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       aria-hidden: 1.2.6
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
+      react-remove-scroll: 2.7.2(@types/react@19.2.18)(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-popper@1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-0NJQ4LFFUuWkE7Oxf0htBKS6zLkkjBH+hM1uk7Ng705ReR8m/uelduy1DBo0PyBXPKVnBA6YBlU94MBGXrSBCw==}
+  /@radix-ui/react-popper@1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-y92/9iIdG1FyNvow0Y0qP5WJZ7oRTAKDVKCy9kJnhZfItwIKYtt0KKQeOhBKfXz0ETizduKwDRbrQHrTQScs5A==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1739,53 +1307,24 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@floating-ui/react-dom': 2.1.7(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-arrow': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-rect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/rect': 1.1.1
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-popper@1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-UsJrrd7w4wuKKTdvd/DNERVlwSlUcyXzjhyDwBk+3aPOsCjOY6ZSbxuw8E6lZTjjfP8Cpd0J8VVkrYUWyGYXyg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@floating-ui/react-dom': 2.1.7(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-arrow': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-rect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
+      '@floating-ui/react-dom': 2.1.9(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-arrow': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-rect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
       '@radix-ui/rect': 1.1.3
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-portal@1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-vKQLcWypUnwZVvfV7UkGahH2g6ySe8M8R+zYBwPrv5byZ9QAW6cQVvNKo7GgmD+p8aYb6D9JBuvy8/WhOno2wQ==}
+  /@radix-ui/react-portal@1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-ygFIJ+QXNKf72AAq8x3hd22+ws7pYpmc1S0E4kW0Je+qD4qEb3FvxeKVQRGrVDMOkQaYfsB+Qs+KDh81a1gAOw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1797,16 +1336,16 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-portal@1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-bpIxvq03if6UNwXZ+HTK71JLh4APvnXntDc6XOX8UVq4XQOVl7lwok0AvIl+b8zgCw3fSaVTZMpAPPagXbKmHQ==}
+  /@radix-ui/react-presence@1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-Z7/atlkab36Z8hm+V8zYc79r0W+tiOB1o0tun1By94JyadkekLQmtEAnBSP9XKXBNQsStoStS4+gX10JpS5akA==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1818,16 +1357,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-presence@1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-3wyzCQ6+ubRA+D4uv9m95JYLXxmOHp05qjrkjeA7uKHHtjpPggQzc6DAb0URl7j67oR0K2foO4ip27TiX037Bw==}
+  /@radix-ui/react-primitive@2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-RzcaSw5aKl0OPhBJ1wbolKSF6nVvN+0o/QLeIVq+09npiPdnCkPggy2k6gI6yetxl4ZyrlBPYFgCB8BQJZwXcQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1839,15 +1377,15 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-presence@1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-/jfEwNDdQVBCNvjkGit4h6pMOzq8bHkopq458dPt2lMjx+eBQUohZNG9A7DtO/O5ukSbxuaNGXMjHicgwy6rQQ==}
+  /@radix-ui/react-progress@1.1.17(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-A+RfMARd9u+UaVbLQDmysGlAGuaFQrF+uxF3v+7+xU8tkKoYxtCSs+6HHnIKJVgam2sGZ7c0hLRPWSnmnFyGJQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1859,118 +1397,16 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-primitive@2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-MucOnzh6hR5mid6VpkbglRAMYMjKLqRnGBbjXkzjK52fuQDd1qbkx78a5P40mkcnVXJdEVxm26E9OPAiUq7nBg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-primitive@2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-m9gTwRkhy2lvCPe6QJp4d3G1TYEUHn/FzJUtq9MjH46an1wJU+GdoGC5VLof8RX8Ft/DlpshApkhswDLZzHIcQ==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-primitive@2.1.4(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-9hQc4+GNVtJAIEPEqlYqW5RiYdrr8ea5XQ0ZOnD6fgru+83kqT15mq2OCcbe8KnjRZl5vF3ks69AKz3kh1jrhg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-slot': 1.2.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-progress@1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-5XnomAsoZZCY+KNTxbIghpGqPruZvKFNlvcAljVAOdDRDsH4/OZQxhtwo5wdtoDM5R6MhJBb2sPnDuRFep3lzg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-progress@1.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-+gISHcSPUJ7ktBy9RnTqbdKW78bcGke3t6taawyZ71pio1JewwGSJizycs7rLhGTvMJYCQB1DBK4KQsxs7U8dA==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-context': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.4(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-radio-group@1.4.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-cgYFEkntCxppHZgtSZ+7vh0wbZQ+IC7PPMw8DSnRG27B6kDd32/Zw0OJt7dGDigCoprMuWHjg2PvUn3PYvPFoQ==}
+  /@radix-ui/react-radio-group@1.4.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-MaOOZ1TlaYoOeSDnHeVIONss0rp5ngOTY5HJY+Xkq16PAopnj21iU9r9m0O/77cBYu8Ksaq3BcjlaOdpbfbzzg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -1983,50 +1419,22 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-roving-focus@1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-7A6S9jSgm/S+7MdtNDSb+IU859vQqJ/QAtcYQcfFC6W8RS4IxIZDldLR0xqCFZ6DCyrQLjLPsxtTNch5jVA4lA==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-collection': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-roving-focus@1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-V9jI6hDjT7l3jsCQD9bLNvDLM3tH/gdbOTp7Tefp3hbbgCGQoK7tUvrWiRlcoBHIZ809ElXwNQwVo0B98LuTXQ==}
+  /@radix-ui/react-roving-focus@1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-+ZjbkU21fTSJdn4nq/IFKUdKS+JSZHORBZ8ma1f6YpOMNzRt0xf3sVAWVK/+ktHCajPy4HYDsyHXZIZmOMxzIA==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2039,52 +1447,24 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-scroll-area@1.2.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-tAXIa1g3sM5CGpVT0uIbUx/U3Gs5N8T52IICuCtObaos1S8fzsrPXG5WObkQN3S6NVl6wKgPhAIiBGbWnvc97A==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/number': 1.1.1
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-scroll-area@1.2.18(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Zn5Cd171wxsO3Dfg8HaW6RifTb9CYTKQJHs/G4+LN1GfmJpaQMZQyQxMprVPHpaz7QY4l9BxK2JwQuzHsXC8nA==}
+  /@radix-ui/react-scroll-area@1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-1WSuAXaahfeLmFX/dfYo/ftg1filMfaWjVye2d40VlgtWMMq/NOVocfukZEXkG+dwrAoNI+qpo9bjhNSJ/XC8g==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2098,21 +1478,22 @@ packages:
     dependencies:
       '@radix-ui/number': 1.1.3
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-select@2.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-I30RydO+bnn2PQztvo25tswPH+wFBjehVGtmagkU78yMdwTwVf12wnAOF+AeP8S2N8xD+5UPbGhkUfPyvT+mwQ==}
+  /@radix-ui/react-select@2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-Tz9xOpV9Rbca5WTk7pzmfAECV3zb++uScT9mP3Px7ziAhH7SlK53b0uwMw8VubPhRK4x7uDKCy5fwevT8I0fAg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2124,35 +1505,56 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/number': 1.1.1
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-collection': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/number': 1.1.3
+      '@radix-ui/primitive': 1.1.7
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       aria-hidden: 1.2.6
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
+      react-remove-scroll: 2.7.2(@types/react@19.2.18)(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-select@2.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-WFGImkmbzcfxeIwq/+4HvRN0pizBwbwQUED4I13ezQsDdfl38ZntN6TmR8XaSzPBqoCToe8rF75j6NPNDSzhbg==}
+  /@radix-ui/react-separator@1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-vugPjGQBXbJVVkftKqv/BVeZ108SyRGT3FR1oa14az6qyLgX8jPhnZYHgdSB7Gw8UJusqwb9eaK6mluOnsn/Vg==}
+    peerDependencies:
+      '@types/react': '*'
+      '@types/react-dom': '*'
+      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
+      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
+    peerDependenciesMeta:
+      '@types/react':
+        optional: true
+      '@types/react-dom':
+        optional: true
+    dependencies:
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
+      react: 19.2.1
+      react-dom: 19.2.1(react@19.2.1)
+    dev: false
+
+  /@radix-ui/react-slider@1.5.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-ZOBLK7KZvGL02V+4VwVbj8NYM4vPrAtyZdrgHSv+AMS9VL+5yAdnQF3DZT4Y8cbCp1Yn4xdp92orvez8fJa77w==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2166,104 +1568,23 @@ packages:
     dependencies:
       '@radix-ui/number': 1.1.3
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      aria-hidden: 1.2.6
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-      react-remove-scroll: 2.7.2(@types/react@19.2.10)(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-separator@1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-jOLO4lssEzWpoDu7G+Ze4VjwMRUBt291pnZD0gmalREZipnTX3wadQo7Fy48GCTfe14/YRN6rw/rOJqrE85Wxw==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-separator@1.1.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-sDvqVY4itsKwwSMEe0jtKgfTh+72Sy3gPmQpjqcQneqQ4PFmr/1I0YA+2/puilhggCe2gJcx5EBAYFkWkdpa5g==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-primitive': 2.1.4(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-slider@1.4.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-mTSLf1GC/C0moWjTbvCM6Qn/gBjvlFt1azuWF2v7MN5C3Zq2U2J2lN3ZEYkpujuOU5Ro7A28wkviSxaKnG0BYg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/number': 1.1.3
-      '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-slot@1.2.3(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-aeNmHnBxbi2St0au6VBVC7JXFlhLlOnvIIlePNniyUNAClzmtAUEY8/pBiK3iHjufOlwA+c20/8jngo7xcrg8A==}
+  /@radix-ui/react-slot@1.4.0(@types/react@19.2.18)(react@19.2.1):
+    resolution: {integrity: sha512-IuMVHZ+Ah5fkctu3X3qeFrV/ZrBcuWuDTFUUPfS/L5hvoTjzhp/HdWGE14CsQombPdEidsa33HVlYZXi49PZ1Q==}
     peerDependencies:
       '@types/react': '*'
       react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
@@ -2271,67 +1592,13 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-slot@1.2.4(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-Jl+bCv8HxKnlTLVrcDE8zTMJ09R9/ukw4qBs/oZClOfoQk/cOTbDn+NceXfV7j09YPVQUryJPHurafcSg6EVKA==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-slot@1.3.3(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-qx7oqnYbxnK9kYI9m317qmFmEgo6ywqWvbTogdj7cL9p3/yx4M48p7Rnw5z3H890cL/ow/EeWJsuTykeZVXP5Q==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-switch@1.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-bByzr1+ep1zk4VubeEVViV592vu2lHE2BZY5OnzehZqOOgogN80+mNtCqPkhn2gklJqOpxWgPoYTSnhBCqpOXQ==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-previous': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-switch@1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-48tB/4dn2UVLBCYhTu9AuR63IHl73l/qLbLgxd86noTUor4/K4LFDAcYjK+isP5313qxaFpjPVogE7+Y0/V3Kw==}
+  /@radix-ui/react-switch@1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-rr+zuoAeHWK3m6H4HgjOUiu3VsrQ/9JkjwY/uMnTgHJGPe9WNzDP/MOuIV7S+65p42DIxjGL0hAHktqBq82+2A==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2344,46 +1611,19 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-tabs@1.1.13(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-7xdcatg7/U+7+Udyoj2zodtI9H/IIopqo+YOIcZOq1nJwXWBZ9p8xiu5llXlekDbZkca79a/fozEYQXIA4sW6A==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-tabs@1.1.21(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-UKxJlZid7FVtsk/WTxj4i4uSEgj2Au+KBbS7SQyTlzMhhn+86Cz3tISZdTa87bfEfcuvZezf2ZsxD4xuEKtkog==}
+  /@radix-ui/react-tabs@1.1.22(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-GmVH+P3wx1ZPMuH7WajQzygZIbqgnsb5h8z4Ow3sQMuDqgcqISSBNom6MLVNJvfrkuJTtqfXc3ZolHjX+w0CHg==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2396,21 +1636,21 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-toast@1.2.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-ofhyAsYaocRGOs/n0XWdUOSVzEAG6BfrMVM8z0c0kLEWY38w/0WuMFPTJP/HVaZPYkMvHZoKIIhNcjbTCBILPg==}
+  /@radix-ui/react-toast@1.2.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-XR9pDEz9CH0ez98zPXQZOpF46anvbarkoOpjXpPSX7gm21vEl5Tzoe1PSQd3EY1OTOHac4lNq+TPx/2Z7jObdQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2423,51 +1663,25 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-toggle-group@1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-5umnS0T8JQzQT6HbPyO7Hh9dgd82NmS36DQr+X/YJ9ctFNCiiQd6IJAYYZ33LUwm8M+taCz5t2ui29fHZc4Y6Q==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toggle': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-toggle-group@1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-OtnwuSVjd1Ofi+AdnvhsjQdyuhCDwYs1w9RyB5BN/OavXOVQo42SYqQjwUnbPnaiPFBpQ9aX70dWeee+v2oBLA==}
+  /@radix-ui/react-toggle-group@1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-O+Sjiv64DfnByHYtnKEXmn63uUskU/X61vjf69X1v8JFn/ZvuNNLdJTBfYurVTSDv4/RJnJGtGgb/uvuS/rRJw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2480,42 +1694,20 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toggle': 1.1.18(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toggle': 1.1.19(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-toggle@1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-lS1odchhFTeZv3xwHH31YPObmJn8gOg7Lq12inrr0+BH/l3Tsq32VfjqH1oh80ARM3mlkfMic15n0kg4sD1poQ==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-toggle@1.1.18(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-7lonPlKfSacd20GlOBx2ltuVKz9oqWYZz+oMQyOltw6t1y2nyftj2ZmwwUHYn49kqfDWcp8dNZm5NgV+5Z+mug==}
+  /@radix-ui/react-toggle@1.1.19(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-l7FQHtuZjUJ0vKLdhKu5LzZ8Jw6K60QEC1tC22QNLuXN2PkydSoqEhKTCqwxBFZJ52NTCjzudDvl9sUsZeUbPw==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2528,16 +1720,16 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-toolbar@1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-Ph0IvtYw4VB12ZnZg+YtrGs8yJQsnizwo/zu0R4Y/nWugtJzA7Pg1eWeuDR9+LSqn+xjamss+UOSOJJJ4gx8jw==}
+  /@radix-ui/react-toolbar@1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-5gyzL6B5/zjBwU3ATBWAVZ4xXQC5F8E9CB4BmYC1UvZ08pN2OJQoooeIwJtZZ/2rz4sTiQmbGfdDvmU73wRJCQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2550,20 +1742,20 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-separator': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toggle-group': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-separator': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toggle-group': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-tooltip@1.2.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-6EamKFRRnlpdadndbZ6LMwycfwkwPte1B42hs6QA0gYhjaOKqW4PZ4pjaW9UrlDX5eVt/OjncE7BFTPL5nmZhg==}
+  /@radix-ui/react-tooltip@1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-4HM4b1Cft1CBXCGDRDG3S/fh8GedBuVOMNh145wMIfTbVqcuv+tyj03nZqYgFevWJMYo58H0s+8P6s6NaUHWtA==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2576,69 +1768,25 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-tooltip@1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-tY7sVt1yL9ozIxvmbtN5qtmH2krXcBCfjEiCgKGLqunJHvgvZG2Pcl2oQ3kbcZARb1BGEHdkLzcYGO8ynVlieg==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/primitive': 1.1.3
-      '@radix-ui/react-compose-refs': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.1.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-popper': 1.2.8(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.9(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.5(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.2.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-use-callback-ref@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-FkBMwD+qbGQeMu1cOHnuGB6x4yzPjho8ap5WtbEJ26umhgqVXbhekKUQO+hZEL1vU92a3wHwdp0HAcqAUF5iDg==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-callback-ref@1.1.4(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-callback-ref@1.1.4(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-R6OUY2e2fA6Yn6s+VSx5KBV6Nx8LQEhu+cz7LCej18rQ1HLyg9PSC9jP/ZNx0o6FAIK9c0F1kHylzSxKsdlkrQ==}
     peerDependencies:
       '@types/react': '*'
@@ -2647,26 +1795,11 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-controllable-state@1.2.2(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-BjasUjixPFdS+NKkypcyyN5Pmg83Olst0+c6vGov0diwTEo6mgdqVR6hxcEgFuh4QrAs7Rc+9KuGJ9TVCj0Zzg==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-use-effect-event': 0.0.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-controllable-state@1.2.6(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-controllable-state@1.2.6(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-uEQJGT97ZA/TgP/Hydw47lHu+/vQj6z/0jA+WeTbK1o9Rx45GImjpD0tc3W5ad3D6XTSR6e1yEO0FvGq6WQfVQ==}
     peerDependencies:
       '@types/react': '*'
@@ -2676,27 +1809,13 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-effect-event@0.0.2(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-Qp8WbZOBe+blgpuUT+lw2xheLP8q0oatc9UpmiemEICxGvFLYmHm9QowVZGHtJlGbS6A6yJ3iViad/2cVjnOiA==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-effect-event@0.0.5(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-effect-event@0.0.5(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-7cshFL8HGS/7HEiHH+9kL9HBwp2sa9yX18Knwek6KYWmXwM7pegMgta2AXMQKI+rq3JnfSj9x8wYqFMTdG1Jgg==}
     peerDependencies:
       '@types/react': '*'
@@ -2705,26 +1824,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-escape-keydown@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-Il0+boE7w/XebUHyBjroE+DbByORGR9KKmITzbR7MyQ4akpORYP/ZmbhAr0DG7RmmBqoOnZdy2QlvajJ2QA59g==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-use-callback-ref': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-escape-keydown@1.1.5(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-escape-keydown@1.1.5(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-ge3ipobwSXTj4JyVtswQ7qZj0ZHdtbGuOno/LrgAAeSxtsJ6Vs4Gz5IkPH2bmqpjcLUFoqGhA/mueuIf63UXlA==}
     peerDependencies:
       '@types/react': '*'
@@ -2733,26 +1838,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-is-hydrated@0.1.0(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-U+UORVEq+cTnRIaostJv9AGdV3G6Y+zbVd+12e18jQ5A3c0xL03IhnHuiU4UV69wolOQp5GfR58NW/EgdQhwOA==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-      use-sync-external-store: 1.6.0(react@19.2.1)
-    dev: false
-
-  /@radix-ui/react-use-is-hydrated@0.1.3(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-is-hydrated@0.1.3(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-umO/aJ+82CpOnhDZUTbILCQf7kU/g0iv+oGs/Q8jw7IkhWBzaEP4sA268PhFAJTFetbwp3ICc6ktpI4TqtxcIw==}
     peerDependencies:
       '@types/react': '*'
@@ -2761,24 +1852,11 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-layout-effect@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-RbJRS4UWQFkzHTTwVymMTUv8EqYhOp8dOOviLj2ugtTiXRaRQS7GLGxZTLL1jWhMeoSCf5zmcZkqTl9IiYfXcQ==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-layout-effect@1.1.4(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-layout-effect@1.1.4(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-K20DkRkUwDnxEYMBPcg3Y6voLkEy5p5QQmszZgLngKKiC7dzBR/aEuK3w1qlx2JWDUNH6FluahYdgR3BP+QbYw==}
     peerDependencies:
       '@types/react': '*'
@@ -2787,24 +1865,11 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-previous@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-2dHfToCj/pzca2Ck724OZ5L0EVrr3eHRNsG/b3xQJLA2hZpVCS99bLAX+hm1IHXDEnzU6by5z/5MIY794/a8NQ==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-previous@1.1.4(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-previous@1.1.4(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-XoSLhbRbqxFtgJoi2fNHA3C6pDlY34x508vUpUGoFZfvePfHXHbE1lC4FYFMnJWgiCRroSTw6fOsXQoVS9RwZg==}
     peerDependencies:
       '@types/react': '*'
@@ -2813,25 +1878,11 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-rect@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-QTYuDesS0VtuHNNvMh+CjlKJ4LJickCMUAqjlE3+j8w+RlRpwyX3apEQKGFzbZGdo7XNG1tXa+bQqIE7HIXT2w==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/rect': 1.1.1
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-use-rect@1.1.4(@types/react@19.2.10)(react@19.2.1):
+  /@radix-ui/react-use-rect@1.1.4(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-cSOCh6JlkmfjLyNcLiu2nB4v+nm+dkZ+Q5KHWk/soo4U7ZLiEQFKHK9/YmtBHjfCEaU43IBKQOc4/uJmCaiCTQ==}
     peerDependencies:
       '@types/react': '*'
@@ -2841,12 +1892,12 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/rect': 1.1.3
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-size@1.1.1(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-ewrXRDTAqAXlkl6t/fkXWNAhFX9I+CkKlw6zjEwk86RSPKwZr3xpBRso655aqYafwtnbpHLj6toFzmd6xdVptQ==}
+  /@radix-ui/react-use-size@1.1.5(@types/react@19.2.18)(react@19.2.1):
+    resolution: {integrity: sha512-V1McDQkMGzgivwqdu7u96APs7oi0Jvj9LTzgq3WWrDZqxWdAqU9yZnZQqk5OizGFR8badUQwZsrlao7MOlWYOA==}
     peerDependencies:
       '@types/react': '*'
       react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
@@ -2854,27 +1905,13 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /@radix-ui/react-use-size@1.1.4(@types/react@19.2.10)(react@19.2.1):
-    resolution: {integrity: sha512-D3anSY15EJoxrihpsXI6SMrmmonnQtR2ni7arO+Lfdg3O95b9hNXxONk8jA5C8ANdF/h5HMAxejgs8PWJ6rlhw==}
-    peerDependencies:
-      '@types/react': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-    dependencies:
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@types/react': 19.2.10
-      react: 19.2.1
-    dev: false
-
-  /@radix-ui/react-visually-hidden@1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-NFS86RYYZb4/exihaESBGOpMJFz8MGLAfu3mOBSGByVnVPC9JPASfYubxd/8KbkQK0sYAv8lVQDEQukDX/qXvQ==}
+  /@radix-ui/react-visually-hidden@1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-/iPLVHRBzKWvekN+5xaDmXw80f+nleYYceomVORrGjuJe0LF2SCbSH4f/LG4hu32q92QSuPZx4hmSACYBVBfzQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -2886,43 +1923,36 @@ packages:
       '@types/react-dom':
         optional: true
     dependencies:
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /@radix-ui/react-visually-hidden@1.2.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-pzJq12tEaaIhqjbzpCuv/OypJY/BPavOofm+dbab+MHLajy277+1lLm6JFcGgF5eskJ6mquGirhXY2GD/8u8Ug==}
-    peerDependencies:
-      '@types/react': '*'
-      '@types/react-dom': '*'
-      react: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-      react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc
-    peerDependenciesMeta:
-      '@types/react':
-        optional: true
-      '@types/react-dom':
-        optional: true
-    dependencies:
-      '@radix-ui/react-primitive': 2.1.3(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
-      react: 19.2.1
-      react-dom: 19.2.1(react@19.2.1)
-    dev: false
-
-  /@radix-ui/rect@1.1.1:
-    resolution: {integrity: sha512-HPwpGIzkl28mWyZqG52jiqDJ12waP11Pa1lGoiyUkIEuMLBP0oeK/C89esbXrxsky5we7dfd8U58nm0SgAWpVw==}
-    dev: false
-
   /@radix-ui/rect@1.1.3:
     resolution: {integrity: sha512-JtyZR+mqgBibTo8xea3B6ZRmzZiM/YeVBtUkas6zMuXjAlfIFIW2FgqeM9eLyvEaYX66vr6DJMK+4U6LV0KhNw==}
     dev: false
 
-  /@rolldown/binding-android-arm64@1.0.3:
-    resolution: {integrity: sha512-454rs7jHngixp/NMxd5srYD57OnzSlZ/eFTETjORQHLwJG1lRtmNOJcBerZlfu4GjKqeq8aCCIQrMdHyhI51Hw==}
+  /@remixicon/react@4.9.0(react@19.2.1):
+    resolution: {integrity: sha512-5/jLDD4DtKxH2B4QVXTobvV1C2uL8ab9D5yAYNtFt+w80O0Ys1xFOrspqROL3fjrZi+7ElFUWE37hBfaAl6U+Q==}
+    peerDependencies:
+      react: '>=18.2.0'
+    dependencies:
+      react: 19.2.1
+    dev: false
+
+  /@rolldown/binding-android-arm-eabi@1.2.13:
+    resolution: {integrity: sha512-Vu/NICrgdRvUOeaJZIyB2FP63s6fO/4ud5l8dXmxDx9yDyn+mN0y+gsvsVviSli+T361Aqo0h46EuMoobrwawA==}
+    engines: {node: ^20.19.0 || >=22.12.0}
+    cpu: [arm]
+    os: [android]
+    requiresBuild: true
+    dev: false
+    optional: true
+
+  /@rolldown/binding-android-arm64@1.2.13:
+    resolution: {integrity: sha512-6LEoIrGy5vvCd2boNnv2soPSUZj/PRvhp8R/t9KlX+mIZxThIE/ZqxrGXxmuNX0FuvQFvLdZr7S4HrYpYBVhEQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [android]
@@ -2930,8 +1960,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-darwin-arm64@1.0.3:
-    resolution: {integrity: sha512-PcAhP+ynjURNyy8SKGl5DQP94aGuB/7JrXJb/t7P+hanXvQVMWzUvRRhBAcg/lNRadBhoUPqSoP4xw5tR/KBEA==}
+  /@rolldown/binding-darwin-arm64@1.2.13:
+    resolution: {integrity: sha512-CaJSRzlQ6Q+zkDMVA9OkVxBjVoPRCDEE3v0FSJ9kIQBgRjFApA3Si4ijhbNVUji7ZJt+PfO0e46qy4nPInHUPA==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [darwin]
@@ -2939,8 +1969,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-darwin-x64@1.0.3:
-    resolution: {integrity: sha512-9YpfeUvSE2RS7wysJ81uOZkXJz7f7Q55H2Gvp3VEw/EsahqDtrphrZ0EwDLK5vvKOzaCrBsjF8JmnMLcUt78Gg==}
+  /@rolldown/binding-darwin-x64@1.2.13:
+    resolution: {integrity: sha512-e4r340I5hxSzJkD3YJBxgzq/rJ3wwxwdLLb2FPt1Zpyem20nrQPT11DDEEkazwxaHumH8wpglsp/Ozgj/wT6mw==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [x64]
     os: [darwin]
@@ -2948,8 +1978,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-freebsd-x64@1.0.3:
-    resolution: {integrity: sha512-yB1IlAsSNHncV6SCTL27/MVGR5htvQsoGxIv5KMGXALp+Ll1wYsn+x98M9MW7qa+NdSbvrrY7ANI4wLJ0n1e6g==}
+  /@rolldown/binding-freebsd-x64@1.2.13:
+    resolution: {integrity: sha512-hoBWDZ4/K/mJKVMbsgUd1jfpA2DiOrJ54mpegObU9TBHBBrtX4FLIXJJQRDPZ/2Dn91GvPob+Z2hwKJzf4E5SQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [x64]
     os: [freebsd]
@@ -2957,8 +1987,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-arm-gnueabihf@1.0.3:
-    resolution: {integrity: sha512-Yi30IVAAfLUCy2MseFjbB1jAMDl1VMCAas5StnYp8da9+CKvMd2H2cbEjWcw5NPaPqzvYkVIaF1nNUG+b7u/sw==}
+  /@rolldown/binding-linux-arm-gnueabihf@1.2.13:
+    resolution: {integrity: sha512-oF9fZu3vTWnTiVAkDFxU9PJLt1nh20K9RRpuHsR7ZnE2L7XookBbDBVLojCVSLGbvtSlxeUNnjAj6DbNsKyzbg==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm]
     os: [linux]
@@ -2966,68 +1996,62 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-arm64-gnu@1.0.3:
-    resolution: {integrity: sha512-jsO7R8To+AdlYgUmN5sHSCZbfhtMBkO0WUx8iORQnPcMMdgr7qM2DQmMwgabs3GhNztdmoKkMKQFHD6DTMCIQw==}
+  /@rolldown/binding-linux-arm64-gnu@1.2.13:
+    resolution: {integrity: sha512-DTpb/+jMiqWMPgphn4ksW1zW+XLJ6dJJ9FMvnIk9I291LhCctMPhMYJrP5t9p6C9qa3KXadDdUEZHBaPh3kNxQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-arm64-musl@1.0.3:
-    resolution: {integrity: sha512-VWkUHwWriDciit80wleYwKILoR/KMvxh/IdwS/paX+ZgpuRpCrKLUdadJbc0NpBEiyhpYawsJ73j9aCvOH+f7Q==}
+  /@rolldown/binding-linux-arm64-musl@1.2.13:
+    resolution: {integrity: sha512-xU+Kvv9N2A3SY35zFh1FWWNjxbadE/BKeqnkXe7vScU4GZsw757BlVmDLITZEq8DPpk6m7uoDYkK7P+x/uU5Ow==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-ppc64-gnu@1.0.3:
-    resolution: {integrity: sha512-5f1laC0SlIR0yDbFCd8acUhvJIag6N3zC5P7oUPN6wX0aOma+uKJ0wBDH5aq7I1PVI2ttTlhJwzwRIBnLiSGEg==}
+  /@rolldown/binding-linux-ppc64-gnu@1.2.13:
+    resolution: {integrity: sha512-XMk/hnhh25oQ756Lclb1Q11h4ll0xC0Esa5w4Vzp+mAE2S3AGJgeM0SonrIgY8fHhjarb6hcoTuwoFG0VVP2ig==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [ppc64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-s390x-gnu@1.0.3:
-    resolution: {integrity: sha512-Iq4ko0r4XsgbrF/LunNgHtAGLRRVE2kXonAXQ/MV0mC6jQpMOhW1SvtZja2EhC/kd05++bP78dsqBeIQyYJ6Yg==}
+  /@rolldown/binding-linux-s390x-gnu@1.2.13:
+    resolution: {integrity: sha512-U1BoTHYmlywE2n3HUWjzZ+8EQsKr+OwpxYuDWaubx42tUL8QSL7nOqRz/NVHanibbFP3M1Rxuprz/wNcbMKjqw==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [s390x]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-x64-gnu@1.0.3:
-    resolution: {integrity: sha512-B8m6tD5+/N5FeNQFbKlLA/2yVq9ycQP1SeedyEYYKWBNR3ZQbkvIUcNnDNM03lO1l5F2roiiFJGgvoLLyZXtSg==}
+  /@rolldown/binding-linux-x64-gnu@1.2.13:
+    resolution: {integrity: sha512-J4cm80Pc6wEwV5AZ42o62sIQgDk+Y7R2s646jIwuvO+4V46QB58N4vVuMPWCBaNgursXLpmUWJV//WQAkPvHBQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [x64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-linux-x64-musl@1.0.3:
-    resolution: {integrity: sha512-pSdpdUJHkuCxun9LE7jvgUB9qsRgaiyNNCX7m/AvHTcq67AiT/Yhoxvw5zPfhrM8k/BfP8ce/hMOpthKDpEUow==}
+  /@rolldown/binding-linux-x64-musl@1.2.13:
+    resolution: {integrity: sha512-W0euZS40U4V6JkmEnVCIX9mcsL6cbXQ4X8c2w+MMOIjFg68TXUW+u3RT+2e1t1pv9mg/EP3pDtK3POXdQ7Rxyw==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [x64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@rolldown/binding-openharmony-arm64@1.0.3:
-    resolution: {integrity: sha512-OXXS3RKJgX2uLwM+gYyuH5omcH8fL1LJs96pZGgtetVCahON57+d4SJHzTgZiOjxgGkSnpXpOsWuPDGAKAigEg==}
+  /@rolldown/binding-openharmony-arm64@1.2.13:
+    resolution: {integrity: sha512-do/AF/4n9MRshia3a5aXoy8xM2c6mgfT6D7H/dY6my5z+hfCIR4ab7mk4IRj++nsEQpKy39kd0DQnDq+r692tw==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [openharmony]
@@ -3035,20 +2059,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-wasm32-wasi@1.0.3:
-    resolution: {integrity: sha512-JTtb8BWFynicNSoPrehsCzBtOKjZ6jhMiPFEmOiuXg1Fl8dn2KHQob+GuPSGR0dryQa1PQJbzjF3dqO/whhjLg==}
-    engines: {node: ^20.19.0 || >=22.12.0}
-    cpu: [wasm32]
-    requiresBuild: true
-    dependencies:
-      '@emnapi/core': 1.10.0
-      '@emnapi/runtime': 1.10.0
-      '@napi-rs/wasm-runtime': 1.1.4(@emnapi/core@1.10.0)(@emnapi/runtime@1.10.0)
-    dev: false
-    optional: true
-
-  /@rolldown/binding-win32-arm64-msvc@1.0.3:
-    resolution: {integrity: sha512-gEdFFEN70A/jxb2svrWsN3aDL7OUtmvlOy+6fa2jxG8K0wQ1ZbdeLGnidov6Yu5/733dI5ySfzFlQ/cb0bSz1g==}
+  /@rolldown/binding-win32-arm64-msvc@1.2.13:
+    resolution: {integrity: sha512-6FtUf2cJZobutJdxnGYX8z0vtbgqU0b2v8vfBhkXb9wp2MppMFtkunE+IaOntyi01rkIQoPW2GHABtfzsC0BRQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [arm64]
     os: [win32]
@@ -3056,8 +2068,8 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/binding-win32-x64-msvc@1.0.3:
-    resolution: {integrity: sha512-eXB7CHuaQdqmJcc3koCNtNPmT/bj2gc999kUFgBxG8Ac0NdgXc4rkCHhqrgrhN3zddvvvrgzj1e90SuSfmyIXA==}
+  /@rolldown/binding-win32-x64-msvc@1.2.13:
+    resolution: {integrity: sha512-UKhFaZTeTcbKrb3iU82m17jVcPEAoqDkCN5MBqgPyZSwp/cQmib21crTS0DdH7T1B7wZRWEI89YQAOJMet5yhg==}
     engines: {node: ^20.19.0 || >=22.12.0}
     cpu: [x64]
     os: [win32]
@@ -3065,10 +2077,6 @@ packages:
     dev: false
     optional: true
 
-  /@rolldown/pluginutils@1.0.0-rc.7:
-    resolution: {integrity: sha512-qujRfC8sFVInYSPPMLQByRh7zhwkGFS4+tyMQ83srV1qrxL4g8E2tyxVVyxd0+8QeBM1mIk9KbWxkegRr76XzA==}
-    dev: false
-
   /@rolldown/pluginutils@1.0.1:
     resolution: {integrity: sha512-2j9bGt5Jh8hj+vPtgzPtl72j0yRxHAyumoo6TNfAjsLB04UtpSvPbPcDcBMxz7n+9CYB0c1GxQFxYRg2jimqGw==}
     dev: false
@@ -3081,105 +2089,101 @@ packages:
     resolution: {integrity: sha512-e7Mew686owMaPJVNNLs55PUvgz371nKgwsc4vxE49zsODpJEnxgxRo2y/OKrqueavXgZNMDVj3DdHFlaSAeU8g==}
     dev: false
 
-  /@tailwindcss/node@4.1.18:
-    resolution: {integrity: sha512-DoR7U1P7iYhw16qJ49fgXUlry1t4CpXeErJHnQ44JgTSKMaZUdf17cfn5mHchfJ4KRBZRFA/Coo+MUF5+gOaCQ==}
+  /@tailwindcss/node@4.3.3:
+    resolution: {integrity: sha512-/T8IKEsf9VTU6tLjgC7+sv2mOPtQxzE2jMw7u4Tt40Tx+QSZxpzh95/H6cMKoja9XuW7iMdLJYBB0o9G1CaAgg==}
     dependencies:
       '@jridgewell/remapping': 2.3.5
-      enhanced-resolve: 5.18.4
-      jiti: 2.6.1
-      lightningcss: 1.30.2
+      enhanced-resolve: 5.26.0
+      jiti: 2.7.0
+      lightningcss: 1.32.0
       magic-string: 0.30.21
-      source-map-js: 1.2.1
-      tailwindcss: 4.1.18
+      source-map-js: 1.2.2
+      tailwindcss: 4.3.3
     dev: false
 
-  /@tailwindcss/oxide-android-arm64@4.1.18:
-    resolution: {integrity: sha512-dJHz7+Ugr9U/diKJA0W6N/6/cjI+ZTAoxPf9Iz9BFRF2GzEX8IvXxFIi/dZBloVJX/MZGvRuFA9rqwdiIEZQ0Q==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-android-arm64@4.3.3:
+    resolution: {integrity: sha512-Y85A2gmPSkl5Ve5qR86GL4HT509cFqQh1aes9p3sSkyTPwt0Pppf3GkwGe4JPACcRYjgJIEhQgM6dBClnr0NYw==}
+    engines: {node: '>= 20'}
     cpu: [arm64]
     os: [android]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-darwin-arm64@4.1.18:
-    resolution: {integrity: sha512-Gc2q4Qhs660bhjyBSKgq6BYvwDz4G+BuyJ5H1xfhmDR3D8HnHCmT/BSkvSL0vQLy/nkMLY20PQ2OoYMO15Jd0A==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-darwin-arm64@4.3.3:
+    resolution: {integrity: sha512-BiaWatpBcERQFDlOjRDpIVXuFK5PJez5SA4JMg6VYZdBYU+qKfV/vqjcIs+IYmtitf1xYQZTwXvU/8y4lfZUGw==}
+    engines: {node: '>= 20'}
     cpu: [arm64]
     os: [darwin]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-darwin-x64@4.1.18:
-    resolution: {integrity: sha512-FL5oxr2xQsFrc3X9o1fjHKBYBMD1QZNyc1Xzw/h5Qu4XnEBi3dZn96HcHm41c/euGV+GRiXFfh2hUCyKi/e+yw==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-darwin-x64@4.3.3:
+    resolution: {integrity: sha512-fAeUqfV5ndhxRwai8cXGzdLvul9utWOmeTkv69unv4ZXixjn61Z+p9lCWdwOwA3TYboG3BwdVuN/RDjhBRl0mw==}
+    engines: {node: '>= 20'}
     cpu: [x64]
     os: [darwin]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-freebsd-x64@4.1.18:
-    resolution: {integrity: sha512-Fj+RHgu5bDodmV1dM9yAxlfJwkkWvLiRjbhuO2LEtwtlYlBgiAT4x/j5wQr1tC3SANAgD+0YcmWVrj8R9trVMA==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-freebsd-x64@4.3.3:
+    resolution: {integrity: sha512-iyf5bV6+wnAlflVeEy7R25dupxTNECZN5QMI0qNT6eT+EgaGdZcKhGkr5SdoaWiLJ3spLqIY9VCeSGrwmtg4kw==}
+    engines: {node: '>= 20'}
     cpu: [x64]
     os: [freebsd]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-linux-arm-gnueabihf@4.1.18:
-    resolution: {integrity: sha512-Fp+Wzk/Ws4dZn+LV2Nqx3IilnhH51YZoRaYHQsVq3RQvEl+71VGKFpkfHrLM/Li+kt5c0DJe/bHXK1eHgDmdiA==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-linux-arm-gnueabihf@4.3.3:
+    resolution: {integrity: sha512-aAYUprJAJQWWbRrPvtjdroZ56Md+JM8pMiopS6xGEwDfLhqj+2ver2p4nU4Mb3CRqcMmNBjo8KkUgcxhkzVQGQ==}
+    engines: {node: '>= 20'}
     cpu: [arm]
     os: [linux]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-linux-arm64-gnu@4.1.18:
-    resolution: {integrity: sha512-S0n3jboLysNbh55Vrt7pk9wgpyTTPD0fdQeh7wQfMqLPM/Hrxi+dVsLsPrycQjGKEQk85Kgbx+6+QnYNiHalnw==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-linux-arm64-gnu@4.3.3:
+    resolution: {integrity: sha512-nDxldcEENOxZRzC2uu9jrutZdAAQtb+8WWDCSnWL1zvBk1+FN+x6MtDViPB5AJMfttVCUhehGWus3XBPgatM/w==}
+    engines: {node: '>= 20'}
     cpu: [arm64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-linux-arm64-musl@4.1.18:
-    resolution: {integrity: sha512-1px92582HkPQlaaCkdRcio71p8bc8i/ap5807tPRDK/uw953cauQBT8c5tVGkOwrHMfc2Yh6UuxaH4vtTjGvHg==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-linux-arm64-musl@4.3.3:
+    resolution: {integrity: sha512-Md44bD6veX/PC5iyF8cDVnw4HBIANZepRZZ7a8DQOvkfo5WUBwcp6iAuCUz23u+4SUkhJlD3eL7hNdW8ezd/kA==}
+    engines: {node: '>= 20'}
     cpu: [arm64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-linux-x64-gnu@4.1.18:
-    resolution: {integrity: sha512-v3gyT0ivkfBLoZGF9LyHmts0Isc8jHZyVcbzio6Wpzifg/+5ZJpDiRiUhDLkcr7f/r38SWNe7ucxmGW3j3Kb/g==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-linux-x64-gnu@4.3.3:
+    resolution: {integrity: sha512-tx7us1muwOKAKWao2v/GaafFeQboE6aj88vC6ziN2NCGcRm8gWUhwjzg+YdVB1e4boAtdtma4L43onunI6NS4w==}
+    engines: {node: '>= 20'}
     cpu: [x64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-linux-x64-musl@4.1.18:
-    resolution: {integrity: sha512-bhJ2y2OQNlcRwwgOAGMY0xTFStt4/wyU6pvI6LSuZpRgKQwxTec0/3Scu91O8ir7qCR3AuepQKLU/kX99FouqQ==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-linux-x64-musl@4.3.3:
+    resolution: {integrity: sha512-SJxX60smvHgasZoBy11dX6YRjXJFovwWBoedhbQPOBzgFWBHGB+TVPWB9BxzR7TTxU8FQZAI2AyiNCMzFm8Img==}
+    engines: {node: '>= 20'}
     cpu: [x64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-wasm32-wasi@4.1.18:
-    resolution: {integrity: sha512-LffYTvPjODiP6PT16oNeUQJzNVyJl1cjIebq/rWWBF+3eDst5JGEFSc5cWxyRCJ0Mxl+KyIkqRxk1XPEs9x8TA==}
+  /@tailwindcss/oxide-wasm32-wasi@4.3.3:
+    resolution: {integrity: sha512-jx1+rPhY/5Ympkktd656HBWEBLxP7dH06losBLjjf5vgCODXvi9KhtftWcMIwTFIDqBr7cRnQkdLnAG+IOlGvQ==}
     engines: {node: '>=14.0.0'}
     cpu: [wasm32]
     requiresBuild: true
@@ -3193,50 +2197,50 @@ packages:
       - '@emnapi/wasi-threads'
       - tslib
 
-  /@tailwindcss/oxide-win32-arm64-msvc@4.1.18:
-    resolution: {integrity: sha512-HjSA7mr9HmC8fu6bdsZvZ+dhjyGCLdotjVOgLA2vEqxEBZaQo9YTX4kwgEvPCpRh8o4uWc4J/wEoFzhEmjvPbA==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-win32-arm64-msvc@4.3.3:
+    resolution: {integrity: sha512-3rc292Ca2ceK6Ulcc/bAVnTs/3nDtoPhyEKlgPv+yQJQi/JS/AMJlqzxvlDacL1nekbrcf6bTqp/jV4qgnPxNQ==}
+    engines: {node: '>= 20'}
     cpu: [arm64]
     os: [win32]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide-win32-x64-msvc@4.1.18:
-    resolution: {integrity: sha512-bJWbyYpUlqamC8dpR7pfjA0I7vdF6t5VpUGMWRkXVE3AXgIZjYUYAK7II1GNaxR8J1SSrSrppRar8G++JekE3Q==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide-win32-x64-msvc@4.3.3:
+    resolution: {integrity: sha512-yJ0pwIVc/nYeGoV02WtsN8KYyLQv7kyI2wDnkezyJlGGjkd4QLwDGAwl47YpPJeuI0M0ObaXGSPjvWDPeTPggw==}
+    engines: {node: '>= 20'}
     cpu: [x64]
     os: [win32]
     requiresBuild: true
     dev: false
     optional: true
 
-  /@tailwindcss/oxide@4.1.18:
-    resolution: {integrity: sha512-EgCR5tTS5bUSKQgzeMClT6iCY3ToqE1y+ZB0AKldj809QXk1Y+3jB0upOYZrn9aGIzPtUsP7sX4QQ4XtjBB95A==}
-    engines: {node: '>= 10'}
+  /@tailwindcss/oxide@4.3.3:
+    resolution: {integrity: sha512-krXjAikiaFSPaK/FkAQT5UTx3VormQaiZ5hBFlJZ9UFQGB/rwg1MZIhHAG9smMQRTdyJxP6Qt5MwMtdyU5FWrA==}
+    engines: {node: '>= 20'}
     optionalDependencies:
-      '@tailwindcss/oxide-android-arm64': 4.1.18
-      '@tailwindcss/oxide-darwin-arm64': 4.1.18
-      '@tailwindcss/oxide-darwin-x64': 4.1.18
-      '@tailwindcss/oxide-freebsd-x64': 4.1.18
-      '@tailwindcss/oxide-linux-arm-gnueabihf': 4.1.18
-      '@tailwindcss/oxide-linux-arm64-gnu': 4.1.18
-      '@tailwindcss/oxide-linux-arm64-musl': 4.1.18
-      '@tailwindcss/oxide-linux-x64-gnu': 4.1.18
-      '@tailwindcss/oxide-linux-x64-musl': 4.1.18
-      '@tailwindcss/oxide-wasm32-wasi': 4.1.18
-      '@tailwindcss/oxide-win32-arm64-msvc': 4.1.18
-      '@tailwindcss/oxide-win32-x64-msvc': 4.1.18
+      '@tailwindcss/oxide-android-arm64': 4.3.3
+      '@tailwindcss/oxide-darwin-arm64': 4.3.3
+      '@tailwindcss/oxide-darwin-x64': 4.3.3
+      '@tailwindcss/oxide-freebsd-x64': 4.3.3
+      '@tailwindcss/oxide-linux-arm-gnueabihf': 4.3.3
+      '@tailwindcss/oxide-linux-arm64-gnu': 4.3.3
+      '@tailwindcss/oxide-linux-arm64-musl': 4.3.3
+      '@tailwindcss/oxide-linux-x64-gnu': 4.3.3
+      '@tailwindcss/oxide-linux-x64-musl': 4.3.3
+      '@tailwindcss/oxide-wasm32-wasi': 4.3.3
+      '@tailwindcss/oxide-win32-arm64-msvc': 4.3.3
+      '@tailwindcss/oxide-win32-x64-msvc': 4.3.3
     dev: false
 
-  /@tailwindcss/postcss@4.1.18:
-    resolution: {integrity: sha512-Ce0GFnzAOuPyfV5SxjXGn0CubwGcuDB0zcdaPuCSzAa/2vII24JTkH+I6jcbXLb1ctjZMZZI6OjDaLPJQL1S0g==}
+  /@tailwindcss/postcss@4.3.3:
+    resolution: {integrity: sha512-JTSZZGQi1AyKirbLN3azmjVzef92tcX7h+iSqPdaeStyFpGpDlKvvpxeOE8njhbUanbRwr3z8DyzhICWnMtQeg==}
     dependencies:
-      '@alloc/quick-lru': 5.2.0
-      '@tailwindcss/node': 4.1.18
-      '@tailwindcss/oxide': 4.1.18
-      postcss: 8.5.15
-      tailwindcss: 4.1.18
+      '@alloc/quick-lru': 5.3.0
+      '@tailwindcss/node': 4.3.3
+      '@tailwindcss/oxide': 4.3.3
+      postcss: 8.5.29
+      tailwindcss: 4.3.3
     dev: false
 
   /@tanstack/react-table@8.21.3(react-dom@19.2.1)(react@19.2.1):
@@ -3256,14 +2260,6 @@ packages:
     engines: {node: '>=12'}
     dev: false
 
-  /@tybys/wasm-util@0.10.1:
-    resolution: {integrity: sha512-9tTaPJLSiejZKx+Bmog4uSubteqTvFrVrURwkmHixBo0G4seD0zUxp98E1DzUBJxLQ3NPwXrGKDiVjwx/DpPsg==}
-    requiresBuild: true
-    dependencies:
-      tslib: 2.8.1
-    dev: false
-    optional: true
-
   /@types/d3-array@3.2.2:
     resolution: {integrity: sha512-hOLWVbm7uRza0BYXpIIW5pxfrKe0W+D5lrFiAEYR+pb6w3N2SwSMaJbXdUfSEv+dT4MfHBLtn5js0LAWaO6otw==}
     dev: false
@@ -3292,8 +2288,8 @@ packages:
       '@types/d3-time': 3.0.4
     dev: false
 
-  /@types/d3-shape@3.1.8:
-    resolution: {integrity: sha512-lae0iWfcDeR7qt7rA88BNiqdvPS5pFVPpo5OfjElwNaT2yyekbM0C9vK+yqBqEmHr6lDkRnYNoTBYlAgJa7a4w==}
+  /@types/d3-shape@3.2.0:
+    resolution: {integrity: sha512-kVd74ta9eof3eJOvbNd1vGKS/XERRyQbT26Og63hIsvDO84cjD5gEOhsXf26w3FSoNlPVz84DOFcKv/oou+fMw==}
     dependencies:
       '@types/d3-path': 3.1.1
     dev: false
@@ -3306,21 +2302,21 @@ packages:
     resolution: {integrity: sha512-Ps3T8E8dZDam6fUyNiMkekK3XUsaUEik+idO9/YjPtfj2qruF8tFBXS7XhtE4iIXBLxhmLjP3SXpLhVf21I9Lw==}
     dev: false
 
-  /@types/debug@4.1.12:
-    resolution: {integrity: sha512-vIChWdVG3LG1SMxEvI/AK+FWJthlrqlTu7fbrlywTkkaONwk/UAGaULXRlf8vkzFBLVm0zkMdCquhL5aOjhXPQ==}
+  /@types/debug@4.1.13:
+    resolution: {integrity: sha512-KSVgmQmzMwPlmtljOomayoR89W4FynCAi3E8PPs7vmDVPe84hT+vGPKkJfThkmXs0x0jAaa9U8uW8bbfyS2fWw==}
     dependencies:
       '@types/ms': 2.1.0
 
   /@types/estree-jsx@1.0.5:
     resolution: {integrity: sha512-52CcUVNFyfb1A2ALocQw/Dd1BQFNmSdkuC3BkZ6iqhdMfQz7JWOFRuJFloOzjk+6WijU56m9oKXFAXc7o3Towg==}
     dependencies:
-      '@types/estree': 1.0.8
+      '@types/estree': 1.0.9
 
-  /@types/estree@1.0.8:
-    resolution: {integrity: sha512-dWHzHa2WqEXI/O1E9OjrocMTKJl2mSrEolh1Iomrv6U+JuNwaHXsXx9bLu5gG7BUWFIN0skIQJQ/L1rIex4X6w==}
+  /@types/estree@1.0.9:
+    resolution: {integrity: sha512-GhdPgy1el4/ImP05X05Uw4cw2/M93BCUmnEvWZNStlCzEKME4Fkk+YpoA5OiHNQmoS7Cafb8Xa3Pya8m1Qrzeg==}
 
-  /@types/hast@3.0.4:
-    resolution: {integrity: sha512-WPs+bbQw5aCj+x6laNGWLH3wviHtoCv/P3+otBhbOhJgG8qtpdAMlTCxLtsTWA7LH1Oh/bFCHsBn0TPS5m30EQ==}
+  /@types/hast@3.0.5:
+    resolution: {integrity: sha512-rp/ezSWaD1m44dPKICGhiskI13nVr7qTloFwDa/IYkhhf5nzwP+zIQcIJh3WIFSBOy/H1PzB40jPjMDksN4F+g==}
     dependencies:
       '@types/unist': 3.0.3
 
@@ -3328,8 +2324,8 @@ packages:
     resolution: {integrity: sha512-5+fP8P8MFNC+AyZCDxrB2pkZFPGzqQWUzpSeuuVLvm8VMcorNYavBqoFcxK8bQz4Qsbn4oUEEem4wDLfcysGHA==}
     dev: true
 
-  /@types/lodash@4.17.23:
-    resolution: {integrity: sha512-RDvF6wTulMPjrNdCoYRC8gNR880JNGT8uB+REUpC2Ns4pRqQJhGz90wh7rgdXDPpCczF3VGktDuFGVnz8zP7HA==}
+  /@types/lodash@4.17.26:
+    resolution: {integrity: sha512-48NWfiY6oNcKOh47Z4h+sIvwcQR91h6KPhaYKgeMIrkT2jV82Gy90SErj60vW8jQbeMSOSbhIv3bZ7JoqPPVhA==}
     dev: true
 
   /@types/mdast@4.0.4:
@@ -3340,36 +2336,36 @@ packages:
   /@types/ms@2.1.0:
     resolution: {integrity: sha512-GsCCIZDE/p3i96vtEqx+7dBUGXrc7zeSK3wwPHIaRThS+9OhWIXRqzs4d6k1SVU8g91DrNRWxWUGhp5KXQb2VA==}
 
-  /@types/node@20.19.30:
-    resolution: {integrity: sha512-WJtwWJu7UdlvzEAUm484QNg5eAoq5QR08KDNx7g45Usrs2NtOPiX8ugDqmKdXkyL03rBqU5dYNYVQetEpBHq2g==}
+  /@types/node@20.19.43:
+    resolution: {integrity: sha512-6oYBAi5ikg4Pl+kGsoYtawUMBT2zZMCvPNF7pVLnHZfd1zf38DRiWn/gT01RYCdUqkv7Fhr+C9ot4/tb+2sVvA==}
     dependencies:
       undici-types: 6.21.0
 
-  /@types/prismjs@1.26.5:
-    resolution: {integrity: sha512-AUZTa7hQ2KY5L7AmtSiqxlhWxb4ina0yd8hNbl4TWuqnv/pFP0nDMb3YrfSBf4hJVGLh2YEIBfKaBW/9UEl6IQ==}
+  /@types/prismjs@1.26.6:
+    resolution: {integrity: sha512-vqlvI7qlMvcCBbVe0AKAb4f97//Hy0EBTaiW8AalRnG/xAN5zOiWWyrNqNXeq8+KAuvRewjCVY1+IPxk4RdNYw==}
     dev: false
 
   /@types/qrcode@1.5.6:
     resolution: {integrity: sha512-te7NQcV2BOvdj2b1hCAHzAoMNuj65kNBMz0KBaxM6c3VGBOhU0dURQKOtH8CFNI/dsKkwlv32p26qYQTWoB5bw==}
     dependencies:
-      '@types/node': 20.19.30
+      '@types/node': 20.19.43
     dev: true
 
-  /@types/react-dom@19.2.3(@types/react@19.2.10):
-    resolution: {integrity: sha512-jp2L/eY6fn+KgVVQAOqYItbF0VY/YApe5Mz2F0aykSO8gx31bYCZyvSeYxCHKvzHG5eZjc+zyaS5BrBWya2+kQ==}
+  /@types/react-dom@19.2.7(@types/react@19.2.18):
+    resolution: {integrity: sha512-I8bPpDLcHBv1qiIiXDCy71Rt8eQDKJP0sMSWJphDdAcdqiJ1sGpZamavoEIRZmYzjia9LuEb2HlYdDpmoENpvQ==}
     peerDependencies:
       '@types/react': ^19.2.0
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
 
   /@types/react-syntax-highlighter@15.5.13:
     resolution: {integrity: sha512-uLGJ87j6Sz8UaBAooU0T6lWJ0dBmjZgN1PZTrj05TNql2/XpC6+4HhMT5syIdFUUt+FASfCeLLv4kBygNU+8qA==}
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
     dev: true
 
-  /@types/react@19.2.10:
-    resolution: {integrity: sha512-WPigyYuGhgZ/cTPRXB2EwUw+XvsRA3GqHlsP4qteqrnnjDrApbS7MxcGr/hke5iUoeB7E/gQtrs9I37zAJ0Vjw==}
+  /@types/react@19.2.18:
+    resolution: {integrity: sha512-AnzbBERsrLKtk2XSfTbYRLjQPdy116Sty4q+T+Bp3IC4l6jNBvreVPAHmpq9qhXQM7CXZPjLVmGMw9sy+hxQ3w==}
     dependencies:
       csstype: 3.2.3
 
@@ -3380,177 +2376,180 @@ packages:
   /@types/unist@3.0.3:
     resolution: {integrity: sha512-ko/gIFJRv177XgZsZcBwnqJN5x/Gien8qNOn0D5bQU/zAzVf9Zt3BlcUiLqhV9y4ARk0GbT3tnUiPNgnTXzc/Q==}
 
-  /@typescript-eslint/eslint-plugin@8.54.0(@typescript-eslint/parser@8.54.0)(eslint@9.39.2)(typescript@5.9.3):
-    resolution: {integrity: sha512-hAAP5io/7csFStuOmR782YmTthKBJ9ND3WVL60hcOjvtGFb+HJxH4O5huAcmcZ9v9G8P+JETiZ/G1B8MALnWZQ==}
+  /@typescript-eslint/eslint-plugin@8.71.1(@typescript-eslint/parser@8.71.1)(eslint@9.39.5)(typescript@5.9.3):
+    resolution: {integrity: sha512-nNlBf6HuotqMgZvSkWD5hNdemiBIxQ2NmrlgfcUIeZRs+BbakjscCQRq/Hb74EFVnFiSWWdL12JruqwBy3ERDw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      '@typescript-eslint/parser': ^8.54.0
-      eslint: ^8.57.0 || ^9.0.0
-      typescript: '>=4.8.4 <6.0.0'
+      '@typescript-eslint/parser': ^8.71.1
+      eslint: ^8.57.0 || ^9.0.0 || ^10.0.0
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
       '@eslint-community/regexpp': 4.12.2
-      '@typescript-eslint/parser': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
-      '@typescript-eslint/scope-manager': 8.54.0
-      '@typescript-eslint/type-utils': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
-      '@typescript-eslint/utils': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
-      '@typescript-eslint/visitor-keys': 8.54.0
-      eslint: 9.39.2
-      ignore: 7.0.5
+      '@typescript-eslint/parser': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
+      '@typescript-eslint/scope-manager': 8.71.1
+      '@typescript-eslint/type-utils': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
+      '@typescript-eslint/utils': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
+      '@typescript-eslint/visitor-keys': 8.71.1
+      eslint: 9.39.5
+      ignore: 7.0.12
       natural-compare: 1.4.0
-      ts-api-utils: 2.4.0(typescript@5.9.3)
+      ts-api-utils: 2.5.0(typescript@5.9.3)
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/parser@8.54.0(eslint@9.39.2)(typescript@5.9.3):
-    resolution: {integrity: sha512-BtE0k6cjwjLZoZixN0t5AKP0kSzlGu7FctRXYuPAm//aaiZhmfq1JwdYpYr1brzEspYyFeF+8XF5j2VK6oalrA==}
+  /@typescript-eslint/parser@8.71.1(eslint@9.39.5)(typescript@5.9.3):
+    resolution: {integrity: sha512-oM/5sAqz/l1v/zYK1qTOPrn2+JIrBGC1V6uNYucRO5gh6eT9vWJ/RhPbGE59byvZEUQ4VP6XkLBaS5Jti5jpvw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      eslint: ^8.57.0 || ^9.0.0
-      typescript: '>=4.8.4 <6.0.0'
+      eslint: ^8.57.0 || ^9.0.0 || ^10.0.0
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@typescript-eslint/scope-manager': 8.54.0
-      '@typescript-eslint/types': 8.54.0
-      '@typescript-eslint/typescript-estree': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/visitor-keys': 8.54.0
+      '@typescript-eslint/scope-manager': 8.71.1
+      '@typescript-eslint/types': 8.71.1
+      '@typescript-eslint/typescript-estree': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/visitor-keys': 8.71.1
       debug: 4.4.3
-      eslint: 9.39.2
+      eslint: 9.39.5
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/project-service@8.54.0(typescript@5.9.3):
-    resolution: {integrity: sha512-YPf+rvJ1s7MyiWM4uTRhE4DvBXrEV+d8oC3P9Y2eT7S+HBS0clybdMIPnhiATi9vZOYDc7OQ1L/i6ga6NFYK/g==}
+  /@typescript-eslint/project-service@8.71.1(typescript@5.9.3):
+    resolution: {integrity: sha512-Oijc9RUsohHjncg6iGi7gCtjnEjYt9FWv9u6ygK8uKcHK5eLjLAkqERQ9BaytbpceOFdk6kifPCS/zAZE+A99g==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      typescript: '>=4.8.4 <6.0.0'
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@typescript-eslint/tsconfig-utils': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/types': 8.54.0
+      '@typescript-eslint/tsconfig-utils': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/types': 8.71.1
       debug: 4.4.3
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/scope-manager@8.54.0:
-    resolution: {integrity: sha512-27rYVQku26j/PbHYcVfRPonmOlVI6gihHtXFbTdB5sb6qA0wdAQAbyXFVarQ5t4HRojIz64IV90YtsjQSSGlQg==}
+  /@typescript-eslint/scope-manager@8.71.1:
+    resolution: {integrity: sha512-1YUDZXdTnXLlob03SFHBipdyMBsDVm7zoSz0ytTJ39dk5J1TgsTK03VlR/dLXG/RLLUrPcYbD4rp23BmHxHP0w==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dependencies:
-      '@typescript-eslint/types': 8.54.0
-      '@typescript-eslint/visitor-keys': 8.54.0
+      '@typescript-eslint/types': 8.71.1
+      '@typescript-eslint/visitor-keys': 8.71.1
     dev: true
 
-  /@typescript-eslint/tsconfig-utils@8.54.0(typescript@5.9.3):
-    resolution: {integrity: sha512-dRgOyT2hPk/JwxNMZDsIXDgyl9axdJI3ogZ2XWhBPsnZUv+hPesa5iuhdYt2gzwA9t8RE5ytOJ6xB0moV0Ujvw==}
+  /@typescript-eslint/tsconfig-utils@8.71.1(typescript@5.9.3):
+    resolution: {integrity: sha512-oTvkml5SxXhgg+WWKhod5qWQICDXtOB+/Hi9VdPKFAANJ2b7YOgMffgvzXuqog7R6A/JtrzRf128dsxummyWhg==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      typescript: '>=4.8.4 <6.0.0'
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
       typescript: 5.9.3
     dev: true
 
-  /@typescript-eslint/type-utils@8.54.0(eslint@9.39.2)(typescript@5.9.3):
-    resolution: {integrity: sha512-hiLguxJWHjjwL6xMBwD903ciAwd7DmK30Y9Axs/etOkftC3ZNN9K44IuRD/EB08amu+Zw6W37x9RecLkOo3pMA==}
+  /@typescript-eslint/type-utils@8.71.1(eslint@9.39.5)(typescript@5.9.3):
+    resolution: {integrity: sha512-kH3t3jZCYv2PZSBEWPKTariLrCaoF1CLoPitYFBJOhAGEwWnSRAtGEifrfrpBa4K9JlwWNZotgA5w3/wLSk/FQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      eslint: ^8.57.0 || ^9.0.0
-      typescript: '>=4.8.4 <6.0.0'
+      eslint: ^8.57.0 || ^9.0.0 || ^10.0.0
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@typescript-eslint/types': 8.54.0
-      '@typescript-eslint/typescript-estree': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/utils': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
+      '@typescript-eslint/types': 8.71.1
+      '@typescript-eslint/typescript-estree': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/utils': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
       debug: 4.4.3
-      eslint: 9.39.2
-      ts-api-utils: 2.4.0(typescript@5.9.3)
+      eslint: 9.39.5
+      ts-api-utils: 2.5.0(typescript@5.9.3)
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/types@8.54.0:
-    resolution: {integrity: sha512-PDUI9R1BVjqu7AUDsRBbKMtwmjWcn4J3le+5LpcFgWULN3LvHC5rkc9gCVxbrsrGmO1jfPybN5s6h4Jy+OnkAA==}
+  /@typescript-eslint/types@8.71.1:
+    resolution: {integrity: sha512-ybYnPTUwg3VQQOFEiBPLs8Jxl1ry2vXdn1kVKguomBbvxpetdu3MKTNDqFNUrlG8ylP9MpZosmXMFYlGQHk1oQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dev: true
 
-  /@typescript-eslint/typescript-estree@8.54.0(typescript@5.9.3):
-    resolution: {integrity: sha512-BUwcskRaPvTk6fzVWgDPdUndLjB87KYDrN5EYGetnktoeAvPtO4ONHlAZDnj5VFnUANg0Sjm7j4usBlnoVMHwA==}
+  /@typescript-eslint/typescript-estree@8.71.1(typescript@5.9.3):
+    resolution: {integrity: sha512-pSKSEK1JJpHs6KiKVtABCE4iJK5d7FLFzDXQNBQ2oRY7LEd3UwOt3TZemTgu9syTIf56dJWlTn/pdmrO9oKUXQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      typescript: '>=4.8.4 <6.0.0'
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@typescript-eslint/project-service': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/tsconfig-utils': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/types': 8.54.0
-      '@typescript-eslint/visitor-keys': 8.54.0
+      '@typescript-eslint/project-service': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/tsconfig-utils': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/types': 8.71.1
+      '@typescript-eslint/visitor-keys': 8.71.1
       debug: 4.4.3
-      minimatch: 9.0.7
-      semver: 7.7.3
-      tinyglobby: 0.2.15
-      ts-api-utils: 2.4.0(typescript@5.9.3)
+      minimatch: 10.2.6
+      semver: 7.8.5
+      tinyglobby: 0.2.17
+      ts-api-utils: 2.5.0(typescript@5.9.3)
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/utils@8.54.0(eslint@9.39.2)(typescript@5.9.3):
-    resolution: {integrity: sha512-9Cnda8GS57AQakvRyG0PTejJNlA2xhvyNtEVIMlDWOOeEyBkYWhGPnfrIAnqxLMTSTo6q8g12XVjjev5l1NvMA==}
+  /@typescript-eslint/utils@8.71.1(eslint@9.39.5)(typescript@5.9.3):
+    resolution: {integrity: sha512-CBjT6gfAz3DW2j9Q4moFnqt208Aq+506sVGXfpRcyAeqn0EpayL0mjdWoCO2t3RC7GewNrFxdeehuEaPg6DMzQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      eslint: ^8.57.0 || ^9.0.0
-      typescript: '>=4.8.4 <6.0.0'
+      eslint: ^8.57.0 || ^9.0.0 || ^10.0.0
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@eslint-community/eslint-utils': 4.9.1(eslint@9.39.2)
-      '@typescript-eslint/scope-manager': 8.54.0
-      '@typescript-eslint/types': 8.54.0
-      '@typescript-eslint/typescript-estree': 8.54.0(typescript@5.9.3)
-      eslint: 9.39.2
+      '@eslint-community/eslint-utils': 4.10.1(eslint@9.39.5)
+      '@typescript-eslint/scope-manager': 8.71.1
+      '@typescript-eslint/types': 8.71.1
+      '@typescript-eslint/typescript-estree': 8.71.1(typescript@5.9.3)
+      eslint: 9.39.5
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
     dev: true
 
-  /@typescript-eslint/visitor-keys@8.54.0:
-    resolution: {integrity: sha512-VFlhGSl4opC0bprJiItPQ1RfUhGDIBokcPwaFH4yiBCaNPeld/9VeXbiPO1cLyorQi1G1vL+ecBk1x8o1axORA==}
+  /@typescript-eslint/visitor-keys@8.71.1:
+    resolution: {integrity: sha512-0GxiUGqMU0qXJt58cBsiG3DP8lUmBaFo2vChtUX+CT7PnR3v/VGLLrzVP4K9r1lNiZXjHYefMBbPunplr09v6w==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dependencies:
-      '@typescript-eslint/types': 8.54.0
-      eslint-visitor-keys: 4.2.1
+      '@typescript-eslint/types': 8.71.1
+      eslint-visitor-keys: 5.0.1
     dev: true
 
-  /@ungap/structured-clone@1.3.0:
-    resolution: {integrity: sha512-WmoN8qaIAo7WTYWbAZuG8PYEhn5fkz7dZrqTBZ7dtt//lL2Gwms1IcnQ5yHqjDfX8Ft5j4YzDM23f87zBfDe9g==}
+  /@ungap/structured-clone@1.4.0:
+    resolution: {integrity: sha512-1mEZtMKPM09vDmQt5y7YvmN2+DFTP7Tg0EWXdic8/C6VRnpb33e4ghisCIE3WZjsE2N8mf+QV1Zqh7ZFYLWInQ==}
     dev: false
 
-  /@vitejs/plugin-react@6.0.1(vite@8.0.16):
-    resolution: {integrity: sha512-l9X/E3cDb+xY3SWzlG1MOGt2usfEHGMNIaegaUGFsLkb3RCn/k8/TOXBcab+OndDI4TBtktT8/9BwwW8Vi9KUQ==}
+  /@vitejs/plugin-react@6.1.2(vite@8.3.4):
+    resolution: {integrity: sha512-fhJNGUdscQa4HkI0xFxwonWyhnVReCPDbXI7FH7jmzpnVCrFQI7PRrhWecO263610vOrs9tr2Y4xcf8uAdFxbw==}
     engines: {node: ^20.19.0 || >=22.12.0}
     peerDependencies:
       '@rolldown/plugin-babel': ^0.1.7 || ^0.2.0
       babel-plugin-react-compiler: ^1.0.0
+      oxc-transform-react: ^0.152.0
       vite: ^8.0.0
     peerDependenciesMeta:
       '@rolldown/plugin-babel':
         optional: true
       babel-plugin-react-compiler:
         optional: true
+      oxc-transform-react:
+        optional: true
     dependencies:
-      '@rolldown/pluginutils': 1.0.0-rc.7
-      vite: 8.0.16(@types/node@20.19.30)
+      '@rolldown/pluginutils': 1.0.1
+      vite: 8.3.4(@types/node@20.19.43)
     dev: false
 
-  /acorn-jsx@5.3.2(acorn@8.15.0):
+  /acorn-jsx@5.3.2(acorn@8.19.0):
     resolution: {integrity: sha512-rq9s+JNhf0IChjtDXxllJ7g41oZk5SlXtp0LHwyA5cejwn7vKmKp4pPri6YEePv2PU65sAsegbXtIinmDFDXgQ==}
     peerDependencies:
       acorn: ^6.0.0 || ^7.0.0 || ^8.0.0
     dependencies:
-      acorn: 8.15.0
+      acorn: 8.19.0
     dev: true
 
-  /acorn@8.15.0:
-    resolution: {integrity: sha512-NZyJarBfL7nWwIq+FDL6Zp/yHEhePMNnnJ0y3qfieCrmNvYct8uvtiV41UvlSe6apAfk0fY1FbWx+NwfmpvtTg==}
+  /acorn@8.19.0:
+    resolution: {integrity: sha512-oJlA3XiRm7Cyk6qFD2Jn8ak9B7jSy0qy00ADO3+8dpT0LSjFihQYv4C02LFCSYJV3Q37xYwwRy5m+IpIiUzqWw==}
     engines: {node: '>=0.4.0'}
     hasBin: true
     dev: true
@@ -3564,8 +2563,8 @@ packages:
       - supports-color
     dev: false
 
-  /ajv@6.12.6:
-    resolution: {integrity: sha512-j3fVLgvTo527anyYyJOGTYJbG+vnnQYvE0m5mmkc1TK+nxAppkCLMIL0aZ4dblVCNoGShhm+kzE4ZUykBoMg4g==}
+  /ajv@6.15.0:
+    resolution: {integrity: sha512-fgFx7Hfoq60ytK2c7DhnF8jIvzYgOMxfugjLOSMHjLIPgenqa7S7oaagATUq99mV6IYvN2tRmC0wnTYX6iPbMw==}
     dependencies:
       fast-deep-equal: 3.1.3
       fast-json-stable-stringify: 2.1.0
@@ -3573,8 +2572,8 @@ packages:
       uri-js: 4.4.1
     dev: true
 
-  /ansi-escapes@7.2.0:
-    resolution: {integrity: sha512-g6LhBsl+GBPRWGWsBtutpzBYuIIdBkLEvad5C/va/74Db018+5TZiyA26cZJAr3Rft5lprVqOIPxf5Vid6tqAw==}
+  /ansi-escapes@7.3.0:
+    resolution: {integrity: sha512-BvU8nYgGQBxcmMuEeUEmNTvrMVjJNSH7RgW24vXexN4Ven6qCvy4TntnvlnwnMLTVlcRQQdbRY8NKnaIoeWDNg==}
     engines: {node: '>=18'}
     dependencies:
       environment: 1.1.0
@@ -3585,8 +2584,8 @@ packages:
     engines: {node: '>=8'}
     dev: false
 
-  /ansi-regex@6.2.2:
-    resolution: {integrity: sha512-Bq3SmSpyFHaWjPk8If9yc6svM8c56dB5BAtW4Qbw5jHTwwXXcTLoRMkpDJp6VL0XzlWaCHTXrkFURMYmD0sLqg==}
+  /ansi-regex@6.4.0:
+    resolution: {integrity: sha512-KzTVk2tCWAHtYrvvvaP8bJKJq2pVinhLcGEQdtLIYPbmNGNyYe8QwNaTUYQp2J7/vIsUKt5QCqAfUkYyG9DkOw==}
     engines: {node: '>=12'}
     dev: true
 
@@ -3620,16 +2619,16 @@ packages:
       is-array-buffer: 3.0.5
     dev: true
 
-  /array-includes@3.1.9:
-    resolution: {integrity: sha512-FmeCCAenzH0KH381SPT5FZmiA/TmpndpcaShhfgEN9eCVjnFBqq3l1xrI42y8+PPLI6hypzou4GXw00WHmPBLQ==}
+  /array-includes@3.2.0:
+    resolution: {integrity: sha512-VXY5eFRarnXcYxwBjJzPmEhH55+rmP79/+ueDhi0F+TuqfHCItagIHqxeUZrmgrOPa31QTh9H85DjX3FfJ0FTg==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-abstract: 1.24.1
-      es-object-atoms: 1.1.1
-      get-intrinsic: 1.3.0
+      es-abstract: 1.24.2
+      es-object-atoms: 1.1.2
+      es-shim-unscopables: 1.1.0
       is-string: 1.1.1
       math-intrinsics: 1.1.0
     dev: true
@@ -3638,11 +2637,11 @@ packages:
     resolution: {integrity: sha512-CVvd6FHg1Z3POpBLxO6E6zr+rSKEQ9L6rZHAaY7lLfhKsWYUBBOuMs0e9o24oopj6H+geRCX0YJ+TJLBK2eHyQ==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       es-shim-unscopables: 1.1.0
     dev: true
 
@@ -3650,9 +2649,9 @@ packages:
     resolution: {integrity: sha512-rwG/ja1neyLqCuGZ5YYrznA62D4mZXg0i1cIskIUKSiqF3Cje9/wXAls9B9s1Wa2fomMsIv8czB8jZcPmxCXFg==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-shim-unscopables: 1.1.0
     dev: true
 
@@ -3660,9 +2659,9 @@ packages:
     resolution: {integrity: sha512-Y7Wt51eKJSyi80hFrJCePGGNo5ktJCslFuboqJsbf57CCPcm5zztluPlc4/aD8sWsKvlwatezpV4U1efk8kpjg==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-shim-unscopables: 1.1.0
     dev: true
 
@@ -3670,9 +2669,9 @@ packages:
     resolution: {integrity: sha512-p6Fx8B7b7ZhL/gmUsAy0D15WhvDccw3mnGNbZpi3pmeJdxtWsj2jEaI4Y6oo3XiHfzuSgPwKc04MYt6KgvC/wA==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
       es-shim-unscopables: 1.1.0
     dev: true
@@ -3682,9 +2681,9 @@ packages:
     engines: {node: '>= 0.4'}
     dependencies:
       array-buffer-byte-length: 1.0.2
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
       get-intrinsic: 1.3.0
       is-array-buffer: 3.0.5
@@ -3706,10 +2705,10 @@ packages:
       possible-typed-array-names: 1.1.0
     dev: true
 
-  /axios@1.17.0:
-    resolution: {integrity: sha512-J8SwNxprqqpbfenehxWYXE7CW+wM1BB4w3+N+g+/Wx40xM4rsLrfPmHHxSWIxJLYDgSY/HqlFPIYb2/S3rxafw==}
+  /axios@1.20.0:
+    resolution: {integrity: sha512-r8aOh8j9cGKpgQAqpzrUHnSIc6a59Y3Xf/cv8sy1DrHCkZHzQGEuoq1tARk6qSyDdtQGSDgpb9kFlruzPvrgwg==}
     dependencies:
-      follow-redirects: 1.16.0
+      follow-redirects: 1.16.1
       form-data: 4.0.6
       https-proxy-agent: 5.0.1
       proxy-from-env: 2.1.0
@@ -3731,16 +2730,16 @@ packages:
     engines: {node: 18 || 20 || >=22}
     dev: true
 
-  /brace-expansion@1.1.12:
-    resolution: {integrity: sha512-9T9UjW3r0UW5c1Q7GTwllptXwhvYmEzFhzMfZ9H7FQWt+uZePjZPjBP/W1ZEyZ1twGWom5/56TF4lPcqjnDHcg==}
+  /brace-expansion@1.1.21:
+    resolution: {integrity: sha512-9zeA+KLZNNzglF2TPKRQEDyx6Yby7daAkuy8MiPzpXPsYDWi/DRM8jmwUDxokQjYqBpv5DgPiwD4h4ZZSy1Ujw==}
     dependencies:
       balanced-match: 1.0.2
       concat-map: 0.0.1
     dev: true
 
-  /brace-expansion@5.0.6:
-    resolution: {integrity: sha512-kLpxurY4Z4r9sgMsyG0Z9uzsBlgiU/EFKhj/h91/8yHu0edo7XuixOIH3VcJ8kkxs6/jPzoI6U9Vj3WqbMQ94g==}
-    engines: {node: 18 || 20 || >=22}
+  /brace-expansion@5.0.12:
+    resolution: {integrity: sha512-YovQ3rzhaLMIrDjNDMkNS01tea93qhEhG5xy8f6+R0l+dw3Ki+5sCoIoI942iuLZTHWogWktgwVDhU09iNEimQ==}
+    engines: {node: 20 || >=22}
     dependencies:
       balanced-match: 4.0.4
     dev: true
@@ -3759,8 +2758,8 @@ packages:
       es-errors: 1.3.0
       function-bind: 1.1.2
 
-  /call-bind@1.0.8:
-    resolution: {integrity: sha512-oKlSFMcMwpUg2ednkhQ454wfWiU/ul3CkJe/PEHcTKuiX6RpbehUiFMXu13HalGZxfUwCQzZG747YXBn1im9ww==}
+  /call-bind@1.0.9:
+    resolution: {integrity: sha512-a/hy+pNsFUTR+Iz8TCJvXudKVLAnz/DyeSUo10I5yvFDQJBFU2s9uqQpoSrJlroHUKoKqzg+epxyP9lqFdzfBQ==}
     engines: {node: '>= 0.4'}
     dependencies:
       call-bind-apply-helpers: 1.0.2
@@ -3854,16 +2853,16 @@ packages:
     engines: {node: '>=6'}
     dev: false
 
-  /cmdk@1.1.1(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
+  /cmdk@1.1.1(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
     resolution: {integrity: sha512-Vsv7kFaXm+ptHDMZ7izaRsP70GgrW9NBNGswt9OZaVBLlE0SNpDq8eu/VGXyF9r7M0azK3Wy7OlYXsuyYLFzHg==}
     peerDependencies:
       react: ^18 || ^19 || ^19.0.0-rc
       react-dom: ^18 || ^19 || ^19.0.0-rc
     dependencies:
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dialog': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-id': 1.1.1(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dialog': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-id': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     transitivePeerDependencies:
@@ -4106,7 +3105,7 @@ packages:
   /dom-helpers@5.2.1:
     resolution: {integrity: sha512-nRCa7CK3VTrM2NmGkIy4cbK7IZlgBE/PYMn55rrXefr5xXDP0LdtfPnblFDoVdcAfslJ7or6iqAUnx0CCGIWQA==}
     dependencies:
-      '@babel/runtime': 7.28.6
+      '@babel/runtime': 7.29.10
       csstype: 3.2.3
     dev: false
 
@@ -4126,12 +3125,12 @@ packages:
     resolution: {integrity: sha512-MSjYzcWNOA0ewAHpz0MxpYFvwg6yjy1NG3xteoqz644VCo/RPgnr1/GGt+ic3iJTzQ8Eu3TdM14SawnVUmGE6A==}
     dev: false
 
-  /enhanced-resolve@5.18.4:
-    resolution: {integrity: sha512-LgQMM4WXU3QI+SYgEc2liRgznaD5ojbmY3sb8LxyguVkIg5FxdpTkvk72te2R38/TGKxH634oLxXRGY6d7AP+Q==}
+  /enhanced-resolve@5.26.0:
+    resolution: {integrity: sha512-9vhedylFonb2YGogzUKX6+Ja72gOJbN1QHAqdrvqLwhdl/QWbKopzoUC9EbQNVsAns/bx/4uyqalrQAoy1IByw==}
     engines: {node: '>=10.13.0'}
     dependencies:
       graceful-fs: 4.2.11
-      tapable: 2.3.0
+      tapable: 2.3.3
     dev: false
 
   /entities@6.0.1:
@@ -4144,24 +3143,34 @@ packages:
     engines: {node: '>=18'}
     dev: true
 
-  /es-abstract@1.24.1:
-    resolution: {integrity: sha512-zHXBLhP+QehSSbsS9Pt23Gg964240DPd6QCf8WpkqEXxQ7fhdZzYsocOr5u7apWonsS5EjZDmTF+/slGMyasvw==}
+  /es-abstract-get@1.0.0:
+    resolution: {integrity: sha512-6PMWXpdhshVvFp+FoWYs1EvG1Nj0tvk0dZM+XcK0xMEM1czRVcP6ohqPWHy6qPagSpC8j4+p89WXlT+xXJs/fg==}
+    engines: {node: '>= 0.4'}
+    dependencies:
+      es-errors: 1.3.0
+      es-object-atoms: 1.1.2
+      is-callable: 1.2.7
+      object-inspect: 1.13.4
+    dev: true
+
+  /es-abstract@1.24.2:
+    resolution: {integrity: sha512-2FpH9Q5i2RRwyEP1AylXe6nYLR5OhaJTZwmlcP0dL/+JCbgg7yyEo/sEK6HeGZRf3dFpWwThaRHVApXSkW3xeg==}
     engines: {node: '>= 0.4'}
     dependencies:
       array-buffer-byte-length: 1.0.2
       arraybuffer.prototype.slice: 1.0.4
       available-typed-arrays: 1.0.7
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       data-view-buffer: 1.0.2
       data-view-byte-length: 1.0.2
       data-view-byte-offset: 1.0.1
       es-define-property: 1.0.1
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       es-set-tostringtag: 2.1.0
-      es-to-primitive: 1.3.0
-      function.prototype.name: 1.1.8
+      es-to-primitive: 1.3.4
+      function.prototype.name: 1.2.0
       get-intrinsic: 1.3.0
       get-proto: 1.0.1
       get-symbol-description: 1.1.0
@@ -4170,7 +3179,7 @@ packages:
       has-property-descriptors: 1.0.2
       has-proto: 1.2.0
       has-symbols: 1.1.0
-      hasown: 2.0.2
+      hasown: 2.0.4
       internal-slot: 1.1.0
       is-array-buffer: 3.0.5
       is-callable: 1.2.7
@@ -4186,22 +3195,22 @@ packages:
       object-inspect: 1.13.4
       object-keys: 1.1.1
       object.assign: 4.1.7
-      own-keys: 1.0.1
+      own-keys: 1.0.2
       regexp.prototype.flags: 1.5.4
-      safe-array-concat: 1.1.3
+      safe-array-concat: 1.1.4
       safe-push-apply: 1.0.0
       safe-regex-test: 1.1.0
       set-proto: 1.0.0
       stop-iteration-iterator: 1.1.0
-      string.prototype.trim: 1.2.10
-      string.prototype.trimend: 1.0.9
+      string.prototype.trim: 1.2.11
+      string.prototype.trimend: 1.0.10
       string.prototype.trimstart: 1.0.8
       typed-array-buffer: 1.0.3
       typed-array-byte-length: 1.0.3
-      typed-array-byte-offset: 1.0.4
-      typed-array-length: 1.0.7
+      typed-array-byte-offset: 1.0.5
+      typed-array-length: 1.0.8
       unbox-primitive: 1.1.0
-      which-typed-array: 1.1.20
+      which-typed-array: 1.1.24
     dev: true
 
   /es-define-property@1.0.1:
@@ -4212,14 +3221,14 @@ packages:
     resolution: {integrity: sha512-Zf5H2Kxt2xjTvbJvP2ZWLEICxA6j+hAmMzIlypy4xcBg1vKVnx89Wy0GbS+kf5cwCVFFzdCFh2XSCFNULS6csw==}
     engines: {node: '>= 0.4'}
 
-  /es-iterator-helpers@1.3.1:
-    resolution: {integrity: sha512-zWwRvqWiuBPr0muUG/78cW3aHROFCNIQ3zpmYDpwdbnt2m+xlNyRWpHBpa2lJjSBit7BQ+RXA1iwbSmu5yJ/EQ==}
+  /es-iterator-helpers@1.4.0:
+    resolution: {integrity: sha512-c/A0P0oxkACDc+cKWw8evLXK83oBKgn0qPOqCYT4x9uolpCIJAcYvJC9QYKNDRPsTeGyCrQ326jrvgZWdCdK5Q==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
       es-set-tostringtag: 2.1.0
       function-bind: 1.1.2
@@ -4232,11 +3241,10 @@ packages:
       internal-slot: 1.1.0
       iterator.prototype: 1.1.5
       math-intrinsics: 1.1.0
-      safe-array-concat: 1.1.3
     dev: true
 
-  /es-object-atoms@1.1.1:
-    resolution: {integrity: sha512-FGgH2h8zKNim9ljj7dankFPcICIK9Cp5bm+c2gQSYePhpaG5+esrLODihIorn+Pe6FGJzWhXQotPv73jTaldXA==}
+  /es-object-atoms@1.1.2:
+    resolution: {integrity: sha512-HWcBoN6NileqtSydK2FqHbS/LoDd2pqrnQHLyJzBj4kOp/ky2MWMN694xOfkK8/SnUsW2DH7EfyVlydKCsm1Zw==}
     engines: {node: '>= 0.4'}
     dependencies:
       es-errors: 1.3.0
@@ -4248,19 +3256,22 @@ packages:
       es-errors: 1.3.0
       get-intrinsic: 1.3.0
       has-tostringtag: 1.0.2
-      hasown: 2.0.2
+      hasown: 2.0.4
 
   /es-shim-unscopables@1.1.0:
     resolution: {integrity: sha512-d9T8ucsEhh8Bi1woXCf+TIKDIROLG5WCkxg8geBCbvk22kzwC5G2OnXVMO6FUsvQlgUUXQ2itephWDLqDzbeCw==}
     engines: {node: '>= 0.4'}
     dependencies:
-      hasown: 2.0.2
+      hasown: 2.0.4
     dev: true
 
-  /es-to-primitive@1.3.0:
-    resolution: {integrity: sha512-w+5mJ3GuFL+NjVtJlvydShqE1eN3h3PbI7/5LAsYJP/2qtuMXjfL2LpHSRqo4b4eSF5K/DH1JXKUAHSB2UW50g==}
+  /es-to-primitive@1.3.4:
+    resolution: {integrity: sha512-yPDz7wqpg1/mmHLmS3tcfTfbw5f1eryXvyghYBffGdERwe+mV7ZcWzTR8LR17Kvqt3qfPurjlonmnq3MKXIOXw==}
     engines: {node: '>= 0.4'}
     dependencies:
+      es-abstract-get: 1.0.0
+      es-define-property: 1.0.1
+      es-errors: 1.3.0
       is-callable: 1.2.7
       is-date-object: 1.1.0
       is-symbol: 1.1.1
@@ -4276,17 +3287,17 @@ packages:
     engines: {node: '>=12'}
     dev: false
 
-  /eslint-config-prettier@10.1.8(eslint@9.39.2):
+  /eslint-config-prettier@10.1.8(eslint@9.39.5):
     resolution: {integrity: sha512-82GZUjRS0p/jganf6q1rEO25VSoHH0hKPCTrgillPjdI/3bgBhAE1QzHrHTizjpRvy6pGAvKjDJtk2pF9NDq8w==}
     hasBin: true
     peerDependencies:
       eslint: '>=7.0.0'
     dependencies:
-      eslint: 9.39.2
+      eslint: 9.39.5
     dev: true
 
-  /eslint-plugin-prettier@5.5.5(eslint-config-prettier@10.1.8)(eslint@9.39.2)(prettier@3.8.1):
-    resolution: {integrity: sha512-hscXkbqUZ2sPithAuLm5MXL+Wph+U7wHngPBv9OMWwlP8iaflyxpjTYZkmdgB4/vPIhemRlBEoLrH7UC1n7aUw==}
+  /eslint-plugin-prettier@5.5.6(eslint-config-prettier@10.1.8)(eslint@9.39.5)(prettier@3.9.10):
+    resolution: {integrity: sha512-ifetmTcxWfz+4qRW3pH/ujdTq2jQIj59AxJMIN26K5avYgU8dxycUETQonWiW+wPrYXA0j3Try0l1CnwVQtDqQ==}
     engines: {node: ^14.18.0 || >=16.0.0}
     peerDependencies:
       '@types/eslint': '>=8.0.0'
@@ -4299,46 +3310,46 @@ packages:
       eslint-config-prettier:
         optional: true
     dependencies:
-      eslint: 9.39.2
-      eslint-config-prettier: 10.1.8(eslint@9.39.2)
-      prettier: 3.8.1
+      eslint: 9.39.5
+      eslint-config-prettier: 10.1.8(eslint@9.39.5)
+      prettier: 3.9.10
       prettier-linter-helpers: 1.0.1
-      synckit: 0.11.12
+      synckit: 0.11.13
     dev: true
 
-  /eslint-plugin-react-hooks@5.2.0(eslint@9.39.2):
+  /eslint-plugin-react-hooks@5.2.0(eslint@9.39.5):
     resolution: {integrity: sha512-+f15FfK64YQwZdJNELETdn5ibXEUQmW1DZL6KXhNnc2heoy/sg9VJJeT7n8TlMWouzWqSWavFkIhHyIbIAEapg==}
     engines: {node: '>=10'}
     peerDependencies:
       eslint: ^3.0.0 || ^4.0.0 || ^5.0.0 || ^6.0.0 || ^7.0.0 || ^8.0.0-0 || ^9.0.0
     dependencies:
-      eslint: 9.39.2
+      eslint: 9.39.5
     dev: true
 
-  /eslint-plugin-react@7.37.5(eslint@9.39.2):
+  /eslint-plugin-react@7.37.5(eslint@9.39.5):
     resolution: {integrity: sha512-Qteup0SqU15kdocexFNAJMvCJEfa2xUKNV4CC1xsVMrIIqEy3SQ/rqyxCWNzfrd3/ldy6HMlD2e0JDVpDg2qIA==}
     engines: {node: '>=4'}
     peerDependencies:
       eslint: ^3 || ^4 || ^5 || ^6 || ^7 || ^8 || ^9.7
     dependencies:
-      array-includes: 3.1.9
+      array-includes: 3.2.0
       array.prototype.findlast: 1.2.5
       array.prototype.flatmap: 1.3.3
       array.prototype.tosorted: 1.1.4
       doctrine: 2.1.0
-      es-iterator-helpers: 1.3.1
-      eslint: 9.39.2
+      es-iterator-helpers: 1.4.0
+      eslint: 9.39.5
       estraverse: 5.3.0
-      hasown: 2.0.2
+      hasown: 2.0.4
       jsx-ast-utils: 3.3.5
       minimatch: 3.1.3
       object.entries: 1.1.9
       object.fromentries: 2.0.8
       object.values: 1.2.1
       prop-types: 15.8.1
-      resolve: 2.0.0-next.6
+      resolve: 2.0.0-next.7
       semver: 6.3.1
-      string.prototype.matchall: 4.0.12
+      string.prototype.matchall: 4.1.0
       string.prototype.repeat: 1.0.0
     dev: true
 
@@ -4360,9 +3371,15 @@ packages:
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dev: true
 
-  /eslint@9.39.2:
-    resolution: {integrity: sha512-LEyamqS7W5HB3ujJyvi0HQK/dtVINZvd5mAAp9eT5S/ujByGjiZLCzPcHVzuXbpJDJF/cxwHlfceVUDZ2lnSTw==}
+  /eslint-visitor-keys@5.0.1:
+    resolution: {integrity: sha512-tD40eHxA35h0PEIZNeIjkHoDR4YjjJp34biM0mDvplBe//mB+IHCqHDGV7pxF+7MklTvighcCPPZC7ynWyjdTA==}
+    engines: {node: ^20.19.0 || ^22.13.0 || >=24}
+    dev: true
+
+  /eslint@9.39.5:
+    resolution: {integrity: sha512-DgZS62aPLXKlnxILS/AYCoRvHaZeXceIzlXPkkGGzJWSow1aEk0lbTlxUSlyjC8jcaKxAdOnTDz+o1JFSBsyjw==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
+    deprecated: This version is no longer supported. Please see https://eslint.org/version-support for other options.
     hasBin: true
     peerDependencies:
       jiti: '*'
@@ -4370,19 +3387,19 @@ packages:
       jiti:
         optional: true
     dependencies:
-      '@eslint-community/eslint-utils': 4.9.1(eslint@9.39.2)
+      '@eslint-community/eslint-utils': 4.10.1(eslint@9.39.5)
       '@eslint-community/regexpp': 4.12.2
-      '@eslint/config-array': 0.21.1
+      '@eslint/config-array': 0.21.2
       '@eslint/config-helpers': 0.4.2
       '@eslint/core': 0.17.0
-      '@eslint/eslintrc': 3.3.3
-      '@eslint/js': 9.39.2
+      '@eslint/eslintrc': 3.3.7
+      '@eslint/js': 9.39.5
       '@eslint/plugin-kit': 0.4.1
-      '@humanfs/node': 0.16.7
+      '@humanfs/node': 0.16.8
       '@humanwhocodes/module-importer': 1.0.1
       '@humanwhocodes/retry': 0.4.3
-      '@types/estree': 1.0.8
-      ajv: 6.12.6
+      '@types/estree': 1.0.9
+      ajv: 6.15.0
       chalk: 4.1.2
       cross-spawn: 7.0.6
       debug: 4.4.3
@@ -4401,7 +3418,7 @@ packages:
       is-glob: 4.0.3
       json-stable-stringify-without-jsonify: 1.0.1
       lodash.merge: 4.6.2
-      minimatch: 3.1.3
+      minimatch: 3.1.5
       natural-compare: 1.4.0
       optionator: 0.9.4
     transitivePeerDependencies:
@@ -4412,8 +3429,8 @@ packages:
     resolution: {integrity: sha512-j6PAQ2uUr79PZhBjP5C5fhl8e39FmRnOjsD5lGnWrFU8i2G776tBK7+nP8KuQUTTyAZUwfQqXAgrVH5MbH9CYQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     dependencies:
-      acorn: 8.15.0
-      acorn-jsx: 5.3.2(acorn@8.15.0)
+      acorn: 8.19.0
+      acorn-jsx: 5.3.2(acorn@8.19.0)
       eslint-visitor-keys: 4.2.1
     dev: true
 
@@ -4480,8 +3497,8 @@ packages:
     resolution: {integrity: sha512-VxPP4NqbUjj6MaAOafWeUn2cXWLcCtljklUtZf0Ind4XQ+QPtmA0b18zZy0jIQx+ExRVCR/ZQpBmik5lXshNsw==}
     dev: true
 
-  /fast-equals@5.4.0:
-    resolution: {integrity: sha512-jt2DW/aNFNwke7AUd+Z+e6pz39KO5rzdbbFCg2sGafS4mk13MI7Z8O5z9cADNn5lhGODIgLwug6TZO2ctf7kcw==}
+  /fast-equals@5.4.4:
+    resolution: {integrity: sha512-d9/I4rfMfUlO9HOkQqLJgfyRM4siqk1SF3Gjxm4OAMZzV+kCKHe1t/bHxn3xtCPHs3HCQFcvqzkTuanzqECQxw==}
     engines: {node: '>=6.0.0'}
     dev: false
 
@@ -4499,7 +3516,7 @@ packages:
       format: 0.2.2
     dev: false
 
-  /fdir@6.5.0(picomatch@4.0.4):
+  /fdir@6.5.0(picomatch@4.0.7):
     resolution: {integrity: sha512-tIbYtZbucOs0BRGqPJkshJUYdL+SDH7dVM8gjy+ERp3WAUjLEFJE+02kanyHtwjWOnwrKYBiwAmM0p4kLJAnXg==}
     engines: {node: '>=12.0.0'}
     peerDependencies:
@@ -4508,7 +3525,7 @@ packages:
       picomatch:
         optional: true
     dependencies:
-      picomatch: 4.0.4
+      picomatch: 4.0.7
 
   /file-entry-cache@8.0.0:
     resolution: {integrity: sha512-XXTUwCvisa5oacNGRP9SfNtYBNAMi+RPwBFmblZEF7N7swHYQS6/Zfk7SRwx4D5j3CH211YNRco1DEMNVfZCnQ==}
@@ -4544,16 +3561,16 @@ packages:
     resolution: {integrity: sha512-f7ccFPK3SXFHpx15UIGyRJ/FJQctuKZ0zVuN3frBo4HnK3cay9VEW0R6yPYFHC0AgqhukPzKjq22t5DmAyqGyw==}
     engines: {node: '>=16'}
     dependencies:
-      flatted: 3.4.2
+      flatted: 3.4.4
       keyv: 4.5.4
     dev: true
 
-  /flatted@3.4.2:
-    resolution: {integrity: sha512-PjDse7RzhcPkIJwy5t7KPWQSZ9cAbzQXcafsetQoD7sOJRQlGikNbx7yZp2OotDnJyrDcbyRq3Ttb18iYOqkxA==}
+  /flatted@3.4.4:
+    resolution: {integrity: sha512-5+ybhBZANEJxaH3X5evAFatUxLfEHSr7n6kYJ+1Qd0mUqr4eu9gIf6GDbWHf8RJijHrjjO8G+la14SlL2SeS1Q==}
     dev: true
 
-  /follow-redirects@1.16.0:
-    resolution: {integrity: sha512-y5rN/uOsadFT/JfYwhxRS5R7Qce+g3zG97+JrtFZlC9klX/W5hD7iiLzScI4nZqUS7DNUdhPgw4xI8W2LuXlUw==}
+  /follow-redirects@1.16.1:
+    resolution: {integrity: sha512-FNvFGzoMLWmE6Yj9spb/zjd7yiNCHiAW9/Tg9CXrQ8wuu32HtlJOwWO11OJafl5FfY3DxTdQ0vj42zU1kvv5jg==}
     engines: {node: '>=4.0'}
     peerDependencies:
       debug: '*'
@@ -4585,14 +3602,6 @@ packages:
     engines: {node: '>=0.4.x'}
     dev: false
 
-  /fsevents@2.3.2:
-    resolution: {integrity: sha512-xiqMQR4xAeHTuB9uWm+fFRcIOgKBMiOBP+eXiyT7jsgVCq1bkVygt00oASowB7EdtpOHaaPgKt812P9ab+DDKA==}
-    engines: {node: ^8.16.0 || ^10.6.0 || >=11.0.0}
-    os: [darwin]
-    requiresBuild: true
-    dev: true
-    optional: true
-
   /fsevents@2.3.3:
     resolution: {integrity: sha512-5xoDfX+fL7faATnagmWPpbFtwh/R77WmMMqqHGS65C3vvB0YHrgF+B1YmZ3441tMj5n63k0212XNoJwzlhffQw==}
     engines: {node: ^8.16.0 || ^10.6.0 || >=11.0.0}
@@ -4604,16 +3613,19 @@ packages:
   /function-bind@1.1.2:
     resolution: {integrity: sha512-7XHNxH7qX9xG5mIwxkhumTox/MIRNcOgDrxWsMt2pAr23WHp6MrRlN7FBSFpCpr+oVO0F744iUgR82nJMfG2SA==}
 
-  /function.prototype.name@1.1.8:
-    resolution: {integrity: sha512-e5iwyodOHhbMr/yNrc7fDYG4qlbIvI5gajyzPnb5TCwyhjApznQh1BMFou9b30SevY43gCJKXycoCBjMbsuW0Q==}
+  /function.prototype.name@1.2.0:
+    resolution: {integrity: sha512-jObKIik1P2QjPHP5nz5BaOtUlfgS0fWo8IUByNXkM+o+02sJOi94em77GwJKQSJ3gfPHdgzLNrHc1uokV4P/ew==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
-      define-properties: 1.2.1
+      es-define-property: 1.0.1
+      es-errors: 1.3.0
       functions-have-names: 1.2.3
-      hasown: 2.0.2
+      has-property-descriptors: 1.0.2
+      hasown: 2.0.4
       is-callable: 1.2.7
+      is-document.all: 1.0.0
     dev: true
 
   /functions-have-names@1.2.3:
@@ -4630,8 +3642,8 @@ packages:
     engines: {node: 6.* || 8.* || >= 10.*}
     dev: false
 
-  /get-east-asian-width@1.4.0:
-    resolution: {integrity: sha512-QZjmEOC+IT1uk6Rx0sX22V6uHWVwbdbxf1faPqJ1QhLdGgsRGCZoyaQBm/piRdJy/D2um6hM1UP7ZEeQ4EkP+Q==}
+  /get-east-asian-width@1.7.0:
+    resolution: {integrity: sha512-XjH1AECxf0giL2V1aU8vKyRR2ppRUb5c0EvT7zuJTokQ74bNo52zOtghqdWIqrhUD79fo3x0WfKZdOqxF6LG1Q==}
     engines: {node: '>=18'}
     dev: true
 
@@ -4642,12 +3654,12 @@ packages:
       call-bind-apply-helpers: 1.0.2
       es-define-property: 1.0.1
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       function-bind: 1.1.2
       get-proto: 1.0.1
       gopd: 1.2.0
       has-symbols: 1.1.0
-      hasown: 2.0.2
+      hasown: 2.0.4
       math-intrinsics: 1.1.0
 
   /get-nonce@1.0.1:
@@ -4660,7 +3672,7 @@ packages:
     engines: {node: '>= 0.4'}
     dependencies:
       dunder-proto: 1.0.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
 
   /get-stream@8.0.1:
     resolution: {integrity: sha512-VaUJspBffn/LMCJVoMvSAdmscJyS1auj5Zulnn5UoYcY531UWmdwhRWkcGKnGU93m5HSXP9LP2usOryrBtQowA==}
@@ -4741,27 +3753,20 @@ packages:
     dependencies:
       has-symbols: 1.1.0
 
-  /hasown@2.0.2:
-    resolution: {integrity: sha512-0hJU9SCPvmMzIBdZFqNPXWa6dqh7WdH0cII9y+CyS8rG3nL48Bclra9HmKhVVUHyPWNH5Y7xDwAB7bfgSjkUMQ==}
-    engines: {node: '>= 0.4'}
-    dependencies:
-      function-bind: 1.1.2
-
   /hasown@2.0.4:
     resolution: {integrity: sha512-T2UbfbBEF32wiepXIsMlTW9+dDYC6wMh/t/vYA4tuOMKqWz/n3vr1NFSxQiyP+zk2mXsoMA/i/7qV6LKut1t1A==}
     engines: {node: '>= 0.4'}
     dependencies:
       function-bind: 1.1.2
-    dev: false
 
   /hast-util-from-parse5@8.0.3:
     resolution: {integrity: sha512-3kxEVkEKt0zvcZ3hCRYI8rqrgwtlIOFMWkbclACvjlDw8Li9S2hk/d51OI0nr/gIpdMHNepwgOKqZ/sy0Clpyg==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/unist': 3.0.3
       devlop: 1.1.0
       hastscript: 9.0.1
-      property-information: 7.1.0
+      property-information: 7.2.0
       vfile: 6.0.3
       vfile-location: 5.0.3
       web-namespaces: 2.0.1
@@ -4770,27 +3775,27 @@ packages:
   /hast-util-heading-rank@3.0.0:
     resolution: {integrity: sha512-EJKb8oMUXVHcWZTDepnr+WNbfnXKFNf9duMesmr4S8SXTJBJ9M4Yok08pu9vxdJwdlGRhVumk9mEhkEvKGifwA==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
     dev: false
 
   /hast-util-is-element@3.0.0:
     resolution: {integrity: sha512-Val9mnv2IWpLbNPqc/pUem+a7Ipj2aHacCwgNfTiK0vJKl0LF+4Ba4+v1oPHFpf3bLYmreq0/l3Gud9S5OH42g==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
     dev: false
 
   /hast-util-parse-selector@4.0.0:
     resolution: {integrity: sha512-wkQCkSYoOGCRKERFWcxMVMOcYE2K1AaNLU8DXS9arxnLOUEWbOXKXiJUNzEpqZ3JOKpnha3jkFrumEjVliDe7A==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
     dev: false
 
   /hast-util-raw@9.1.0:
     resolution: {integrity: sha512-Y8/SBAHkZGoNkpzqqfCldijcuUKh7/su31kEBp67cFY09Wy0mTRgtsLYsiIxMJxlu0f6AA5SUTbDR8K0rxnbUw==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/unist': 3.0.3
-      '@ungap/structured-clone': 1.3.0
+      '@ungap/structured-clone': 1.4.0
       hast-util-from-parse5: 8.0.3
       hast-util-to-parse5: 8.0.1
       html-void-elements: 3.0.0
@@ -4806,16 +3811,16 @@ packages:
   /hast-util-sanitize@5.0.2:
     resolution: {integrity: sha512-3yTWghByc50aGS7JlGhk61SPenfE/p1oaFeNwkOOyrscaOkMGrcW9+Cy/QAIOBpZxP1yqDIzFMR0+Np0i0+usg==}
     dependencies:
-      '@types/hast': 3.0.4
-      '@ungap/structured-clone': 1.3.0
+      '@types/hast': 3.0.5
+      '@ungap/structured-clone': 1.4.0
       unist-util-position: 5.0.0
     dev: false
 
   /hast-util-to-jsx-runtime@2.3.6:
     resolution: {integrity: sha512-zl6s8LwNyo1P9uw+XJGvZtdFF1GdAkOg8ujOw+4Pyb76874fLps4ueHXDhXWdk6YHQ6OgUtinliG7RsYvCbbBg==}
     dependencies:
-      '@types/estree': 1.0.8
-      '@types/hast': 3.0.4
+      '@types/estree': 1.0.9
+      '@types/hast': 3.0.5
       '@types/unist': 3.0.3
       comma-separated-tokens: 2.0.3
       devlop: 1.1.0
@@ -4824,7 +3829,7 @@ packages:
       mdast-util-mdx-expression: 2.0.1
       mdast-util-mdx-jsx: 3.2.0
       mdast-util-mdxjs-esm: 2.0.1
-      property-information: 7.1.0
+      property-information: 7.2.0
       space-separated-tokens: 2.0.2
       style-to-js: 1.1.21
       unist-util-position: 5.0.0
@@ -4836,10 +3841,10 @@ packages:
   /hast-util-to-parse5@8.0.1:
     resolution: {integrity: sha512-MlWT6Pjt4CG9lFCjiz4BH7l9wmrMkfkJYCxFwKQic8+RTZgWPuWxwAfjJElsXkex7DJjfSJsQIt931ilUgmwdA==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       comma-separated-tokens: 2.0.3
       devlop: 1.1.0
-      property-information: 7.1.0
+      property-information: 7.2.0
       space-separated-tokens: 2.0.2
       web-namespaces: 2.0.1
       zwitch: 2.0.4
@@ -4848,13 +3853,13 @@ packages:
   /hast-util-to-string@3.0.1:
     resolution: {integrity: sha512-XelQVTDWvqcl3axRfI0xSeoVKzyIFPwsAGSLIsKdJKQMXDYJS4WYrBNF/8J7RdhIcFI2BOHgAifggsvsxp/3+A==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
     dev: false
 
   /hast-util-to-text@4.0.2:
     resolution: {integrity: sha512-KK6y/BN8lbaq654j7JgBydev7wuNMcID54lkRav1P0CaE1e47P72AWWPiGKXTJU271ooYzcvTAn/Zt0REnvc7A==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/unist': 3.0.3
       hast-util-is-element: 3.0.0
       unist-util-find-after: 5.0.0
@@ -4863,16 +3868,16 @@ packages:
   /hast-util-whitespace@3.0.0:
     resolution: {integrity: sha512-88JUN06ipLwsnv+dVn+OIYOvAuvBMy/Qoi6O7mQHxdPXpjy+Cd6xRkWwux7DKO+4sYILtLBRIKgsdpS2gQc7qw==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
     dev: false
 
   /hastscript@9.0.1:
     resolution: {integrity: sha512-g7df9rMFX/SPi34tyGCyUBREQoKkapwdY/T04Qn9TDWfHhAYt4/I0gMVirzK5wEzeUqIjEB+LXC/ypb7Aqno5w==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       comma-separated-tokens: 2.0.3
       hast-util-parse-selector: 4.0.0
-      property-information: 7.1.0
+      property-information: 7.2.0
       space-separated-tokens: 2.0.2
     dev: false
 
@@ -4880,8 +3885,13 @@ packages:
     resolution: {integrity: sha512-tzcUFauisWKNHaRkN4Wjl/ZA07gENAjFl3J/c480dprkGTg5EQstgaNFqBfUqCq54kZRIEcreTsAgF/m2quD7A==}
     dev: false
 
-  /highlight.js@11.11.1:
-    resolution: {integrity: sha512-Xwwo44whKBVCYoliBQwaPvtd/2tYFkRQtXDWj1nackaV2JPXx3L0+Jvd8/qCJ2p+ML0/XVkJ2q+Mr+UVdpJK5w==}
+  /highlight.js@11.11.2:
+    resolution: {integrity: sha512-oaXMACAU0kzOMXBjWpNcX+vlwSBCIAiZ9BHa7gA15NOTtT2L/l8OSZDuqS2XppOhZBPJ7hm4o8ep2kyuip2uEQ==}
+    engines: {node: '>=12.0.0'}
+    dev: false
+
+  /highlight.js@11.12.0:
+    resolution: {integrity: sha512-nbfWpyRMcMrPMmDwJB+dhX/eiaPKtc2RB+0QZskqJ3WjRA/FDS0e9hZrx8EC/lbEv8gXy98FcDbNa/dspAaJMg==}
     engines: {node: '>=12.0.0'}
     dev: false
 
@@ -4889,8 +3899,8 @@ packages:
     resolution: {integrity: sha512-PDEfEF102G23vHmPhLyPboFCD+BkMGu+GuJe2d9/eH4FsCwvgBpnc9n0pGE+ffKdph38s6foEZiEjdgHdzp+IA==}
     dev: false
 
-  /html-parse-stringify@3.0.1:
-    resolution: {integrity: sha512-KknJ50kTInJ7qIScF3jeaFRpMpE8/lfiTdzf/twXyPBLAGrLRTmkz3AdTnKeh40X8k9L2fdYwEp/42WGXIRGcg==}
+  /html-parse-stringify@3.1.0:
+    resolution: {integrity: sha512-E0oAXcELOtsXe+BmpJ2EZyedbldPpriV5vICzEuo6xjC/D1lDukOI7KrpfQGF2Qc4wWEy0nk3bFORS2K5ZAhFQ==}
     dependencies:
       void-elements: 3.1.0
     dev: false
@@ -4918,21 +3928,21 @@ packages:
     engines: {node: '>=16.17.0'}
     dev: true
 
-  /i18next-browser-languagedetector@8.2.0:
-    resolution: {integrity: sha512-P+3zEKLnOF0qmiesW383vsLdtQVyKtCNA9cjSoKCppTKPQVfKd2W8hbVo5ZhNJKDqeM7BOcvNoKJOjpHh4Js9g==}
+  /i18next-browser-languagedetector@8.2.1:
+    resolution: {integrity: sha512-bZg8+4bdmaOiApD7N7BPT9W8MLZG+nPTOFlLiJiT8uzKXFjhxw4v2ierCXOwB5sFDMtuA5G4kgYZ0AznZxQ/cw==}
     dependencies:
-      '@babel/runtime': 7.28.6
+      '@babel/runtime': 7.29.10
     dev: false
 
-  /i18next@25.8.0(typescript@5.9.3):
-    resolution: {integrity: sha512-urrg4HMFFMQZ2bbKRK7IZ8/CTE7D8H4JRlAwqA2ZwDRFfdd0K/4cdbNNLgfn9mo+I/h9wJu61qJzH7jCFAhUZQ==}
+  /i18next@25.10.10(typescript@5.9.3):
+    resolution: {integrity: sha512-cqUW2Z3EkRx7NqSyywjkgCLK7KLCL6IFVFcONG7nVYIJ3ekZ1/N5jUsihHV6Bq37NfhgtczxJcxduELtjTwkuQ==}
     peerDependencies:
-      typescript: ^5
+      typescript: ^5 || ^6
     peerDependenciesMeta:
       typescript:
         optional: true
     dependencies:
-      '@babel/runtime': 7.28.6
+      '@babel/runtime': 7.29.10
       typescript: 5.9.3
     dev: false
 
@@ -4941,8 +3951,8 @@ packages:
     engines: {node: '>= 4'}
     dev: true
 
-  /ignore@7.0.5:
-    resolution: {integrity: sha512-Hs59xBNfUIunMFgWAbGX5cq6893IbWg4KnrjbYwX3tx0ztorVgTDA6B2sxf8ejHJ4wz8BqGUMYlnzNBer5NvGg==}
+  /ignore@7.0.12:
+    resolution: {integrity: sha512-/8UvqAPU9DGTI9k4mxtf49U37Isfwr8Uts96+SBHIkFxPJnHS0Ew4f00sM4Scd8V8EjM0jUNtVDdv1kPdt35lg==}
     engines: {node: '>= 4'}
     dev: true
 
@@ -4963,8 +3973,8 @@ packages:
     resolution: {integrity: sha512-Nb2ctOyNR8DqQoR0OwRG95uNWIC0C1lCgf5Naz5H6Ji72KZ8OcFZLz2P5sNgwlyoJ8Yif11oMuYs5pBQa86csA==}
     dev: false
 
-  /input-otp@1.4.2(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-l3jWwYNvrEa6NTCt7BECfCm48GvwuZzkoeG3gBL2w4CHeOXW3eKFmf9UNYkNfYc3mxMrthMnxjIE07MT0zLBQA==}
+  /input-otp@1.5.0(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-3AcfdW1sNG0FmSA5hHMBXG7jNW5CdcdKs8ln8JOmn6S2SRkoXoJx+2UwC+ZvjflfnOdHJx0SfQ9/YjlxVYLtGQ==}
     peerDependencies:
       react: ^16.8 || ^17.0 || ^18.0 || ^19.0.0 || ^19.0.0-rc
       react-dom: ^16.8 || ^17.0 || ^18.0 || ^19.0.0 || ^19.0.0-rc
@@ -4978,8 +3988,8 @@ packages:
     engines: {node: '>= 0.4'}
     dependencies:
       es-errors: 1.3.0
-      hasown: 2.0.2
-      side-channel: 1.1.0
+      hasown: 2.0.4
+      side-channel: 1.1.1
     dev: true
 
   /internmap@2.0.3:
@@ -5002,7 +4012,7 @@ packages:
     resolution: {integrity: sha512-DDfANUiiG2wC1qawP66qlTugJeL5HyzMpfr8lLK+jMQirGzNod0B12cFB/9q838Ru27sBwfw78/rdoU7RERz6A==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       get-intrinsic: 1.3.0
     dev: true
@@ -5038,11 +4048,11 @@ packages:
     engines: {node: '>= 0.4'}
     dev: true
 
-  /is-core-module@2.16.1:
-    resolution: {integrity: sha512-UfoeMA6fIJ8wTYFEUjelnaGI67v6+N7qXJEvQuIGa99l4xsCruSYOVSQ0uPANn4dAzm8lkYPaKLrrijLq7x23w==}
+  /is-core-module@2.17.0:
+    resolution: {integrity: sha512-J/vG0zBCbIKOQFfufSwyXdMrsohyJIUNkrnmo6WZGzoM7tr/lsbfW5b2BvisL6zsyMzK9UxV9L6c7AoFbyXHOA==}
     engines: {node: '>= 0.4'}
     dependencies:
-      hasown: 2.0.2
+      hasown: 2.0.4
     dev: true
 
   /is-data-view@1.0.2:
@@ -5066,6 +4076,13 @@ packages:
     resolution: {integrity: sha512-AAB9hiomQs5DXWcRB1rqsxGUstbRroFOPPVAomNk/3XHR5JyEZChOyTWe2oayKnsSsr/kcGqF+z6yuH6HHpN0A==}
     dev: false
 
+  /is-document.all@1.0.0:
+    resolution: {integrity: sha512-+XSoyS05OdBbhFuELhgTCpFNHkpBOJqtsZfUFFpe5QTw+9Sjbh8zitxhQkYAo6wV7e1Vb8cAPvpCk9jGam/82g==}
+    engines: {node: '>= 0.4'}
+    dependencies:
+      call-bound: 1.0.4
+    dev: true
+
   /is-extglob@2.1.1:
     resolution: {integrity: sha512-SbKbANkN603Vi4jEZv49LeVJMn4yGwsbzZworEoyEiutsN3nJYdbO36zfhGJ6QEDpOZIFkDtnq5JRxmvl3jsoQ==}
     engines: {node: '>=0.10.0'}
@@ -5092,7 +4109,7 @@ packages:
     resolution: {integrity: sha512-5XHYaSyiqADb4RnZ1Bdad6cPp8Toise4TzEjcOYDHZkTCbKgiUl7WTUCpNWHuxmDt91wnsZBc9xinNzopv3JMQ==}
     engines: {node: '>=18'}
     dependencies:
-      get-east-asian-width: 1.4.0
+      get-east-asian-width: 1.7.0
     dev: true
 
   /is-generator-function@1.1.2:
@@ -5152,7 +4169,7 @@ packages:
       call-bound: 1.0.4
       gopd: 1.2.0
       has-tostringtag: 1.0.2
-      hasown: 2.0.2
+      hasown: 2.0.4
     dev: true
 
   /is-set@2.0.3:
@@ -5193,7 +4210,7 @@ packages:
     resolution: {integrity: sha512-p3EcsicXjit7SaskXHs1hA91QxgTw46Fv6EFKKGS5DRFLD8yKnohjF3hxoju94b/OcMZoQukzpPpBE9uLVKzgQ==}
     engines: {node: '>= 0.4'}
     dependencies:
-      which-typed-array: 1.1.20
+      which-typed-array: 1.1.24
     dev: true
 
   /is-weakmap@2.0.2:
@@ -5229,23 +4246,23 @@ packages:
     engines: {node: '>= 0.4'}
     dependencies:
       define-data-property: 1.1.4
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       get-intrinsic: 1.3.0
       get-proto: 1.0.1
       has-symbols: 1.1.0
       set-function-name: 2.0.2
     dev: true
 
-  /jiti@2.6.1:
-    resolution: {integrity: sha512-ekilCSN1jwRvIbgeg/57YFh8qQDNbwDb9xT/qu2DAHbFFZUicIl4ygVaAvzveMhMVr3LnpSKTNnwt8PoOfmKhQ==}
+  /jiti@2.7.0:
+    resolution: {integrity: sha512-AC/7JofJvZGrrneWNaEnJeOLUx+JlGt7tNa0wZiRPT4MY1wmfKjt2+6O2p2uz2+skll8OZZmJMNqeke7kKbNgQ==}
     hasBin: true
     dev: false
 
   /js-tokens@4.0.0:
     resolution: {integrity: sha512-RdJUflcE3cUzKiMqQgsCu06FPu9UdIJO0beYbPhHN4k6apgJtifcoCtT9bcxOpYBtpD2kCM6Sbzg4CausW/PKQ==}
 
-  /js-yaml@4.2.0:
-    resolution: {integrity: sha512-ePWsvanv0DWuDRsW8dnt+R4jQ31SCRCQ7hhNcPXZPsoBZiemuZNYGf7adZdqX2D86j6rvKp3RpCxVTSb8WQlOw==}
+  /js-yaml@4.3.2:
+    resolution: {integrity: sha512-SFNOvSJ+Dgf/9An904Yx+CgSlIPCkIpao4qo51lpee25TIRejdH3rhR4EZMGoNx3/TP3O+wzWuiTFl4sqbltzA==}
     hasBin: true
     dependencies:
       argparse: 2.0.1
@@ -5267,7 +4284,7 @@ packages:
     resolution: {integrity: sha512-ZZow9HBI5O6EPgSJLUb8n2NKgmVWTwCvHGwFuJlMjvLFqlGG6pjirPhtdsseaLZjSibD8eegzmYpUZwoIlj2cQ==}
     engines: {node: '>=4.0'}
     dependencies:
-      array-includes: 3.1.9
+      array-includes: 3.2.0
       array.prototype.flat: 1.3.3
       object.assign: 4.1.7
       object.values: 1.2.1
@@ -5287,15 +4304,6 @@ packages:
       type-check: 0.4.0
     dev: true
 
-  /lightningcss-android-arm64@1.30.2:
-    resolution: {integrity: sha512-BH9sEdOCahSgmkVhBLeU7Hc9DWeZ1Eb6wNS6Da8igvUwAe0sqROHddIlvU06q3WyXVEOYDZ6ykBZQnjTbmo4+A==}
-    engines: {node: '>= 12.0.0'}
-    cpu: [arm64]
-    os: [android]
-    requiresBuild: true
-    dev: false
-    optional: true
-
   /lightningcss-android-arm64@1.32.0:
     resolution: {integrity: sha512-YK7/ClTt4kAK0vo6w3X+Pnm0D2cf2vPHbhOXdoNti1Ga0al1P4TBZhwjATvjNwLEBCnKvjJc2jQgHXH0NEwlAg==}
     engines: {node: '>= 12.0.0'}
@@ -5305,11 +4313,11 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-darwin-arm64@1.30.2:
-    resolution: {integrity: sha512-ylTcDJBN3Hp21TdhRT5zBOIi73P6/W0qwvlFEk22fkdXchtNTOU4Qc37SkzV+EKYxLouZ6M4LG9NfZ1qkhhBWA==}
+  /lightningcss-android-arm64@1.33.0:
+    resolution: {integrity: sha512-gEpRTalKdosp4Bb8qWtc2iOgE5SeIHlpS1up9bFq2wAyYhl1UdTObYiHe98zEM9SQvSoqQZ1IQD0JNpg3Ml5pg==}
     engines: {node: '>= 12.0.0'}
     cpu: [arm64]
-    os: [darwin]
+    os: [android]
     requiresBuild: true
     dev: false
     optional: true
@@ -5323,10 +4331,10 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-darwin-x64@1.30.2:
-    resolution: {integrity: sha512-oBZgKchomuDYxr7ilwLcyms6BCyLn0z8J0+ZZmfpjwg9fRVZIR5/GMXd7r9RH94iDhld3UmSjBM6nXWM2TfZTQ==}
+  /lightningcss-darwin-arm64@1.33.0:
+    resolution: {integrity: sha512-Sciaz8eenNTKn9b3t7+xr0ipTp9YxKQY4npwQ3mrRuL0BAVHBLyZxofhaKBAVtzmtRZ/zTyo0/to4B1uWG/Djg==}
     engines: {node: '>= 12.0.0'}
-    cpu: [x64]
+    cpu: [arm64]
     os: [darwin]
     requiresBuild: true
     dev: false
@@ -5341,11 +4349,11 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-freebsd-x64@1.30.2:
-    resolution: {integrity: sha512-c2bH6xTrf4BDpK8MoGG4Bd6zAMZDAXS569UxCAGcA7IKbHNMlhGQ89eRmvpIUGfKWNVdbhSbkQaWhEoMGmGslA==}
+  /lightningcss-darwin-x64@1.33.0:
+    resolution: {integrity: sha512-Z5UPAxzrjlWNNyGy6i65cJzzvgJ5D3T6wMvs+gWpY9d7qRhANrxqAp6LhxIgZhWEw18RfJTGcRxjuLIBr+m8XQ==}
     engines: {node: '>= 12.0.0'}
     cpu: [x64]
-    os: [freebsd]
+    os: [darwin]
     requiresBuild: true
     dev: false
     optional: true
@@ -5359,11 +4367,11 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-linux-arm-gnueabihf@1.30.2:
-    resolution: {integrity: sha512-eVdpxh4wYcm0PofJIZVuYuLiqBIakQ9uFZmipf6LF/HRj5Bgm0eb3qL/mr1smyXIS1twwOxNWndd8z0E374hiA==}
+  /lightningcss-freebsd-x64@1.33.0:
+    resolution: {integrity: sha512-QQM/Ti/hQajJwCY+RiWuCZ9sdtI/XQk7nDK5vC8kkdwixezOlDgvDx7+RT+QjK6FcFT4MpsuoBnHIo/O3StRRg==}
     engines: {node: '>= 12.0.0'}
-    cpu: [arm]
-    os: [linux]
+    cpu: [x64]
+    os: [freebsd]
     requiresBuild: true
     dev: false
     optional: true
@@ -5377,12 +4385,11 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-linux-arm64-gnu@1.30.2:
-    resolution: {integrity: sha512-UK65WJAbwIJbiBFXpxrbTNArtfuznvxAJw4Q2ZGlU8kPeDIWEX1dg3rn2veBVUylA2Ezg89ktszWbaQnxD/e3A==}
+  /lightningcss-linux-arm-gnueabihf@1.33.0:
+    resolution: {integrity: sha512-N7FVBe6iS24MlM6R/4RBTxGhQheZGs7tiQ9U32UtF75NzP5Q7xWPRqLBCKxlRQRk3rY1jCIPLzx7WzOhuUIRLQ==}
     engines: {node: '>= 12.0.0'}
-    cpu: [arm64]
+    cpu: [arm]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
@@ -5392,17 +4399,15 @@ packages:
     engines: {node: '>= 12.0.0'}
     cpu: [arm64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /lightningcss-linux-arm64-musl@1.30.2:
-    resolution: {integrity: sha512-5Vh9dGeblpTxWHpOx8iauV02popZDsCYMPIgiuw97OJ5uaDsL86cnqSFs5LZkG3ghHoX5isLgWzMs+eD1YzrnA==}
+  /lightningcss-linux-arm64-gnu@1.33.0:
+    resolution: {integrity: sha512-j2v/itmy4HlNxlc6voKXYgBqNi0Ng2LShg4z7GufpEgs05P+2suBVyi9I6YHq5uoVFx9ETin3eCEhLVyXGQnKg==}
     engines: {node: '>= 12.0.0'}
     cpu: [arm64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
@@ -5412,17 +4417,15 @@ packages:
     engines: {node: '>= 12.0.0'}
     cpu: [arm64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /lightningcss-linux-x64-gnu@1.30.2:
-    resolution: {integrity: sha512-Cfd46gdmj1vQ+lR6VRTTadNHu6ALuw2pKR9lYq4FnhvgBc4zWY1EtZcAc6EffShbb1MFrIPfLDXD6Xprbnni4w==}
+  /lightningcss-linux-arm64-musl@1.33.0:
+    resolution: {integrity: sha512-yiO5ROMuYQgXbC60yjZU5CYSFZGKXL0HFATXt9mHJn1+zW55oCtMI9NfcVhYLMFDL7gV7oBPon/EmMMGg2OvtQ==}
     engines: {node: '>= 12.0.0'}
-    cpu: [x64]
+    cpu: [arm64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
@@ -5432,17 +4435,15 @@ packages:
     engines: {node: '>= 12.0.0'}
     cpu: [x64]
     os: [linux]
-    libc: [glibc]
     requiresBuild: true
     dev: false
     optional: true
 
-  /lightningcss-linux-x64-musl@1.30.2:
-    resolution: {integrity: sha512-XJaLUUFXb6/QG2lGIW6aIk6jKdtjtcffUT0NKvIqhSBY3hh9Ch+1LCeH80dR9q9LBjG3ewbDjnumefsLsP6aiA==}
+  /lightningcss-linux-x64-gnu@1.33.0:
+    resolution: {integrity: sha512-ar+Ju7LmcN0Jo4FpL4hpFybwNG9/3A/Br5KW2n2jyODg3MEZXaDYADdemoNS+BDNfMgKvylJLj4S5tyRActuAg==}
     engines: {node: '>= 12.0.0'}
     cpu: [x64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
@@ -5452,16 +4453,15 @@ packages:
     engines: {node: '>= 12.0.0'}
     cpu: [x64]
     os: [linux]
-    libc: [musl]
     requiresBuild: true
     dev: false
     optional: true
 
-  /lightningcss-win32-arm64-msvc@1.30.2:
-    resolution: {integrity: sha512-FZn+vaj7zLv//D/192WFFVA0RgHawIcHqLX9xuWiQt7P0PtdFEVaxgF9rjM/IRYHQXNnk61/H/gb2Ei+kUQ4xQ==}
+  /lightningcss-linux-x64-musl@1.33.0:
+    resolution: {integrity: sha512-RYiYbkokw0trfKqqzfF55lginwEPrD3OJDfTuJzFs1MK6iFnDenaz1fqLLtX4ITG3OktJQXOeTaw1awrBAlZPw==}
     engines: {node: '>= 12.0.0'}
-    cpu: [arm64]
-    os: [win32]
+    cpu: [x64]
+    os: [linux]
     requiresBuild: true
     dev: false
     optional: true
@@ -5475,10 +4475,10 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss-win32-x64-msvc@1.30.2:
-    resolution: {integrity: sha512-5g1yc73p+iAkid5phb4oVFMB45417DkRevRbt/El/gKXJk4jid+vPFF/AXbxn05Aky8PapwzZrdJShv5C0avjw==}
+  /lightningcss-win32-arm64-msvc@1.33.0:
+    resolution: {integrity: sha512-1K+MPfLSFVpphzpdbfkhlWk6wBrTObBzS2T6db10PNOZgR9GoVsAWzwNyuhUYYbTp23j+4RrncfujZ4uAzXvwA==}
     engines: {node: '>= 12.0.0'}
-    cpu: [x64]
+    cpu: [arm64]
     os: [win32]
     requiresBuild: true
     dev: false
@@ -5493,24 +4493,14 @@ packages:
     dev: false
     optional: true
 
-  /lightningcss@1.30.2:
-    resolution: {integrity: sha512-utfs7Pr5uJyyvDETitgsaqSyjCb2qNRAtuqUeWIAKztsOYdcACf2KtARYXg2pSvhkt+9NfoaNY7fxjl6nuMjIQ==}
+  /lightningcss-win32-x64-msvc@1.33.0:
+    resolution: {integrity: sha512-OlEICDx/Xl0FqSp4bry8zFnCvGpig3Gl4gCquvYwHuqJKEC1+n9NgDniFvqHGmMv1ZkqDJrDqKKSykTDX+ehuA==}
     engines: {node: '>= 12.0.0'}
-    dependencies:
-      detect-libc: 2.1.2
-    optionalDependencies:
-      lightningcss-android-arm64: 1.30.2
-      lightningcss-darwin-arm64: 1.30.2
-      lightningcss-darwin-x64: 1.30.2
-      lightningcss-freebsd-x64: 1.30.2
-      lightningcss-linux-arm-gnueabihf: 1.30.2
-      lightningcss-linux-arm64-gnu: 1.30.2
-      lightningcss-linux-arm64-musl: 1.30.2
-      lightningcss-linux-x64-gnu: 1.30.2
-      lightningcss-linux-x64-musl: 1.30.2
-      lightningcss-win32-arm64-msvc: 1.30.2
-      lightningcss-win32-x64-msvc: 1.30.2
+    cpu: [x64]
+    os: [win32]
+    requiresBuild: true
     dev: false
+    optional: true
 
   /lightningcss@1.32.0:
     resolution: {integrity: sha512-NXYBzinNrblfraPGyrbPoD19C1h9lfI/1mzgWYvXUTe414Gz/X1FD2XBZSZM7rRTrMA8JL3OtAaGifrIKhQ5yQ==}
@@ -5531,6 +4521,25 @@ packages:
       lightningcss-win32-x64-msvc: 1.32.0
     dev: false
 
+  /lightningcss@1.33.0:
+    resolution: {integrity: sha512-WkUDrojuJs0xkgGf2udWxa3yGBRxPtxUkB79i6aCZLRgc7PM8fZe9TosfPDcvEpQZbuFASnHYmRLBLUbmLOIIA==}
+    engines: {node: '>= 12.0.0'}
+    dependencies:
+      detect-libc: 2.1.2
+    optionalDependencies:
+      lightningcss-android-arm64: 1.33.0
+      lightningcss-darwin-arm64: 1.33.0
+      lightningcss-darwin-x64: 1.33.0
+      lightningcss-freebsd-x64: 1.33.0
+      lightningcss-linux-arm-gnueabihf: 1.33.0
+      lightningcss-linux-arm64-gnu: 1.33.0
+      lightningcss-linux-arm64-musl: 1.33.0
+      lightningcss-linux-x64-gnu: 1.33.0
+      lightningcss-linux-x64-musl: 1.33.0
+      lightningcss-win32-arm64-msvc: 1.33.0
+      lightningcss-win32-x64-msvc: 1.33.0
+    dev: false
+
   /lilconfig@3.1.3:
     resolution: {integrity: sha512-/vlFKAoH5Cgt3Ie+JLhRbwOsCQePABiU3tJ1egGvyQ+33R/vcwM2Zl2QR/LzjsBeItPt3oSVXapn+m4nQDvpzw==}
     engines: {node: '>=14'}
@@ -5548,9 +4557,9 @@ packages:
       lilconfig: 3.1.3
       listr2: 8.3.3
       micromatch: 4.0.8
-      pidtree: 0.6.0
+      pidtree: 0.6.1
       string-argv: 0.3.2
-      yaml: 2.8.2
+      yaml: 2.9.1
     transitivePeerDependencies:
       - supports-color
     dev: true
@@ -5593,10 +4602,10 @@ packages:
     resolution: {integrity: sha512-9ie8ItPR6tjY5uYJh8K/Zrv/RMZ5VOlOWvtZdEHYSTFKZfIBPQa9tOAEeAWhd+AnIneLJ22w5fjOYtoutpWq5w==}
     engines: {node: '>=18'}
     dependencies:
-      ansi-escapes: 7.2.0
+      ansi-escapes: 7.3.0
       cli-cursor: 5.0.0
       slice-ansi: 7.1.2
-      strip-ansi: 7.1.2
+      strip-ansi: 7.2.0
       wrap-ansi: 9.0.2
     dev: true
 
@@ -5620,9 +4629,9 @@ packages:
   /lowlight@3.3.0:
     resolution: {integrity: sha512-0JNhgFoPvP6U6lE/UdVsSq99tn6DhjjpAj5MxG49ewd2mOBVtwWYIT8ClyABhq198aXXODMU6Ox8DrGy/CpTZQ==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       devlop: 1.1.0
-      highlight.js: 11.11.1
+      highlight.js: 11.11.2
     dev: false
 
   /lucide-react@0.507.0(react@19.2.1):
@@ -5636,7 +4645,7 @@ packages:
   /magic-string@0.30.21:
     resolution: {integrity: sha512-vd2F4YUyEXKGcLHoq+TEyCjxueSeHnFxyyjNp80yg0XV4vUhnDer/lvvlqM/arB5bXQN5K2/3oinyCRyx8T2CQ==}
     dependencies:
-      '@jridgewell/sourcemap-codec': 1.5.5
+      '@jridgewell/sourcemap-codec': 1.6.0
     dev: false
 
   /markdown-table@3.0.4:
@@ -5647,8 +4656,8 @@ packages:
     resolution: {integrity: sha512-/IXtbwEk5HTPyEwyKX6hGkYXxM9nbj64B+ilVJnC/R6B0pH5G4V3b0pVbL7DBj4tkhBAppbQUlf6F6Xl9LHu1g==}
     engines: {node: '>= 0.4'}
 
-  /mdast-util-find-and-replace@3.0.2:
-    resolution: {integrity: sha512-Tmd1Vg/m3Xz43afeNxDIhWRtFZgM2VLyaf4vSTYwudTyeuTneoL3qtWMA5jeLyz/O1vDJmmV4QuScFCA2tBPwg==}
+  /mdast-util-find-and-replace@3.0.3:
+    resolution: {integrity: sha512-xPgpDNl0/OXHsI7TlaIs22lWnH3KlpvZdXG1ET/m/YT2Hhdkx8lJV4kLphE6l9simyM4KPDFBwas08BgEob6Jw==}
     dependencies:
       '@types/mdast': 4.0.4
       escape-string-regexp: 5.0.0
@@ -5656,20 +4665,20 @@ packages:
       unist-util-visit-parents: 6.0.2
     dev: false
 
-  /mdast-util-from-markdown@2.0.2:
-    resolution: {integrity: sha512-uZhTV/8NBuw0WHkPTrCqDOl0zVe1BIng5ZtHoDk49ME1qqcjYmmLmOf0gELgcRMxN4w2iuIeVso5/6QymSrgmA==}
+  /mdast-util-from-markdown@2.1.0:
+    resolution: {integrity: sha512-KRLgmPaP2H1ul2bpoJkOS+9pAcDAAj9Nu1mqLnh8NLeJG5tlvj42TH80PIGBpROWlRXLk0lDrWNuZCFy/tsB4A==}
     dependencies:
       '@types/mdast': 4.0.4
       '@types/unist': 3.0.3
       decode-named-character-reference: 1.3.0
       devlop: 1.1.0
       mdast-util-to-string: 4.0.0
-      micromark: 4.0.2
+      micromark: 4.0.3
       micromark-util-decode-numeric-character-reference: 2.0.2
       micromark-util-decode-string: 2.0.1
       micromark-util-normalize-identifier: 2.0.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
       unist-util-stringify-position: 4.0.0
     transitivePeerDependencies:
       - supports-color
@@ -5681,7 +4690,7 @@ packages:
       '@types/mdast': 4.0.4
       ccount: 2.0.1
       devlop: 1.1.0
-      mdast-util-find-and-replace: 3.0.2
+      mdast-util-find-and-replace: 3.0.3
       micromark-util-character: 2.1.1
     dev: false
 
@@ -5690,19 +4699,19 @@ packages:
     dependencies:
       '@types/mdast': 4.0.4
       devlop: 1.1.0
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
       micromark-util-normalize-identifier: 2.0.1
     transitivePeerDependencies:
       - supports-color
     dev: false
 
-  /mdast-util-gfm-strikethrough@2.0.0:
-    resolution: {integrity: sha512-mKKb915TF+OC5ptj5bJ7WFRPdYtuHv0yTRxK2tJvi+BDqbkiG7h7u/9SI89nRAYcmap2xHQL9D+QG/6wSrTtXg==}
+  /mdast-util-gfm-strikethrough@2.0.1:
+    resolution: {integrity: sha512-OuJHqvr455pwu2OaOrir7dbzqs4jlWKOtlu9L6GjcIjjQwNyw02VntQsr/Ek6/A13vgnXfUhBekWRUP1OkNivw==}
     dependencies:
       '@types/mdast': 4.0.4
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5713,8 +4722,8 @@ packages:
       '@types/mdast': 4.0.4
       devlop: 1.1.0
       markdown-table: 3.0.4
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5724,8 +4733,8 @@ packages:
     dependencies:
       '@types/mdast': 4.0.4
       devlop: 1.1.0
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5733,13 +4742,13 @@ packages:
   /mdast-util-gfm@3.1.0:
     resolution: {integrity: sha512-0ulfdQOM3ysHhCJ1p06l0b0VKlhU0wuQs3thxZQagjcjPrlFRqY215uZGHHJan9GEAXd9MbfPjFJz+qMkVR6zQ==}
     dependencies:
-      mdast-util-from-markdown: 2.0.2
+      mdast-util-from-markdown: 2.1.0
       mdast-util-gfm-autolink-literal: 2.0.1
       mdast-util-gfm-footnote: 2.1.0
-      mdast-util-gfm-strikethrough: 2.0.0
+      mdast-util-gfm-strikethrough: 2.0.1
       mdast-util-gfm-table: 2.0.0
       mdast-util-gfm-task-list-item: 2.0.0
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5748,11 +4757,11 @@ packages:
     resolution: {integrity: sha512-J6f+9hUp+ldTZqKRSg7Vw5V6MqjATc+3E4gf3CFNcuZNWD8XdyI6zQ8GqH7f8169MM6P7hMBRDVGnn7oHB9kXQ==}
     dependencies:
       '@types/estree-jsx': 1.0.5
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
       devlop: 1.1.0
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5761,13 +4770,13 @@ packages:
     resolution: {integrity: sha512-lj/z8v0r6ZtsN/cGNNtemmmfoLAFZnjMbNyLzBafjzikOM+glrjNHPlf6lQDOTccj9n5b0PPihEBbhneMyGs1Q==}
     dependencies:
       '@types/estree-jsx': 1.0.5
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
       '@types/unist': 3.0.3
       ccount: 2.0.1
       devlop: 1.1.0
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
       parse-entities: 4.0.2
       stringify-entities: 4.0.4
       unist-util-stringify-position: 4.0.0
@@ -5780,11 +4789,11 @@ packages:
     resolution: {integrity: sha512-EcmOpxsZ96CvlP03NghtH1EsLtr0n9Tm4lPUJUBccV9RwUOneqSycg19n5HGzCf+10LozMRSObtVr3ee1WoHtg==}
     dependencies:
       '@types/estree-jsx': 1.0.5
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
       devlop: 1.1.0
-      mdast-util-from-markdown: 2.0.2
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-from-markdown: 2.1.0
+      mdast-util-to-markdown: 2.2.0
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -5799,9 +4808,9 @@ packages:
   /mdast-util-to-hast@13.2.1:
     resolution: {integrity: sha512-cctsq2wp5vTsLIcaymblUriiTcZd0CwWtCbLvrOzYCDZoWyMNV8sZ7krj09FSnsiJi3WVsHLM4k6Dq/yaPyCXA==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
-      '@ungap/structured-clone': 1.3.0
+      '@ungap/structured-clone': 1.4.0
       devlop: 1.1.0
       micromark-util-sanitize-uri: 2.0.1
       trim-lines: 3.0.1
@@ -5810,16 +4819,18 @@ packages:
       vfile: 6.0.3
     dev: false
 
-  /mdast-util-to-markdown@2.1.2:
-    resolution: {integrity: sha512-xj68wMTvGXVOKonmog6LwyJKrYXZPvlwabaryTjLh9LuvovB/KAH+kvi8Gjj+7rJjsFi23nkUxRQv1KqSroMqA==}
+  /mdast-util-to-markdown@2.2.0:
+    resolution: {integrity: sha512-2Jn/ADrHFFyyniBMi0bncTbJqU78K3/TEYvzvwrWWzD9MYPUa/Gm4r3nOeKFQg/PDM5BsbDYdc8ackW4NTY6NA==}
     dependencies:
       '@types/mdast': 4.0.4
       '@types/unist': 3.0.3
       longest-streak: 3.1.0
       mdast-util-phrasing: 4.1.0
       mdast-util-to-string: 4.0.0
+      micromark-util-character: 2.1.1
       micromark-util-classify-character: 2.0.1
       micromark-util-decode-string: 2.0.1
+      micromark-util-html-tag-name: 2.0.1
       unist-util-visit: 5.1.0
       zwitch: 2.0.4
     dev: false
@@ -5834,25 +4845,26 @@ packages:
     resolution: {integrity: sha512-abv/qOcuPfk3URPfDzmZU1LKmuw8kT+0nIHvKrKgFrwifol/doWcdA4ZqsWQ8ENrFKkd67Mfpo/LovbIUsbt3w==}
     dev: true
 
-  /micromark-core-commonmark@2.0.3:
-    resolution: {integrity: sha512-RDBrHEMSxVFLg6xvnXmb1Ayr2WzLAWjeSATAoxwKYJV94TeNavgoIdA0a9ytzDSVzBy2YKFK+emCPOEibLeCrg==}
+  /micromark-core-commonmark@2.0.4:
+    resolution: {integrity: sha512-wxEeE8v8XVvOrxn1TZj74qYhAtTQsSqufHVS3uNVyT1MupXxhyaHk8/9xDNPh9BmL77QNDgWTkZ3RsYHB2ry7w==}
     dependencies:
       decode-named-character-reference: 1.3.0
       devlop: 1.1.0
       micromark-factory-destination: 2.0.1
       micromark-factory-label: 2.0.1
-      micromark-factory-space: 2.0.1
+      micromark-factory-space: 2.1.0
       micromark-factory-title: 2.0.1
       micromark-factory-whitespace: 2.0.1
       micromark-util-character: 2.1.1
       micromark-util-chunked: 2.0.1
       micromark-util-classify-character: 2.0.1
+      micromark-util-edit-map: 1.0.0
       micromark-util-html-tag-name: 2.0.1
       micromark-util-normalize-identifier: 2.0.1
       micromark-util-resolve-all: 2.0.1
       micromark-util-subtokenize: 2.1.0
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm-autolink-literal@2.1.0:
@@ -5861,20 +4873,20 @@ packages:
       micromark-util-character: 2.1.1
       micromark-util-sanitize-uri: 2.0.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm-footnote@2.1.0:
     resolution: {integrity: sha512-/yPhxI1ntnDNsiHtzLKYnE3vf9JZ6cAisqVDauhp4CEHxlb4uoOTxOCJ+9s51bIB8U1N1FJ1RXOKTIlD5B/gqw==}
     dependencies:
       devlop: 1.1.0
-      micromark-core-commonmark: 2.0.3
-      micromark-factory-space: 2.0.1
+      micromark-core-commonmark: 2.0.4
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-normalize-identifier: 2.0.1
       micromark-util-sanitize-uri: 2.0.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm-strikethrough@2.1.0:
@@ -5885,33 +4897,33 @@ packages:
       micromark-util-classify-character: 2.0.1
       micromark-util-resolve-all: 2.0.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
-  /micromark-extension-gfm-table@2.1.1:
-    resolution: {integrity: sha512-t2OU/dXXioARrC6yWfJ4hqB7rct14e8f7m0cbI5hUmDyyIlwv5vEtooptH8INkbLzOatzKuVbQmAYcbWoyz6Dg==}
+  /micromark-extension-gfm-table@2.1.2:
+    resolution: {integrity: sha512-pRzm4kDTu0MjlmBkxmS9yYhw60nncfcEwu9NNdPFSQEFXS95ZKyIIyTSHu/o3ReBUrLKYEq+7YaXCRn/bPB4MA==}
     dependencies:
       devlop: 1.1.0
-      micromark-factory-space: 2.0.1
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm-tagfilter@2.0.0:
     resolution: {integrity: sha512-xHlTOmuCSotIA8TW1mDIM6X2O1SiX5P9IuDtqGonFhEK0qgRI4yeC6vMxEV2dgyr2TiD+2PQ10o+cOhdVAcwfg==}
     dependencies:
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm-task-list-item@2.1.0:
     resolution: {integrity: sha512-qIBZhqxqI6fjLDYFTBIa4eivDMnP+OZqsNwmQ3xNLE4Cxwc+zfQEfbs6tzAo2Hjq+bh6q5F+Z8/cksrLFYWQQw==}
     dependencies:
       devlop: 1.1.0
-      micromark-factory-space: 2.0.1
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-extension-gfm@3.0.0:
@@ -5920,11 +4932,11 @@ packages:
       micromark-extension-gfm-autolink-literal: 2.1.0
       micromark-extension-gfm-footnote: 2.1.0
       micromark-extension-gfm-strikethrough: 2.1.0
-      micromark-extension-gfm-table: 2.1.1
+      micromark-extension-gfm-table: 2.1.2
       micromark-extension-gfm-tagfilter: 2.0.0
       micromark-extension-gfm-task-list-item: 2.1.0
       micromark-util-combine-extensions: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-factory-destination@2.0.1:
@@ -5932,7 +4944,7 @@ packages:
     dependencies:
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-factory-label@2.0.1:
@@ -5941,39 +4953,39 @@ packages:
       devlop: 1.1.0
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
-  /micromark-factory-space@2.0.1:
-    resolution: {integrity: sha512-zRkxjtBxxLd2Sc0d+fbnEunsTj46SWXgXciZmHq0kDYGnck/ZSGj9/wULTV95uoeYiK5hRXP2mJ98Uo4cq/LQg==}
+  /micromark-factory-space@2.1.0:
+    resolution: {integrity: sha512-fS8hnLIjnjvdQIj39Geug8wWsR0HrYZ43KShKxNfwT7t2LOHo/LbWZEzEaSOjwtjdCsjoE6syHhonEzW0zv0+Q==}
     dependencies:
       micromark-util-character: 2.1.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-factory-title@2.0.1:
     resolution: {integrity: sha512-5bZ+3CjhAd9eChYTHsjy6TGxpOFSKgKKJPJxr293jTbfry2KDoWkhBb6TcPVB4NmzaPhMs1Frm9AZH7OD4Cjzw==}
     dependencies:
-      micromark-factory-space: 2.0.1
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-factory-whitespace@2.0.1:
     resolution: {integrity: sha512-Ob0nuZ3PKt/n0hORHyvoD9uZhr+Za8sFoP+OnMcnWK5lngSzALgQYKMr9RJVOWLqQYuyn6ulqGWSXdwf6F80lQ==}
     dependencies:
-      micromark-factory-space: 2.0.1
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-character@2.1.1:
     resolution: {integrity: sha512-wv8tdUTJ3thSFFFJKtpYKOYiGP2+v96Hvk4Tu8KpCAsTMs6yi+nVmGh1syvSCsaxz45J6Jbw+9DD6g97+NV67Q==}
     dependencies:
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-chunked@2.0.1:
@@ -5987,14 +4999,14 @@ packages:
     dependencies:
       micromark-util-character: 2.1.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-combine-extensions@2.0.1:
     resolution: {integrity: sha512-OnAnH8Ujmy59JcyZw8JSbK9cGpdVY44NKgSM7E9Eh7DiLS2E9RNQf0dONaGDzEG9yjEl5hcqeIsj4hfRkLH/Bg==}
     dependencies:
       micromark-util-chunked: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-decode-numeric-character-reference@2.0.2:
@@ -6012,6 +5024,12 @@ packages:
       micromark-util-symbol: 2.0.1
     dev: false
 
+  /micromark-util-edit-map@1.0.0:
+    resolution: {integrity: sha512-Pa2ljlsEL6sVwFaYeyrOLSYbQt73JvGbPOYFq+9AElXiSnfI5Q4465RRZ1sHv7lDjDOnGRDDaqPXqClqWK/q1Q==}
+    dependencies:
+      micromark-util-types: 2.0.3
+    dev: false
+
   /micromark-util-encode@2.0.1:
     resolution: {integrity: sha512-c3cVx2y4KqUnwopcO9b/SCdo2O67LwJJ/UyqGfbigahfegL9myoEFoDYZgkT7f36T0bLrM9hZTAaAyH+PCAXjw==}
     dev: false
@@ -6029,7 +5047,7 @@ packages:
   /micromark-util-resolve-all@2.0.1:
     resolution: {integrity: sha512-VdQyxFWFT2/FGJgwQnJYbe1jjQoNTS4RjglmSjTUlpUMa95Htx9NHeYW4rGDJzbjvCsl9eLjMQwGeElsqmzcHg==}
     dependencies:
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-sanitize-uri@2.0.1:
@@ -6046,37 +5064,38 @@ packages:
       devlop: 1.1.0
       micromark-util-chunked: 2.0.1
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     dev: false
 
   /micromark-util-symbol@2.0.1:
     resolution: {integrity: sha512-vs5t8Apaud9N28kgCrRUdEed4UJ+wWNvicHLPxCa9ENlYuAY31M0ETy5y1vA33YoNPDFTghEbnh6efaE8h4x0Q==}
     dev: false
 
-  /micromark-util-types@2.0.2:
-    resolution: {integrity: sha512-Yw0ECSpJoViF1qTU4DC6NwtC4aWGt1EkzaQB8KPPyCRR8z9TWeV0HbEFGTO+ZY1wB22zmxnJqhPyTpOVCpeHTA==}
+  /micromark-util-types@2.0.3:
+    resolution: {integrity: sha512-oxB2Ik03hI0gv+VNn9tnh1t1YEe9MDPptViAEgfdf3YQHsn0pzGTgCdlSCJXcwhqm8phaEuM7zeEu3QQzVBrPg==}
     dev: false
 
-  /micromark@4.0.2:
-    resolution: {integrity: sha512-zpe98Q6kvavpCr1NPVSCMebCKfD7CA2NqZ+rykeNhONIJBpc1tFKt9hucLGwha3jNTNI8lHpctWJWoimVF4PfA==}
+  /micromark@4.0.3:
+    resolution: {integrity: sha512-oGYfQzHSG5dOMovQcJ3fyTmZlWAWpi0XA0sJwJs+i6OT88o1+Jtw/8z0CmdowSLxGhhFK89rQ/oXph/wN02PNw==}
     dependencies:
-      '@types/debug': 4.1.12
+      '@types/debug': 4.1.13
       debug: 4.4.3
       decode-named-character-reference: 1.3.0
       devlop: 1.1.0
-      micromark-core-commonmark: 2.0.3
-      micromark-factory-space: 2.0.1
+      micromark-core-commonmark: 2.0.4
+      micromark-factory-space: 2.1.0
       micromark-util-character: 2.1.1
       micromark-util-chunked: 2.0.1
       micromark-util-combine-extensions: 2.0.1
       micromark-util-decode-numeric-character-reference: 2.0.2
+      micromark-util-edit-map: 1.0.0
       micromark-util-encode: 2.0.1
       micromark-util-normalize-identifier: 2.0.1
       micromark-util-resolve-all: 2.0.1
       micromark-util-sanitize-uri: 2.0.1
       micromark-util-subtokenize: 2.1.0
       micromark-util-symbol: 2.0.1
-      micromark-util-types: 2.0.2
+      micromark-util-types: 2.0.3
     transitivePeerDependencies:
       - supports-color
     dev: false
@@ -6111,24 +5130,30 @@ packages:
     engines: {node: '>=18'}
     dev: true
 
+  /minimatch@10.2.6:
+    resolution: {integrity: sha512-vpLQEs+VLCr1nU0BXS07maYoFwlDAH0gngQuuttxIwutDFEMHq2blX+8vpgxDdK3J1PwjCJiep77OitTZ4Ll1A==}
+    engines: {node: 18 || 20 || >=22}
+    dependencies:
+      brace-expansion: 5.0.12
+    dev: true
+
   /minimatch@3.1.3:
     resolution: {integrity: sha512-M2GCs7Vk83NxkUyQV1bkABc4yxgz9kILhHImZiBPAZ9ybuvCb0/H7lEl5XvIg3g+9d4eNotkZA5IWwYl0tibaA==}
     dependencies:
-      brace-expansion: 1.1.12
+      brace-expansion: 1.1.21
     dev: true
 
-  /minimatch@9.0.7:
-    resolution: {integrity: sha512-MOwgjc8tfrpn5QQEvjijjmDVtMw2oL88ugTevzxQnzRLm6l3fVEF2gzU0kYeYYKD8C66+IdGX6peJ4MyUlUnPg==}
-    engines: {node: '>=16 || 14 >=14.17'}
+  /minimatch@3.1.5:
+    resolution: {integrity: sha512-VgjWUsnnT6n+NUk6eZq77zeFdpW2LWDzP6zFGrCbHXiYNul5Dzqk2HHQ5uFH2DNW5Xbp8+jVzaeNt94ssEEl4w==}
     dependencies:
-      brace-expansion: 5.0.6
+      brace-expansion: 1.1.21
     dev: true
 
   /ms@2.1.3:
     resolution: {integrity: sha512-6FlzubTLZG3J2a/NVCAleEhjzq5oxgHyaCU9yYXvcLsvoVaHJq/s5xXI6/XXP6tz7R9xAOtHnSO/tXtF3WRTlA==}
 
-  /nanoid@3.3.12:
-    resolution: {integrity: sha512-ZB9RH/39qpq5Vu6Y+NmUaFhQR6pp+M2Xt76XBnEwDaGcVAqhlvxrl3B2bKS5D3NH3QR76v3aSrKaF/Kiy7lEtQ==}
+  /nanoid@3.3.20:
+    resolution: {integrity: sha512-uKdg2G3GNCKQn9byYOpxbGqrT2fGO5KRt5J/8b3pok8rT6qxGWF6hxMyJiEYtAf+FVyYuD9hRaDqX5uPFYJ4ZQ==}
     engines: {node: ^10 || ^12 || ^13.7 || ^14 || >=15.0.1}
     hasBin: true
     dev: false
@@ -6137,8 +5162,8 @@ packages:
     resolution: {integrity: sha512-OWND8ei3VtNC9h7V60qff3SVobHr996CTwgxubgyQYEpg290h9J0buyECNNJexkFm5sOajh5G116RYA1c8ZMSw==}
     dev: true
 
-  /node-exports-info@1.6.0:
-    resolution: {integrity: sha512-pyFS63ptit/P5WqUkt+UUfe+4oevH+bFeIiPPdfb0pFeYEu/1ELnJu5l+5EcTKYL5M7zaAa7S8ddywgXypqKCw==}
+  /node-exports-info@1.6.2:
+    resolution: {integrity: sha512-kXs9Go0cah0qHVV2v389IXQLdLCeE1xfFtjOAF+iobu0OIoG1pje8At2vMHyaPMiPMnG/LWP50twML21eMcAag==}
     engines: {node: '>= 0.4'}
     dependencies:
       array.prototype.flatmap: 1.3.3
@@ -6172,10 +5197,10 @@ packages:
     resolution: {integrity: sha512-nK28WOo+QIjBkDduTINE4JkF/UJJKyf2EJxvJKfblDpyg0Q+pkOHNTL0Qwy6NP6FhE/EnzV73BxxqcJaXY9anw==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       has-symbols: 1.1.0
       object-keys: 1.1.1
     dev: true
@@ -6184,30 +5209,30 @@ packages:
     resolution: {integrity: sha512-8u/hfXFRBD1O0hPUjioLhoWFHRmt6tKA4/vZPyckBr18l1KE9uHrFaFaUi8MDRTpi4uak2goyPTSNJLXX2k2Hw==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
     dev: true
 
   /object.fromentries@2.0.8:
     resolution: {integrity: sha512-k6E21FzySsSK5a21KRADBd/NGneRegFO5pLHfdQLpRDETUNJueLXs3WCzyQ3tFRDYgbq3KHGXfTbi2bs8WQ6rQ==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
-      es-object-atoms: 1.1.1
+      es-abstract: 1.24.2
+      es-object-atoms: 1.1.2
     dev: true
 
   /object.values@1.2.1:
     resolution: {integrity: sha512-gXah6aZrcUxjWg2zR2MwouP2eHlCBzdV4pygudehaKXSGW4v2AsRQUK+lwwXhii6KFZcunEnmSUoYp5CXibxtA==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
     dev: true
 
   /onetime@6.0.0:
@@ -6236,10 +5261,11 @@ packages:
       word-wrap: 1.2.5
     dev: true
 
-  /own-keys@1.0.1:
-    resolution: {integrity: sha512-qFOyK5PjiWZd+QQIh+1jhdb9LpxTF0qs7Pm8o5QHYZ0M3vKqSqzsZaEB6oWlxZ+q2sJBMI/Ktgd2N5ZwQoRHfg==}
+  /own-keys@1.0.2:
+    resolution: {integrity: sha512-19YVAg7T+WTrxggPukVq7DjTv6+PJ867TmhCvBsYwmbFCsZd344rq2Ld1p0wo8f8Qrrhgp82c6FJRqdXWtSEhg==}
     engines: {node: '>= 0.4'}
     dependencies:
+      call-bound: 1.0.4
       get-intrinsic: 1.3.0
       object-keys: 1.1.1
       safe-push-apply: 1.0.0
@@ -6330,30 +5356,28 @@ packages:
     engines: {node: '>=8.6'}
     dev: true
 
-  /picomatch@4.0.4:
-    resolution: {integrity: sha512-QP88BAKvMam/3NxH6vj2o21R6MjxZUAd6nlwAS/pnGvN9IVLocLHxGYIzFhg6fUQ+5th6P4dv4eW9jX3DSIj7A==}
+  /picomatch@4.0.7:
+    resolution: {integrity: sha512-qcJu88Q2IWqJsDD529JKMdwGm/dvInW4HvQnRwiH9JtihJvzGOscDtHE3x1pBKeUOTysQ8kVmLnJ2kJu7yhcGA==}
     engines: {node: '>=12'}
 
-  /pidtree@0.6.0:
-    resolution: {integrity: sha512-eG2dWTVw5bzqGRztnHExczNxt5VGsE6OwTeCG3fdUf9KBsZzO3R5OIIIzWR+iZA0NtZ+RDVdaoE2dK1cn6jH4g==}
+  /pidtree@0.6.1:
+    resolution: {integrity: sha512-e0F9AOF1JMrCfBsyJOwU9lNvQ0WtXTq0j/4jk0BQ5JSI9VAybPXmDpPRw/2FQ3e5d3ZFN1mLh7jW99m/jjaptw==}
     engines: {node: '>=0.10'}
     hasBin: true
     dev: true
 
-  /playwright-core@1.61.0:
-    resolution: {integrity: sha512-caX7TrY3Ml6egyDX0WUcTHDxodl/b51y5wJOdCEA36QviK/s2g081hvmGs8eaE3DWb6NYZQ6BjO/QkNRPenoPA==}
-    engines: {node: '>=18'}
+  /playwright-core@1.64.0:
+    resolution: {integrity: sha512-T9r+MZkTECl2+oUcZ26YgZzTrYidTsIhYYkJix+6iDOymOb1FB2RLLSOxRGnZGv5DStt+aduIVbt+E8WygXYpg==}
+    engines: {node: '>=20'}
     hasBin: true
     dev: true
 
-  /playwright@1.61.0:
-    resolution: {integrity: sha512-Z+7BeeqQPRRzklHsVFP4KTGIyMxKUmfeRA4WisM6G3/XW6nwGeX6fX9qYaDa+CiUqpOkb2f6X3nar05R3kSuJQ==}
-    engines: {node: '>=18'}
+  /playwright@1.64.0:
+    resolution: {integrity: sha512-kVzTnYFYEyQ9RGzBjNCd2+yEqWmVbxl3AHPynv3qQ6JTjv4a+kwRcnqQwIpuP3NnrnIrE/Q6wWAIAvMfIE2/Yg==}
+    engines: {node: '>=20'}
     hasBin: true
     dependencies:
-      playwright-core: 1.61.0
-    optionalDependencies:
-      fsevents: 2.3.2
+      playwright-core: 1.64.0
     dev: true
 
   /pngjs@5.0.0:
@@ -6366,13 +5390,13 @@ packages:
     engines: {node: '>= 0.4'}
     dev: true
 
-  /postcss@8.5.15:
-    resolution: {integrity: sha512-FfR8sjd4em2T6fb3I2MwAJU7HWVMr9zba+enmQeeWFfCbm+UOC/0X4DS8XtpUTMwWMGbjKYP7xjfNekzyGmB3A==}
+  /postcss@8.5.29:
+    resolution: {integrity: sha512-49cGhUbXj8Qenv0iTMxA1cFBzxXoctpC9Ujd77t1WcbJIr6nF/eI7g/8MgxrYldFRuAXvja7xQRwavoW7kgrxQ==}
     engines: {node: ^10 || ^12 || >=14}
     dependencies:
-      nanoid: 3.3.12
+      nanoid: 3.3.20
       picocolors: 1.1.1
-      source-map-js: 1.2.1
+      source-map-js: 1.2.2
     dev: false
 
   /prelude-ls@1.2.1:
@@ -6387,8 +5411,8 @@ packages:
       fast-diff: 1.3.0
     dev: true
 
-  /prettier@3.8.1:
-    resolution: {integrity: sha512-UOnG6LftzbdaHZcKoPFtOcCKztrQ57WkHDeRD9t/PTQtmT0NHSeWWepj6pS0z/N7+08BHFDQVUrfmfMRcZwbMg==}
+  /prettier@3.9.10:
+    resolution: {integrity: sha512-skGHTH1Yb9QvaTC4hWnRTzEYtCpApF0UMBPz+t/0ZPzP8ihefo/btKd+QA7pUE0fEwA1SbS/NmhyEcstdXOIdA==}
     engines: {node: '>=14'}
     hasBin: true
     dev: true
@@ -6405,8 +5429,8 @@ packages:
       object-assign: 4.1.1
       react-is: 16.13.1
 
-  /property-information@7.1.0:
-    resolution: {integrity: sha512-TwEZ+X+yCJmYfL7TPUOcvBZ4QfoT5YenQiJuX//0th53DE6w0xxLEtfK3iyryQFddXuvkIk51EEgrJQ0WJkOmQ==}
+  /property-information@7.2.0:
+    resolution: {integrity: sha512-IAtzIB6sUiWaJYrX9smp3V46pBGbBeLFRGdh25kg1334VcBlD8HzhPeNIWQH9zhGmo2itIe25EHt9dQP7G5hmg==}
     dev: false
 
   /proxy-from-env@2.1.0:
@@ -6429,8 +5453,8 @@ packages:
       yargs: 15.4.1
     dev: false
 
-  /radix-ui@1.6.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-QBdhh1arIEUvPC0dQ5+nwWAxt7+N+oP/9jPwjJkGFoSk/sqxg32gJtSXGtFh8frAIcS6oC9cx2Q+7KYCQLOAeA==}
+  /radix-ui@1.7.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-x+z5TV2v6u0QpAQ4dQfNMcvtIdSQ1Op2oqBEyfDXjNldRZbMTP+hcjr5Iie31bep81Axeioa+70Ww7pZDzTEQQ==}
     peerDependencies:
       '@types/react': '*'
       '@types/react-dom': '*'
@@ -6443,62 +5467,62 @@ packages:
         optional: true
     dependencies:
       '@radix-ui/primitive': 1.1.7
-      '@radix-ui/react-accessible-icon': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-accordion': 1.2.20(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-alert-dialog': 1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-arrow': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-aspect-ratio': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-avatar': 1.2.6(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-checkbox': 1.3.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-collapsible': 1.1.20(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-collection': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context': 1.2.2(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-context-menu': 2.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-dialog': 1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-direction': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-dismissable-layer': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-dropdown-menu': 2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-focus-scope': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-form': 0.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-hover-card': 1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-label': 2.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-menu': 2.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-menubar': 1.1.24(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-navigation-menu': 1.2.22(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-one-time-password-field': 0.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-password-toggle-field': 0.1.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-popover': 1.1.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-popper': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-portal': 1.1.17(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-presence': 1.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-primitive': 2.1.10(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-progress': 1.1.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-radio-group': 1.4.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-roving-focus': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-scroll-area': 1.2.18(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-select': 2.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-separator': 1.1.15(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slider': 1.4.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-slot': 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-switch': 1.3.7(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-tabs': 1.1.21(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toast': 1.2.23(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toggle': 1.1.18(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toggle-group': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-toolbar': 1.1.19(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-tooltip': 1.2.16(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-escape-keydown': 1.1.5(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-use-size': 1.1.4(@types/react@19.2.10)(react@19.2.1)
-      '@radix-ui/react-visually-hidden': 1.2.11(@types/react-dom@19.2.3)(@types/react@19.2.10)(react-dom@19.2.1)(react@19.2.1)
-      '@types/react': 19.2.10
-      '@types/react-dom': 19.2.3(@types/react@19.2.10)
+      '@radix-ui/react-accessible-icon': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-accordion': 1.2.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-alert-dialog': 1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-arrow': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-aspect-ratio': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-avatar': 1.2.7(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-checkbox': 1.3.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-collapsible': 1.1.21(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-collection': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-compose-refs': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context': 1.2.2(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-context-menu': 2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-dialog': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-direction': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-dismissable-layer': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-dropdown-menu': 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-focus-guards': 1.1.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-focus-scope': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-form': 0.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-hover-card': 1.1.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-label': 2.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-menu': 2.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-menubar': 1.1.25(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-navigation-menu': 1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-one-time-password-field': 0.1.17(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-password-toggle-field': 0.1.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-popover': 1.2.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-popper': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-portal': 1.1.18(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-presence': 1.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-primitive': 2.1.11(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-progress': 1.1.17(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-radio-group': 1.4.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-roving-focus': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-scroll-area': 1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-select': 2.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-separator': 1.1.16(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slider': 1.5.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-slot': 1.4.0(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-switch': 1.3.8(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-tabs': 1.1.22(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toast': 1.2.24(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toggle': 1.1.19(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toggle-group': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-toolbar': 1.1.20(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-tooltip': 1.3.0(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@radix-ui/react-use-callback-ref': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-controllable-state': 1.2.6(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-effect-event': 0.0.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-escape-keydown': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-is-hydrated': 0.1.3(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-layout-effect': 1.1.4(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-use-size': 1.1.5(@types/react@19.2.18)(react@19.2.1)
+      '@radix-ui/react-visually-hidden': 1.2.12(@types/react-dom@19.2.7)(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1)
+      '@types/react': 19.2.18
+      '@types/react-dom': 19.2.7(@types/react@19.2.18)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
@@ -6512,16 +5536,21 @@ packages:
       scheduler: 0.27.0
     dev: false
 
-  /react-hook-form@7.71.1(react@19.2.1):
-    resolution: {integrity: sha512-9SUJKCGKo8HUSsCO+y0CtqkqI5nNuaDqTxyqPsZPqIwudpj4rCrAz/jZV+jn57bx5gtZKOh3neQu94DXMc+w5w==}
+  /react-hook-form@7.90.0(@types/react@19.2.18)(react@19.2.1):
+    resolution: {integrity: sha512-+NdG0FLzCPhFOp2NNxaUYLF9fyZhNjx9FTBBSYWF71/l6BBlmIc/j6xlp3DbaHWLJbIG7Zd8oznA19xJClU97w==}
     engines: {node: '>=18.0.0'}
     peerDependencies:
+      '@types/react': '*'
       react: ^16.8.0 || ^17 || ^18 || ^19
+    peerDependenciesMeta:
+      '@types/react':
+        optional: true
     dependencies:
+      '@types/react': 19.2.18
       react: 19.2.1
     dev: false
 
-  /react-i18next@15.7.4(i18next@25.8.0)(react-dom@19.2.1)(react@19.2.1)(typescript@5.9.3):
+  /react-i18next@15.7.4(i18next@25.10.10)(react-dom@19.2.1)(react@19.2.1)(typescript@5.9.3):
     resolution: {integrity: sha512-nyU8iKNrI5uDJch0z9+Y5XEr34b0wkyYj3Rp+tfbahxtlswxSCjcUL9H0nqXo9IR3/t5Y5PKIA3fx3MfUyR9Xw==}
     peerDependencies:
       i18next: '>= 23.4.0'
@@ -6537,9 +5566,9 @@ packages:
       typescript:
         optional: true
     dependencies:
-      '@babel/runtime': 7.28.6
-      html-parse-stringify: 3.0.1
-      i18next: 25.8.0(typescript@5.9.3)
+      '@babel/runtime': 7.29.10
+      html-parse-stringify: 3.1.0
+      i18next: 25.10.10(typescript@5.9.3)
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
       typescript: 5.9.3
@@ -6552,15 +5581,15 @@ packages:
     resolution: {integrity: sha512-/LLMVyas0ljjAtoYiPqYiL8VWXzUUdThrmU5+n20DZv+a+ClRoevUzw5JxU+Ieh5/c87ytoTBV9G1FiKfNJdmg==}
     dev: false
 
-  /react-markdown@10.1.0(@types/react@19.2.10)(react@19.2.1):
+  /react-markdown@10.1.0(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-qKxVopLT/TyA6BX3Ue5NwabOsAzm0Q7kAPwq6L+wWDwisYs7R8vZ0nRXqq6rkueboxpkjvLGU9fWifiX/ZZFxQ==}
     peerDependencies:
       '@types/react': '>=18'
       react: '>=18'
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       devlop: 1.1.0
       hast-util-to-jsx-runtime: 2.3.6
       html-url-attributes: 3.0.1
@@ -6585,7 +5614,7 @@ packages:
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /react-remove-scroll-bar@2.3.8(@types/react@19.2.10)(react@19.2.1):
+  /react-remove-scroll-bar@2.3.8(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-9r+yi9+mgU33AKcj6IbT9oRCO78WriSj6t/cF8DWBZJ9aOGPOTEDvdUDz1FwKim7QXWwmHqtdHnRJfhAxEG46Q==}
     engines: {node: '>=10'}
     peerDependencies:
@@ -6595,13 +5624,13 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
-      react-style-singleton: 2.2.3(@types/react@19.2.10)(react@19.2.1)
+      react-style-singleton: 2.2.3(@types/react@19.2.18)(react@19.2.1)
       tslib: 2.8.1
     dev: false
 
-  /react-remove-scroll@2.7.2(@types/react@19.2.10)(react@19.2.1):
+  /react-remove-scroll@2.7.2(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-Iqb9NjCCTt6Hf+vOdNIZGdTiH1QSqr27H/Ek9sv/a97gfueI/5h1s3yRi1nngzMUaOOToin5dI1dXKdXiF+u0Q==}
     engines: {node: '>=10'}
     peerDependencies:
@@ -6611,17 +5640,17 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
-      react-remove-scroll-bar: 2.3.8(@types/react@19.2.10)(react@19.2.1)
-      react-style-singleton: 2.2.3(@types/react@19.2.10)(react@19.2.1)
+      react-remove-scroll-bar: 2.3.8(@types/react@19.2.18)(react@19.2.1)
+      react-style-singleton: 2.2.3(@types/react@19.2.18)(react@19.2.1)
       tslib: 2.8.1
-      use-callback-ref: 1.3.3(@types/react@19.2.10)(react@19.2.1)
-      use-sidecar: 1.1.3(@types/react@19.2.10)(react@19.2.1)
+      use-callback-ref: 1.3.3(@types/react@19.2.18)(react@19.2.1)
+      use-sidecar: 1.1.3(@types/react@19.2.18)(react@19.2.1)
     dev: false
 
-  /react-router-dom@7.17.0(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-fyU2yjGups/hE6Xz0I5ZYbVL8Gx29eCjgpHaRaTaVU+OOAdfRX05KsvyRm0GO8YQwOkhpU3MurW1jyMUJn+zSw==}
+  /react-router-dom@7.18.4(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-yrfmJHIpDG7taCpqKjT1G5B6q3O2K+RN8/fgNf0lTjCwiPbQ0ei6vXX9ZjQR+7ld8Tr7Z5xmyMnZ8YJrphWQUw==}
     engines: {node: '>=20.0.0'}
     peerDependencies:
       react: '>=18'
@@ -6629,11 +5658,11 @@ packages:
     dependencies:
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
-      react-router: 7.17.0(react-dom@19.2.1)(react@19.2.1)
+      react-router: 7.18.4(react-dom@19.2.1)(react@19.2.1)
     dev: false
 
-  /react-router@7.17.0(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-FDELK7rTMlCHO5+reyXsPlmfr7N1F91lPHsWYfMEGQm/KQ+F4JFM8jGoeQDmDvdTs93Fw9aSilH+uKRb4/jXvQ==}
+  /react-router@7.18.4(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-PUPQcMhMGRAslLcvtlPz/kmzBEWPhLdgLFrL7pLNepBL6dX0lWj4WD2cUYVgYCuT3jxvghYFg81cDTj44DhetQ==}
     engines: {node: '>=20.0.0'}
     peerDependencies:
       react: '>=18'
@@ -6654,14 +5683,14 @@ packages:
       react: ^16.8.0 || ^17.0.0 || ^18.0.0 || ^19.0.0
       react-dom: ^16.8.0 || ^17.0.0 || ^18.0.0 || ^19.0.0
     dependencies:
-      fast-equals: 5.4.0
+      fast-equals: 5.4.4
       prop-types: 15.8.1
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
       react-transition-group: 4.4.5(react-dom@19.2.1)(react@19.2.1)
     dev: false
 
-  /react-style-singleton@2.2.3(@types/react@19.2.10)(react@19.2.1):
+  /react-style-singleton@2.2.3(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-b6jSvxvVnyptAiLjbkWLE/lOnR4lfTtDAl+eUC7RZy+QQWc6wRzIV2CE6xBuMmDxc2qIihtDCZD5NPOFl7fRBQ==}
     engines: {node: '>=10'}
     peerDependencies:
@@ -6671,19 +5700,19 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       get-nonce: 1.0.1
       react: 19.2.1
       tslib: 2.8.1
     dev: false
 
-  /react-syntax-highlighter@16.1.0(react@19.2.1):
-    resolution: {integrity: sha512-E40/hBiP5rCNwkeBN1vRP+xow1X0pndinO+z3h7HLsHyjztbyjfzNWNKuAsJj+7DLam9iT4AaaOZnueCU+Nplg==}
+  /react-syntax-highlighter@16.1.2(react@19.2.1):
+    resolution: {integrity: sha512-4dM0+mYtfnhgwYWAvFx56fgyjIYk2yFopfiICFvMDhRfeSGRNd+ngFALdPD9q7llyGAmiE6pzOSOwI8Lf9ca1w==}
     engines: {node: '>= 16.20.2'}
     peerDependencies:
       react: '>= 0.14.0'
     dependencies:
-      '@babel/runtime': 7.28.6
+      '@babel/runtime': 7.29.10
       highlight.js: 10.7.3
       highlightjs-vue: 1.0.0
       lowlight: 1.20.0
@@ -6698,7 +5727,7 @@ packages:
       react: '>=16.6.0'
       react-dom: '>=16.6.0'
     dependencies:
-      '@babel/runtime': 7.28.6
+      '@babel/runtime': 7.29.10
       dom-helpers: 5.2.1
       loose-envify: 1.4.0
       prop-types: 15.8.1
@@ -6720,6 +5749,7 @@ packages:
   /recharts@2.15.4(react-dom@19.2.1)(react@19.2.1):
     resolution: {integrity: sha512-UT/q6fwS3c1dHbXv2uFgYJ9BMFHu3fwnd7AYZaEQhXuYQ4hgsxLvsUXzGdKeZrW5xopzDCvuA2N41WJ88I7zIw==}
     engines: {node: '>=14'}
+    deprecated: 1.x and 2.x branches are no longer active. Bump to Recharts v3 to receive latest features and bugfixes. See https://github.com/recharts/recharts/wiki/3.0-migration-guide
     peerDependencies:
       react: ^16.0.0 || ^17.0.0 || ^18.0.0 || ^19.0.0
       react-dom: ^16.0.0 || ^17.0.0 || ^18.0.0 || ^19.0.0
@@ -6740,11 +5770,11 @@ packages:
     resolution: {integrity: sha512-00o4I+DVrefhv+nX0ulyi3biSHCPDe+yLv5o/p6d/UVlirijB8E16FtfwSAi4g3tcqrQ4lRAqQSoFEZJehYEcw==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       get-intrinsic: 1.3.0
       get-proto: 1.0.1
       which-builtin-type: 1.2.1
@@ -6753,8 +5783,8 @@ packages:
   /refractor@5.0.0:
     resolution: {integrity: sha512-QXOrHQF5jOpjjLfiNk5GFnWhRXvxjUVnlFxkeDmewR5sXkr3iM46Zo+CnRR8B+MDVqkULW4EcLVcRBNOPXHosw==}
     dependencies:
-      '@types/hast': 3.0.4
-      '@types/prismjs': 1.26.5
+      '@types/hast': 3.0.5
+      '@types/prismjs': 1.26.6
       hastscript: 9.0.1
       parse-entities: 4.0.2
     dev: false
@@ -6763,7 +5793,7 @@ packages:
     resolution: {integrity: sha512-dYqgNSZbDwkaJ2ceRd9ojCGjBq+mOm9LmtXnAnEGyHhN/5R7iDW2TRw3h+o/jCFxus3P2LfWIIiwowAjANm7IA==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
       es-errors: 1.3.0
       get-proto: 1.0.1
@@ -6774,8 +5804,8 @@ packages:
   /rehype-autolink-headings@7.1.0:
     resolution: {integrity: sha512-rItO/pSdvnvsP4QRB1pmPiNHUskikqtPojZKJPPPAVx9Hj8i8TwMBhofrrAYRhYOOBZH9tgmG5lPqDLuIWPWmw==}
     dependencies:
-      '@types/hast': 3.0.4
-      '@ungap/structured-clone': 1.3.0
+      '@types/hast': 3.0.5
+      '@ungap/structured-clone': 1.4.0
       hast-util-heading-rank: 3.0.0
       hast-util-is-element: 3.0.0
       unified: 11.0.5
@@ -6785,7 +5815,7 @@ packages:
   /rehype-highlight@7.0.2:
     resolution: {integrity: sha512-k158pK7wdC2qL3M5NcZROZ2tR/l7zOzjxXd5VGdcfIyoijjQqpHd3JKtYSBDpDZ38UI2WJWuFAtkMDxmx5kstA==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       hast-util-to-text: 4.0.2
       lowlight: 3.3.0
       unist-util-visit: 5.1.0
@@ -6795,7 +5825,7 @@ packages:
   /rehype-raw@7.0.0:
     resolution: {integrity: sha512-/aE8hCfKlQeA8LmyeyQvQF3eBiLRGNlfBJEvWH7ivp9sBqs7TNqBL5X3v157rM4IFETqDnIOO+z5M/biZbo9Ww==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       hast-util-raw: 9.1.0
       vfile: 6.0.3
     dev: false
@@ -6803,14 +5833,14 @@ packages:
   /rehype-sanitize@6.0.0:
     resolution: {integrity: sha512-CsnhKNsyI8Tub6L4sm5ZFsme4puGfc6pYylvXo1AeqaGbjOYyzNv3qZPwvs0oMJ39eryyeOdmxwUIo94IpEhqg==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       hast-util-sanitize: 5.0.2
     dev: false
 
   /rehype-slug@6.0.0:
     resolution: {integrity: sha512-lWyvf/jwu+oS5+hL5eClVd3hNdmwM1kAC0BUvEGD19pajQMIzcNUd/k9GsfQ+FfECvX+JE+e9/btsKH0EjJT6A==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       github-slugger: 2.0.0
       hast-util-heading-rank: 3.0.0
       hast-util-to-string: 3.0.1
@@ -6834,8 +5864,8 @@ packages:
     resolution: {integrity: sha512-FCxlKLNGknS5ba/1lmpYijMUzX2esxW5xQqjWxw2eHFfS2MSdaHVINFmhjo+qN1WhZhNimq0dZATN9pH0IDrpA==}
     dependencies:
       '@types/mdast': 4.0.4
-      mdast-util-from-markdown: 2.0.2
-      micromark-util-types: 2.0.2
+      mdast-util-from-markdown: 2.1.0
+      micromark-util-types: 2.0.3
       unified: 11.0.5
     transitivePeerDependencies:
       - supports-color
@@ -6844,7 +5874,7 @@ packages:
   /remark-rehype@11.1.2:
     resolution: {integrity: sha512-Dh7l57ianaEoIpzbp0PC9UKAdCSVklD8E5Rpw7ETfbTl3FqcOOgq5q2LVDhgGCkaBv7p24JXikPdvhhmHvKMsw==}
     dependencies:
-      '@types/hast': 3.0.4
+      '@types/hast': 3.0.5
       '@types/mdast': 4.0.4
       mdast-util-to-hast: 13.2.1
       unified: 11.0.5
@@ -6855,7 +5885,7 @@ packages:
     resolution: {integrity: sha512-1OSmLd3awB/t8qdoEOMazZkNsfVTeY4fTsgzcQFdXNq8ToTN4ZGwrMnlda4K6smTFKD+GRV6O48i6Z4iKgPPpw==}
     dependencies:
       '@types/mdast': 4.0.4
-      mdast-util-to-markdown: 2.1.2
+      mdast-util-to-markdown: 2.2.0
       unified: 11.0.5
     dev: false
 
@@ -6873,14 +5903,14 @@ packages:
     engines: {node: '>=4'}
     dev: true
 
-  /resolve@2.0.0-next.6:
-    resolution: {integrity: sha512-3JmVl5hMGtJ3kMmB3zi3DL25KfkCEyy3Tw7Gmw7z5w8M9WlwoPFnIvwChzu1+cF3iaK3sp18hhPz8ANeimdJfA==}
+  /resolve@2.0.0-next.7:
+    resolution: {integrity: sha512-tqt+NBWwyaMgw3zDsnygx4CByWjQEJHOPMdslYhppaQSJUtL/D4JO9CcBBlhPoI8lz9oJIDXkwXfhF4aWqP8xQ==}
     engines: {node: '>= 0.4'}
     hasBin: true
     dependencies:
       es-errors: 1.3.0
-      is-core-module: 2.16.1
-      node-exports-info: 1.6.0
+      is-core-module: 2.17.0
+      node-exports-info: 1.6.2
       object-keys: 1.1.1
       path-parse: 1.0.7
       supports-preserve-symlinks-flag: 1.0.0
@@ -6898,36 +5928,36 @@ packages:
     resolution: {integrity: sha512-q1b3N5QkRUWUl7iyylaaj3kOpIT0N2i9MqIEQXP73GVsN9cw3fdx8X63cEmWhJGi2PPCF23Ijp7ktmd39rawIA==}
     dev: true
 
-  /rolldown@1.0.3:
-    resolution: {integrity: sha512-i00lAJ2ks1BYr7rjNjKC7BcqAS7nVfiT3QX1SI5aY+AFHblCmaUf9OE9dbdzDvW6dJxbi2ZCZiy9v3CcwOiX3g==}
+  /rolldown@1.2.13:
+    resolution: {integrity: sha512-/5apOkwWxtACcQtZ5WlrAsKS4JwXVThQPJZ3RqIruRY9UG9M8e0VlXN068xum/JRKA4aIUoJ13ghe/kGCvidiQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     hasBin: true
     dependencies:
-      '@oxc-project/types': 0.133.0
+      '@oxc-project/types': 0.153.0
       '@rolldown/pluginutils': 1.0.1
     optionalDependencies:
-      '@rolldown/binding-android-arm64': 1.0.3
-      '@rolldown/binding-darwin-arm64': 1.0.3
-      '@rolldown/binding-darwin-x64': 1.0.3
-      '@rolldown/binding-freebsd-x64': 1.0.3
-      '@rolldown/binding-linux-arm-gnueabihf': 1.0.3
-      '@rolldown/binding-linux-arm64-gnu': 1.0.3
-      '@rolldown/binding-linux-arm64-musl': 1.0.3
-      '@rolldown/binding-linux-ppc64-gnu': 1.0.3
-      '@rolldown/binding-linux-s390x-gnu': 1.0.3
-      '@rolldown/binding-linux-x64-gnu': 1.0.3
-      '@rolldown/binding-linux-x64-musl': 1.0.3
-      '@rolldown/binding-openharmony-arm64': 1.0.3
-      '@rolldown/binding-wasm32-wasi': 1.0.3
-      '@rolldown/binding-win32-arm64-msvc': 1.0.3
-      '@rolldown/binding-win32-x64-msvc': 1.0.3
+      '@rolldown/binding-android-arm-eabi': 1.2.13
+      '@rolldown/binding-android-arm64': 1.2.13
+      '@rolldown/binding-darwin-arm64': 1.2.13
+      '@rolldown/binding-darwin-x64': 1.2.13
+      '@rolldown/binding-freebsd-x64': 1.2.13
+      '@rolldown/binding-linux-arm-gnueabihf': 1.2.13
+      '@rolldown/binding-linux-arm64-gnu': 1.2.13
+      '@rolldown/binding-linux-arm64-musl': 1.2.13
+      '@rolldown/binding-linux-ppc64-gnu': 1.2.13
+      '@rolldown/binding-linux-s390x-gnu': 1.2.13
+      '@rolldown/binding-linux-x64-gnu': 1.2.13
+      '@rolldown/binding-linux-x64-musl': 1.2.13
+      '@rolldown/binding-openharmony-arm64': 1.2.13
+      '@rolldown/binding-win32-arm64-msvc': 1.2.13
+      '@rolldown/binding-win32-x64-msvc': 1.2.13
     dev: false
 
-  /safe-array-concat@1.1.3:
-    resolution: {integrity: sha512-AURm5f0jYEOydBj7VQlVvDrjeFgthDdEF5H1dP+6mNpoXOMo1quQqJ4wvJDyRZ9+pO3kGWoOdmV08cSv2aJV6Q==}
+  /safe-array-concat@1.1.4:
+    resolution: {integrity: sha512-wtZlHyOje6OZTGqAoaDKxFkgRtkF9CnHAVnCHKfuj200wAgL+bSJhdsCD2l0Qx/2ekEXjPWcyKkfGb5CPboslg==}
     engines: {node: '>=0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       get-intrinsic: 1.3.0
       has-symbols: 1.1.0
@@ -6960,8 +5990,8 @@ packages:
     hasBin: true
     dev: true
 
-  /semver@7.7.3:
-    resolution: {integrity: sha512-SdsKMrI9TdgjdweUSR9MweHA4EJ8YxHn8DFaDisvhVlUOe4BF1tLD7GAj0lIqWVl+dPb/rExr0Btby5loQm20Q==}
+  /semver@7.8.5:
+    resolution: {integrity: sha512-Y7/KDsb8LjooZpwaqGyulO6DQlksgCncchHGk+sZIY4SBvUocMBEFH5Ur1fI4dV+Jvl0w6cjvucaIi40puRioA==}
     engines: {node: '>=10'}
     hasBin: true
     dev: true
@@ -7002,7 +6032,7 @@ packages:
     dependencies:
       dunder-proto: 1.0.1
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
     dev: true
 
   /shebang-command@2.0.0:
@@ -7017,8 +6047,8 @@ packages:
     engines: {node: '>=8'}
     dev: true
 
-  /side-channel-list@1.0.0:
-    resolution: {integrity: sha512-FCLHtRD/gnpCiCHEiJLOwdmFP+wzCmDEkc9y7NsYxeF4u7Btsn1ZuwgwJGxImImHicJArLP4R0yX4c2KCrMrTA==}
+  /side-channel-list@1.0.1:
+    resolution: {integrity: sha512-mjn/0bi/oUURjc5Xl7IaWi/OJJJumuoJFQJfDDyO46+hBWsfaVM65TBHq2eoZBhzl9EchxOijpkbRC8SVBQU0w==}
     engines: {node: '>= 0.4'}
     dependencies:
       es-errors: 1.3.0
@@ -7046,13 +6076,13 @@ packages:
       side-channel-map: 1.0.1
     dev: true
 
-  /side-channel@1.1.0:
-    resolution: {integrity: sha512-ZX99e6tRweoUXqR+VBrslhda51Nh5MTQwou5tnUDgbtyM0dBgmhEDtWGP/xbKn6hqfPRHujUNwz5fy/wbbhnpw==}
+  /side-channel@1.1.1:
+    resolution: {integrity: sha512-6x6dK6zJdpTzF4sQeNYxwtvBzf6Eg4GtlesS94HOvTudUeyK2WXAaIfmDgsyslYrRBeFIlsi54AYsFGUuhmvrQ==}
     engines: {node: '>= 0.4'}
     dependencies:
       es-errors: 1.3.0
       object-inspect: 1.13.4
-      side-channel-list: 1.0.0
+      side-channel-list: 1.0.1
       side-channel-map: 1.0.1
       side-channel-weakmap: 1.0.2
     dev: true
@@ -7078,18 +6108,23 @@ packages:
       is-fullwidth-code-point: 5.1.0
     dev: true
 
-  /sonner@2.0.7(react-dom@19.2.1)(react@19.2.1):
-    resolution: {integrity: sha512-W6ZN4p58k8aDKA4XPcx2hpIQXBRAgyiWVkYhT7CvK6D3iAu7xjvVyhQHg2/iaKJZ1XVJ4r7XuwGL+WGEK37i9w==}
+  /sonner@2.0.8(@types/react@19.2.18)(react-dom@19.2.1)(react@19.2.1):
+    resolution: {integrity: sha512-UM/ByIoFra8yzV75n1o0Puu0bw5U/9UNnDacrJNspekBewIfsQ3D6ez1nvlWpt7aTsO6rujQtifBpycwIivqlg==}
     peerDependencies:
+      '@types/react': ^18.0.0 || ^19.0.0
       react: ^18.0.0 || ^19.0.0 || ^19.0.0-rc
       react-dom: ^18.0.0 || ^19.0.0 || ^19.0.0-rc
+    peerDependenciesMeta:
+      '@types/react':
+        optional: true
     dependencies:
+      '@types/react': 19.2.18
       react: 19.2.1
       react-dom: 19.2.1(react@19.2.1)
     dev: false
 
-  /source-map-js@1.2.1:
-    resolution: {integrity: sha512-UXWMKhLOwVKb728IUtQPXxfYU+usdybtUrK/8uGE8CQMvrhOpwvzDBwj0QhSL7MQc7vIsISBG8VQ8+IDQxpfQA==}
+  /source-map-js@1.2.2:
+    resolution: {integrity: sha512-KGj/8Y43x35aZVDtt+J4mK1hoLGHULMYfSkODJNQjNDC3oW1PqPoxMwo0pLUsWM/UEGzON/NxeHywEfNXNP3Vw==}
     engines: {node: '>=0.10.0'}
     dev: false
 
@@ -7124,66 +6159,67 @@ packages:
     engines: {node: '>=18'}
     dependencies:
       emoji-regex: 10.6.0
-      get-east-asian-width: 1.4.0
-      strip-ansi: 7.1.2
+      get-east-asian-width: 1.7.0
+      strip-ansi: 7.2.0
     dev: true
 
-  /string.prototype.matchall@4.0.12:
-    resolution: {integrity: sha512-6CC9uyBL+/48dYizRf7H7VAYCMCNTBeM78x/VTUe9bFEaxBepPJDa1Ow99LqI/1yF7kuy7Q3cQsYMrcjGUcskA==}
+  /string.prototype.matchall@4.1.0:
+    resolution: {integrity: sha512-tHNHTxInrYLCga9O9YGxWA3G9/nnzQw8UGAyqGx3Ar1pSTTzIuM4woFSq4SowkXCjJIwq5sIiQvEfRI9tCH1qQ==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
       es-errors: 1.3.0
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
       get-intrinsic: 1.3.0
       gopd: 1.2.0
       has-symbols: 1.1.0
       internal-slot: 1.1.0
       regexp.prototype.flags: 1.5.4
       set-function-name: 2.0.2
-      side-channel: 1.1.0
+      side-channel: 1.1.1
     dev: true
 
   /string.prototype.repeat@1.0.0:
     resolution: {integrity: sha512-0u/TldDbKD8bFCQ/4f5+mNRrXwZ8hg2w7ZR8wa16e8z9XpePWl3eGEcUD0OXpEH/VJH/2G3gjUtR3ZOiBe2S/w==}
     dependencies:
       define-properties: 1.2.1
-      es-abstract: 1.24.1
+      es-abstract: 1.24.2
     dev: true
 
-  /string.prototype.trim@1.2.10:
-    resolution: {integrity: sha512-Rs66F0P/1kedk5lyYyH9uBzuiI/kNRmwJAR9quK6VOtIpZ2G+hMZd+HQbbv25MgCA6gEffoMZYxlTod4WcdrKA==}
+  /string.prototype.trim@1.2.11:
+    resolution: {integrity: sha512-PwvK7BU+CMTJGYQCTZb5RWXIML92lftJLhQz1tBzgKiqGxJaMlBAa48POXaNAC2s4y8jr3EFqrkF9+44neS46w==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-data-property: 1.1.4
       define-properties: 1.2.1
-      es-abstract: 1.24.1
-      es-object-atoms: 1.1.1
+      es-abstract: 1.24.2
+      es-object-atoms: 1.1.2
       has-property-descriptors: 1.0.2
+      safe-regex-test: 1.1.0
     dev: true
 
-  /string.prototype.trimend@1.0.9:
-    resolution: {integrity: sha512-G7Ok5C6E/j4SGfyLCloXTrngQIQU3PWtXGst3yM7Bea9FRURf1S42ZHlZZtsNque2FN2PoUhfZXYLNWwEr4dLQ==}
+  /string.prototype.trimend@1.0.10:
+    resolution: {integrity: sha512-2+3aDAOmPTmuFwjDnmJG2ctEkQKVki7vOSqaxkv42Mowj1V6PnvuwFCRrR5lChUux1TBskPjfkeTOhqczDMxTw==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       define-properties: 1.2.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
     dev: true
 
   /string.prototype.trimstart@1.0.8:
     resolution: {integrity: sha512-UXSH262CSZY1tfu3G3Secr6uGLCFVPMhIqHjlgCUtCCcgihYc/xKs9djMTMUOb2j1mVSeU8EU6NWc/iQKU6Gfg==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       define-properties: 1.2.1
-      es-object-atoms: 1.1.1
+      es-object-atoms: 1.1.2
     dev: true
 
   /stringify-entities@4.0.4:
@@ -7200,11 +6236,11 @@ packages:
       ansi-regex: 5.0.1
     dev: false
 
-  /strip-ansi@7.1.2:
-    resolution: {integrity: sha512-gmBGslpoQJtgnMAvOVqGZpEz9dyoKTCzy2nfz/n8aIFhN/jCE/rCmcxabB6jOOHV+0WNnylOxaxBQPSvcWklhA==}
+  /strip-ansi@7.2.0:
+    resolution: {integrity: sha512-yDPMNjp4WyfYBkHnjIRLfca1i6KMyGCtsVgoKe/z1+6vukgaENdgGBZt+ZmKPc4gavvEZ5OgHfHdrazhgNyG7w==}
     engines: {node: '>=12'}
     dependencies:
-      ansi-regex: 6.2.2
+      ansi-regex: 6.4.0
     dev: true
 
   /strip-final-newline@3.0.0:
@@ -7241,23 +6277,23 @@ packages:
     engines: {node: '>= 0.4'}
     dev: true
 
-  /synckit@0.11.12:
-    resolution: {integrity: sha512-Bh7QjT8/SuKUIfObSXNHNSK6WHo6J1tHCqJsuaFDP7gP0fkzSfTxI8y85JrppZ0h8l0maIgc2tfuZQ6/t3GtnQ==}
+  /synckit@0.11.13:
+    resolution: {integrity: sha512-eNRKgb3z66Yp3D2CixVujOUvXLFUTij/zVnV8KRyvFdQwpz7I5DS8UfRkTeLzb64u+dkzDSdelE24izu+zSSUg==}
     engines: {node: ^14.18.0 || >=16.0.0}
     dependencies:
-      '@pkgr/core': 0.2.9
+      '@pkgr/core': 0.3.6
     dev: true
 
-  /tailwind-merge@3.4.0:
-    resolution: {integrity: sha512-uSaO4gnW+b3Y2aWoWfFpX62vn2sR3skfhbjsEnaBI81WD1wBLlHZe5sWf0AqjksNdYTbGBEd0UasQMT3SNV15g==}
+  /tailwind-merge@3.7.0:
+    resolution: {integrity: sha512-XPPUyAc+cvspz3lHTcR/QgPfW2A0lv/xQNIjX3HGhLR+Nq2lHaLq5MtTesHn8GUr3W3DguT2KT5x3NVgRtYwmA==}
     dev: false
 
-  /tailwindcss@4.1.18:
-    resolution: {integrity: sha512-4+Z+0yiYyEtUVCScyfHCxOYP06L5Ne+JiHhY2IjR2KWMIWhJOYZKLSGZaP5HkZ8+bY0cxfzwDE5uOmzFXyIwxw==}
+  /tailwindcss@4.3.3:
+    resolution: {integrity: sha512-gOhV3P7ufE62QDGg1zVaTgCR+EtPv92k2nIhVcVKcLmxT1sUBsQGhnZj175j+MqRt4zLF7ic+sCYjfhxMxj7YQ==}
     dev: false
 
-  /tapable@2.3.0:
-    resolution: {integrity: sha512-g9ljZiwki/LfxmQADO3dEY1CbpmXT5Hm2fJ+QaGKwSXUylMybePR7/67YW7jOrrvjEgL1Fmz5kzyAjWVWLlucg==}
+  /tapable@2.3.3:
+    resolution: {integrity: sha512-uxc/zpqFg6x7C8vOE7lh6Lbda8eEL9zmVm/PLeTPBRhh1xCgdWaQ+J1CUieGpIfm2HdtsUpRv+HshiasBMcc6A==}
     engines: {node: '>=6'}
     dev: false
 
@@ -7265,21 +6301,12 @@ packages:
     resolution: {integrity: sha512-+FbBPE1o9QAYvviau/qC5SE3caw21q3xkvWKBtja5vgqOWIHHJ3ioaq1VPfn/Szqctz2bU/oYeKd9/z5BL+PVg==}
     dev: false
 
-  /tinyglobby@0.2.15:
-    resolution: {integrity: sha512-j2Zq4NyQYG5XMST4cbs02Ak8iJUdxRM0XI5QyxXuZOzKOINmWurp3smXu3y5wDcJrptwpSjgXHzIQxR0omXljQ==}
-    engines: {node: '>=12.0.0'}
-    dependencies:
-      fdir: 6.5.0(picomatch@4.0.4)
-      picomatch: 4.0.4
-    dev: true
-
   /tinyglobby@0.2.17:
     resolution: {integrity: sha512-wXR/dYpcqKmfWpEdZjiKJOwCNFndD0DMnrW/cYjVGttEkBfVgcLFHoNrlj47mjOVic9yyNu65alsgF4NQyTa2g==}
     engines: {node: '>=12.0.0'}
     dependencies:
-      fdir: 6.5.0(picomatch@4.0.4)
-      picomatch: 4.0.4
-    dev: false
+      fdir: 6.5.0(picomatch@4.0.7)
+      picomatch: 4.0.7
 
   /to-regex-range@5.0.1:
     resolution: {integrity: sha512-65P7iz6X5yEr1cwcgvQxbbIw7Uk3gOy5dIdtZ4rDveLqhrdJP+Li/Hx6tyK0NEb+2GCyneCMJiGqrADCSNk8sQ==}
@@ -7296,8 +6323,8 @@ packages:
     resolution: {integrity: sha512-tmMpK00BjZiUyVyvrBK7knerNgmgvcV/KLVyuma/SC+TQN167GrMRciANTz09+k3zW8L8t60jWO1GpfkZdjTaw==}
     dev: false
 
-  /ts-api-utils@2.4.0(typescript@5.9.3):
-    resolution: {integrity: sha512-3TaVTaAv2gTiMB35i3FiGJaRfwb3Pyn/j3m/bfAvGe8FB7CF6u+LMYqYlDh7reQf7UNvoTvdfAqHGmPGOSsPmA==}
+  /ts-api-utils@2.5.0(typescript@5.9.3):
+    resolution: {integrity: sha512-OJ/ibxhPlqrMM0UiNHJ/0CKQkoKF243/AEmplt3qpRgkW8VG7IfOS41h7V8TjITqdByHzrjcS/2si+y4lIh8NA==}
     engines: {node: '>=18.12'}
     peerDependencies:
       typescript: '>=4.8.4'
@@ -7333,31 +6360,30 @@ packages:
     resolution: {integrity: sha512-BaXgOuIxz8n8pIq3e7Atg/7s+DpiYrxn4vdot3w9KbnBhcRQq6o3xemQdIfynqSeXeDrF32x+WvfzmOjPiY9lg==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       for-each: 0.3.5
       gopd: 1.2.0
       has-proto: 1.2.0
       is-typed-array: 1.1.15
     dev: true
 
-  /typed-array-byte-offset@1.0.4:
-    resolution: {integrity: sha512-bTlAFB/FBYMcuX81gbL4OcpH5PmlFHqlCCpAl8AlEzMz5k53oNDvN8p1PNOWLEmI2x4orp3raOFB51tv9X+MFQ==}
+  /typed-array-byte-offset@1.0.5:
+    resolution: {integrity: sha512-0FHJvLPqZ7KJzp17O13jfsAjsqazgrxBu2zEK95PmUz8lv2+GjRuxUInCr2Rk9Dms3ihN21zJ929ZO43yJ95QQ==}
     engines: {node: '>= 0.4'}
     dependencies:
       available-typed-arrays: 1.0.7
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       for-each: 0.3.5
       gopd: 1.2.0
-      has-proto: 1.2.0
       is-typed-array: 1.1.15
       reflect.getprototypeof: 1.0.10
     dev: true
 
-  /typed-array-length@1.0.7:
-    resolution: {integrity: sha512-3KS2b+kL7fsuk/eJZ7EQdnEmQoaho/r6KUef7hxvltNA5DR8NAUM+8wJMbJyZ4G9/7i3v5zPBIMN5aybAh2/Jg==}
+  /typed-array-length@1.0.8:
+    resolution: {integrity: sha512-phPGCwqr2+Qo0fwniCE8e4pKnGu/yFb5nD5Y8bf0EEeiI5GklnACYA9GFy/DrAeRrKHXvHn+1SUsOWgJp6RO+g==}
     engines: {node: '>= 0.4'}
     dependencies:
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       for-each: 0.3.5
       gopd: 1.2.0
       is-typed-array: 1.1.15
@@ -7365,18 +6391,18 @@ packages:
       reflect.getprototypeof: 1.0.10
     dev: true
 
-  /typescript-eslint@8.54.0(eslint@9.39.2)(typescript@5.9.3):
-    resolution: {integrity: sha512-CKsJ+g53QpsNPqbzUsfKVgd3Lny4yKZ1pP4qN3jdMOg/sisIDLGyDMezycquXLE5JsEU0wp3dGNdzig0/fmSVQ==}
+  /typescript-eslint@8.71.1(eslint@9.39.5)(typescript@5.9.3):
+    resolution: {integrity: sha512-oiNVPC3/NbV1FpaHU07l8O6ynqG9v4PAiW2WvTrAmZdR0f2HI/gE668ElMLCK7Xbsu5k2fK5C4fXNiq05j7/nQ==}
     engines: {node: ^18.18.0 || ^20.9.0 || >=21.1.0}
     peerDependencies:
-      eslint: ^8.57.0 || ^9.0.0
-      typescript: '>=4.8.4 <6.0.0'
+      eslint: ^8.57.0 || ^9.0.0 || ^10.0.0
+      typescript: '>=4.8.4 <6.1.0'
     dependencies:
-      '@typescript-eslint/eslint-plugin': 8.54.0(@typescript-eslint/parser@8.54.0)(eslint@9.39.2)(typescript@5.9.3)
-      '@typescript-eslint/parser': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
-      '@typescript-eslint/typescript-estree': 8.54.0(typescript@5.9.3)
-      '@typescript-eslint/utils': 8.54.0(eslint@9.39.2)(typescript@5.9.3)
-      eslint: 9.39.2
+      '@typescript-eslint/eslint-plugin': 8.71.1(@typescript-eslint/parser@8.71.1)(eslint@9.39.5)(typescript@5.9.3)
+      '@typescript-eslint/parser': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
+      '@typescript-eslint/typescript-estree': 8.71.1(typescript@5.9.3)
+      '@typescript-eslint/utils': 8.71.1(eslint@9.39.5)(typescript@5.9.3)
+      eslint: 9.39.5
       typescript: 5.9.3
     transitivePeerDependencies:
       - supports-color
@@ -7458,7 +6484,7 @@ packages:
       punycode: 2.3.1
     dev: true
 
-  /use-callback-ref@1.3.3(@types/react@19.2.10)(react@19.2.1):
+  /use-callback-ref@1.3.3(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-jQL3lRnocaFtu3V00JToYz/4QkNWswxijDaCVNZRiRTO3HQDLsdu1ZtmIUvV4yPp+rvWm5j0y0TG/S61cuijTg==}
     engines: {node: '>=10'}
     peerDependencies:
@@ -7468,12 +6494,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       react: 19.2.1
       tslib: 2.8.1
     dev: false
 
-  /use-sidecar@1.1.3(@types/react@19.2.10)(react@19.2.1):
+  /use-sidecar@1.1.3(@types/react@19.2.18)(react@19.2.1):
     resolution: {integrity: sha512-Fedw0aZvkhynoPYlA5WXrMCAMm+nSWdZt6lzJQ7Ok8S6Q+VsHmHpRWndVRJ8Be0ZbkfPc5LRYH+5XrzXcEeLRQ==}
     engines: {node: '>=10'}
     peerDependencies:
@@ -7483,20 +6509,12 @@ packages:
       '@types/react':
         optional: true
     dependencies:
-      '@types/react': 19.2.10
+      '@types/react': 19.2.18
       detect-node-es: 1.1.0
       react: 19.2.1
       tslib: 2.8.1
     dev: false
 
-  /use-sync-external-store@1.6.0(react@19.2.1):
-    resolution: {integrity: sha512-Pp6GSwGP/NrPIrxVFAIkOQeyw8lFenOHijQWkUTrDvrF4ALqylP2C/KCkeS9dpUM3KvYRQhna5vt7IL95+ZQ9w==}
-    peerDependencies:
-      react: ^16.8.0 || ^17.0.0 || ^18.0.0 || ^19.0.0
-    dependencies:
-      react: 19.2.1
-    dev: false
-
   /uuidjs@5.1.0:
     resolution: {integrity: sha512-HAQPtUkr7t5Ud3uCwRcqtBRNagu/2aerrrBQE6PzgSluGijvFF75UaOq22Xw545GGviRjSLhc4c8CaSMI5h4Ng==}
     hasBin: true
@@ -7530,7 +6548,7 @@ packages:
       '@types/d3-ease': 3.0.2
       '@types/d3-interpolate': 3.0.4
       '@types/d3-scale': 4.0.9
-      '@types/d3-shape': 3.1.8
+      '@types/d3-shape': 3.2.0
       '@types/d3-time': 3.0.4
       '@types/d3-timer': 3.0.2
       d3-array: 3.2.4
@@ -7542,13 +6560,13 @@ packages:
       d3-timer: 3.0.1
     dev: false
 
-  /vite@8.0.16(@types/node@20.19.30):
-    resolution: {integrity: sha512-h9bXPmJichP5fLmVQo3PyaGSDE2n3aPuomeAlVRm0JLmt4rY6zmPKd59HYI4LNW8oTK7tlTsuC7l/m7awx9Jcw==}
+  /vite@8.3.4(@types/node@20.19.43):
+    resolution: {integrity: sha512-IftLE8phJhdNRIGNEV4JCCvfNScoZQVfcXaN8TQropWGJtOHcu+iAB9pHcuGvmEb/kddw0LBe/jS8ztdC6xdiQ==}
     engines: {node: ^20.19.0 || >=22.12.0}
     hasBin: true
     peerDependencies:
       '@types/node': ^20.19.0 || >=22.12.0
-      '@vitejs/devtools': ^0.1.18
+      '@vitejs/devtools': ^0.7.1
       esbuild: ^0.27.0 || ^0.28.0
       jiti: '>=1.21.0'
       less: ^4.0.0
@@ -7585,11 +6603,11 @@ packages:
       yaml:
         optional: true
     dependencies:
-      '@types/node': 20.19.30
-      lightningcss: 1.32.0
-      picomatch: 4.0.4
-      postcss: 8.5.15
-      rolldown: 1.0.3
+      '@types/node': 20.19.43
+      lightningcss: 1.33.0
+      picomatch: 4.0.7
+      postcss: 8.5.29
+      rolldown: 1.2.13
       tinyglobby: 0.2.17
     optionalDependencies:
       fsevents: 2.3.3
@@ -7620,7 +6638,7 @@ packages:
     engines: {node: '>= 0.4'}
     dependencies:
       call-bound: 1.0.4
-      function.prototype.name: 1.1.8
+      function.prototype.name: 1.2.0
       has-tostringtag: 1.0.2
       is-async-function: 2.1.1
       is-date-object: 1.1.0
@@ -7631,7 +6649,7 @@ packages:
       isarray: 2.0.5
       which-boxed-primitive: 1.1.1
       which-collection: 1.0.2
-      which-typed-array: 1.1.20
+      which-typed-array: 1.1.24
     dev: true
 
   /which-collection@1.0.2:
@@ -7648,12 +6666,12 @@ packages:
     resolution: {integrity: sha512-iBdZ57RDvnOR9AGBhML2vFZf7h8vmBjhoaZqODJBFWHVtKkDmKuHai3cx5PgVMrX5YDNp27AofYbAwctSS+vhQ==}
     dev: false
 
-  /which-typed-array@1.1.20:
-    resolution: {integrity: sha512-LYfpUkmqwl0h9A2HL09Mms427Q1RZWuOHsukfVcKRq9q95iQxdw0ix1JQrqbcDR9PH1QDwf5Qo8OZb5lksZ8Xg==}
+  /which-typed-array@1.1.24:
+    resolution: {integrity: sha512-wk4Mf4pR5mRP7eYuuTBCIQ9d0ud2Fv2jRLQpfgnRjbOxAFHmjKFValgTpitVKzJJS8ajnYQV2Du1SZ8j6b/EUQ==}
     engines: {node: '>= 0.4'}
     dependencies:
       available-typed-arrays: 1.0.7
-      call-bind: 1.0.8
+      call-bind: 1.0.9
       call-bound: 1.0.4
       for-each: 0.3.5
       get-proto: 1.0.1
@@ -7689,15 +6707,15 @@ packages:
     dependencies:
       ansi-styles: 6.2.3
       string-width: 7.2.0
-      strip-ansi: 7.1.2
+      strip-ansi: 7.2.0
     dev: true
 
   /y18n@4.0.3:
     resolution: {integrity: sha512-JKhqTOwSrqNA1NY5lSztJ1GrBiUodLMmIZuLiDaMRJ+itFd+ABVE8XBjOvIWL+rSqNDC74LCSFmlb/U4UZ4hJQ==}
     dev: false
 
-  /yaml@2.8.2:
-    resolution: {integrity: sha512-mplynKqc1C2hTVYxd0PU2xQAc22TI1vShAYGksCCfxbn/dFwnHTNi1bvYsBTkhdUNtGIf5xNOg938rrSSYvS9A==}
+  /yaml@2.9.1:
+    resolution: {integrity: sha512-3NxN8+78OdzbT7C/WjGsyfPAtJaN3FNDsWxv7Y7mcDsT/oOmgW8BpyQQFFBnvZE3j9Y2Sdz1ULFLezL7Eb2yFw==}
     engines: {node: '>= 14.6'}
     hasBin: true
     dev: true
diff --git a/web/public/social/cloud-v3.png b/web/public/social/cloud-v3.png
new file mode 100644
index 000000000..5ab66f53f
Binary files /dev/null and b/web/public/social/cloud-v3.png differ
diff --git a/web/scripts/check-i18n.mjs b/web/scripts/check-i18n.mjs
index fd426046a..514750725 100755
--- a/web/scripts/check-i18n.mjs
+++ b/web/scripts/check-i18n.mjs
@@ -88,7 +88,11 @@ function extractKeys(filePath) {
 }
 
 function main() {
-  const files = readdirSync(LOCALES_DIR).filter((f) => f.endsWith('.ts'));
+  // Shared translation modules (for example bot-setup.ts) are not locales.
+  // Match language-region and language-script filenames used by the app.
+  const files = readdirSync(LOCALES_DIR).filter((f) =>
+    /^[a-z]{2,3}-(?:[A-Z]{2}|[A-Z][a-z]{3})\.ts$/.test(f),
+  );
 
   if (!files.includes(REFERENCE)) {
     console.error(`Reference file ${REFERENCE} not found in ${LOCALES_DIR}`);
diff --git a/web/src/app/global.css b/web/src/app/global.css
index a1be44406..7458cdc84 100644
--- a/web/src/app/global.css
+++ b/web/src/app/global.css
@@ -195,3 +195,28 @@
 body:has([data-tour-modal='true']) [data-tour-modal='false'] {
   display: none;
 }
+
+/* Reserve native scrollbar space before content overflows. Extending the
+   utilities also covers responsive variants, nested panels and popup menus.
+   Horizontal-only scrollers and custom overlay scroll areas stay unchanged. */
+@utility overflow-auto {
+  scrollbar-gutter: stable;
+}
+
+@utility overflow-y-auto {
+  scrollbar-gutter: stable;
+}
+
+@utility overflow-scroll {
+  scrollbar-gutter: stable;
+}
+
+@utility overflow-y-scroll {
+  scrollbar-gutter: stable;
+}
+
+@layer base {
+  textarea {
+    scrollbar-gutter: stable;
+  }
+}
diff --git a/web/src/app/home/agents/AgentDetailContent.tsx b/web/src/app/home/agents/AgentDetailContent.tsx
index 85184b337..a371b3ef5 100644
--- a/web/src/app/home/agents/AgentDetailContent.tsx
+++ b/web/src/app/home/agents/AgentDetailContent.tsx
@@ -1,11 +1,11 @@
 import EntityLoadState from '@/components/EntityLoadState';
-import { useEffect, useRef, useState } from 'react';
+import { useCallback, useEffect, useRef, useState } from 'react';
 import { useNavigate } from 'react-router-dom';
 import { useTranslation } from 'react-i18next';
 import { toast } from 'sonner';
 import { AlertTriangle, Trash2 } from 'lucide-react';
 import { httpClient } from '@/app/infra/http/HttpClient';
-import { useCurrentWorkspace } from '@/app/infra/http';
+import { useCurrentWorkspace, userInfo } from '@/app/infra/http';
 import { Agent, AgentPlatformTool } from '@/app/infra/entities/api';
 import { useSidebarData } from '@/app/home/components/home-sidebar/SidebarDataContext';
 import ProcessorDetailWorkbench from '@/app/home/components/processor-detail/ProcessorDetailWorkbench';
@@ -33,6 +33,9 @@ import AgentFormComponent, {
   RunnerStatus,
 } from './components/AgentFormComponent';
 
+// Keep the last checked display state across navigation within this app session.
+const runnerStatusCache = new Map<string, RunnerStatus>();
+
 export default function AgentDetailContent({
   id,
   pipelineRevision,
@@ -42,7 +45,7 @@ export default function AgentDetailContent({
 }) {
   const isCreateMode = id === 'new';
   const navigate = useNavigate();
-  const { t } = useTranslation();
+  const { t, i18n } = useTranslation();
   const currentWorkspace = useCurrentWorkspace();
   const canManage =
     currentWorkspace?.permissions.includes('resource.manage') ?? false;
@@ -59,7 +62,43 @@ export default function AgentDetailContent({
   const [basicInfoOpen, setBasicInfoOpen] = useState(false);
   const [deleteConfirmOpen, setDeleteConfirmOpen] = useState(false);
   const [deleting, setDeleting] = useState(false);
-  const [runnerStatus, setRunnerStatus] = useState<RunnerStatus | null>(null);
+  const runnerStatusKey = JSON.stringify([
+    userInfo?.account_uuid,
+    currentWorkspace?.workspace.uuid,
+    id,
+    i18n.language,
+  ]);
+  const [runnerStatusState, setRunnerStatusState] = useState<{
+    key: string;
+    status: RunnerStatus;
+  } | null>(null);
+  const runnerStatus =
+    runnerStatusState?.key === runnerStatusKey
+      ? runnerStatusState.status
+      : (runnerStatusCache.get(runnerStatusKey) ?? null);
+  const updateRunnerStatus = useCallback(
+    (status: RunnerStatus) => {
+      setRunnerStatusState((previous) => {
+        if (status.tone === 'neutral') {
+          const cached =
+            previous?.key === runnerStatusKey
+              ? previous.status
+              : runnerStatusCache.get(runnerStatusKey);
+          if (cached) return { key: runnerStatusKey, status: cached };
+        }
+        return { key: runnerStatusKey, status };
+      });
+      if (status.tone !== 'neutral' && !formDirty) {
+        runnerStatusCache.delete(runnerStatusKey);
+        runnerStatusCache.set(runnerStatusKey, status);
+        if (runnerStatusCache.size > 100) {
+          const oldest = runnerStatusCache.keys().next().value;
+          if (oldest !== undefined) runnerStatusCache.delete(oldest);
+        }
+      }
+    },
+    [runnerStatusKey, formDirty],
+  );
   const [availableEventTypes, setAvailableEventTypes] = useState<string[]>([
     'message.received',
   ]);
@@ -79,10 +118,6 @@ export default function AgentDetailContent({
     return () => setDetailEntityName(null);
   }, [id, isCreateMode, pipelines, setDetailEntityName, t]);
 
-  useEffect(() => {
-    setRunnerStatus(null);
-  }, [id]);
-
   useEffect(() => {
     if (isCreateMode) return;
     let cancelled = false;
@@ -258,7 +293,7 @@ export default function AgentDetailContent({
                 }}
                 onDirtyChange={setFormDirty}
                 onSavingChange={setFormSaving}
-                onRunnerStatusChange={setRunnerStatus}
+                onRunnerStatusChange={updateRunnerStatus}
                 onSupportedEventPatternsChange={setSupportedEventPatterns}
                 onPlatformToolsChange={setPlatformTools}
                 guideEnabled={canManage}
diff --git a/web/src/app/home/agents/PluginProcessorDetailContent.tsx b/web/src/app/home/agents/PluginProcessorDetailContent.tsx
index 53e08f708..938809088 100644
--- a/web/src/app/home/agents/PluginProcessorDetailContent.tsx
+++ b/web/src/app/home/agents/PluginProcessorDetailContent.tsx
@@ -6,7 +6,7 @@ import { useCallback, useEffect, useMemo, useRef, useState } from 'react';
 import { useTranslation } from 'react-i18next';
 import { Link, useSearchParams } from 'react-router-dom';
 import { eventPatternLabel } from '@/app/home/components/event-patterns/event-pattern-groups';
-import { RefreshCw, Trash2, ScrollText, Settings2 } from 'lucide-react';
+import { RefreshCw, Trash2, Clock3, Puzzle } from 'lucide-react';
 import isEqual from 'lodash/isEqual';
 import { toast } from 'sonner';
 import type {
@@ -21,6 +21,7 @@ import { Button } from '@/components/ui/button';
 import { extractI18nObject } from '@/i18n/I18nProvider';
 import { Badge } from '@/components/ui/badge';
 import { Alert, AlertDescription } from '@/components/ui/alert';
+import LoadErrorState from '@/components/LoadErrorState';
 import { ScrollArea } from '@/components/ui/scroll-area';
 import ProcessorDetailWorkbench from '@/app/home/components/processor-detail/ProcessorDetailWorkbench';
 import EntityTitleEditButton from '@/app/home/components/entity-basic-info/EntityTitleEditButton';
@@ -29,6 +30,10 @@ import PluginProcessorTrace, {
   ProcessorPayload,
 } from './components/PluginProcessorTrace';
 import ProcessorRunList from './components/ProcessorRunList';
+import {
+  processorRunDuration,
+  formatRunDuration,
+} from './components/processor-run-timing';
 import PluginProcessorSettings from './components/PluginProcessorSettings';
 import DynamicFormComponent from '@/app/home/components/dynamic-form/DynamicFormComponent';
 
@@ -52,9 +57,19 @@ export default function PluginProcessorDetailContent({
   onSaved: () => void;
 }) {
   const { t } = useTranslation();
-  const [searchParams] = useSearchParams();
-  const [activeTab, setActiveTab] = useState(
-    searchParams.get('tab') === 'logs' ? 'logs' : 'config',
+  const [, setSearchParams] = useSearchParams();
+  const setActiveTab = useCallback(
+    (tab: string) => {
+      setSearchParams(
+        (current) => {
+          const next = new URLSearchParams(current);
+          next.set('tab', tab);
+          return next;
+        },
+        { preventScrollReset: true },
+      );
+    },
+    [setSearchParams],
   );
   const [platformTools, setPlatformTools] = useState<AgentPlatformTool[]>([]);
   const toolLabels = Object.fromEntries(
@@ -74,7 +89,11 @@ export default function PluginProcessorDetailContent({
     componentRef,
     parameters: initialParameters,
   });
+  const [needsInitialSave, setNeedsInitialSave] = useState(
+    Boolean(agent.component_ref) && !agent.supported_event_patterns?.length,
+  );
   const dirty =
+    needsInitialSave ||
     componentRef !== savedConfig.componentRef ||
     !isEqual(parameters, savedConfig.parameters);
   const [runs, setRuns] = useState<ProcessorRun[]>([]);
@@ -132,7 +151,7 @@ export default function PluginProcessorDetailContent({
         : []),
       {
         id: 'logs',
-        target: '[data-guide="event-processor-form-tab-logs"]',
+        target: '[data-guide="event-processor-form-monitoring"]',
         title: t('guidedTour.pluginProcessor.logs.title'),
         description: t('guidedTour.pluginProcessor.logs.description'),
         onEnter: () => setActiveTab('logs'),
@@ -149,7 +168,27 @@ export default function PluginProcessorDetailContent({
           ]
         : []),
     ],
-    [t, available, hasParameters, canOperate],
+    [t, available, hasParameters, canOperate, setActiveTab],
+  );
+
+  const applyInstalledProcessor = useCallback(
+    (component: RunnerDescriptor) => {
+      setComponents((current) => [
+        ...current.filter((item) => item.id !== component.id),
+        component,
+      ]);
+      setComponentRef(component.id);
+      setParameters(
+        Object.fromEntries(
+          (component.config_schema ?? [])
+            .filter((field) => field.default !== undefined)
+            .map((field) => [field.name, field.default]),
+        ),
+      );
+      setActiveTab('config');
+      validate.current = null;
+    },
+    [setActiveTab],
   );
 
   const load = useCallback(async () => {
@@ -324,6 +363,7 @@ export default function PluginProcessorDetailContent({
       toast.success(t('agents.saveSuccess'));
       onSaved();
       setSavedConfig({ componentRef, parameters });
+      setNeedsInitialSave(false);
       await load();
       return true;
     } catch {
@@ -335,20 +375,13 @@ export default function PluginProcessorDetailContent({
   }
 
   const logsContent = (
-    <div className="flex h-full min-h-0 flex-col gap-3">
-      <form
-        id="event-processor-form"
-        onSubmit={(event) => {
-          event.preventDefault();
-          void save();
-        }}
-      />
+    <div className="flex h-full min-h-0 flex-col gap-4 p-4">
       {failed && (
-        <Alert variant="destructive">
-          <AlertDescription>
-            {t('agents.eventProcessor.loadError')}
-          </AlertDescription>
-        </Alert>
+        <LoadErrorState
+          compact
+          title={t('agents.eventProcessor.loadError')}
+          onRetry={refreshLatestRun}
+        />
       )}
       <div className="flex shrink-0 items-center justify-between gap-2">
         <span className="text-sm font-medium">
@@ -356,98 +389,123 @@ export default function PluginProcessorDetailContent({
           <span className="text-muted-foreground">({runs.length})</span>
         </span>
         <Button
-          variant="ghost"
-          size="icon"
+          variant="outline"
+          size="sm"
           aria-label={t('agents.eventProcessor.refresh')}
           onClick={() => void refreshLatestRun()}
         >
           <RefreshCw className="size-4" />
+          {t('agents.eventProcessor.refresh')}
         </Button>
       </div>
-      {runs.length > 0 && (
-        <ProcessorRunList
-          runs={runs}
-          selectedId={selected?.run_id}
-          onSelect={(run) => void openRun(run)}
-          footer={
-            cursor !== null ? (
-              <Button
-                className="w-full"
-                variant="ghost"
-                disabled={pagingRuns}
-                onClick={() => void loadMoreRuns()}
-              >
-                {t('agents.eventProcessor.loadMore')}
-              </Button>
-            ) : undefined
-          }
-        />
-      )}
-      <ScrollArea className="min-h-0 flex-1">
-        <div className="space-y-2 pr-3">
-          {!selected ? (
-            <Alert>
-              <AlertDescription>
-                {loading
-                  ? t('common.loading')
-                  : t('agents.eventProcessor.noRuns')}
-                <Button asChild variant="link" className="h-auto px-0">
-                  <Link to="/home/bots">
-                    {t('agents.eventProcessor.bindBot')}
-                  </Link>
-                </Button>
-              </AlertDescription>
-            </Alert>
-          ) : (
-            <>
-              <div className="border-b pb-2">
-                <p className="text-sm font-medium">
-                  {eventPatternLabel(selected.metadata.event_type ?? '', t)}
-                </p>
-                <p className="text-xs text-muted-foreground">
-                  {new Date(selected.created_at * 1000).toLocaleString()}
-                </p>
-              </div>
-              <Badge
-                variant={
-                  selected.status === 'failed' ? 'destructive' : 'outline'
-                }
-              >
-                {t(`agents.eventProcessor.status_${selected.status}`, {
-                  defaultValue: selected.status,
-                })}
-              </Badge>
-              <ProcessorPayload
-                title={t('agents.eventProcessor.input')}
-                value={selected.metadata.input_event}
-              />
-              {selected.metadata.delivery != null && (
-                <ProcessorPayload
-                  title={t('agents.eventProcessor.destination')}
-                  value={selected.metadata.delivery}
-                />
-              )}
-              <PluginProcessorTrace events={events} toolLabels={toolLabels} />
-              {selected.status === 'failed' && selected.status_reason && (
-                <Alert variant="destructive">
-                  <AlertDescription className="break-words">
-                    {selected.status_reason}
-                  </AlertDescription>
-                </Alert>
-              )}
-              {eventCursor !== null && (
+      <div className="grid min-h-0 flex-1 grid-cols-1 gap-6 lg:grid-cols-[18rem_minmax(0,1fr)]">
+        {runs.length > 0 && (
+          <ProcessorRunList
+            className="max-h-64 overflow-y-auto rounded-lg border lg:max-h-none lg:h-full"
+            runs={runs}
+            selectedId={selected?.run_id}
+            onSelect={(run) => void openRun(run)}
+            footer={
+              cursor !== null ? (
                 <Button
+                  className="w-full"
                   variant="ghost"
-                  disabled={pagingEvents}
-                  onClick={() => void loadMoreEvents()}
+                  disabled={pagingRuns}
+                  onClick={() => void loadMoreRuns()}
                 >
                   {t('agents.eventProcessor.loadMore')}
                 </Button>
-              )}
-            </>
-          )}
-        </div>
-      </ScrollArea>
+              ) : undefined
+            }
+          />
+        )}
+        <ScrollArea className="min-h-0 flex-1">
+          <div className="space-y-4 pr-2">
+            {!selected ? (
+              <Alert>
+                <AlertDescription>
+                  {loading
+                    ? t('common.loading')
+                    : t('agents.eventProcessor.noRuns')}
+                  <Button asChild variant="link" className="h-auto px-0">
+                    <Link to="/home/bots">
+                      {t('agents.eventProcessor.bindBot')}
+                    </Link>
+                  </Button>
+                </AlertDescription>
+              </Alert>
+            ) : (
+              <>
+                <div className="flex flex-wrap items-center justify-between gap-3 border-b pb-3">
+                  <div className="space-y-1.5">
+                    <p className="font-medium">
+                      {eventPatternLabel(selected.metadata.event_type ?? '', t)}
+                    </p>
+                    <p className="text-xs text-muted-foreground">
+                      {new Date(selected.created_at * 1000).toLocaleString()}
+                    </p>
+                  </div>
+                  <div className="flex items-center gap-3">
+                    {processorRunDuration(selected) !== null && (
+                      <span className="flex items-center gap-1.5 text-xs tabular-nums text-muted-foreground">
+                        <Clock3 className="size-3.5" />
+                        {formatRunDuration(processorRunDuration(selected)!)}
+                      </span>
+                    )}
+                    <Badge status={selected.status}>
+                      {t(`agents.eventProcessor.status_${selected.status}`, {
+                        defaultValue: selected.status,
+                      })}
+                    </Badge>
+                  </div>
+                </div>
+                <div className="space-y-2 rounded-lg border p-3">
+                  <h3 className="text-sm font-medium">
+                    {t('agents.eventProcessor.input')}
+                  </h3>
+                  <ProcessorPayload
+                    title={t('agents.monitoring.eventData')}
+                    value={selected.metadata.input_event}
+                  />
+                  {selected.metadata.delivery != null && (
+                    <ProcessorPayload
+                      title={t('agents.eventProcessor.destination')}
+                      value={selected.metadata.delivery}
+                    />
+                  )}
+                </div>
+                <section className="space-y-3">
+                  <h2 className="flex items-center gap-2 text-sm font-medium">
+                    <Puzzle className="size-4 text-amber-600 dark:text-amber-400" />
+                    {t('agents.eventProcessor.type')} ·{' '}
+                    {t('agents.eventProcessor.logsTab')}
+                  </h2>
+                  <PluginProcessorTrace
+                    events={events}
+                    toolLabels={toolLabels}
+                  />
+                </section>
+                {selected.status === 'failed' && selected.status_reason && (
+                  <Alert variant="destructive">
+                    <AlertDescription className="break-words">
+                      {selected.status_reason}
+                    </AlertDescription>
+                  </Alert>
+                )}
+                {eventCursor !== null && (
+                  <Button
+                    variant="ghost"
+                    disabled={pagingEvents}
+                    onClick={() => void loadMoreEvents()}
+                  >
+                    {t('agents.eventProcessor.loadMore')}
+                  </Button>
+                )}
+              </>
+            )}
+          </div>
+        </ScrollArea>
+      </div>
     </div>
   );
 
@@ -456,6 +514,13 @@ export default function PluginProcessorDetailContent({
 
   return (
     <>
+      <form
+        id="event-processor-form"
+        onSubmit={(event) => {
+          event.preventDefault();
+          void save();
+        }}
+      />
       <GuidedTour
         enabled={canManage}
         storageKey="langbot_plugin_processor_setup_guide_v1"
@@ -464,11 +529,22 @@ export default function PluginProcessorDetailContent({
       />
       <ProcessorDetailWorkbench
         title={`${agent.emoji || '🧩'} ${agent.name}`}
+        titleBadge={
+          <Badge
+            variant="outline"
+            className="gap-1.5 border-amber-500/20 bg-amber-500/10 text-amber-700 dark:text-amber-300"
+          >
+            <Puzzle className="size-3.5" />
+            {t('agents.eventProcessor.type')}
+          </Badge>
+        }
         titleAction={
           canManage ? <EntityTitleEditButton onClick={onEdit} /> : undefined
         }
         titleControls={
           <PluginProcessorSettings
+            installScope={`event-processor:${id}`}
+            onInstalled={applyInstalledProcessor}
             components={components}
             value={componentRef}
             disabled={!canManage || saving || loading}
@@ -506,50 +582,41 @@ export default function PluginProcessorDetailContent({
           ) : undefined
         }
         configTitle={t('agents.eventProcessor.type')}
-        configTabs={{
-          value: activeTab,
-          onValueChange: setActiveTab,
-          items: [
-            {
-              value: 'config',
-              label: t('agents.eventProcessor.configTab'),
-              icon: <Settings2 className="size-4" />,
-              content: (
-                <div className="h-full overflow-y-auto">
-                  {!component ? (
-                    <p className="text-sm text-muted-foreground">
-                      {t('agents.eventProcessor.selectComponent')}
-                    </p>
-                  ) : component.config_schema.length === 0 ? (
-                    <p className="text-sm text-muted-foreground">
-                      {t('agents.eventProcessor.noSettings')}
-                    </p>
-                  ) : (
-                    <fieldset disabled={!canManage || saving}>
-                      <DynamicFormComponent
-                        key={componentRef}
-                        itemConfigList={component.config_schema}
-                        initialValues={parameters}
-                        onSubmit={(values) =>
-                          setParameters(values as Record<string, unknown>)
-                        }
-                        onValidate={(fn) => {
-                          validate.current = fn;
-                        }}
-                      />
-                    </fieldset>
-                  )}
-                </div>
-              ),
-            },
-            {
-              value: 'logs',
-              label: t('agents.eventProcessor.logsTab'),
-              icon: <ScrollText className="size-4" />,
-              content: logsContent,
-            },
-          ],
+        configIcon={
+          <Puzzle className="size-4 text-amber-600 dark:text-amber-400" />
+        }
+        monitoring={{
+          label: t('pipelines.monitoring.title'),
+          workbenchLabel: t('pipelines.monitoring.workbench'),
+          content: logsContent,
         }}
+        configContent={
+          <div className="h-full overflow-y-auto">
+            {!component ? (
+              <p className="text-sm text-muted-foreground">
+                {t('agents.eventProcessor.selectComponent')}
+              </p>
+            ) : component.config_schema.length === 0 ? (
+              <p className="text-sm text-muted-foreground">
+                {t('agents.eventProcessor.noSettings')}
+              </p>
+            ) : (
+              <fieldset disabled={!canManage || saving}>
+                <DynamicFormComponent
+                  key={componentRef}
+                  itemConfigList={component.config_schema}
+                  initialValues={parameters}
+                  onSubmit={(values) =>
+                    setParameters(values as Record<string, unknown>)
+                  }
+                  onValidate={(fn) => {
+                    validate.current = fn;
+                  }}
+                />
+              </fieldset>
+            )}
+          </div>
+        }
         debugTitle={canOperate ? t('agents.debugTab') : undefined}
         debugDescription={t('agents.eventProcessor.debugNotice')}
         debugContent={
@@ -568,7 +635,6 @@ export default function PluginProcessorDetailContent({
                 hasUnsavedChanges={dirty}
                 beforeRun={save}
                 onRunFinished={() => {
-                  setActiveTab('logs');
                   void refreshLatestRun();
                 }}
                 supportedEventPatterns={component.supported_event_patterns}
diff --git a/web/src/app/home/agents/components/AgentApiToolPicker.tsx b/web/src/app/home/agents/components/AgentApiToolPicker.tsx
index 6b7032c53..ca32fe68a 100644
--- a/web/src/app/home/agents/components/AgentApiToolPicker.tsx
+++ b/web/src/app/home/agents/components/AgentApiToolPicker.tsx
@@ -206,6 +206,22 @@ export default function AgentApiToolPicker({
       ? platformCatalogAvailable
       : hostCatalogAvailable;
 
+  const isSelected = (tool: ToolEntry) =>
+    (tool.kind === 'platform' ? selectedPlatform : selectedHost).has(tool.name);
+
+  const setGroup = (tools: ToolEntry[], checked: boolean) => {
+    const platform = new Set(platformValue);
+    const host = new Set(hostValue);
+    for (const tool of tools) {
+      const selection = tool.kind === 'platform' ? platform : host;
+      if (checked) selection.add(tool.name);
+      else selection.delete(tool.name);
+    }
+    if (tools.some((tool) => tool.kind === 'platform'))
+      onPlatformChange([...platform]);
+    if (tools.some((tool) => tool.kind === 'host')) onHostChange([...host]);
+  };
+
   return (
     <div className="space-y-4">
       <div className="flex flex-col gap-3 sm:flex-row sm:items-start sm:justify-between">
@@ -275,8 +291,24 @@ export default function AgentApiToolPicker({
         <div className="overflow-hidden rounded-lg border bg-background">
           {groupedEntries.map(([group, groupTools], groupIndex) => (
             <section key={group} className={cn(groupIndex > 0 && 'border-t')}>
-              <div className="bg-muted/30 px-3 py-1.5 text-xs font-medium text-muted-foreground">
-                {groupLabel(group)}
+              <div className="flex items-center justify-between gap-3 bg-muted/30 px-3 py-2 text-xs font-medium text-muted-foreground">
+                <span>{groupLabel(group)}</span>
+                <label className="flex cursor-pointer items-center gap-2">
+                  <Checkbox
+                    checked={
+                      groupTools.every(isSelected)
+                        ? true
+                        : groupTools.some(isSelected)
+                          ? 'indeterminate'
+                          : false
+                    }
+                    onCheckedChange={(checked) =>
+                      setGroup(groupTools, checked === true)
+                    }
+                    aria-label={`${groupLabel(group)} · ${t('agents.apiToolsSelectAll')}`}
+                  />
+                  {t('agents.apiToolsSelectAll')}
+                </label>
               </div>
               <div className="divide-y">
                 {groupTools.map((tool) => {
@@ -287,8 +319,7 @@ export default function AgentApiToolPicker({
                   const expanded = expandedTool === tool.key;
                   const parameterNames = Object.keys(
                     (tool.parameters.properties as
-                      | Record<string, unknown>
-                      | undefined) ?? {},
+                      Record<string, unknown> | undefined) ?? {},
                   );
                   return (
                     <div key={tool.key} className="bg-background px-3 py-2.5">
diff --git a/web/src/app/home/agents/components/AgentCreateContent.tsx b/web/src/app/home/agents/components/AgentCreateContent.tsx
index 9c5fcf852..c7d96248f 100644
--- a/web/src/app/home/agents/components/AgentCreateContent.tsx
+++ b/web/src/app/home/agents/components/AgentCreateContent.tsx
@@ -1,4 +1,4 @@
-import { useState } from 'react';
+import { useCallback, useEffect, useRef, useState } from 'react';
 import { useForm } from 'react-hook-form';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { z } from 'zod';
@@ -27,17 +27,55 @@ import {
 import { Input } from '@/components/ui/input';
 import EmojiPicker from '@/components/ui/emoji-picker';
 import ProcessorTypeDiagram from './ProcessorTypeDiagram';
+import RunnerSelect from './RunnerSelect';
+import type { GetAgentMetadataResponseData } from '@/app/infra/entities/api';
+import { getDefaultValues } from '@/app/home/components/dynamic-form/DynamicFormItemConfig';
+import { extractI18nObject } from '@/i18n/I18nProvider';
 
 export default function AgentCreateContent({
   onCreated,
+  allowedKinds,
+  embedded = false,
 }: {
   onCreated: (agentId: string) => void;
+  allowedKinds?: AgentKind[];
+  embedded?: boolean;
 }) {
   const { t } = useTranslation();
-  const [kind, setKind] = useState<AgentKind>('agent');
+  const [kind, setKind] = useState<AgentKind>(allowedKinds?.[0] ?? 'pipeline');
+  const [metadata, setMetadata] = useState<GetAgentMetadataResponseData>();
+  const [runner, setRunner] = useState('');
+  const [loading, setLoading] = useState(true);
+  const createdPipeline = useRef<string | null>(null);
+  const loadMetadata = useCallback(async () => {
+    setLoading(true);
+    try {
+      setMetadata(await httpClient.getAgentMetadata());
+    } catch {
+      toast.error(t('agents.createError'));
+    } finally {
+      setLoading(false);
+    }
+  }, [t]);
+  useEffect(() => {
+    void loadMetadata();
+  }, [loadMetadata]);
+  const runnerOptions =
+    kind === 'event_processor'
+      ? (metadata?.event_processors ?? []).map((item) => ({
+          name: item.id,
+          label: { en_US: item.id, zh_Hans: item.id, ...item.label },
+        }))
+      : (metadata?.runner_config?.stages
+          .find((stage) => stage.name === 'runner')
+          ?.config.find((item) => item.name === 'id')?.options ?? []);
+  const selectorLabel = t(
+    kind === 'event_processor'
+      ? 'agents.selectProcessorPlugin'
+      : 'agents.selectAgentRunner',
+  );
   const formSchema = z.object({
     name: z.string().min(1, { message: t('agents.nameRequired') }),
-    description: z.string().optional(),
     emoji: z.string().optional(),
   });
   type FormValues = z.infer<typeof formSchema>;
@@ -45,11 +83,25 @@ export default function AgentCreateContent({
     resolver: zodResolver(formSchema),
     defaultValues: {
       name: '',
-      description: '',
-      emoji: '🤖',
+      emoji: '⚙️',
     },
   });
 
+  function selectRunner(value: string, name?: string) {
+    setRunner(value);
+    if (!form.getValues('name').trim()) {
+      const option = runnerOptions.find((item) => item.name === value);
+      const displayName =
+        name || (option ? extractI18nObject(option.label) : '');
+      if (displayName) {
+        form.setValue('name', displayName, {
+          shouldDirty: true,
+          shouldValidate: true,
+        });
+      }
+    }
+  }
+
   function handleKindChange(nextKind: AgentKind) {
     const previousDefaultEmoji =
       kind === 'pipeline' ? '⚙️' : kind === 'event_processor' ? '🧩' : '🤖';
@@ -59,6 +111,9 @@ export default function AgentCreateContent({
         : nextKind === 'event_processor'
           ? '🧩'
           : '🤖';
+    if ((kind === 'event_processor') !== (nextKind === 'event_processor')) {
+      setRunner('');
+    }
     setKind(nextKind);
     const currentEmoji = form.getValues('emoji');
     if (!currentEmoji || currentEmoji === previousDefaultEmoji) {
@@ -67,28 +122,93 @@ export default function AgentCreateContent({
   }
 
   async function handleSubmit(values: FormValues) {
-    return httpClient
-      .createAgent({
-        kind,
-        name: values.name,
-        description: values.description ?? '',
-        emoji:
-          values.emoji ||
-          (kind === 'pipeline'
-            ? '⚙️'
-            : kind === 'event_processor'
-              ? '🧩'
-              : '🤖'),
-      })
-      .then((resp) => {
+    if (!runner) {
+      toast.error(selectorLabel);
+      return;
+    }
+    const runnerStage = metadata?.runner_config?.stages.find(
+      (stage) => stage.name === runner,
+    );
+    const parameters = runnerStage ? getDefaultValues(runnerStage.config) : {};
+    const config = {
+      runner: { id: runner },
+      runner_config: { [runner]: parameters },
+    };
+    try {
+      if (
+        kind === 'pipeline' &&
+        runnerStage?.config.some((field) =>
+          [
+            'llm-model-selector',
+            'select-llm-model',
+            'model-fallback-selector',
+          ].includes(field.type),
+        )
+      ) {
+        const { uuid } = await httpClient.getDefaultModel();
+        if (uuid) {
+          for (const field of runnerStage.config) {
+            if (
+              ['llm-model-selector', 'select-llm-model'].includes(field.type) &&
+              !parameters[field.name]
+            )
+              parameters[field.name] = uuid;
+            if (
+              field.type === 'model-fallback-selector' &&
+              !parameters[field.name]?.primary
+            ) {
+              parameters[field.name] = {
+                ...parameters[field.name],
+                primary: uuid,
+                fallbacks: parameters[field.name]?.fallbacks ?? [],
+              };
+            }
+          }
+        }
+      }
+      if (kind === 'pipeline') {
+        // Keep the created ID if configuration fails, so retry does not create duplicates.
+        const uuid =
+          createdPipeline.current ??
+          (
+            await httpClient.createAgent({
+              kind,
+              name: values.name,
+              description: '',
+              emoji: values.emoji,
+            })
+          ).uuid;
+        createdPipeline.current = uuid;
+        const { pipeline } = await httpClient.getPipeline(uuid);
+        await httpClient.updatePipeline(uuid, {
+          name: values.name,
+          emoji: values.emoji,
+          config: {
+            ...pipeline.config,
+            ai: { ...pipeline.config.ai, ...config },
+          },
+        });
         toast.success(t('agents.createSuccess'));
-        onCreated(resp.uuid);
-      })
-      .catch((err) => {
-        toast.error(t('agents.createError') + err.msg);
-      });
+        onCreated(uuid);
+      } else {
+        const response = await httpClient.createAgent({
+          kind,
+          name: values.name,
+          description: '',
+          emoji: values.emoji,
+          component_ref: runner,
+          ...(kind === 'event_processor' ? {} : { config }),
+        });
+        toast.success(t('agents.createSuccess'));
+        onCreated(response.uuid);
+      }
+    } catch (error) {
+      toast.error(
+        t('agents.createError') +
+          String((error as { msg?: string }).msg ?? error),
+      );
+    }
   }
-
   const typeOptions = [
     {
       kind: 'pipeline' as const,
@@ -111,14 +231,22 @@ export default function AgentCreateContent({
   ];
   return (
     <div className="flex h-full flex-col">
-      <div className="flex items-center justify-between pb-4 shrink-0">
-        <h1 className="text-xl font-semibold">
-          {t('agents.eventProcessor.createPageTitle')}
-        </h1>
+      <div
+        className={
+          embedded
+            ? 'order-last flex justify-end border-t pt-4 shrink-0'
+            : 'flex items-center justify-between pb-4 shrink-0'
+        }
+      >
+        {!embedded && (
+          <h1 className="text-xl font-semibold">
+            {t('agents.eventProcessor.createPageTitle')}
+          </h1>
+        )}
         <Button
           type="submit"
           form="agent-create-form"
-          disabled={form.formState.isSubmitting}
+          disabled={form.formState.isSubmitting || loading || !runner}
         >
           {t('common.submit')}
         </Button>
@@ -139,9 +267,11 @@ export default function AgentCreateContent({
                   >
                     {t('agents.chooseType')}
                   </h2>
-                  <p className="mt-1 text-sm text-muted-foreground">
-                    {t('agents.chooseTypeDescription')}
-                  </p>
+                  {!embedded && (
+                    <p className="mt-1 text-sm text-muted-foreground">
+                      {t('agents.chooseTypeDescription')}
+                    </p>
+                  )}
                 </div>
 
                 <ToggleGroup
@@ -154,30 +284,36 @@ export default function AgentCreateContent({
                   spacing={3}
                   className="grid w-full gap-3 sm:grid-cols-2 lg:grid-cols-1"
                 >
-                  {typeOptions.map((option) => {
-                    const Icon = option.icon;
-                    return (
-                      <ToggleGroupItem
-                        key={option.kind}
-                        value={option.kind}
-                        data-processor-kind={option.kind}
-                        aria-label={`${option.title} ${option.description}`}
-                        className="h-auto min-h-28 w-full items-start justify-start gap-3 rounded-lg border px-4 py-4 text-left whitespace-normal shadow-none hover:bg-muted/40 data-[state=on]:border-[#2288ee]/50 data-[state=on]:bg-blue-50/60 data-[state=on]:text-foreground data-[state=on]:shadow-none dark:data-[state=on]:border-blue-500/50 dark:data-[state=on]:bg-blue-500/10"
-                      >
-                        <span className="flex size-9 shrink-0 items-center justify-center rounded-md border bg-background text-[#2288ee] shadow-xs">
-                          <Icon className="size-4" />
-                        </span>
-                        <span className="min-w-0 space-y-1.5">
-                          <span className="block text-sm font-medium text-foreground">
-                            {option.title}
-                          </span>
-                          <span className="block text-sm font-normal leading-relaxed text-muted-foreground">
-                            {option.description}
-                          </span>
-                        </span>
-                      </ToggleGroupItem>
-                    );
-                  })}
+                  {typeOptions
+                    .filter(
+                      (option) =>
+                        !allowedKinds || allowedKinds.includes(option.kind),
+                    )
+                    .map((option) => {
+                      const Icon = option.icon;
+                      return (
+                        <div key={option.kind} className="relative w-full">
+                          <ToggleGroupItem
+                            value={option.kind}
+                            data-processor-kind={option.kind}
+                            aria-label={`${option.title} ${option.description}`}
+                            className="h-auto min-h-28 w-full items-start justify-start gap-3 rounded-lg border px-4 py-4 text-left whitespace-normal shadow-none hover:bg-muted/40 data-[state=on]:border-[#2288ee]/50 data-[state=on]:bg-blue-50/60 data-[state=on]:text-foreground data-[state=on]:shadow-none dark:data-[state=on]:border-blue-500/50 dark:data-[state=on]:bg-blue-500/10"
+                          >
+                            <span className="flex size-9 shrink-0 items-center justify-center rounded-md border bg-background text-[#2288ee] shadow-xs">
+                              <Icon className="size-4" />
+                            </span>
+                            <span className="min-w-0 space-y-1.5">
+                              <span className="block text-sm font-medium text-foreground">
+                                {option.title}
+                              </span>
+                              <span className="block text-sm font-normal leading-relaxed text-muted-foreground">
+                                {option.description}
+                              </span>
+                            </span>
+                          </ToggleGroupItem>
+                        </div>
+                      );
+                    })}
                 </ToggleGroup>
               </section>
 
@@ -185,7 +321,7 @@ export default function AgentCreateContent({
                 <CardHeader>
                   <CardTitle>{t('agents.basicInfo')}</CardTitle>
                   <CardDescription>
-                    {t('agents.basicInfoDescription')}
+                    {t('agents.createBasicInfoDescription')}
                   </CardDescription>
                 </CardHeader>
                 <CardContent>
@@ -195,6 +331,58 @@ export default function AgentCreateContent({
                       onSubmit={form.handleSubmit(handleSubmit)}
                       className="space-y-4"
                     >
+                      <div className="space-y-2">
+                        <p className="text-sm font-medium">{selectorLabel}</p>
+                        <RunnerSelect
+                          key={kind === 'event_processor' ? 'event' : 'agent'}
+                          usage={kind === 'event_processor' ? 'event' : 'agent'}
+                          options={runnerOptions}
+                          label={selectorLabel}
+                          value={runner}
+                          onValueChange={selectRunner}
+                          disabled={loading || form.formState.isSubmitting}
+                          installScope={`processor-create-${kind === 'event_processor' ? 'event' : 'agent'}`}
+                          onInstalled={(installed) => {
+                            setMetadata((previous) =>
+                              previous
+                                ? {
+                                    ...previous,
+                                    runner_config: installed.configTab,
+                                  }
+                                : previous,
+                            );
+                            selectRunner(
+                              installed.runner.name,
+                              extractI18nObject(installed.runner.label),
+                            );
+                          }}
+                          onEventProcessorInstalled={(component) => {
+                            setMetadata((previous) =>
+                              previous
+                                ? {
+                                    ...previous,
+                                    event_processors: [
+                                      ...(
+                                        previous.event_processors ?? []
+                                      ).filter(
+                                        (item) => item.id !== component.id,
+                                      ),
+                                      component,
+                                    ],
+                                  }
+                                : previous,
+                            );
+                            selectRunner(
+                              component.id,
+                              extractI18nObject({
+                                en_US: component.id,
+                                zh_Hans: component.id,
+                                ...component.label,
+                              }),
+                            );
+                          }}
+                        />
+                      </div>
                       <div className="flex gap-4 items-start">
                         <FormField
                           control={form.control}
@@ -229,27 +417,19 @@ export default function AgentCreateContent({
                           )}
                         />
                       </div>
-
-                      <FormField
-                        control={form.control}
-                        name="description"
-                        render={({ field }) => (
-                          <FormItem>
-                            <FormLabel>{t('common.description')}</FormLabel>
-                            <FormControl>
-                              <Input {...field} value={field.value ?? ''} />
-                            </FormControl>
-                            <FormMessage />
-                          </FormItem>
-                        )}
-                      />
                     </form>
                   </Form>
                 </CardContent>
               </Card>
             </div>
 
-            <Card className="min-h-[600px] overflow-hidden py-0 dark:border-white/16 lg:min-h-[680px]">
+            <Card
+              className={
+                embedded
+                  ? 'min-h-[360px] overflow-hidden py-0 dark:border-white/16 lg:min-h-[480px]'
+                  : 'min-h-[600px] overflow-hidden py-0 dark:border-white/16 lg:min-h-[680px]'
+              }
+            >
               <CardContent className="flex h-full items-center p-0">
                 <ProcessorTypeDiagram kind={kind} />
               </CardContent>
diff --git a/web/src/app/home/agents/components/AgentFormComponent.tsx b/web/src/app/home/agents/components/AgentFormComponent.tsx
index 6f7f93014..cfb9656ea 100644
--- a/web/src/app/home/agents/components/AgentFormComponent.tsx
+++ b/web/src/app/home/agents/components/AgentFormComponent.tsx
@@ -1,3 +1,4 @@
+import { useSnapshotDirty } from '@/app/infra/hooks/useSnapshotDirty';
 import EntityLoadState from '@/components/EntityLoadState';
 import {
   forwardRef,
@@ -9,7 +10,7 @@ import {
   useRef,
   useState,
 } from 'react';
-import { useForm } from 'react-hook-form';
+import { useForm, useWatch } from 'react-hook-form';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { z } from 'zod';
 import { useTranslation } from 'react-i18next';
@@ -157,7 +158,6 @@ function AgentFormComponent(
   const [activeSection, setActiveSection] =
     useState<AgentConfigSection>('runner');
   const isSavingRef = useRef(false);
-  const hasUnsavedChangesRef = useRef(false);
   const loadedHostToolPolicyRef = useRef<string[] | undefined>(undefined);
 
   const formSchema = z.object({
@@ -197,18 +197,20 @@ function AgentFormComponent(
 
   const savedSnapshotRef = useRef('');
   const initializedStagesRef = useRef<Set<string>>(new Set());
-  const watchedValues = form.watch();
-  const hasUnsavedChanges = (() => {
-    if (!savedSnapshotRef.current) return false;
-    return JSON.stringify(watchedValues) !== savedSnapshotRef.current;
-  })();
-  hasUnsavedChangesRef.current = hasUnsavedChanges;
+  const {
+    isDirty: hasUnsavedChanges,
+    dirtyRef: hasUnsavedChangesRef,
+    refreshDirty,
+  } = useSnapshotDirty(form, savedSnapshotRef);
 
   useEffect(() => {
     onDirtyChange?.(hasUnsavedChanges);
   }, [hasUnsavedChanges, onDirtyChange]);
 
-  const supportedEventPatterns = form.watch('supported_event_patterns');
+  const supportedEventPatterns = useWatch({
+    control: form.control,
+    name: 'supported_event_patterns',
+  });
   useEffect(() => {
     onSupportedEventPatternsChange?.(supportedEventPatterns);
   }, [onSupportedEventPatternsChange, supportedEventPatterns]);
@@ -279,6 +281,7 @@ function AgentFormComponent(
         };
         form.reset(loadedValues);
         savedSnapshotRef.current = JSON.stringify(loadedValues);
+        refreshDirty();
         initializedStagesRef.current.clear();
         setInitialDataLoaded(true);
       })
@@ -290,7 +293,7 @@ function AgentFormComponent(
     return () => {
       cancelled = true;
     };
-  }, [agentId, form, t, loadAttempt]);
+  }, [agentId, form, t, loadAttempt, refreshDirty]);
 
   useEffect(() => {
     if (!initialDataLoaded || !readPendingRunnerInstall(runnerInstallScope)) {
@@ -340,7 +343,11 @@ function AgentFormComponent(
     void loadPluginSystemStatus();
   }, [loadPluginSystemStatus]);
 
-  const currentRunner = (form.watch('runner') as Record<string, any>)?.id;
+  const currentRunner = useWatch({
+    control: form.control,
+    name: 'runner',
+    compute: (runner) => runner?.id,
+  });
   const runnerOptions = useMemo(() => {
     const runnerStage = runnerConfigSchema?.stages.find(
       (stage) => stage.name === 'runner',
@@ -358,24 +365,38 @@ function AgentFormComponent(
   const activeRunnerStage = runnerConfigSchema?.stages.find(
     (stage) => stage.name === currentRunner,
   );
-  const runnerConfigValues = form.watch('runner_config') as Record<
+  const runnerConfigValues = form.getValues('runner_config') as Record<
     string,
     Record<string, unknown>
   >;
-  const activeRunnerValues = useMemo(
-    () => runnerConfigValues?.[currentRunner] ?? {},
-    [currentRunner, runnerConfigValues],
-  );
-  const missingRunnerFields = useMemo(
-    () =>
-      (activeRunnerStage?.config ?? []).filter(
-        (field) =>
-          field.required &&
-          isRunnerFieldVisible(field, activeRunnerValues) &&
-          isRequiredRunnerValueMissing(activeRunnerValues[field.name]),
-      ),
-    [activeRunnerStage, activeRunnerValues],
+  useWatch({
+    control: form.control,
+    compute: (values) =>
+      (activeRunnerStage?.config ?? [])
+        .filter(
+          (field) =>
+            field.required &&
+            isRunnerFieldVisible(
+              field,
+              values.runner_config?.[currentRunner] ?? {},
+            ) &&
+            isRequiredRunnerValueMissing(
+              values.runner_config?.[currentRunner]?.[field.name],
+            ),
+        )
+        .map((field) => field.name)
+        .join('\n'),
+  });
+  const activeRunnerValues = runnerConfigValues?.[currentRunner] ?? {};
+  const missingRunnerFields = (activeRunnerStage?.config ?? []).filter(
+    (field) =>
+      field.required &&
+      isRunnerFieldVisible(field, activeRunnerValues) &&
+      isRequiredRunnerValueMissing(activeRunnerValues[field.name]),
   );
+  const missingRunnerFieldNames = missingRunnerFields
+    .map((field) => extractI18nObject(field.label))
+    .join(', ');
   const primarySections: Array<{
     name: AgentConfigSection;
     label: string;
@@ -448,9 +469,7 @@ function AgentFormComponent(
       return {
         label: t('agents.runnerConfigIncomplete'),
         description: t('agents.runnerConfigIncompleteDescription', {
-          fields: missingRunnerFields
-            .map((field) => extractI18nObject(field.label))
-            .join(', '),
+          fields: missingRunnerFieldNames,
         }),
         tone: 'warning',
       };
@@ -471,7 +490,8 @@ function AgentFormComponent(
     pluginStatusLoading,
     pluginSystemStatus,
     runnerOptions.length,
-    missingRunnerFields,
+    missingRunnerFieldNames,
+    missingRunnerFields.length,
     selectedRunnerOption,
     t,
   ]);
@@ -520,11 +540,12 @@ function AgentFormComponent(
     onRunnerStatusChange?.(runnerStatus);
   }, [onRunnerStatusChange, runnerStatus]);
 
-  function updateSnapshotIfInitial(stageKey: string) {
+  function updateSnapshotIfInitial(stageKey: string, wasDirty: boolean) {
     if (!initializedStagesRef.current.has(stageKey)) {
       initializedStagesRef.current.add(stageKey);
-      if (!hasUnsavedChanges) {
+      if (!wasDirty) {
         savedSnapshotRef.current = JSON.stringify(form.getValues());
+        refreshDirty();
       }
     }
   }
@@ -534,9 +555,10 @@ function AgentFormComponent(
     stageName: string,
     values: object,
   ) {
+    const wasDirty = hasUnsavedChangesRef.current;
     if (formName === 'runner') {
       form.setValue('runner', values, { shouldDirty: true });
-      updateSnapshotIfInitial(`runner.${stageName}`);
+      updateSnapshotIfInitial(`runner.${stageName}`, wasDirty);
       return;
     }
 
@@ -550,7 +572,7 @@ function AgentFormComponent(
       },
       { shouldDirty: true },
     );
-    updateSnapshotIfInitial(`runner_config.${stageName}`);
+    updateSnapshotIfInitial(`runner_config.${stageName}`, wasDirty);
   }
 
   function renderDynamicStage(stage: PipelineConfigStage) {
@@ -558,8 +580,8 @@ function AgentFormComponent(
     if (!isRunnerSelector && stage.name !== currentRunner) return null;
 
     const initialValues = isRunnerSelector
-      ? (form.watch('runner') as Record<string, unknown>) || {}
-      : ((form.watch('runner_config') as Record<string, any>) || {})[
+      ? (form.getValues('runner') as Record<string, unknown>) || {}
+      : ((form.getValues('runner_config') as Record<string, any>) || {})[
           stage.name
         ] || {};
 
@@ -640,6 +662,7 @@ function AgentFormComponent(
           loadedHostToolPolicyRef.current = [...values.allowed_tools];
         }
         savedSnapshotRef.current = submittedSnapshot;
+        refreshDirty();
         onFinish(agent);
         toast.success(t('agents.saveSuccess'));
         return true;
@@ -655,7 +678,14 @@ function AgentFormComponent(
         onSavingChange?.(false);
       }
     },
-    [agentId, hostToolCatalogAvailable, onFinish, onSavingChange, t],
+    [
+      agentId,
+      hostToolCatalogAvailable,
+      onFinish,
+      onSavingChange,
+      refreshDirty,
+      t,
+    ],
   );
 
   function handleSubmit(values: FormValues) {
@@ -682,6 +712,7 @@ function AgentFormComponent(
             emoji: values.emoji || '🤖',
           };
           savedSnapshotRef.current = JSON.stringify(snapshot);
+          refreshDirty();
         }
       },
       async save() {
@@ -693,7 +724,14 @@ function AgentFormComponent(
         return (await saveValues(form.getValues())) ?? false;
       },
     }),
-    [form, initialDataLoaded, loadFailed, saveValues],
+    [
+      form,
+      initialDataLoaded,
+      loadFailed,
+      saveValues,
+      hasUnsavedChangesRef,
+      refreshDirty,
+    ],
   );
 
   if (loadFailed)
diff --git a/web/src/app/home/agents/components/AgentMonitoringTab.tsx b/web/src/app/home/agents/components/AgentMonitoringTab.tsx
index 34c712240..9a6a8063c 100644
--- a/web/src/app/home/agents/components/AgentMonitoringTab.tsx
+++ b/web/src/app/home/agents/components/AgentMonitoringTab.tsx
@@ -1,6 +1,7 @@
 import { useCallback, useEffect, useRef, useState } from 'react';
 import { useTranslation } from 'react-i18next';
-import { Brain, ChevronRight, RefreshCw, Wrench } from 'lucide-react';
+import { ChevronRight, RefreshCw } from 'lucide-react';
+import AgentRunTimeline, { isReplyStep, isMockStep } from './AgentRunTimeline';
 import { httpClient } from '@/app/infra/http/HttpClient';
 import type {
   AgentPlatformTool,
@@ -13,6 +14,7 @@ import { Button } from '@/components/ui/button';
 import { Badge } from '@/components/ui/badge';
 import { Card, CardContent, CardHeader, CardTitle } from '@/components/ui/card';
 import { Alert, AlertDescription } from '@/components/ui/alert';
+import LoadErrorState from '@/components/LoadErrorState';
 import {
   Collapsible,
   CollapsibleTrigger,
@@ -206,7 +208,17 @@ export default function AgentMonitoringTab({
   const labels = Object.fromEntries(
     platformTools.map((tool) => [tool.name, extractI18nObject(tool.label)]),
   );
-  const steps = executionSteps(events);
+  const steps = executionSteps(events).filter(
+    (step) => step.kind === 'tool' || step.text.trim() || step.reasoning.trim(),
+  );
+  const toolSteps = steps.filter((step) => step.kind === 'tool');
+  const sentCount = toolSteps.filter(
+    (step) =>
+      isReplyStep(step) && !isMockStep(step) && step.status === 'completed',
+  ).length;
+  const sendFailed = toolSteps.some(
+    (step) => isReplyStep(step) && step.status === 'failed',
+  );
   const duration = selected ? processorRunDuration(selected) : null;
   const failureReason =
     selected?.status_reason ||
@@ -257,11 +269,13 @@ export default function AgentMonitoringTab({
         </Button>
       </div>
       {listError && (
-        <Alert variant="destructive">
-          <AlertDescription>{t('monitoring.loadError')}</AlertDescription>
-        </Alert>
+        <LoadErrorState
+          compact
+          title={t('monitoring.loadError')}
+          onRetry={refresh}
+        />
       )}
-      <div className="grid min-h-0 flex-1 gap-4 lg:grid-cols-[minmax(16rem,0.8fr)_minmax(0,2fr)]">
+      <div className="grid min-h-0 flex-1 gap-6 lg:grid-cols-[18rem_minmax(0,1fr)]">
         <section
           className="flex min-h-0 min-w-0 flex-col gap-2"
           aria-label={t('agents.eventProcessor.runs')}
@@ -279,13 +293,14 @@ export default function AgentMonitoringTab({
                 variant="ghost"
                 aria-pressed={selectedId === run.run_id}
                 onClick={() => setSelectedId(run.run_id)}
-                className="h-auto w-full flex-col items-stretch gap-2 rounded-none border-b p-3 text-left font-normal last:border-b-0 aria-pressed:bg-accent"
+                className="h-auto w-full flex-col items-stretch gap-2 rounded-none border-b border-l-2 border-l-transparent p-4 text-left font-normal last:border-b-0 aria-pressed:border-l-primary aria-pressed:bg-primary/5"
               >
                 <span className="flex items-center justify-between gap-2">
                   <span className="truncate font-medium">
                     {eventPatternLabel(run.metadata.event_type ?? '', t)}
                   </span>
                   <Badge
+                    status={run.status}
                     variant={
                       failedStatuses.has(run.status) ? 'destructive' : 'outline'
                     }
@@ -303,7 +318,7 @@ export default function AgentMonitoringTab({
                     {new Date(run.created_at * 1000).toLocaleString()}
                   </time>
                   {debugRun(run) && (
-                    <Badge variant="secondary">{t('agents.debugTab')}</Badge>
+                    <Badge status="debug">{t('agents.debugTab')}</Badge>
                   )}
                   {processorRunDuration(run) !== null && (
                     <span>{formatRunDuration(processorRunDuration(run)!)}</span>
@@ -343,6 +358,7 @@ export default function AgentMonitoringTab({
                       {eventPatternLabel(selected.metadata.event_type ?? '', t)}
                     </h2>
                     <Badge
+                      status={selected.status}
                       variant={
                         failedStatuses.has(selected.status)
                           ? 'destructive'
@@ -352,7 +368,7 @@ export default function AgentMonitoringTab({
                       {stateLabel(selected)}
                     </Badge>
                     {debugRun(selected) && (
-                      <Badge variant="secondary">{t('agents.debugTab')}</Badge>
+                      <Badge status="debug">{t('agents.debugTab')}</Badge>
                     )}
                   </div>
                   <div className="flex flex-wrap gap-x-4 gap-y-1 text-xs text-muted-foreground">
@@ -373,9 +389,37 @@ export default function AgentMonitoringTab({
                       </span>
                     )}
                   </div>
-                  <p className="break-all text-xs text-muted-foreground">
-                    {selected.runner_id}
-                  </p>
+                </div>
+                <div className="grid grid-cols-1 divide-y rounded-lg border bg-muted/20 sm:grid-cols-3 sm:divide-x sm:divide-y-0">
+                  <div className="space-y-1 p-4">
+                    <p className="text-xs text-muted-foreground">
+                      {t('agents.monitoring.runStatus')}
+                    </p>
+                    <p className="text-sm font-semibold">
+                      {stateLabel(selected)}
+                    </p>
+                  </div>
+                  <div className="space-y-1 p-4">
+                    <p className="text-xs text-muted-foreground">
+                      {t('agents.monitoring.toolCalls')}
+                    </p>
+                    <p className="text-sm font-semibold tabular-nums">
+                      {toolSteps.length}
+                    </p>
+                  </div>
+                  <div className="space-y-1 p-4">
+                    <p className="text-xs text-muted-foreground">
+                      {t('agents.monitoring.delivery')}
+                    </p>
+                    <p
+                      className={`text-sm font-semibold ${sendFailed ? 'text-destructive' : ''}`}
+                    >
+                      {t(
+                        `agents.monitoring.${sendFailed ? 'sendFailed' : sentCount ? 'sendCount' : 'noSend'}`,
+                        { count: sentCount },
+                      )}
+                    </p>
+                  </div>
                 </div>
                 {failedStatuses.has(selected.status) && failureReason && (
                   <Alert variant="destructive">
@@ -418,72 +462,22 @@ export default function AgentMonitoringTab({
                 </Card>
                 <h3 className="text-sm font-medium">
                   {t('agents.monitoring.execution')}
+                  <span className="ml-2 text-muted-foreground">
+                    {steps.length}
+                  </span>
                 </h3>
-                {steps.map((step, index) => (
-                  <Card key={index} className="gap-2 py-3">
-                    <CardHeader className="px-3">
-                      <CardTitle className="flex min-w-0 items-center gap-2 text-sm">
-                        {step.kind === 'tool' ? (
-                          <>
-                            <Wrench className="size-4 shrink-0" />
-                            <span className="break-all">
-                              {labels[step.name] || step.name}
-                            </span>
-                            <Badge
-                              variant={
-                                step.status === 'failed'
-                                  ? 'destructive'
-                                  : 'outline'
-                              }
-                            >
-                              {t(
-                                `agents.debugTool${step.status === 'running' ? (activeStatuses.has(selected.status) || eventCursor !== null ? 'Running' : 'Interrupted') : step.result && typeof step.result === 'object' && 'mock' in step.result && step.result.mock === true ? (step.status === 'failed' ? 'MockFailed' : 'Simulated') : step.status === 'failed' ? 'Failed' : 'Completed'}`,
-                              )}
-                            </Badge>
-                          </>
-                        ) : (
-                          <>
-                            <Brain className="size-4" />
-                            {t('agents.debugTextOutput')}
-                          </>
-                        )}
-                      </CardTitle>
-                    </CardHeader>
-                    <CardContent className="space-y-2 px-3">
-                      {step.kind === 'message' ? (
-                        <>
-                          {step.text && (
-                            <p className={textClass}>{step.text}</p>
-                          )}
-                          {step.reasoning && (
-                            <ProcessorPayload
-                              title={t('agents.debugReasoning')}
-                              value={step.reasoning}
-                            />
-                          )}
-                        </>
-                      ) : (
-                        <>
-                          {step.error && (
-                            <p className={`${textClass} text-destructive`}>
-                              {step.error}
-                            </p>
-                          )}
-                          <ProcessorPayload
-                            title={t('agents.debugToolArguments')}
-                            value={step.parameters}
-                          />
-                          {step.result !== undefined && (
-                            <ProcessorPayload
-                              title={t('agents.debugToolResult')}
-                              value={step.result}
-                            />
-                          )}
-                        </>
-                      )}
-                    </CardContent>
-                  </Card>
-                ))}
+                {!steps.length && (
+                  <p className="rounded-lg border border-dashed p-6 text-sm text-muted-foreground">
+                    {t('agents.monitoring.noSteps')}
+                  </p>
+                )}
+                <AgentRunTimeline
+                  steps={steps}
+                  labels={labels}
+                  active={
+                    activeStatuses.has(selected.status) || eventCursor !== null
+                  }
+                />
                 {events
                   .filter((event) => event.type === 'processor.log')
                   .map((event) => (
@@ -520,6 +514,7 @@ export default function AgentMonitoringTab({
                     >
                       {json({
                         run_id: selected.run_id,
+                        runner_id: selected.runner_id,
                         usage: selected.usage,
                         events,
                       })}
@@ -530,17 +525,10 @@ export default function AgentMonitoringTab({
             )
           )}
           {traceError && (
-            <Alert variant="destructive">
-              <AlertDescription>
-                {t('monitoring.loadError')}
-                <Button
-                  variant="link"
-                  onClick={() => setTraceRevision((value) => value + 1)}
-                >
-                  {t('common.retry')}
-                </Button>
-              </AlertDescription>
-            </Alert>
+            <LoadErrorState
+              title={t('monitoring.loadError')}
+              onRetry={() => setTraceRevision((value) => value + 1)}
+            />
           )}
         </section>
       </div>
diff --git a/web/src/app/home/agents/components/AgentRunTimeline.tsx b/web/src/app/home/agents/components/AgentRunTimeline.tsx
new file mode 100644
index 000000000..228140309
--- /dev/null
+++ b/web/src/app/home/agents/components/AgentRunTimeline.tsx
@@ -0,0 +1,174 @@
+import { useTranslation } from 'react-i18next';
+import { MessageSquare, Send, Wrench } from 'lucide-react';
+import { Badge } from '@/components/ui/badge';
+import { ProcessorPayload } from './PluginProcessorTrace';
+import type { ExecutionStep } from './debug-execution';
+import { toolDisplay, toolSummary } from './tool-summary';
+
+export function isReplyStep(step: ExecutionStep) {
+  return step.kind === 'tool' && toolDisplay(step.name)?.kind === 'reply';
+}
+
+export function isMockStep(step: ExecutionStep) {
+  return (
+    step.kind === 'tool' &&
+    !!step.result &&
+    typeof step.result === 'object' &&
+    'mock' in step.result &&
+    step.result.mock === true
+  );
+}
+
+export default function AgentRunTimeline({
+  steps,
+  labels,
+  active,
+}: {
+  steps: ExecutionStep[];
+  labels: Record<string, string>;
+  active: boolean;
+}) {
+  const { t } = useTranslation();
+  return (
+    <ol className="space-y-0">
+      {steps.map((step, index) => {
+        const reply = isReplyStep(step);
+        const mock = isMockStep(step);
+        const params =
+          step.kind === 'tool' &&
+          step.parameters &&
+          typeof step.parameters === 'object'
+            ? (step.parameters as Record<string, unknown>)
+            : {};
+        const summary = toolSummary(
+          step.kind === 'tool' ? step.name : '',
+          params,
+        );
+        const replyText = reply ? summary.text : '';
+        const actionText = reply ? '' : summary.text;
+        const Icon =
+          step.kind === 'message' ? MessageSquare : reply ? Send : Wrench;
+        const status = step.kind === 'tool' ? step.status : null;
+        const statusKey =
+          status === 'running'
+            ? active
+              ? 'Running'
+              : 'Interrupted'
+            : mock
+              ? status === 'failed'
+                ? 'MockFailed'
+                : 'Simulated'
+              : status === 'failed'
+                ? 'Failed'
+                : 'Completed';
+        return (
+          <li
+            key={step.kind === 'tool' ? step.id : `message-${index}`}
+            className="relative flex min-w-0 gap-3 pb-5 last:pb-0"
+          >
+            {index < steps.length - 1 && (
+              <span
+                className="absolute bottom-0 left-4 top-8 w-px bg-border"
+                aria-hidden="true"
+              />
+            )}
+            <span className="relative flex size-8 shrink-0 items-center justify-center rounded-full border bg-background text-xs font-medium text-muted-foreground">
+              {index + 1}
+            </span>
+            <div className="min-w-0 flex-1 rounded-lg border bg-background">
+              <div className="flex flex-wrap items-center gap-2 border-b bg-muted/30 px-4 py-3">
+                <Icon className="size-4 shrink-0 text-muted-foreground" />
+                <h4 className="min-w-0 flex-1 break-words text-sm font-medium">
+                  {step.kind === 'message'
+                    ? t('agents.debugTextOutput')
+                    : labels[step.name] || step.name}
+                </h4>
+                {status && (
+                  <Badge
+                    status={
+                      status === 'failed'
+                        ? 'failed'
+                        : mock
+                          ? 'simulated'
+                          : status === 'running' && !active
+                            ? 'cancelled'
+                            : status
+                    }
+                  >
+                    {t(`agents.debugTool${statusKey}`)}
+                  </Badge>
+                )}
+              </div>
+              <div className="space-y-3 p-4 text-sm">
+                {step.kind === 'message' ? (
+                  <>
+                    {step.text && (
+                      <p className="whitespace-pre-wrap break-words leading-7 [overflow-wrap:anywhere]">
+                        {step.text}
+                      </p>
+                    )}
+                    {step.reasoning && (
+                      <ProcessorPayload
+                        title={t('agents.debugReasoning')}
+                        value={step.reasoning}
+                      />
+                    )}
+                  </>
+                ) : (
+                  <>
+                    {replyText && (
+                      <p className="whitespace-pre-wrap break-words leading-7 [overflow-wrap:anywhere]">
+                        {replyText}
+                      </p>
+                    )}
+                    {!reply && actionText && (
+                      <p className="whitespace-pre-wrap break-words rounded-md bg-muted/50 p-3 font-mono text-xs leading-6 [overflow-wrap:anywhere]">
+                        {actionText}
+                      </p>
+                    )}
+                    {summary.recipient && (
+                      <p className="break-all text-xs text-muted-foreground">
+                        {t('agents.monitoring.recipient')}: {summary.recipient}
+                      </p>
+                    )}
+                    {step.error && (
+                      <p
+                        role="alert"
+                        className="whitespace-pre-wrap break-words text-destructive"
+                      >
+                        {step.error}
+                      </p>
+                    )}
+                    {status === 'completed' && (mock || !reply) && (
+                      <p className="text-xs text-muted-foreground">
+                        {t(
+                          `agents.monitoring.${mock ? 'simulatedAction' : 'actionCompleted'}`,
+                        )}
+                      </p>
+                    )}
+                    {(step.parameters != null || step.result != null) && (
+                      <div className="rounded-md border border-dashed text-muted-foreground">
+                        {step.parameters != null && (
+                          <ProcessorPayload
+                            title={t('agents.debugToolArguments')}
+                            value={step.parameters}
+                          />
+                        )}
+                        {step.result != null && (
+                          <ProcessorPayload
+                            title={t('agents.debugToolResult')}
+                            value={step.result}
+                          />
+                        )}
+                      </div>
+                    )}
+                  </>
+                )}
+              </div>
+            </div>
+          </li>
+        );
+      })}
+    </ol>
+  );
+}
diff --git a/web/src/app/home/agents/components/PluginProcessorSettings.tsx b/web/src/app/home/agents/components/PluginProcessorSettings.tsx
index 34bc94ce8..127152392 100644
--- a/web/src/app/home/agents/components/PluginProcessorSettings.tsx
+++ b/web/src/app/home/agents/components/PluginProcessorSettings.tsx
@@ -1,128 +1,50 @@
-import { Puzzle } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
-import { Link } from 'react-router-dom';
 import type { RunnerDescriptor } from '@/app/infra/entities/api';
-import { httpClient } from '@/app/infra/http';
-import { extractI18nObject } from '@/i18n/I18nProvider';
-import {
-  Select,
-  SelectContent,
-  SelectItem,
-  SelectTrigger,
-  SelectValue,
-} from '@/components/ui/select';
-import { Label } from '@/components/ui/label';
-import { Button } from '@/components/ui/button';
-
-function ProcessorComponentContent({
-  component,
-  option = false,
-}: {
-  component: RunnerDescriptor;
-  option?: boolean;
-}) {
-  const label = extractI18nObject({
-    en_US: component.id,
-    zh_Hans: component.id,
-    ...component.label,
-  });
-  const pluginId = `${component.plugin_author}/${component.plugin_name}`;
-  return (
-    <span
-      className={
-        option
-          ? 'grid w-full min-w-0 grid-cols-[1.75rem_minmax(0,1fr)] items-center gap-x-2 text-left'
-          : 'flex min-w-0 items-center gap-2'
-      }
-    >
-      <img
-        src={httpClient.getPluginIconURL(
-          component.plugin_author,
-          component.plugin_name,
-        )}
-        alt=""
-        className={
-          option
-            ? 'row-span-2 size-7 shrink-0 rounded-md object-cover'
-            : 'size-5 shrink-0 rounded object-cover'
-        }
-      />
-      <span className={option ? 'truncate font-medium leading-5' : 'truncate'}>
-        {label}
-      </span>
-      {option && (
-        <span
-          className="truncate text-xs leading-4 text-muted-foreground"
-          title={pluginId}
-        >
-          {pluginId}
-        </span>
-      )}
-    </span>
-  );
-}
+import RunnerSelect from './RunnerSelect';
 
 export default function PluginProcessorSettings({
   components,
   value,
   onChange,
   disabled = false,
+  installScope,
+  onInstalled,
 }: {
   components: RunnerDescriptor[];
   value: string;
   onChange: (value: string) => void;
   disabled?: boolean;
+  installScope: string;
+  onInstalled: (component: RunnerDescriptor) => void;
 }) {
   const { t } = useTranslation();
-  const selected = components.find((item) => item.id === value);
+  const options = components.map((component) => ({
+    name: component.id,
+    label: { en_US: component.id, zh_Hans: component.id, ...component.label },
+  }));
+  if (value && !components.some((component) => component.id === value))
+    options.push({
+      name: value,
+      label: {
+        en_US: t('agents.eventProcessor.unavailable'),
+        zh_Hans: t('agents.eventProcessor.unavailable'),
+      },
+    });
   return (
-    <div className="flex min-w-0 items-center gap-2">
-      <Label className="sr-only" htmlFor="event-processor-component">
-        {t('agents.eventProcessor.component')}
-      </Label>
-      <Select value={value} disabled={disabled} onValueChange={onChange}>
-        <SelectTrigger
-          id="event-processor-component"
-          className="w-[13.2rem] max-w-[calc(100vw-8rem)] bg-[#ffffff] dark:bg-[#2a2a2e]"
-        >
-          {selected ? (
-            <ProcessorComponentContent component={selected} />
-          ) : (
-            <span className="flex min-w-0 items-center gap-2">
-              <Puzzle className="size-4 shrink-0 text-muted-foreground" />
-              <SelectValue
-                placeholder={t('agents.eventProcessor.selectComponent')}
-              />
-            </span>
-          )}
-        </SelectTrigger>
-        <SelectContent className="max-h-72 w-[var(--radix-select-trigger-width)] max-w-[calc(100vw-2rem)]">
-          {value && !selected && (
-            <SelectItem value={value}>
-              {t('agents.eventProcessor.unavailable')}
-            </SelectItem>
-          )}
-          {components.map((component) => (
-            <SelectItem
-              key={component.id}
-              value={component.id}
-              className="py-1.5 [&>span:last-child]:min-w-0 [&>span:last-child]:flex-1"
-            >
-              <ProcessorComponentContent component={component} option />
-            </SelectItem>
-          ))}
-          {components.length === 0 && (
-            <div className="p-2 text-sm text-muted-foreground">
-              {t('agents.eventProcessor.noComponents')}
-              <Button asChild variant="link" className="h-auto px-0">
-                <Link to="/home/add-extension?type=plugin&component=Runner&runner_usage=event">
-                  {t('agents.eventProcessor.installPlugin')}
-                </Link>
-              </Button>
-            </div>
-          )}
-        </SelectContent>
-      </Select>
+    <div
+      id="event-processor-component"
+      className="w-[20rem] max-w-[calc(100vw-8rem)]"
+    >
+      <RunnerSelect
+        usage="event"
+        options={options}
+        value={value}
+        label={t('agents.eventProcessor.component')}
+        disabled={disabled}
+        installScope={installScope}
+        onValueChange={onChange}
+        onEventProcessorInstalled={onInstalled}
+      />
     </div>
   );
 }
diff --git a/web/src/app/home/agents/components/PluginProcessorTrace.tsx b/web/src/app/home/agents/components/PluginProcessorTrace.tsx
index 40eac2329..4ef9106de 100644
--- a/web/src/app/home/agents/components/PluginProcessorTrace.tsx
+++ b/web/src/app/home/agents/components/PluginProcessorTrace.tsx
@@ -1,7 +1,6 @@
 import type { ReactNode } from 'react';
 import { useTranslation } from 'react-i18next';
-import { ChevronRight } from 'lucide-react';
-import { Alert, AlertDescription } from '@/components/ui/alert';
+import { ChevronRight, ScrollText } from 'lucide-react';
 import { Badge } from '@/components/ui/badge';
 import { Button } from '@/components/ui/button';
 import {
@@ -47,38 +46,50 @@ export default function PluginProcessorTrace({
 }) {
   const { t } = useTranslation();
   return (
-    <div className="space-y-2">
+    <div className="space-y-4">
       {events.map((event, index) =>
         event.type === 'processor.log' ? (
-          <Alert
+          <div
             key={event.sequence ?? index}
-            variant={event.data.level === 'error' ? 'destructive' : 'default'}
+            className="min-w-0 overflow-hidden rounded-lg border bg-background"
           >
-            <AlertDescription className="flex min-w-0 items-start gap-2">
-              <Badge variant="outline">{String(event.data.level)}</Badge>
-              <span className="min-w-0 whitespace-pre-wrap break-words [overflow-wrap:anywhere]">
-                {String(event.data.text)}
+            <div className="flex items-center gap-2 border-b bg-muted/30 px-4 py-3">
+              <ScrollText className="size-4 text-muted-foreground" />
+              <span className="flex-1 text-sm font-medium">
+                {t('agents.eventProcessor.logsTab')}
               </span>
-            </AlertDescription>
-          </Alert>
+              <Badge status={String(event.data.level)}>
+                {String(event.data.level)}
+              </Badge>
+            </div>
+            <p
+              className={`min-w-0 whitespace-pre-wrap break-words p-4 text-sm leading-7 [overflow-wrap:anywhere] ${event.data.level === 'error' ? 'text-destructive' : ''}`}
+            >
+              {String(event.data.text)}
+            </p>
+          </div>
         ) : (
-          <ProcessorPayload
+          <div
             key={event.sequence ?? index}
-            title={
-              <>
-                {t(
-                  `agents.eventProcessor.trace_${event.type.replaceAll('.', '_')}`,
-                  { defaultValue: event.type },
-                )}
-                {typeof event.data.tool_name === 'string' && (
-                  <span className="ml-2 text-muted-foreground">
-                    {toolLabels[event.data.tool_name] || event.data.tool_name}
-                  </span>
-                )}
-              </>
-            }
-            value={event.data}
-          />
+            className="rounded-lg border bg-background p-3"
+          >
+            <ProcessorPayload
+              title={
+                <>
+                  {t(
+                    `agents.eventProcessor.trace_${event.type.replaceAll('.', '_')}`,
+                    { defaultValue: event.type },
+                  )}
+                  {typeof event.data.tool_name === 'string' && (
+                    <span className="ml-2 text-muted-foreground">
+                      {toolLabels[event.data.tool_name] || event.data.tool_name}
+                    </span>
+                  )}
+                </>
+              }
+              value={event.data}
+            />
+          </div>
         ),
       )}
     </div>
diff --git a/web/src/app/home/agents/components/ProcessorRunList.tsx b/web/src/app/home/agents/components/ProcessorRunList.tsx
index c9abec543..01684426f 100644
--- a/web/src/app/home/agents/components/ProcessorRunList.tsx
+++ b/web/src/app/home/agents/components/ProcessorRunList.tsx
@@ -15,18 +15,22 @@ export default function ProcessorRunList({
   selectedId,
   onSelect,
   footer,
+  className,
 }: {
   runs: ProcessorRun[];
   selectedId?: string;
   onSelect: (run: ProcessorRun) => void;
   footer?: ReactNode;
+  className?: string;
 }) {
   const { t } = useTranslation();
   return (
     <div
       role="group"
       aria-label={t('agents.eventProcessor.runs')}
-      className="max-h-56 shrink-0 overflow-y-auto rounded-md border"
+      className={
+        className ?? 'max-h-56 shrink-0 overflow-y-auto rounded-md border'
+      }
     >
       {runs.map((run) => {
         const duration = processorRunDuration(run);
@@ -41,7 +45,7 @@ export default function ProcessorRunList({
             onClick={() => onSelect(run)}
             aria-label={`${label} ${event} ${new Date(run.created_at * 1000).toLocaleString()}`}
             title={event}
-            className="flex h-auto min-h-12 w-full justify-start gap-3 rounded-none border-b px-3 py-2 text-left text-sm font-normal last:border-b-0 aria-pressed:bg-accent"
+            className="flex h-auto min-h-12 w-full justify-start gap-3 rounded-none border-b border-l-2 border-l-transparent p-4 text-left text-sm font-normal last:border-b-0 aria-pressed:border-l-primary aria-pressed:bg-primary/5"
           >
             <span className="min-w-0 flex-1 space-y-1">
               <span className="block truncate font-medium">{label}</span>
@@ -54,6 +58,7 @@ export default function ProcessorRunList({
             </span>
             <span className="flex shrink-0 flex-col items-end gap-1">
               <Badge
+                status={run.status}
                 variant={
                   run.status === 'failed' || run.status === 'timeout'
                     ? 'destructive'
diff --git a/web/src/app/home/agents/components/RunnerSelect.tsx b/web/src/app/home/agents/components/RunnerSelect.tsx
index cb706091c..3953a8e70 100644
--- a/web/src/app/home/agents/components/RunnerSelect.tsx
+++ b/web/src/app/home/agents/components/RunnerSelect.tsx
@@ -1,15 +1,19 @@
 import { useCallback, useEffect, useMemo, useState } from 'react';
-import { Bot, ExternalLink, Loader2, Store } from 'lucide-react';
+import { Bot, Puzzle, ExternalLink, Loader2, Store } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
 import { toast } from 'sonner';
 
-import { getCloudServiceClientSync, httpClient } from '@/app/infra/http';
+import { getCloudServiceClientSync } from '@/app/infra/http';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 import type { IDynamicFormItemOption } from '@/app/infra/entities/form/dynamic';
 import type { PluginV4 } from '@/app/infra/entities/plugin';
+import type { RunnerDescriptor } from '@/app/infra/entities/api';
 import {
   RunnerMarketplaceError,
   getErrorMessage,
   installMarketplaceRunner,
+  installMarketplaceEventProcessor,
+  resumePendingEventProcessorInstall,
   loadRunnerCatalog,
   marketplacePluginId,
   runnerPluginPrefix,
@@ -51,21 +55,25 @@ function installErrorMessage(
   return getErrorMessage(error) || t('wizard.aiEngine.installFailed');
 }
 
-function installedRunnerIconURL(option: IDynamicFormItemOption) {
-  return option.name.startsWith('plugin:')
-    ? (() => {
-        const match = option.name.match(/^plugin:([^/]+)\/([^/]+)(?:\/|$)/);
-        return match ? httpClient.getPluginIconURL(match[1], match[2]) : null;
-      })()
-    : null;
+function pluginOptionParts(option: IDynamicFormItemOption): {
+  author: string;
+  name: string;
+} | null {
+  if (!option.name.startsWith('plugin:')) return null;
+  const match = option.name.match(/^plugin:([^/]+)\/([^/]+)(?:\/|$)/);
+  return match ? { author: match[1], name: match[2] } : null;
 }
 
 function InstalledRunnerContent({
   option,
+  eventProcessor = false,
 }: {
   option: IDynamicFormItemOption;
+  eventProcessor?: boolean;
 }) {
-  const iconURL = installedRunnerIconURL(option);
+  const Icon = eventProcessor ? Puzzle : Bot;
+  const parts = pluginOptionParts(option);
+  const iconURL = useInstalledPluginIcon(parts?.author, parts?.name);
 
   return (
     <span className="flex min-w-0 items-center gap-2">
@@ -76,7 +84,7 @@ function InstalledRunnerContent({
           className="size-5 shrink-0 rounded object-cover"
         />
       ) : (
-        <Bot className="size-4 shrink-0 text-muted-foreground" />
+        <Icon className="size-4 shrink-0 text-muted-foreground" />
       )}
       <span className="truncate">{extractI18nObject(option.label)}</span>
     </span>
@@ -86,11 +94,15 @@ function InstalledRunnerContent({
 function InstalledRunnerOptionContent({
   option,
   description,
+  eventProcessor = false,
 }: {
   option: IDynamicFormItemOption;
   description: string;
+  eventProcessor?: boolean;
 }) {
-  const iconURL = installedRunnerIconURL(option);
+  const Icon = eventProcessor ? Puzzle : Bot;
+  const parts = pluginOptionParts(option);
+  const iconURL = useInstalledPluginIcon(parts?.author, parts?.name);
 
   return (
     <span className="grid w-full min-w-0 grid-cols-[1.75rem_minmax(0,1fr)] items-center gap-x-2 text-left">
@@ -101,7 +113,7 @@ function InstalledRunnerOptionContent({
           className="row-span-2 size-7 shrink-0 rounded-md object-cover"
         />
       ) : (
-        <Bot className="row-span-2 size-5 justify-self-center text-muted-foreground" />
+        <Icon className="row-span-2 size-5 justify-self-center text-muted-foreground" />
       )}
       <span className="truncate font-medium leading-5">
         {extractI18nObject(option.label)}
@@ -199,13 +211,19 @@ export default function RunnerSelect({
   onValueChange,
   installScope,
   onInstalled,
+  usage = 'agent',
+  disabled = false,
+  onEventProcessorInstalled,
 }: {
   options: IDynamicFormItemOption[];
   label: string;
   value: string;
   onValueChange: (value: string) => void;
   installScope: string;
-  onInstalled: (installed: InstalledRunner) => void;
+  onInstalled?: (installed: InstalledRunner) => void;
+  usage?: 'agent' | 'event';
+  disabled?: boolean;
+  onEventProcessorInstalled?: (component: RunnerDescriptor) => void;
 }) {
   const { t } = useTranslation();
   const { addTask, tasks } = usePluginInstallTasks();
@@ -227,7 +245,7 @@ export default function RunnerSelect({
     setCatalogLoading(true);
     setCatalogError(false);
     try {
-      const catalog = await loadRunnerCatalog('agent');
+      const catalog = await loadRunnerCatalog(usage);
       setMarketplaceRunners(catalog.marketplaceRunners);
       setInstalledPluginIds(catalog.installedPluginIds);
       setInstalledPluginDescriptions(catalog.installedPluginDescriptions);
@@ -237,7 +255,37 @@ export default function RunnerSelect({
     } finally {
       setCatalogLoading(false);
     }
-  }, []);
+  }, [usage]);
+
+  useEffect(() => {
+    if (
+      usage !== 'event' ||
+      disabled ||
+      !readPendingRunnerInstall(installScope)
+    )
+      return;
+    let cancelled = false;
+    void resumePendingEventProcessorInstall(installScope)
+      .then((component) => {
+        if (!cancelled && component) {
+          onEventProcessorInstalled?.(component);
+          void loadCatalog();
+        }
+      })
+      .catch((error) => {
+        if (!cancelled) setInstallError(installErrorMessage(error, t));
+      });
+    return () => {
+      cancelled = true;
+    };
+  }, [
+    installScope,
+    usage,
+    disabled,
+    onEventProcessorInstalled,
+    loadCatalog,
+    t,
+  ]);
 
   useEffect(() => {
     void loadCatalog();
@@ -286,22 +334,28 @@ export default function RunnerSelect({
   const handleInstall = useCallback(
     async (plugin: PluginV4) => {
       const pluginId = marketplacePluginId(plugin);
-      if (pendingInstall || installingPluginId) return;
+      if (disabled || pendingInstall || installingPluginId) return;
 
       setInstallingPluginId(pluginId);
       setInstallError(null);
       try {
-        const installed = await installMarketplaceRunner(plugin, {
+        const installOptions = {
           scope: installScope,
-          onTaskCreated: (taskId) =>
+          onTaskCreated: (taskId: number) =>
             addTask({
               taskId,
               pluginName: marketplacePluginId(plugin),
               source: 'marketplace',
               extensionType: 'plugin',
             }),
-        });
-        onInstalled(installed);
+        };
+        if (usage === 'event') {
+          onEventProcessorInstalled?.(
+            await installMarketplaceEventProcessor(plugin, installOptions),
+          );
+        } else {
+          onInstalled?.(await installMarketplaceRunner(plugin, installOptions));
+        }
         await loadCatalog();
         toast.success(
           t('agents.runnerInstallSuccess', {
@@ -326,12 +380,16 @@ export default function RunnerSelect({
       onInstalled,
       pendingInstall,
       t,
+      usage,
+      disabled,
+      onEventProcessorInstalled,
     ],
   );
 
   return (
     <div className="w-full max-w-[22rem] space-y-2">
       <Select
+        disabled={disabled}
         value={value}
         onValueChange={handleValueChange}
         onOpenChange={(open) => {
@@ -343,17 +401,34 @@ export default function RunnerSelect({
           className="w-full bg-[#ffffff] dark:bg-[#2a2a2e]"
         >
           {selectedOption ? (
-            <InstalledRunnerContent option={selectedOption} />
+            <InstalledRunnerContent
+              option={selectedOption}
+              eventProcessor={usage === 'event'}
+            />
           ) : (
-            <SelectValue placeholder={t('common.select')} />
+            <SelectValue
+              placeholder={t(
+                usage === 'event'
+                  ? 'agents.eventProcessor.selectComponent'
+                  : 'common.select',
+              )}
+            />
           )}
         </SelectTrigger>
         <SelectContent className="z-[70] max-h-72 w-[var(--radix-select-trigger-width)] max-w-[calc(100vw-2rem)]">
           <SelectGroup>
             <SelectLabel className="px-2 py-1 text-[11px] font-medium">
               <span className="inline-flex items-center gap-1.5">
-                <Bot className="size-3.5" />
-                {t('agents.installedRunners')}
+                {usage === 'event' ? (
+                  <Puzzle className="size-3.5" />
+                ) : (
+                  <Bot className="size-3.5" />
+                )}
+                {t(
+                  usage === 'event'
+                    ? 'plugins.installed'
+                    : 'agents.installedRunners',
+                )}
               </span>
             </SelectLabel>
             {options.length > 0 ? (
@@ -371,6 +446,7 @@ export default function RunnerSelect({
                   >
                     <InstalledRunnerOptionContent
                       option={option}
+                      eventProcessor={usage === 'event'}
                       description={description}
                     />
                   </SelectItem>
@@ -378,7 +454,11 @@ export default function RunnerSelect({
               })
             ) : (
               <div className="px-2 py-1.5 text-xs text-muted-foreground">
-                {t('agents.noInstalledRunners')}
+                {t(
+                  usage === 'event'
+                    ? 'agents.eventProcessor.noComponents'
+                    : 'agents.noInstalledRunners',
+                )}
               </div>
             )}
           </SelectGroup>
@@ -389,10 +469,14 @@ export default function RunnerSelect({
             <SelectLabel className="flex items-center justify-between gap-2 px-2 py-1 text-[11px] font-medium">
               <span className="inline-flex min-w-0 items-center gap-1.5">
                 <Store className="size-3.5" />
-                {t('agents.marketplaceRunners')}
+                {t(
+                  usage === 'event'
+                    ? 'plugins.marketplace'
+                    : 'agents.marketplaceRunners',
+                )}
               </span>
               <a
-                href="https://space.langbot.app/market?type=plugin&component=Runner&runner_usage=agent"
+                href={`https://space.langbot.app/market?type=plugin&component=Runner&runner_usage=${usage}`}
                 target="_blank"
                 rel="noreferrer"
                 className="inline-flex shrink-0 items-center gap-1 rounded border px-1.5 py-0.5 text-[10px] font-medium text-foreground hover:bg-accent"
@@ -437,6 +521,22 @@ export default function RunnerSelect({
               </div>
             )}
           </SelectGroup>
+          <SelectSeparator />
+          <a
+            href="https://langbot.app/docs/zh/plugin/dev/components/runner"
+            target="_blank"
+            rel="noopener noreferrer"
+            className="flex items-center gap-2 rounded-sm px-2 py-2 text-xs text-muted-foreground hover:bg-accent hover:text-foreground"
+            onPointerDown={(event) => event.stopPropagation()}
+            onClick={(event) => event.stopPropagation()}
+          >
+            {t(
+              usage === 'event'
+                ? 'bots.routingHelp.docs'
+                : 'bots.routingHelp.runnerDocs',
+            )}
+            <ExternalLink className="size-3" />
+          </a>
         </SelectContent>
       </Select>
 
diff --git a/web/src/app/home/agents/components/debug-event-data.ts b/web/src/app/home/agents/components/debug-event-data.ts
index fc96cb811..59c110f59 100644
--- a/web/src/app/home/agents/components/debug-event-data.ts
+++ b/web/src/app/home/agents/components/debug-event-data.ts
@@ -90,6 +90,90 @@ const DEBUG_EVENT_DEFINITIONS: Record<string, DebugEventDefinition> = {
     messageField: 'text',
   },
   'message.deleted': deletion,
+  'dingtalk.card_action': {
+    fields: [user],
+    defaults: {
+      space_id: 'dtv1.card//IM_GROUP.debug-group',
+      card_instance_id: 'debug-card',
+      action: 'confirm',
+      data: {},
+    },
+  },
+  'wecombot.enter_chat': {
+    fields: [user],
+    defaults: {
+      chat_id: 'debug-user',
+      chat_type: 'private',
+      message_id: 'debug-entry',
+      data: {},
+    },
+  },
+  'wecombot.template_card_event': {
+    fields: [user],
+    defaults: {
+      chat_id: 'debug-user',
+      chat_type: 'private',
+      message_id: 'debug-card-click',
+      task_id: 'debug-card',
+      event_key: 'confirm',
+      card_type: 'button_interaction',
+      selected_items: [],
+      data: {},
+    },
+  },
+  'wecomcs.enter_session': {
+    fields: [user],
+    defaults: {
+      open_kfid: 'debug-kf',
+      external_userid: 'debug-user',
+      chat_id: 'debug-user|debug-kf',
+      scene: 'debug',
+      scene_param: '',
+      welcome_code: '',
+      wechat_channels: {},
+    },
+  },
+  'wecomcs.msg_send_fail': {
+    fields: [user],
+    defaults: {
+      open_kfid: 'debug-kf',
+      external_userid: 'debug-user',
+      chat_id: 'debug-user|debug-kf',
+      fail_msgid: 'debug-message',
+      fail_type: 4,
+    },
+  },
+  'wecomcs.servicer_status_change': {
+    fields: [],
+    defaults: {
+      open_kfid: 'debug-kf',
+      servicer_userid: 'debug-servicer',
+      status: 2,
+      stop_type: 0,
+    },
+  },
+  'wecomcs.session_status_change': {
+    fields: [user],
+    defaults: {
+      open_kfid: 'debug-kf',
+      external_userid: 'debug-user',
+      chat_id: 'debug-user|debug-kf',
+      change_type: 2,
+      old_servicer_userid: 'debug-servicer-1',
+      new_servicer_userid: 'debug-servicer-2',
+      msg_code: '',
+    },
+  },
+  'wecomcs.reject_customer_msg_switch_change': {
+    fields: [user],
+    defaults: {
+      open_kfid: 'debug-kf',
+      external_userid: 'debug-user',
+      chat_id: 'debug-user|debug-kf',
+      servicer_userid: 'debug-servicer',
+      reject_switch: 1,
+    },
+  },
   'message.recalled': deletion,
   'message.reaction': {
     fields: [
diff --git a/web/src/app/home/agents/components/tool-summary.ts b/web/src/app/home/agents/components/tool-summary.ts
new file mode 100644
index 000000000..33c3d9b01
--- /dev/null
+++ b/web/src/app/home/agents/components/tool-summary.ts
@@ -0,0 +1,73 @@
+type ToolDisplay = {
+  fields: readonly string[];
+  kind?: 'reply';
+  recipient?: string;
+};
+
+// Explicit display declarations for built-in tools. Unknown tools have no
+// summary: their parameters remain available in the payload disclosure.
+const TOOL_DISPLAYS: Readonly<Record<string, ToolDisplay>> = {
+  exec: { fields: ['command'] },
+  read: { fields: ['path'] },
+  write: { fields: ['path'] },
+  edit: { fields: ['path'] },
+  glob: { fields: ['pattern', 'path'] },
+  grep: { fields: ['pattern', 'path'] },
+  event_reply: { fields: ['text'], kind: 'reply' },
+  platform_send_message: {
+    fields: ['text'],
+    kind: 'reply',
+    recipient: 'target_id',
+  },
+  event_delete_message: { fields: [] },
+  event_get_actor: { fields: [] },
+  event_get_group: { fields: [] },
+  event_get_group_member: { fields: [] },
+  event_mute_member: { fields: ['duration'] },
+  event_unmute_member: { fields: [] },
+  event_kick_member: { fields: [] },
+  event_respond_friend_request: { fields: ['approve', 'remark'] },
+  event_respond_group_invite: { fields: ['approve'] },
+  platform_get_message: { fields: ['chat_id', 'message_id'] },
+  platform_delete_message: { fields: ['chat_id', 'message_id'] },
+  platform_get_group_info: { fields: ['group_id'] },
+  platform_get_group_list: { fields: [] },
+  platform_get_group_member_list: { fields: ['group_id'] },
+  platform_get_group_member_info: { fields: ['group_id', 'user_id'] },
+  platform_set_group_name: { fields: ['group_id', 'name'] },
+  platform_mute_member: { fields: ['group_id', 'user_id', 'duration'] },
+  platform_unmute_member: { fields: ['group_id', 'user_id'] },
+  platform_kick_member: { fields: ['group_id', 'user_id'] },
+  platform_leave_group: { fields: ['group_id'] },
+  platform_get_user_info: { fields: ['user_id'] },
+  platform_get_friend_list: { fields: [] },
+};
+
+export function toolDisplay(name: string): ToolDisplay | undefined {
+  return Object.prototype.hasOwnProperty.call(TOOL_DISPLAYS, name)
+    ? TOOL_DISPLAYS[name]
+    : undefined;
+}
+
+export function toolSummary(name: string, parameters: Record<string, unknown>) {
+  const display = toolDisplay(name);
+  const fields = (display?.fields ?? []).flatMap((field) => {
+    const value = parameters[field];
+    if (!['string', 'number', 'boolean'].includes(typeof value) || value === '')
+      return [];
+    return [{ field, text: String(value) }];
+  });
+  return {
+    text: fields
+      .map(({ field, text }) =>
+        display?.kind === 'reply' || display?.fields.length === 1
+          ? text
+          : `${field}: ${text}`,
+      )
+      .join('\n'),
+    recipient:
+      display?.recipient && typeof parameters[display.recipient] === 'string'
+        ? String(parameters[display.recipient])
+        : '',
+  };
+}
diff --git a/web/src/app/home/agents/runner-marketplace.ts b/web/src/app/home/agents/runner-marketplace.ts
index a2cb498f4..ebec6c3e9 100644
--- a/web/src/app/home/agents/runner-marketplace.ts
+++ b/web/src/app/home/agents/runner-marketplace.ts
@@ -8,7 +8,7 @@ import {
   type PluginV4,
   type RunnerUsage,
 } from '@/app/infra/entities/plugin';
-import type { AsyncTask } from '@/app/infra/entities/api';
+import type { AsyncTask, RunnerDescriptor } from '@/app/infra/entities/api';
 import type { I18nObject } from '@/app/infra/entities/common';
 
 export const RUNNER_COMPONENT_FILTER = 'Runner';
@@ -20,9 +20,7 @@ const RUNNER_INSTALL_INTENT_KEY_PREFIX = 'langbot-runner-install';
 const RUNNER_INSTALL_INTENT_EVENT = 'langbot-runner-install-change';
 
 export type RunnerMarketplaceErrorCode =
-  | 'version-unavailable'
-  | 'install-timeout'
-  | 'registration-timeout';
+  'version-unavailable' | 'install-timeout' | 'registration-timeout';
 
 export class RunnerMarketplaceError extends Error {
   constructor(public readonly code: RunnerMarketplaceErrorCode) {
@@ -232,6 +230,14 @@ export async function installMarketplaceRunner(
   plugin: PluginV4,
   options: InstallRunnerOptions,
 ): Promise<InstalledRunner> {
+  const pending = await startRunnerInstall(plugin, options);
+  return finishRunnerInstall(pending, options.onProgress);
+}
+
+async function startRunnerInstall(
+  plugin: PluginV4,
+  options: InstallRunnerOptions,
+): Promise<PendingRunnerInstall> {
   if (!plugin.latest_version) {
     throw new RunnerMarketplaceError('version-unavailable');
   }
@@ -252,7 +258,7 @@ export async function installMarketplaceRunner(
   };
   writePendingRunnerInstall(pending);
   options.onTaskCreated?.(taskId);
-  return finishRunnerInstall(pending, options.onProgress);
+  return pending;
 }
 
 function extractPluginLabel(plugin: PluginV4) {
@@ -271,6 +277,14 @@ async function finishRunnerInstall(
   pending: PendingRunnerInstall,
   onProgress?: (task: AsyncTask) => void,
 ): Promise<InstalledRunner> {
+  await waitForRunnerInstall(pending, onProgress);
+  return waitForRunnerRegistration(pending);
+}
+
+async function waitForRunnerInstall(
+  pending: PendingRunnerInstall,
+  onProgress?: (task: AsyncTask) => void,
+) {
   // A refreshed page receives a fresh observation window. The backend task is
   // authoritative; `startedAt` is display metadata, not a reason to abandon a
   // still-running installation immediately after recovery.
@@ -293,7 +307,11 @@ async function finishRunnerInstall(
   if (!installCompleted) {
     throw new RunnerMarketplaceError('install-timeout');
   }
+}
 
+async function waitForRunnerRegistration(
+  pending: PendingRunnerInstall,
+): Promise<InstalledRunner> {
   const registrationDeadline = Date.now() + RUNNER_REGISTRATION_TIMEOUT_MS;
   const prefix = runnerPluginPrefix({
     author: pending.pluginAuthor,
@@ -333,3 +351,45 @@ export async function resumePendingRunnerInstall(
   if (!pending) return null;
   return finishRunnerInstall(pending, onProgress);
 }
+
+export async function installMarketplaceEventProcessor(
+  plugin: PluginV4,
+  options: InstallRunnerOptions,
+): Promise<RunnerDescriptor> {
+  const pending = await startRunnerInstall(plugin, options);
+  return finishEventProcessorInstall(pending, options.onProgress);
+}
+
+export async function resumePendingEventProcessorInstall(
+  scope: string,
+): Promise<RunnerDescriptor | null> {
+  const pending = readPendingRunnerInstall(scope);
+  return pending ? finishEventProcessorInstall(pending) : null;
+}
+
+async function finishEventProcessorInstall(
+  pending: PendingRunnerInstall,
+  onProgress?: (task: AsyncTask) => void,
+): Promise<RunnerDescriptor> {
+  await waitForRunnerInstall(pending, onProgress);
+  const deadline = Date.now() + RUNNER_REGISTRATION_TIMEOUT_MS;
+  while (Date.now() < deadline) {
+    const metadata = await httpClient.getAgentMetadata();
+    // Only the event registry is authoritative; Agent-only Runners are invalid.
+    const candidates = (metadata.event_processors ?? []).filter(
+      (component) =>
+        component.plugin_author === pending.pluginAuthor &&
+        component.plugin_name === pending.pluginName,
+    );
+    const component =
+      candidates.find((component) => component.id.endsWith('/default')) ??
+      candidates[0];
+    if (component) {
+      clearPendingRunnerInstall(pending.scope, pending.taskId);
+      return component;
+    }
+    await wait(1000);
+  }
+  clearPendingRunnerInstall(pending.scope, pending.taskId);
+  throw new RunnerMarketplaceError('registration-timeout');
+}
diff --git a/web/src/app/home/bots/BotDetailContent.tsx b/web/src/app/home/bots/BotDetailContent.tsx
index 6ed4ca775..de9e9bdf1 100644
--- a/web/src/app/home/bots/BotDetailContent.tsx
+++ b/web/src/app/home/bots/BotDetailContent.tsx
@@ -1,10 +1,10 @@
 import EntityLoadState from '@/components/EntityLoadState';
 import { useState, useEffect, useRef, useCallback } from 'react';
-import { useNavigate } from 'react-router-dom';
+import { useNavigate, useSearchParams } from 'react-router-dom';
 import { Tabs, TabsList, TabsTrigger, TabsContent } from '@/components/ui/tabs';
 import { Button } from '@/components/ui/button';
 import { Switch } from '@/components/ui/switch';
-import { Label } from '@/components/ui/label';
+import { Badge } from '@/components/ui/badge';
 import {
   Dialog,
   DialogContent,
@@ -22,7 +22,14 @@ import type { BotSessionMonitorHandle } from '@/app/home/bots/components/bot-ses
 import { httpClient } from '@/app/infra/http/HttpClient';
 import { useSidebarData } from '@/app/home/components/home-sidebar/SidebarDataContext';
 import { useTranslation } from 'react-i18next';
-import { Settings, FileText, Users, RefreshCw, Trash2 } from 'lucide-react';
+import {
+  Settings,
+  FileText,
+  Users,
+  RefreshCw,
+  Trash2,
+  Loader2,
+} from 'lucide-react';
 import { cn } from '@/lib/utils';
 import { toast } from 'sonner';
 import { showBotError } from './bot-error';
@@ -56,7 +63,23 @@ export default function BotDetailContent({ id }: { id: string }) {
     return () => setDetailEntityName(null);
   }, [id, isCreateMode, bots, setDetailEntityName, t]);
 
-  const [activeTab, setActiveTab] = useState('config');
+  const [searchParams, setSearchParams] = useSearchParams();
+  const requestedTab = searchParams.get('tab');
+  const activeTab =
+    !isCreateMode &&
+    canViewMonitoring &&
+    ['logs', 'sessions'].includes(requestedTab || '')
+      ? requestedTab!
+      : 'config';
+  const setActiveTab = (tab: string) =>
+    setSearchParams(
+      (previous) => {
+        const next = new URLSearchParams(previous);
+        next.set('tab', tab);
+        return next;
+      },
+      { replace: true },
+    );
   const [adapterLabel, setAdapterLabel] = useState('');
   const [showDeleteConfirm, setShowDeleteConfirm] = useState(false);
   const [basicInfoOpen, setBasicInfoOpen] = useState(false);
@@ -72,6 +95,7 @@ export default function BotDetailContent({ id }: { id: string }) {
 
   // Enable state managed here so the header switch works
   const [botEnabled, setBotEnabled] = useState(true);
+  const [isTogglingEnable, setIsTogglingEnable] = useState(false);
   const [enableLoaded, setEnableLoaded] = useState(false);
 
   // Fetch bot enable state
@@ -91,20 +115,27 @@ export default function BotDetailContent({ id }: { id: string }) {
 
   const handleEnableToggle = useCallback(
     async (checked: boolean) => {
+      if (isTogglingEnable) return;
       const prev = botEnabled;
+      setIsTogglingEnable(true);
       setBotEnabled(checked);
       try {
         await httpClient.updateBot(id, { enable: checked });
         setBot((current) =>
           current ? { ...current, enable: checked } : current,
         );
+        toast.success(
+          t(checked ? 'bots.enableConfirmed' : 'bots.disableConfirmed'),
+        );
         refreshBots();
       } catch (error) {
         setBotEnabled(prev);
         showBotError(error, t('bots.setBotEnableError'), t);
+      } finally {
+        setIsTogglingEnable(false);
       }
     },
-    [id, botEnabled, refreshBots, t],
+    [id, botEnabled, isTogglingEnable, refreshBots, t],
   );
 
   function handleFormSubmit() {
@@ -210,36 +241,68 @@ export default function BotDetailContent({ id }: { id: string }) {
             </div>
             {enableLoaded && (
               <div className="flex items-center gap-2">
+                <Badge
+                  status={
+                    isTogglingEnable
+                      ? 'pending'
+                      : botEnabled
+                        ? 'enabled'
+                        : 'disabled'
+                  }
+                  className="gap-1.5 rounded-full"
+                  role="status"
+                  aria-live="polite"
+                >
+                  {isTogglingEnable ? (
+                    <Loader2 className="animate-spin" aria-hidden="true" />
+                  ) : (
+                    <span
+                      className="size-1.5 rounded-full bg-current"
+                      aria-hidden="true"
+                    />
+                  )}
+                  {t(
+                    isTogglingEnable
+                      ? botEnabled
+                        ? 'bots.enabling'
+                        : 'bots.disabling'
+                      : botEnabled
+                        ? 'bots.enableConfirmed'
+                        : 'bots.disableConfirmed',
+                  )}
+                </Badge>
                 <Switch
                   id="bot-enable-switch"
                   checked={botEnabled}
                   onCheckedChange={handleEnableToggle}
-                  disabled={!canManage}
+                  disabled={!canManage || isTogglingEnable}
+                  aria-label={t('common.enable')}
+                  aria-busy={isTogglingEnable}
                 />
-                <Label
-                  htmlFor="bot-enable-switch"
-                  className="text-sm text-muted-foreground cursor-pointer"
-                >
-                  {t('common.enable')}
-                </Label>
               </div>
             )}
           </div>
           {canManage && (
             <div className="flex shrink-0 items-center gap-2">
-              <AdapterEventDebugDialog
-                key={id}
-                botId={id}
-                adapterLabel={adapterLabel}
-              />
+              {activeTab === 'config' && (
+                <AdapterEventDebugDialog
+                  key={id}
+                  botId={id}
+                  adapterLabel={adapterLabel}
+                />
+              )}
               <Button
                 type="submit"
                 form="bot-form"
-                disabled={!formDirty}
+                disabled={!enableLoaded || !formDirty || isTogglingEnable}
                 className={activeTab !== 'config' ? 'invisible' : ''}
                 data-guide="bot-config-save"
               >
-                {t('common.save')}
+                {t(
+                  formDirty && !botEnabled
+                    ? 'bots.saveAndEnable'
+                    : 'common.save',
+                )}
               </Button>
               <Button
                 type="button"
diff --git a/web/src/app/home/bots/components/bot-form/BotForm.tsx b/web/src/app/home/bots/components/bot-form/BotForm.tsx
index d8b6f2123..0ae7d6dff 100644
--- a/web/src/app/home/bots/components/bot-form/BotForm.tsx
+++ b/web/src/app/home/bots/components/bot-form/BotForm.tsx
@@ -27,18 +27,22 @@ import { Agent, Bot } from '@/app/infra/entities/api';
 import { getAdapterDocUrl } from '@/app/infra/entities/adapter-docs';
 import {
   Cable,
+  TriangleAlert,
   Check,
   ExternalLink,
   ChevronDown,
   ChevronRight,
   Webhook,
+  Search,
 } from 'lucide-react';
 import { cn } from '@/lib/utils';
 import EventBindingsEditor from './EventBindingsEditor';
+import EventRoutingHelp from './EventRoutingHelp';
+import LegacyAdapterHelp from './LegacyAdapterHelp';
 import PluginProcessorBindings from './PluginProcessorBindings';
 
 import { zodResolver } from '@hookform/resolvers/zod';
-import { useForm } from 'react-hook-form';
+import { useForm, useWatch } from 'react-hook-form';
 import { z } from 'zod';
 import { toast } from 'sonner';
 import { useTranslation } from 'react-i18next';
@@ -152,7 +156,8 @@ function detectConnectionMode(
   if (!webhookField.show_if) return 'webhook';
   return conditionMatches(
     webhookField.show_if,
-    values[webhookField.show_if.field],
+    values[webhookField.show_if.field] ??
+      config.find((item) => item.name === webhookField.show_if?.field)?.default,
   )
     ? 'webhook'
     : 'persistent';
@@ -273,7 +278,26 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
   useEffect(() => {
     onAdapterLabelChange?.(adapterLabel);
   }, [adapterLabel, onAdapterLabelChange]);
-  const currentAdapterConfig = form.watch('adapter_config');
+  const currentAdapterConfig = form.getValues('adapter_config');
+  const embedTitle = useWatch({
+    control: form.control,
+    name: 'adapter_config',
+    compute: (config) => config?.title,
+  });
+  const dynamicFormSystemContext = useMemo(
+    () => ({
+      bot_uuid: initBotId,
+      webhook_url: webhookUrl,
+      onebot_ws_url: webhookUrl
+        ? webhookUrl.replace(/^http/, 'ws').replace(/\/$/, '') +
+          '/onebot/v11/ws'
+        : '',
+      extra_webhook_url: extraWebhookUrl,
+      adapter_config: { title: embedTitle },
+      outbound_ips: systemInfo.outbound_ips,
+    }),
+    [initBotId, webhookUrl, extraWebhookUrl, embedTitle],
+  );
 
   // Group adapters by category for the Select dropdown. Legacy adapters are
   // split out and shown in a collapsed group at the bottom so they're
@@ -290,6 +314,16 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
     () => groupByCategory(activeAdapters),
     [activeAdapters],
   );
+  const [adapterSearch, setAdapterSearch] = useState('');
+  const adapterQuery = adapterSearch.trim().toLocaleLowerCase();
+  const matchesAdapterSearch = (adapter: IChooseAdapterEntity) =>
+    `${adapter.label} ${adapter.value} ${adapterDescriptionList[adapter.value] ?? ''}`
+      .toLocaleLowerCase()
+      .includes(adapterQuery);
+  const galleryGroups = groupByCategory(
+    activeAdapters.filter(matchesAdapterSearch),
+  );
+  const galleryLegacyAdapters = legacyAdapters.filter(matchesAdapterSearch);
 
   // Whether the collapsed legacy adapter group is expanded in the Select.
   const [showLegacyAdapters, setShowLegacyAdapters] = useState(false);
@@ -464,8 +498,7 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
         .then((res) => {
           const bot = res.bot;
           const runtimeValues = bot.adapter_runtime_values as
-            | Record<string, unknown>
-            | undefined;
+            Record<string, unknown> | undefined;
           resolve({
             adapter: bot.adapter,
             description: bot.description,
@@ -475,11 +508,9 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
             event_bindings: bot.event_bindings ?? [],
             plugin_processors: bot.plugin_processors ?? [],
             webhook_full_url: runtimeValues?.webhook_full_url as
-              | string
-              | undefined,
+              string | undefined,
             extra_webhook_full_url: runtimeValues?.extra_webhook_full_url as
-              | string
-              | undefined,
+              string | undefined,
           });
         })
         .catch((err) => {
@@ -522,6 +553,17 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
   }
 
   function selectAdapter(adapterName: string) {
+    if (!form.getValues('name')?.trim()) {
+      const selectedAdapter = adapterNameList.find(
+        (adapter) => adapter.value === adapterName,
+      );
+      if (selectedAdapter) {
+        form.setValue('name', selectedAdapter.label, {
+          shouldDirty: true,
+          shouldValidate: true,
+        });
+      }
+    }
     form.setValue('adapter', adapterName, {
       shouldDirty: !isInitializing.current,
       shouldValidate: true,
@@ -552,6 +594,14 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
     [adapterNameToDynamicConfigMap, currentAdapter],
   );
 
+  const connectionModeControllerNames = useMemo(() => {
+    if (supportedConnectionModes.length <= 1) return [];
+    const webhookField = dynamicFormConfigList.find(
+      (item) => item.type === DynamicFormItemType.WEBHOOK_URL,
+    );
+    return webhookField?.show_if?.field ? [webhookField.show_if.field] : [];
+  }, [dynamicFormConfigList, supportedConnectionModes.length]);
+
   const botGuideSteps = useMemo<GuidedTourStep[]>(() => {
     const steps: GuidedTourStep[] = [];
 
@@ -649,7 +699,7 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
       description: form.getValues().description ?? '',
       adapter: form.getValues().adapter,
       adapter_config: form.getValues().adapter_config,
-      enable: form.getValues().enable,
+      enable: true,
       event_bindings: form.getValues().event_bindings ?? [],
       plugin_processors: form.getValues().plugin_processors ?? [],
     };
@@ -657,7 +707,7 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
       .updateBot(initBotId, updateBot)
       .then(() => {
         // Reset dirty baseline to current values so isDirty becomes false
-        form.reset(form.getValues());
+        form.reset({ ...form.getValues(), enable: true });
         onFormSubmit(form.getValues());
         toast.success(t('bots.saveSuccess'));
       })
@@ -669,6 +719,18 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
       });
   }
 
+  const hasProcessing =
+    (form.watch('event_bindings') ?? []).some(
+      (binding) =>
+        binding.enabled !== false &&
+        !!binding.event_pattern &&
+        (binding.target_type === 'discard' ||
+          agentNameList.some((agent) => agent.uuid === binding.target_uuid)),
+    ) ||
+    (form.watch('plugin_processors') ?? []).some(
+      (processor) => processor.enabled !== false && !!processor.processor_uuid,
+    );
+
   if (loadFailed)
     return (
       <EntityLoadState error onRetry={() => setLoadAttempt((n) => n + 1)} />
@@ -740,17 +802,44 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
                 data-testid="adapter-gallery"
                 className="min-w-0 space-y-8 py-2"
               >
-                <div className="max-w-2xl">
-                  <h2 className="text-xl font-semibold">
-                    {t('guidedTour.bot.adapter.title')}
-                  </h2>
+                <div>
+                  <div className="flex flex-wrap items-center justify-between gap-3">
+                    <h2 className="text-xl font-semibold">
+                      {t('guidedTour.bot.adapter.title')}
+                    </h2>
+                    <div className="relative w-full sm:w-64">
+                      <Search className="pointer-events-none absolute left-3 top-1/2 size-4 -translate-y-1/2 text-muted-foreground" />
+                      <Input
+                        type="search"
+                        value={adapterSearch}
+                        onChange={(event) =>
+                          setAdapterSearch(event.target.value)
+                        }
+                        onKeyDown={(event) => {
+                          if (event.key === 'Enter') event.preventDefault();
+                        }}
+                        placeholder={t('bots.searchAdapters')}
+                        aria-label={t('bots.searchAdapters')}
+                        className="pl-9"
+                      />
+                    </div>
+                  </div>
                   <p className="mt-2 text-sm leading-6 text-muted-foreground">
                     {t('guidedTour.bot.adapter.description')}
                   </p>
                 </div>
 
                 <div className="space-y-8">
-                  {groupedAdapters.map((group) => (
+                  {galleryGroups.length === 0 &&
+                    galleryLegacyAdapters.length === 0 && (
+                      <p
+                        className="py-8 text-center text-sm text-muted-foreground"
+                        role="status"
+                      >
+                        {t('bots.noMatchingAdapters')}
+                      </p>
+                    )}
+                  {galleryGroups.map((group) => (
                     <section
                       key={group.categoryId ?? 'uncategorized'}
                       data-adapter-category={
@@ -805,30 +894,35 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
                     </section>
                   ))}
 
-                  {legacyAdapters.length > 0 && (
+                  {galleryLegacyAdapters.length > 0 && (
                     <section
                       className="border-t pt-5"
                       data-adapter-category="legacy"
                     >
-                      <button
-                        type="button"
-                        onClick={() => setShowLegacyAdapters((value) => !value)}
-                        className="flex w-full items-center gap-2 py-2 text-left text-sm font-medium text-muted-foreground hover:text-foreground"
-                        aria-expanded={showLegacyAdapters}
-                      >
-                        {showLegacyAdapters ? (
-                          <ChevronDown className="size-4" />
-                        ) : (
-                          <ChevronRight className="size-4" />
-                        )}
-                        {t('bots.legacyAdapters')}
-                        <span className="rounded bg-muted px-1.5 py-0.5 text-[10px]">
-                          {legacyAdapters.length}
-                        </span>
-                      </button>
-                      {showLegacyAdapters && (
+                      <div className="flex items-center gap-1">
+                        <button
+                          type="button"
+                          onClick={() =>
+                            setShowLegacyAdapters((value) => !value)
+                          }
+                          className="flex items-center gap-2 py-2 text-left text-sm font-medium text-muted-foreground hover:text-foreground"
+                          aria-expanded={showLegacyAdapters || !!adapterQuery}
+                        >
+                          {showLegacyAdapters || adapterQuery ? (
+                            <ChevronDown className="size-4" />
+                          ) : (
+                            <ChevronRight className="size-4" />
+                          )}
+                          {t('bots.legacyAdapters')}
+                          <span className="rounded bg-muted px-1.5 py-0.5 text-[10px]">
+                            {galleryLegacyAdapters.length}
+                          </span>
+                        </button>
+                        <LegacyAdapterHelp />
+                      </div>
+                      {(showLegacyAdapters || !!adapterQuery) && (
                         <div className="mt-3 grid grid-cols-1 gap-3 sm:grid-cols-2 lg:grid-cols-3">
-                          {legacyAdapters.map((item) => (
+                          {galleryLegacyAdapters.map((item) => (
                             <button
                               key={`legacy:${item.value}`}
                               type="button"
@@ -1140,19 +1234,14 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
                   <div data-guide="bot-adapter-parameters">
                     <DynamicFormComponent
                       itemConfigList={dynamicFormConfigList}
+                      hiddenItemNames={connectionModeControllerNames}
                       initialValues={currentAdapterConfig}
                       onSubmit={(values) => {
                         form.setValue('adapter_config', values, {
                           shouldDirty: !isInitializing.current,
                         });
                       }}
-                      systemContext={{
-                        webhook_url: webhookUrl,
-                        extra_webhook_url: extraWebhookUrl,
-                        bot_uuid: initBotId || '',
-                        adapter_config: form.getValues('adapter_config') || {},
-                        outbound_ips: systemInfo.outbound_ips,
-                      }}
+                      systemContext={dynamicFormSystemContext}
                     />
                   </div>
                 )}
@@ -1166,11 +1255,27 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
               data-guide="bot-routing"
               className={cn(
                 'min-w-0',
+                !hasProcessing && 'pt-0',
                 initBotId && 'lg:min-h-0 lg:overflow-hidden',
               )}
             >
+              {!hasProcessing && (
+                <div
+                  role="status"
+                  className="flex shrink-0 items-start gap-2.5 rounded-t-xl border-b border-amber-200 bg-amber-50 px-6 py-3 text-sm leading-relaxed text-amber-800 dark:border-amber-800/50 dark:bg-amber-950/40 dark:text-amber-200"
+                >
+                  <TriangleAlert
+                    className="mt-0.5 size-4 shrink-0"
+                    aria-hidden="true"
+                  />
+                  <p>{t('botSetup.warning')}</p>
+                </div>
+              )}
               <CardHeader>
-                <CardTitle>{t('bots.eventRouting')}</CardTitle>
+                <CardTitle className="flex items-center gap-1.5">
+                  {t('bots.eventRouting')}
+                  <EventRoutingHelp />
+                </CardTitle>
                 <CardDescription>
                   {t('bots.eventRoutingDescription')}
                 </CardDescription>
@@ -1183,6 +1288,12 @@ const BotForm = forwardRef<BotFormHandle, BotFormProps>(function BotForm(
               >
                 <EventBindingsEditor
                   form={form}
+                  onAgentCreated={(agent) =>
+                    setAgentNameList((items) => [
+                      ...items.filter((item) => item.uuid !== agent.uuid),
+                      agent,
+                    ])
+                  }
                   botId={initBotId}
                   supportedEvents={adapterSupportedEvents[currentAdapter] || []}
                   agentOptions={agentNameList.filter(
diff --git a/web/src/app/home/bots/components/bot-form/EventBindingsEditor.module.css b/web/src/app/home/bots/components/bot-form/EventBindingsEditor.module.css
new file mode 100644
index 000000000..0318c496a
--- /dev/null
+++ b/web/src/app/home/bots/components/bot-form/EventBindingsEditor.module.css
@@ -0,0 +1,32 @@
+.emptyRouteHint {
+  --route-glow-soft: 6%;
+  --route-glow-strong: 30%;
+  animation: route-hint-glow 3.6s ease-in-out infinite;
+}
+
+:global(.dark) .emptyRouteHint {
+  --route-glow-soft: 4%;
+  --route-glow-strong: 22%;
+}
+
+.emptyRouteHint:hover,
+.emptyRouteHint:focus-visible,
+.emptyRouteHint[data-state='open'] {
+  animation: none;
+}
+
+@keyframes route-hint-glow {
+  0%,
+  100% {
+    box-shadow: 0 0 6px 1px color-mix(in srgb, var(--color-amber-400) var(--route-glow-soft), transparent);
+  }
+  50% {
+    box-shadow: 0 0 14px 4px color-mix(in srgb, var(--color-amber-400) var(--route-glow-strong), transparent);
+  }
+}
+
+@media (prefers-reduced-motion: reduce) {
+  .emptyRouteHint {
+    animation: none;
+  }
+}
diff --git a/web/src/app/home/bots/components/bot-form/EventBindingsEditor.tsx b/web/src/app/home/bots/components/bot-form/EventBindingsEditor.tsx
index a750c02c6..06fcae3d7 100644
--- a/web/src/app/home/bots/components/bot-form/EventBindingsEditor.tsx
+++ b/web/src/app/home/bots/components/bot-form/EventBindingsEditor.tsx
@@ -9,6 +9,9 @@ import {
   useState,
 } from 'react';
 import { useTranslation } from 'react-i18next';
+import { Link } from 'react-router-dom';
+import { cn } from '@/lib/utils';
+import styles from './EventBindingsEditor.module.css';
 import type { TFunction } from 'i18next';
 import { UseFormReturn } from 'react-hook-form';
 import {
@@ -30,6 +33,7 @@ import {
   Play,
   RefreshCw,
   Shield,
+  Settings2,
   Trash2,
   UserCheck,
   UserMinus,
@@ -39,6 +43,7 @@ import {
   XCircle,
 } from 'lucide-react';
 import { Button } from '@/components/ui/button';
+import { Switch } from '@/components/ui/switch';
 import { Badge } from '@/components/ui/badge';
 import {
   Popover,
@@ -115,6 +120,9 @@ import {
   BotEventRouteStatus,
 } from '@/app/infra/entities/api';
 import { backendClient } from '@/app/infra/http';
+import AgentCreateContent from '@/app/home/agents/components/AgentCreateContent';
+import { useSidebarData } from '@/app/home/components/home-sidebar/SidebarDataContext';
+import { toast } from 'sonner';
 import {
   eventGroupLabel,
   eventNamespaces,
@@ -132,20 +140,13 @@ interface EventBindingsEditorProps {
   botId?: string;
   supportedEvents: string[];
   agentOptions: Agent[];
+  onAgentCreated: (agent: Agent) => void;
 }
 
 type FilterField =
-  | 'chat_type'
-  | 'chat_id'
-  | 'message_text'
-  | 'message_element_types';
+  'chat_type' | 'chat_id' | 'message_text' | 'message_element_types';
 type FilterOperator =
-  | 'eq'
-  | 'neq'
-  | 'contains'
-  | 'not_contains'
-  | 'starts_with'
-  | 'regex';
+  'eq' | 'neq' | 'contains' | 'not_contains' | 'starts_with' | 'regex';
 
 interface FilterRow {
   field: FilterField;
@@ -218,7 +219,7 @@ const BEHAVIOR_PRESETS = [
 // ── helpers ───────────────────────────────────────────────────────────────────
 
 function isMessageEventPattern(p: string) {
-  return p === 'message.*' || p.startsWith('message.');
+  return p === 'message.received';
 }
 
 interface RouteConflict {
@@ -327,7 +328,14 @@ function eventLabel(event: string, t: TFunction) {
 
 function eventDescription(event: string, t: TFunction) {
   if (event === '*') return t('bots.eventDescriptions.all');
-  if (event.endsWith('.*')) return t('bots.eventDescriptions.namespace');
+  if (event.endsWith('.*')) {
+    const group = event.slice(0, -2);
+    const key = `bots.eventDescriptions.namespace_${group}`;
+    const description = t(key);
+    return description === key
+      ? t('bots.eventDescriptions.namespace', { group })
+      : description;
+  }
   const key = `bots.eventDescriptions.${event.replace(/\./g, '_')}`;
   const description = t(key);
   return description === key ? t('bots.eventDescriptions.custom') : description;
@@ -485,15 +493,23 @@ function TargetCombobox({
   binding,
   agentOptions,
   onUpdate,
+  onAgentCreated,
 }: {
   binding: EventBinding;
   agentOptions: Agent[];
   onUpdate: (patch: Partial<EventBinding>) => void;
+  onAgentCreated: (agent: Agent) => void;
 }) {
   const { t } = useTranslation();
   const [open, setOpen] = useState(false);
+  const [creating, setCreating] = useState(false);
+  const { refreshPipelines } = useSidebarData();
   const pipelineAllowed = isMessageEventPattern(binding.event_pattern);
   const targetType = binding.target_type || 'agent';
+  const selectedTarget =
+    targetType !== 'discard'
+      ? agentOptions.find((agent) => agent.uuid === binding.target_uuid)
+      : undefined;
 
   const current =
     targetType === 'discard'
@@ -544,63 +560,147 @@ function TargetCombobox({
   }
 
   return (
-    <Popover open={open} onOpenChange={setOpen}>
-      <PopoverTrigger asChild>
-        <Button
-          variant="outline"
-          role="combobox"
-          aria-expanded={open}
-          className="h-8 w-[200px] justify-between text-sm font-normal px-3"
+    <div className="relative w-[200px] shrink-0">
+      <Popover open={open} onOpenChange={setOpen}>
+        <PopoverTrigger asChild>
+          <Button
+            variant="outline"
+            role="combobox"
+            aria-expanded={open}
+            className={cn(
+              'h-8 w-full justify-between text-sm font-normal px-3',
+              selectedTarget && 'pr-10',
+            )}
+          >
+            <span className="truncate">{currentLabel()}</span>
+            <ChevronsUpDown className="ml-1 h-3.5 w-3.5 shrink-0 opacity-50" />
+          </Button>
+        </PopoverTrigger>
+        <PopoverContent className="w-[240px] p-0" align="start">
+          <Command>
+            <CommandInput
+              placeholder={t('bots.searchTarget')}
+              className="h-8"
+            />
+            <CommandList>
+              <CommandEmpty>{t('bots.noTargetFound')}</CommandEmpty>
+              {agents.length > 0 && (
+                <CommandGroup heading={t('bots.targetAgent')}>
+                  {agents.map((a) => (
+                    <CommandItem
+                      key={a.uuid}
+                      value={`agent:${a.uuid}:${a.name}`}
+                      onSelect={() =>
+                        select(encodeTarget('agent', a.uuid || ''))
+                      }
+                    >
+                      <Bot className="mr-2 size-3.5 shrink-0" />
+                      <span className="truncate">{targetLabel(a)}</span>
+                      {current === encodeTarget('agent', a.uuid || '') && (
+                        <Check className="ml-auto size-3.5 shrink-0" />
+                      )}
+                    </CommandItem>
+                  ))}
+                </CommandGroup>
+              )}
+              {pipelines.length > 0 && (
+                <CommandGroup heading={t('bots.targetPipeline')}>
+                  {pipelines.map((a) => (
+                    <CommandItem
+                      key={a.uuid}
+                      value={`pipeline:${a.uuid}:${a.name}`}
+                      onSelect={() =>
+                        select(encodeTarget('pipeline', a.uuid || ''))
+                      }
+                    >
+                      <Workflow className="mr-2 size-3.5 shrink-0" />
+                      <span className="truncate">{targetLabel(a)}</span>
+                      {current === encodeTarget('pipeline', a.uuid || '') && (
+                        <Check className="ml-auto size-3.5 shrink-0" />
+                      )}
+                    </CommandItem>
+                  ))}
+                </CommandGroup>
+              )}
+            </CommandList>
+          </Command>
+          <div className="border-t p-1">
+            <Button
+              type="button"
+              variant="ghost"
+              className="w-full justify-start"
+              onClick={() => {
+                setOpen(false);
+                setCreating(true);
+              }}
+            >
+              <Plus className="mr-2 size-4" />
+              {t('botSetup.create')}
+            </Button>
+          </div>
+        </PopoverContent>
+      </Popover>
+      <Dialog open={creating} onOpenChange={setCreating}>
+        <DialogContent
+          className="flex max-h-[90dvh] w-[calc(100%-2rem)] flex-col overflow-y-auto sm:max-w-6xl"
+          onSubmit={(event) => event.stopPropagation()}
         >
-          <span className="truncate">{currentLabel()}</span>
-          <ChevronsUpDown className="ml-1 h-3.5 w-3.5 shrink-0 opacity-50" />
-        </Button>
-      </PopoverTrigger>
-      <PopoverContent className="w-[240px] p-0" align="start">
-        <Command>
-          <CommandInput placeholder={t('bots.searchTarget')} className="h-8" />
-          <CommandList>
-            <CommandEmpty>{t('bots.noTargetFound')}</CommandEmpty>
-            {agents.length > 0 && (
-              <CommandGroup heading={t('bots.targetAgent')}>
-                {agents.map((a) => (
-                  <CommandItem
-                    key={a.uuid}
-                    value={`agent:${a.uuid}:${a.name}`}
-                    onSelect={() => select(encodeTarget('agent', a.uuid || ''))}
-                  >
-                    <Bot className="mr-2 size-3.5 shrink-0" />
-                    <span className="truncate">{targetLabel(a)}</span>
-                    {current === encodeTarget('agent', a.uuid || '') && (
-                      <Check className="ml-auto size-3.5 shrink-0" />
-                    )}
-                  </CommandItem>
-                ))}
-              </CommandGroup>
-            )}
-            {pipelines.length > 0 && (
-              <CommandGroup heading={t('bots.targetPipeline')}>
-                {pipelines.map((a) => (
-                  <CommandItem
-                    key={a.uuid}
-                    value={`pipeline:${a.uuid}:${a.name}`}
-                    onSelect={() =>
-                      select(encodeTarget('pipeline', a.uuid || ''))
-                    }
-                  >
-                    <Workflow className="mr-2 size-3.5 shrink-0" />
-                    <span className="truncate">{targetLabel(a)}</span>
-                    {current === encodeTarget('pipeline', a.uuid || '') && (
-                      <Check className="ml-auto size-3.5 shrink-0" />
-                    )}
-                  </CommandItem>
-                ))}
-              </CommandGroup>
-            )}
-          </CommandList>
-        </Command>
-      </PopoverContent>
-    </Popover>
+          <DialogHeader>
+            <DialogTitle>{t('botSetup.create')}</DialogTitle>
+            <DialogDescription>
+              {t('botSetup.createDescription')}
+            </DialogDescription>
+          </DialogHeader>
+          <AgentCreateContent
+            embedded
+            allowedKinds={pipelineAllowed ? ['pipeline', 'agent'] : ['agent']}
+            onCreated={async (id) => {
+              void refreshPipelines();
+              try {
+                const { agent } = await backendClient.getAgent(id);
+                onAgentCreated(agent);
+                const compatible =
+                  agent.kind === 'pipeline'
+                    ? pipelineAllowed
+                    : agentSupportsEventPattern(agent, binding.event_pattern);
+                if (compatible) {
+                  onUpdate({ target_type: agent.kind, target_uuid: id });
+                } else {
+                  toast.warning(t('botSetup.incompatible'));
+                }
+                setCreating(false);
+              } catch {
+                toast.error(t('agents.createError'));
+              }
+            }}
+          />
+        </DialogContent>
+      </Dialog>
+      {selectedTarget && (
+        <Tooltip>
+          <TooltipTrigger asChild>
+            <Button
+              asChild
+              variant="ghost"
+              size="icon"
+              className="absolute right-1 top-1 size-6"
+            >
+              <Link
+                to={`/home/agents?id=${encodeURIComponent(selectedTarget.uuid || '')}&tab=config`}
+                target="_blank"
+                rel="noopener noreferrer"
+                aria-label={t('bots.pluginSubscriptions.configure')}
+              >
+                <Settings2 className="size-3.5" />
+              </Link>
+            </Button>
+          </TooltipTrigger>
+          <TooltipContent>
+            {t('bots.pluginSubscriptions.configure')}
+          </TooltipContent>
+        </Tooltip>
+      )}
+    </div>
   );
 }
 
@@ -1179,9 +1279,10 @@ function RouteDryRunDialog({
 interface BindingCardProps {
   binding: EventBinding;
   globalIndex: number;
-  routeStatus?: BotEventRouteStatus;
+
   eventOptions: string[];
   agentOptions: Agent[];
+  onAgentCreated: (agent: Agent) => void;
   expandedIds: Set<string>;
   onToggleExpand: (id: string) => void;
   onUpdate: (globalIndex: number, patch: Partial<EventBinding>) => void;
@@ -1193,9 +1294,9 @@ interface BindingCardProps {
 function BindingCardContent({
   binding,
   globalIndex,
-  routeStatus,
   eventOptions,
   agentOptions,
+  onAgentCreated,
   expandedIds,
   onToggleExpand,
   onUpdate,
@@ -1209,12 +1310,10 @@ function BindingCardContent({
   const isExpanded = expandedIds.has(id);
   const filterCount = (binding.filters as FilterRow[] | undefined)?.length ?? 0;
   const pipelineAllowed = isMessageEventPattern(binding.event_pattern);
-  const statusTime = formatRouteStatusTime(routeStatus?.timestamp);
-  const statusDetail = routeStatusDetail(routeStatus, t);
 
   return (
     <div
-      className={`rounded-lg border bg-card ${
+      className={`rounded-lg border ${isEnabled ? 'bg-card' : 'bg-muted/30 border-dashed'} ${
         isOverlay ? 'pointer-events-none shadow-lg ring-1 ring-primary/20' : ''
       }`}
       data-drag-overlay={isOverlay ? 'true' : undefined}
@@ -1233,6 +1332,7 @@ function BindingCardContent({
             <GripVertical className="h-4 w-4" />
           </button>
         )}
+        {!isEnabled && <span className="h-4 w-4 shrink-0" aria-hidden="true" />}
 
         <Badge
           variant="secondary"
@@ -1309,6 +1409,7 @@ function BindingCardContent({
         <TargetCombobox
           binding={binding}
           agentOptions={agentOptions}
+          onAgentCreated={onAgentCreated}
           onUpdate={(patch) => onUpdate(globalIndex, patch)}
         />
 
@@ -1319,16 +1420,11 @@ function BindingCardContent({
         )}
 
         <div className="ml-auto flex shrink-0 items-center gap-1">
-          {/* disable/enable toggle */}
-          <Button
-            type="button"
-            variant="ghost"
-            size="sm"
-            className="h-8 px-2 text-xs text-muted-foreground"
-            onClick={() => onUpdate(globalIndex, { enabled: !isEnabled })}
-          >
-            {isEnabled ? t('bots.disable') : t('bots.enable')}
-          </Button>
+          <Switch
+            checked={isEnabled}
+            onCheckedChange={(enabled) => onUpdate(globalIndex, { enabled })}
+            aria-label={`${t('bots.enable')} · ${t('bots.dryRunRuleIndex', { index: globalIndex + 1 })}`}
+          />
 
           <Button
             type="button"
@@ -1342,23 +1438,6 @@ function BindingCardContent({
         </div>
       </div>
 
-      <div className="flex items-center gap-2 border-t px-3 py-1.5">
-        <Badge
-          variant="outline"
-          className={`rounded-md px-2 py-0.5 text-[10px] font-medium ${routeStatusBadgeClass(
-            routeStatus?.last_status,
-          )}`}
-        >
-          {routeStatusLabel(routeStatus?.last_status, t)}
-        </Badge>
-        <span
-          className="min-w-0 truncate text-[11px] text-muted-foreground"
-          title={statusTime ? `${statusDetail} · ${statusTime}` : statusDetail}
-        >
-          {statusDetail || statusTime}
-        </span>
-      </div>
-
       {/* conditions panel */}
       {isExpanded && (
         <div className="border-t px-3 py-2.5">
@@ -1421,14 +1500,15 @@ export default function EventBindingsEditor({
   botId,
   supportedEvents,
   agentOptions,
+  onAgentCreated,
 }: EventBindingsEditorProps) {
   const { t } = useTranslation();
   const watchedBindings: EventBinding[] | undefined =
     form.watch('event_bindings');
   const bindings = useMemo(() => watchedBindings ?? [], [watchedBindings]);
   const [expandedIds, setExpandedIds] = useState<Set<string>>(new Set());
-  const [disabledSectionOpen, setDisabledSectionOpen] = useState(false);
   const [activeId, setActiveId] = useState<string | null>(null);
+  const [recordsOpen, setRecordsOpen] = useState(false);
   const [routeStatuses, setRouteStatuses] = useState<BotEventRouteStatus[]>([]);
   const [routeStatusLoading, setRouteStatusLoading] = useState(false);
   const [routeStatusError, setRouteStatusError] = useState<string | null>(null);
@@ -1455,13 +1535,7 @@ export default function EventBindingsEditor({
     () => (supportedEvents.length > 0 ? supportedEvents : DEFAULT_EVENTS),
     [supportedEvents],
   );
-  const routeStatusByBinding = useMemo(() => {
-    const map = new Map<string, BotEventRouteStatus>();
-    routeStatuses.forEach((status) => {
-      if (status.binding_id) map.set(String(status.binding_id), status);
-    });
-    return map;
-  }, [routeStatuses]);
+
   const routeConflicts = useMemo(
     () => findRouteConflicts(bindings),
     [bindings],
@@ -1496,7 +1570,15 @@ export default function EventBindingsEditor({
     setRouteStatusError(null);
     try {
       const response = await backendClient.getBotEventRouteStatuses(botId);
-      setRouteStatuses(response.routes || []);
+      setRouteStatuses(
+        [
+          ...response.routes,
+          ...response.unmatched_events,
+          ...response.stale_routes,
+        ]
+          .filter((record) => record.last_status)
+          .sort((a, b) => (b.timestamp ?? 0) - (a.timestamp ?? 0)),
+      );
     } catch (error) {
       console.error('Failed to refresh Bot event route status', error);
       setRouteStatusError(t('bots.routeStatusRefreshFailed'));
@@ -1506,11 +1588,15 @@ export default function EventBindingsEditor({
   }, [botId, t]);
 
   useEffect(() => {
-    refreshRouteStatuses();
-  }, [refreshRouteStatuses]);
+    if (recordsOpen) void refreshRouteStatuses();
+  }, [recordsOpen, refreshRouteStatuses]);
 
   function updateBindings(next: EventBinding[]) {
-    form.setValue('event_bindings', next, { shouldDirty: true });
+    const ordered = [
+      ...next.filter((binding) => binding.enabled ?? true),
+      ...next.filter((binding) => !(binding.enabled ?? true)),
+    ];
+    form.setValue('event_bindings', ordered, { shouldDirty: true });
   }
 
   function addBinding(eventPattern = 'message.received') {
@@ -1661,7 +1747,7 @@ export default function EventBindingsEditor({
           strategy={verticalListSortingStrategy}
         >
           <div className="space-y-2">
-            {enabledBindings.length === 0 && (
+            {bindings.length === 0 && (
               <div className="rounded-lg border border-dashed p-5 text-center text-sm text-muted-foreground">
                 {t('bots.noEventBindings')}
               </div>
@@ -1674,13 +1760,9 @@ export default function EventBindingsEditor({
                   sortableId={idsRef.current[sortIdx]}
                   binding={binding}
                   globalIndex={globalIdx}
-                  routeStatus={
-                    binding.id
-                      ? routeStatusByBinding.get(String(binding.id))
-                      : undefined
-                  }
                   eventOptions={eventOptions}
                   agentOptions={agentOptions}
+                  onAgentCreated={onAgentCreated}
                   expandedIds={expandedIds}
                   onToggleExpand={toggleExpand}
                   onUpdate={updateBinding}
@@ -1688,6 +1770,20 @@ export default function EventBindingsEditor({
                 />
               );
             })}
+            {disabledBindings.map(({ b, i }) => (
+              <BindingCardContent
+                key={b.id ?? `disabled-${i}`}
+                binding={b}
+                globalIndex={i}
+                eventOptions={eventOptions}
+                agentOptions={agentOptions}
+                onAgentCreated={onAgentCreated}
+                expandedIds={expandedIds}
+                onToggleExpand={toggleExpand}
+                onUpdate={updateBinding}
+                onRemove={removeBinding}
+              />
+            ))}
           </div>
         </SortableContext>
         <DragOverlay adjustScale={false} dropAnimation={null}>
@@ -1695,13 +1791,9 @@ export default function EventBindingsEditor({
             <BindingCardContent
               binding={activeBinding}
               globalIndex={activeGlobalIdx}
-              routeStatus={
-                activeBinding.id
-                  ? routeStatusByBinding.get(String(activeBinding.id))
-                  : undefined
-              }
               eventOptions={eventOptions}
               agentOptions={agentOptions}
+              onAgentCreated={onAgentCreated}
               expandedIds={expandedIds}
               onToggleExpand={toggleExpand}
               onUpdate={updateBinding}
@@ -1715,7 +1807,14 @@ export default function EventBindingsEditor({
       <div className="flex flex-wrap gap-2">
         <DropdownMenu>
           <DropdownMenuTrigger asChild>
-            <Button type="button" variant="outline" size="sm">
+            <Button
+              type="button"
+              variant="outline"
+              size="sm"
+              className={
+                bindings.length === 0 ? styles.emptyRouteHint : undefined
+              }
+            >
               <Plus className="h-4 w-4 mr-1" />
               {t('bots.addBehavior')}
               <ChevronDown className="ml-1 h-3.5 w-3.5" />
@@ -1791,68 +1890,100 @@ export default function EventBindingsEditor({
           eventOptions={dryRunEventOptions}
           agentOptions={agentOptions}
         />
-        <Tooltip>
-          <TooltipTrigger asChild>
-            <Button
-              type="button"
-              variant="ghost"
-              size="icon"
-              className={`size-8 ${routeStatusError ? 'text-destructive' : 'text-muted-foreground'}`}
-              aria-label={t('bots.refreshRouteStatus')}
-              onClick={refreshRouteStatuses}
-              disabled={!botId || routeStatusLoading}
-            >
-              <RefreshCw
-                className={`h-4 w-4 ${routeStatusLoading ? 'animate-spin' : ''}`}
-              />
-            </Button>
-          </TooltipTrigger>
-          <TooltipContent>
-            {routeStatusError || t('bots.refreshRouteStatus')}
-          </TooltipContent>
-        </Tooltip>
-      </div>
-
-      {/* disabled section */}
-      {disabledBindings.length > 0 && (
-        <div className="rounded-lg border border-dashed">
-          <button
-            type="button"
-            className="flex w-full items-center gap-2 px-3 py-2 text-sm text-muted-foreground hover:text-foreground"
-            onClick={() => setDisabledSectionOpen((v) => !v)}
-          >
-            {disabledSectionOpen ? (
-              <ChevronDown className="h-4 w-4" />
-            ) : (
-              <ChevronRight className="h-4 w-4" />
-            )}
-            {t('bots.disabledBindings')}
-            <Badge variant="outline" className="ml-1 text-xs">
-              {disabledBindings.length}
-            </Badge>
-          </button>
-          {disabledSectionOpen && (
-            <div className="border-t p-2 space-y-2">
-              {disabledBindings.map(({ b, i }) => (
-                <BindingCardContent
-                  key={b.id ?? i}
-                  binding={b}
-                  globalIndex={i}
-                  routeStatus={
-                    b.id ? routeStatusByBinding.get(String(b.id)) : undefined
-                  }
-                  eventOptions={eventOptions}
-                  agentOptions={agentOptions}
-                  expandedIds={expandedIds}
-                  onToggleExpand={toggleExpand}
-                  onUpdate={updateBinding}
-                  onRemove={removeBinding}
+        <Button
+          type="button"
+          variant="outline"
+          size="sm"
+          disabled={!botId}
+          onClick={() => setRecordsOpen(true)}
+        >
+          <ListChecks className="mr-1 size-4" />
+          {t('bots.matchRecords')}
+        </Button>
+        <Dialog open={recordsOpen} onOpenChange={setRecordsOpen}>
+          <DialogContent className="sm:max-w-2xl">
+            <DialogHeader>
+              <DialogTitle>{t('bots.matchRecords')}</DialogTitle>
+              <DialogDescription>
+                {t('bots.matchRecordsDescription')}
+              </DialogDescription>
+            </DialogHeader>
+            <div className="flex justify-end">
+              <Button
+                type="button"
+                variant="outline"
+                size="sm"
+                onClick={refreshRouteStatuses}
+                disabled={routeStatusLoading}
+              >
+                <RefreshCw
+                  className={`mr-2 size-4 ${routeStatusLoading ? 'animate-spin' : ''}`}
                 />
-              ))}
+                {t('monitoring.refreshData')}
+              </Button>
             </div>
-          )}
-        </div>
-      )}
+            {routeStatusError && (
+              <p role="alert" className="text-sm text-destructive">
+                {routeStatusError}
+              </p>
+            )}
+            <div className="max-h-[60vh] space-y-3 overflow-y-auto">
+              {routeStatuses.map((record, index) => (
+                <div
+                  key={`${record.binding_id}:${record.seq_id}:${index}`}
+                  className="space-y-2 rounded-lg border p-3"
+                >
+                  <div className="flex flex-wrap items-center gap-2">
+                    <span className="flex-1 text-sm font-medium">
+                      {eventLabel(
+                        record.event_type || record.event_pattern || '*',
+                        t,
+                      )}
+                    </span>
+                    <Badge
+                      variant="outline"
+                      className={routeStatusBadgeClass(record.last_status)}
+                    >
+                      {routeStatusLabel(record.last_status, t)}
+                    </Badge>
+                  </div>
+                  <div className="text-xs text-muted-foreground">
+                    {formatRouteStatusTime(record.timestamp)}
+                  </div>
+                  {record.target_type && (
+                    <div className="break-words text-sm">
+                      {record.target_type}
+                      {record.target_uuid
+                        ? ` · ${agentOptions.find((agent) => agent.uuid === record.target_uuid)?.name || record.target_uuid}`
+                        : ''}
+                    </div>
+                  )}
+                  <p className="text-sm text-muted-foreground">
+                    {routeStatusDetail(record, t)}
+                  </p>
+                  <details className="text-xs text-muted-foreground">
+                    <summary className="cursor-pointer">
+                      {t('monitoring.unified.metadata')}
+                    </summary>
+                    <pre className="mt-2 whitespace-pre-wrap break-words [overflow-wrap:anywhere]">
+                      {JSON.stringify(record, null, 2)}
+                    </pre>
+                  </details>
+                </div>
+              ))}
+              {!routeStatuses.length && (
+                <p className="py-8 text-center text-sm text-muted-foreground">
+                  {t(
+                    routeStatusLoading
+                      ? 'common.loading'
+                      : 'monitoring.unified.noRecords',
+                  )}
+                </p>
+              )}
+            </div>
+          </DialogContent>
+        </Dialog>
+      </div>
     </div>
   );
 }
diff --git a/web/src/app/home/bots/components/bot-form/EventRoutingHelp.css b/web/src/app/home/bots/components/bot-form/EventRoutingHelp.css
new file mode 100644
index 000000000..cfd3b20b8
--- /dev/null
+++ b/web/src/app/home/bots/components/bot-form/EventRoutingHelp.css
@@ -0,0 +1,23 @@
+.routing-help-dot {
+  animation: route-travel 4.8s ease-in-out infinite;
+  animation-delay: var(--route-delay);
+  opacity: 0;
+}
+.routing-help-target {
+  animation: route-receive 4.8s ease-in-out infinite;
+  animation-delay: var(--route-delay);
+}
+.routing-help-paused * { animation-play-state: paused; }
+@keyframes route-travel {
+  0%, 4% { opacity: 0; transform: translateX(0); }
+  8% { opacity: 1; }
+  25% { opacity: 1; transform: translateX(40px); }
+  30%, 100% { opacity: 0; transform: translateX(40px); }
+}
+@keyframes route-receive {
+  0%, 24%, 45%, 100% { transform: translateY(0); }
+  30%, 36% { transform: translateY(-2px); }
+}
+@media (prefers-reduced-motion: reduce) {
+  .routing-help-dot, .routing-help-target { animation: none; }
+}
diff --git a/web/src/app/home/bots/components/bot-form/EventRoutingHelp.tsx b/web/src/app/home/bots/components/bot-form/EventRoutingHelp.tsx
new file mode 100644
index 000000000..07c4430f0
--- /dev/null
+++ b/web/src/app/home/bots/components/bot-form/EventRoutingHelp.tsx
@@ -0,0 +1,113 @@
+'use client';
+
+import { useState, type CSSProperties } from 'react';
+import { useTranslation } from 'react-i18next';
+import {
+  Popover,
+  PopoverContent,
+  PopoverTrigger,
+} from '@/components/ui/popover';
+import './EventRoutingHelp.css';
+
+const routeColors = [
+  'text-sky-700 dark:text-sky-300 [&_.route-node]:border-sky-200 [&_.route-node]:bg-sky-50 dark:[&_.route-node]:border-sky-800 dark:[&_.route-node]:bg-sky-950/40',
+  'text-amber-700 dark:text-amber-300 [&_.route-node]:border-amber-200 [&_.route-node]:bg-amber-50 dark:[&_.route-node]:border-amber-800 dark:[&_.route-node]:bg-amber-950/40',
+  'text-emerald-700 dark:text-emerald-300 [&_.route-node]:border-emerald-200 [&_.route-node]:bg-emerald-50 dark:[&_.route-node]:border-emerald-800 dark:[&_.route-node]:bg-emerald-950/40',
+];
+
+export default function EventRoutingHelp() {
+  const { t } = useTranslation();
+  const [paused, setPaused] = useState(false);
+  const text = (key: string) => t(`bots.routingHelp.${key}`);
+  return (
+    <Popover>
+      <PopoverTrigger asChild>
+        <button
+          type="button"
+          aria-label={text('title')}
+          className="inline-flex size-6 items-center justify-center rounded-full text-muted-foreground hover:bg-muted hover:text-foreground focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring"
+        >
+          <svg
+            width="16"
+            height="16"
+            viewBox="0 0 24 24"
+            fill="none"
+            stroke="currentColor"
+            strokeWidth="1.7"
+            aria-hidden="true"
+          >
+            <circle cx="12" cy="12" r="9" />
+            <path d="M12 11v6" />
+            <circle cx="12" cy="7.5" r=".7" fill="currentColor" />
+          </svg>
+        </button>
+      </PopoverTrigger>
+      <PopoverContent
+        align="start"
+        className="w-[min(32rem,calc(100vw-2rem))] max-h-[80vh] overflow-y-auto p-5"
+      >
+        <h3 className="text-sm font-semibold">{text('title')}</h3>
+        <p className="mt-2 text-sm leading-6 text-muted-foreground">
+          {text('intro')}
+        </p>
+        <div className="my-4 overflow-hidden rounded-xl border bg-muted/20 p-3">
+          <div className="mb-3 grid grid-cols-2 text-xs text-muted-foreground">
+            <span>{text('events')}</span>
+            <span className="text-right">{text('agents')}</span>
+          </div>
+          <div
+            className={`routing-help space-y-3 ${paused ? 'routing-help-paused' : ''}`}
+          >
+            {['message', 'friend', 'group'].map((kind, index) => (
+              <div
+                key={kind}
+                className={`grid grid-cols-[minmax(0,1fr)_48px_minmax(0,1fr)] items-center ${routeColors[index]}`}
+                style={{ '--route-delay': `${index * 1.6}s` } as CSSProperties}
+              >
+                <div className="route-node rounded-lg border px-2 py-3 text-center text-xs leading-5">
+                  {text(kind)}
+                </div>
+                <div className="relative h-8" aria-hidden="true">
+                  <svg viewBox="0 0 48 32" className="h-full w-full opacity-30">
+                    <path
+                      d="M0 16H45m-5-4 5 4-5 4"
+                      fill="none"
+                      stroke="currentColor"
+                      strokeWidth="1.5"
+                    />
+                  </svg>
+                  <span className="routing-help-dot absolute left-0 top-[13px] size-1.5 rounded-full bg-current" />
+                </div>
+                <div className="route-node routing-help-target rounded-lg border px-2 py-3 text-center text-xs leading-5">
+                  {text(`${kind}Action`)}
+                </div>
+              </div>
+            ))}
+          </div>
+          <div className="mt-3 flex items-center justify-between gap-3 text-xs text-muted-foreground">
+            <span>{text('rules')}</span>
+            <button
+              type="button"
+              className="shrink-0 rounded px-1 underline underline-offset-4 focus-visible:ring-2 focus-visible:ring-ring motion-reduce:hidden"
+              onClick={() => setPaused(!paused)}
+            >
+              {text(paused ? 'play' : 'pause')}
+            </button>
+          </div>
+        </div>
+        <p className="text-sm leading-6">{text('instructions')}</p>
+        <div className="mt-4 border-t pt-3 text-xs leading-6 text-muted-foreground">
+          <p>{text('plugin')}</p>
+          <a
+            href="https://langbot.app/docs/zh/plugin/dev/components/runner"
+            target="_blank"
+            rel="noopener noreferrer"
+            className="font-medium text-foreground underline underline-offset-4"
+          >
+            {text('docs')} ↗
+          </a>
+        </div>
+      </PopoverContent>
+    </Popover>
+  );
+}
diff --git a/web/src/app/home/bots/components/bot-form/LegacyAdapterHelp.tsx b/web/src/app/home/bots/components/bot-form/LegacyAdapterHelp.tsx
new file mode 100644
index 000000000..360b19a3d
--- /dev/null
+++ b/web/src/app/home/bots/components/bot-form/LegacyAdapterHelp.tsx
@@ -0,0 +1,41 @@
+import { useTranslation } from 'react-i18next';
+import {
+  Popover,
+  PopoverContent,
+  PopoverTrigger,
+} from '@/components/ui/popover';
+
+export default function LegacyAdapterHelp() {
+  const { t } = useTranslation();
+  return (
+    <Popover>
+      <PopoverTrigger asChild>
+        <button
+          type="button"
+          aria-label={t('bots.legacyAdapters')}
+          className="inline-flex size-6 shrink-0 items-center justify-center rounded-full text-muted-foreground hover:bg-muted hover:text-foreground focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring"
+        >
+          <svg
+            width="16"
+            height="16"
+            viewBox="0 0 24 24"
+            fill="none"
+            stroke="currentColor"
+            strokeWidth="1.7"
+            aria-hidden="true"
+          >
+            <circle cx="12" cy="12" r="9" />
+            <path d="M12 11v6" />
+            <circle cx="12" cy="7.5" r=".7" fill="currentColor" />
+          </svg>
+        </button>
+      </PopoverTrigger>
+      <PopoverContent align="start" className="w-80 max-w-[calc(100vw-2rem)]">
+        <h4 className="mb-2 text-sm font-medium">{t('bots.legacyAdapters')}</h4>
+        <p className="whitespace-pre-line text-sm leading-6 text-muted-foreground">
+          {t('bots.legacyAdaptersHint')}
+        </p>
+      </PopoverContent>
+    </Popover>
+  );
+}
diff --git a/web/src/app/home/bots/components/bot-session/BotSessionMonitor.tsx b/web/src/app/home/bots/components/bot-session/BotSessionMonitor.tsx
index 582cfb720..bdce4047f 100644
--- a/web/src/app/home/bots/components/bot-session/BotSessionMonitor.tsx
+++ b/web/src/app/home/bots/components/bot-session/BotSessionMonitor.tsx
@@ -1,3 +1,14 @@
+import { useCurrentWorkspace } from '@/app/infra/http';
+import { Button } from '@/components/ui/button';
+import {
+  Dialog,
+  DialogContent,
+  DialogHeader,
+  DialogTitle,
+  DialogDescription,
+  DialogFooter,
+} from '@/components/ui/dialog';
+import LoadErrorState from '@/components/LoadErrorState';
 import React, {
   useState,
   useEffect,
@@ -12,6 +23,7 @@ import { httpClient } from '@/app/infra/http/HttpClient';
 import { ScrollArea } from '@/components/ui/scroll-area';
 import { cn } from '@/lib/utils';
 import {
+  RotateCcw,
   Ban,
   Bot,
   Copy,
@@ -142,6 +154,23 @@ const BotSessionMonitor = forwardRef<
   BotSessionMonitorProps
 >(function BotSessionMonitor({ botId }, ref) {
   const { t } = useTranslation();
+  const workspace = useCurrentWorkspace();
+  const canReset = workspace?.permissions.includes('resource.manage') ?? false;
+  const [resetTarget, setResetTarget] = useState<string | null>(null);
+  const [resetting, setResetting] = useState(false);
+  const resetContext = async () => {
+    if (!resetTarget || resetting) return;
+    setResetting(true);
+    try {
+      await httpClient.resetSessionContext(botId, resetTarget);
+      toast.success(t('bots.sessionMonitor.resetSuccess'));
+      setResetTarget(null);
+    } catch {
+      toast.error(t('bots.sessionMonitor.resetError'));
+    } finally {
+      setResetting(false);
+    }
+  };
   const [sessions, setSessions] = useState<SessionInfo[]>([]);
   const [sessionTotal, setSessionTotal] = useState(0);
   const [sessionPage, setSessionPage] = useState(0);
@@ -319,7 +348,7 @@ const BotSessionMonitor = forwardRef<
 
         // Collect user message IDs for feedback matching
         const userMsgIds = new Set(
-          sorted.filter((m) => !m.role || m.role === 'user').map((m) => m.id),
+          sorted.filter(isUserMessage).map((m) => m.id),
         );
 
         if (userMsgIds.size > 0) {
@@ -389,9 +418,10 @@ const BotSessionMonitor = forwardRef<
   }, [selectedSessionId, messagePage, loadMessages]);
 
   useEffect(() => {
+    if (loadingMessages || messageError) return;
     if (messages.length === 0 && toolCalls.length === 0) return;
-    // Wait for DOM to render the new messages before scrolling
-    requestAnimationFrame(() => {
+    // The timeline is hidden until all message data has finished loading.
+    const frame = requestAnimationFrame(() => {
       const container = messagesContainerRef.current;
       if (container) {
         const viewport = container.querySelector(
@@ -401,7 +431,8 @@ const BotSessionMonitor = forwardRef<
         scrollTarget.scrollTop = scrollTarget.scrollHeight;
       }
     });
-  }, [messages, toolCalls]);
+    return () => cancelAnimationFrame(frame);
+  }, [loadingMessages, messageError, messages, toolCalls]);
 
   const parseMessageChain = (content: string): MessageChainComponent[] => {
     try {
@@ -409,17 +440,44 @@ const BotSessionMonitor = forwardRef<
       if (Array.isArray(parsed)) {
         return parsed as MessageChainComponent[];
       }
+      if (parsed && typeof parsed === 'object') {
+        const chain = parsed.message_chain ?? parsed.event?.message_chain;
+        if (Array.isArray(chain)) return chain as MessageChainComponent[];
+        if (Array.isArray(chain?.root))
+          return chain.root as MessageChainComponent[];
+        if (Array.isArray(parsed.contents)) {
+          const components: MessageChainComponent[] = [];
+          for (const part of parsed.contents) {
+            if (part.type === 'text' && typeof part.text === 'string') {
+              components.push({ type: 'Plain', text: part.text } as Plain);
+            } else if (part.type === 'image' && part.image_url) {
+              components.push({
+                type: 'Image',
+                url: part.image_url,
+              } as MessageChainComponent);
+            }
+          }
+          if (components.length) return components;
+        }
+        if (typeof parsed.text === 'string') {
+          return [{ type: 'Plain', text: parsed.text } as Plain];
+        }
+      }
     } catch {
       // Not JSON, return as plain text
     }
     return [{ type: 'Plain', text: content } as Plain];
   };
 
-  const isUserMessage = (msg: SessionMessage): boolean => {
+  function isUserMessage(msg: SessionMessage): boolean {
+    // Older processing-error records inherited the default user role.
+    if (msg.status === 'error' && msg.message_content.startsWith('Error: ')) {
+      return false;
+    }
     if (msg.role === 'assistant') return false;
     if (msg.role === 'user') return true;
     return !msg.runner_name;
-  };
+  }
 
   const renderMessageComponent = (
     component: MessageChainComponent,
@@ -664,7 +722,7 @@ const BotSessionMonitor = forwardRef<
         {/* Left Panel: Session List */}
         <div className="max-h-48 md:max-h-none md:w-60 flex-shrink-0 border-b md:border-b-0 md:border-r flex flex-col min-h-0">
           {/* Admin header */}
-          <div className="px-2 py-1.5 border-b shrink-0 flex items-center justify-between">
+          <div className="px-2 py-1.5 border-b shrink-0 flex items-center justify-between gap-4">
             <button
               type="button"
               className="inline-flex items-center gap-1.5 text-sm font-medium hover:text-foreground transition-colors"
@@ -738,19 +796,11 @@ const BotSessionMonitor = forwardRef<
                 {t('bots.sessionMonitor.loading')}
               </div>
             ) : sessionError ? (
-              <div
-                role="alert"
-                className="p-3 space-y-2 text-sm text-destructive"
-              >
-                <p>{t('monitoring.loadError')}</p>
-                <button
-                  type="button"
-                  onClick={loadSessions}
-                  className="rounded border px-2 py-1 text-foreground"
-                >
-                  {t('common.retry')}
-                </button>
-              </div>
+              <LoadErrorState
+                compact
+                title={t('monitoring.loadError')}
+                onRetry={loadSessions}
+              />
             ) : sessions.length === 0 ? (
               <div className="text-center text-muted-foreground py-12 text-sm">
                 {t('bots.sessionMonitor.noSessions')}
@@ -850,7 +900,7 @@ const BotSessionMonitor = forwardRef<
                       selectedSession?.user_id ||
                       selectedSessionId.slice(0, 20)}
                   </div>
-                  <div className="flex items-center gap-1.5 text-xs text-muted-foreground mt-0.5">
+                  <div className="flex flex-wrap items-center gap-1.5 text-xs text-muted-foreground mt-0.5">
                     {parseSessionType(selectedSessionId) && (
                       <span>{parseSessionType(selectedSessionId)}</span>
                     )}
@@ -913,6 +963,18 @@ const BotSessionMonitor = forwardRef<
                     )}
                   </div>
                 </div>
+                {canReset && (
+                  <Button
+                    variant="outline"
+                    size="sm"
+                    className="h-7 shrink-0 gap-1.5"
+                    disabled={resetting}
+                    onClick={() => setResetTarget(selectedSessionId)}
+                  >
+                    <RotateCcw className="size-3.5" />
+                    {t('bots.sessionMonitor.resetContext')}
+                  </Button>
+                )}
               </div>
 
               {/* Messages Area */}
@@ -922,45 +984,25 @@ const BotSessionMonitor = forwardRef<
               >
                 <div className="space-y-4">
                   {analysisError && !loadingMessages && (
-                    <div
-                      role="alert"
-                      className="text-sm text-destructive space-y-2"
-                    >
-                      <p>
-                        {t('monitoring.toolCalls.title')}:{' '}
-                        {t('monitoring.loadError')}
-                      </p>
-                      <button
-                        type="button"
-                        onClick={() =>
-                          loadMessages(selectedSessionId, messagePage)
-                        }
-                        className="rounded border px-2 py-1 text-foreground"
-                      >
-                        {t('common.retry')}
-                      </button>
-                    </div>
+                    <LoadErrorState
+                      compact
+                      title={`${t('monitoring.toolCalls.title')}: ${t('monitoring.loadError')}`}
+                      onRetry={() =>
+                        loadMessages(selectedSessionId, messagePage)
+                      }
+                    />
                   )}
                   {loadingMessages ? (
                     <div className="text-center text-muted-foreground py-12 text-sm">
                       {t('bots.sessionMonitor.loading')}
                     </div>
                   ) : messageError ? (
-                    <div
-                      role="alert"
-                      className="text-sm text-destructive space-y-2"
-                    >
-                      <p>{t('monitoring.loadError')}</p>
-                      <button
-                        type="button"
-                        onClick={() =>
-                          loadMessages(selectedSessionId, messagePage)
-                        }
-                        className="rounded border px-2 py-1 text-foreground"
-                      >
-                        {t('common.retry')}
-                      </button>
-                    </div>
+                    <LoadErrorState
+                      title={t('monitoring.loadError')}
+                      onRetry={() =>
+                        loadMessages(selectedSessionId, messagePage)
+                      }
+                    />
                   ) : timelineItems.length === 0 ? (
                     <div className="text-center text-muted-foreground py-12 text-sm">
                       {t('bots.sessionMonitor.noMessages')}
@@ -1181,38 +1223,73 @@ const BotSessionMonitor = forwardRef<
                   )}
                 </div>
               </ScrollArea>
-              <div className="h-9 border-t px-3 flex items-center justify-center gap-3 shrink-0 text-xs">
-                <button
-                  type="button"
-                  disabled={messagePage === 0 || loadingMessages}
-                  onClick={() =>
-                    setMessagePage((page) => Math.max(0, page - 1))
-                  }
-                  className="inline-flex items-center gap-1 rounded px-2 py-1 hover:bg-accent disabled:opacity-40"
-                >
-                  <ChevronLeft className="size-3.5" />
-                  {t('common.previous')}
-                </button>
-                <span className="tabular-nums text-muted-foreground">
-                  {messagePage + 1} / {messagePageCount} · {messageTotal}
-                </span>
-                <button
-                  type="button"
-                  disabled={
-                    messagePage + 1 >= messagePageCount || loadingMessages
-                  }
-                  onClick={() => setMessagePage((page) => page + 1)}
-                  className="inline-flex items-center gap-1 rounded px-2 py-1 hover:bg-accent disabled:opacity-40"
-                >
-                  {t('common.next')}
-                  <ChevronRight className="size-3.5" />
-                </button>
-              </div>
+              {messagePageCount > 1 && (
+                <div className="h-9 border-t px-3 flex items-center justify-center gap-3 shrink-0 text-xs">
+                  <button
+                    type="button"
+                    disabled={messagePage === 0 || loadingMessages}
+                    onClick={() =>
+                      setMessagePage((page) => Math.max(0, page - 1))
+                    }
+                    className="inline-flex items-center gap-1 rounded px-2 py-1 hover:bg-accent disabled:opacity-40"
+                  >
+                    <ChevronLeft className="size-3.5" />
+                    {t('operationTrace.previousPage')}
+                  </button>
+                  <span className="tabular-nums text-muted-foreground">
+                    {messagePage + 1} / {messagePageCount} · {messageTotal}
+                  </span>
+                  <button
+                    type="button"
+                    disabled={
+                      messagePage + 1 >= messagePageCount || loadingMessages
+                    }
+                    onClick={() => setMessagePage((page) => page + 1)}
+                    className="inline-flex items-center gap-1 rounded px-2 py-1 hover:bg-accent disabled:opacity-40"
+                  >
+                    {t('operationTrace.nextPage')}
+                    <ChevronRight className="size-3.5" />
+                  </button>
+                </div>
+              )}
             </>
           )}
         </div>
       </div>
 
+      <Dialog
+        open={resetTarget !== null}
+        onOpenChange={(open) => {
+          if (!open && !resetting) setResetTarget(null);
+        }}
+      >
+        <DialogContent>
+          <DialogHeader>
+            <DialogTitle>{t('bots.sessionMonitor.resetContext')}</DialogTitle>
+            <DialogDescription>
+              {t('bots.sessionMonitor.resetDescription')}
+            </DialogDescription>
+          </DialogHeader>
+          <DialogFooter>
+            <Button
+              variant="outline"
+              disabled={resetting}
+              onClick={() => setResetTarget(null)}
+            >
+              {t('common.cancel')}
+            </Button>
+            <Button disabled={resetting} onClick={resetContext}>
+              <RotateCcw
+                className={cn(
+                  'size-4',
+                  resetting && 'animate-spin motion-reduce:animate-none',
+                )}
+              />
+              {t('bots.sessionMonitor.resetContext')}
+            </Button>
+          </DialogFooter>
+        </DialogContent>
+      </Dialog>
       <BotAdminsDialog
         botId={botId}
         open={adminsDialogOpen}
diff --git a/web/src/app/home/components/AssistantToolResult.tsx b/web/src/app/home/components/AssistantToolResult.tsx
index 4eabe60c1..c5ee065bf 100644
--- a/web/src/app/home/components/AssistantToolResult.tsx
+++ b/web/src/app/home/components/AssistantToolResult.tsx
@@ -4,6 +4,7 @@ import {
   ChevronDown,
   CircleAlert,
   MinusCircle,
+  LoaderCircle,
 } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
 
@@ -16,11 +17,11 @@ export type AssistantTool = {
 export default function AssistantToolResult({
   tool,
   content,
-  defaultCollapsed = false,
+  defaultCollapsed = true,
 }: {
   tool?: AssistantTool;
   content: string;
-  /** Turn finished: collapse to one line until the user expands it again. */
+  /** Keep tool steps quiet until the user requests details. */
   defaultCollapsed?: boolean;
 }) {
   const { t } = useTranslation();
@@ -36,11 +37,17 @@ export default function AssistantToolResult({
     result && typeof result === 'object' && !Array.isArray(result)
       ? (result as Record<string, unknown>)
       : {};
+  const running = data.status === 'running';
   const failed = !!data.error;
   const denied = data.status === 'denied';
   const partial = !!data.truncated;
-  const Icon =
-    failed || partial ? CircleAlert : denied ? MinusCircle : CheckCircle2;
+  const Icon = running
+    ? LoaderCircle
+    : failed || partial
+      ? CircleAlert
+      : denied
+        ? MinusCircle
+        : CheckCircle2;
   const items = Array.isArray(result)
     ? result
     : Array.isArray(data.items)
@@ -56,13 +63,15 @@ export default function AssistantToolResult({
       : t(`assistant.operations.${tool?.name}`, {
           defaultValue: t('assistant.toolResult'),
         });
-  const status = failed
-    ? 'failed'
-    : denied
-      ? 'denied'
-      : partial
-        ? 'partial'
-        : 'completed';
+  const status = running
+    ? 'toolRunning'
+    : failed
+      ? 'failed'
+      : denied
+        ? 'denied'
+        : partial
+          ? 'partial'
+          : 'completed';
   const url =
     typeof data.url === 'string' &&
     /^\/home\/(pipelines|knowledge)\?id=[\w-]+$/.test(data.url)
@@ -79,7 +88,7 @@ export default function AssistantToolResult({
     return (
       <button
         type="button"
-        className="flex w-full items-center gap-2 rounded-lg border bg-background px-3 py-1.5 text-left text-xs hover:bg-muted"
+        className="flex w-full items-center gap-1.5 rounded-md px-0.5 py-1 text-left text-[11px] leading-4 text-muted-foreground hover:text-foreground focus-visible:outline-2 focus-visible:outline-primary"
         aria-expanded={false}
         onClick={() => {
           setManual(true);
@@ -87,25 +96,25 @@ export default function AssistantToolResult({
         }}
       >
         <Icon
-          className={`size-3.5 shrink-0 ${failed ? 'text-destructive' : 'text-muted-foreground'}`}
+          className={`size-3 shrink-0 ${running ? 'animate-spin motion-reduce:animate-none' : ''} ${failed ? 'text-destructive' : 'text-muted-foreground'}`}
         />
-        <span className="truncate font-medium">{label}</span>
+        <span className="min-w-0 truncate">{label}</span>
         <span className="ml-auto shrink-0 text-muted-foreground">
           {tool && t(`assistant.${status}`)}
         </span>
-        <ChevronDown className="size-3.5 shrink-0 text-muted-foreground" />
+        <ChevronDown className="size-3 shrink-0 text-muted-foreground" />
       </button>
     );
   }
 
   return (
-    <section className="space-y-2 rounded-xl border bg-background p-3 text-sm">
-      <div className="flex items-center gap-2">
+    <section className="min-w-0 space-y-1.5 px-0.5 py-1 text-[11px] leading-4 text-muted-foreground">
+      <div className="flex items-center gap-1.5">
         <Icon
-          className={`size-4 shrink-0 ${failed ? 'text-destructive' : 'text-muted-foreground'}`}
+          className={`size-3 shrink-0 ${running ? 'animate-spin motion-reduce:animate-none' : ''} ${failed ? 'text-destructive' : 'text-muted-foreground'}`}
         />
-        <span className="font-medium">{label}</span>
-        <span className="ml-auto text-xs text-muted-foreground">
+        <span className="min-w-0 flex-1 break-words">{label}</span>
+        <span className="ml-auto shrink-0 text-[11px] text-muted-foreground">
           {tool && t(`assistant.${status}`)}
         </span>
         <button
@@ -118,10 +127,12 @@ export default function AssistantToolResult({
             setCollapsed(true);
           }}
         >
-          <ChevronDown className="size-3.5 rotate-180" />
+          <ChevronDown className="size-3 rotate-180" />
         </button>
       </div>
-      {failed ? (
+      {running ? (
+        <p>{t('assistant.toolRunning')}</p>
+      ) : failed ? (
         <p className="text-destructive">{t('assistant.operationFailed')}</p>
       ) : denied ? (
         <p className="text-muted-foreground">
@@ -136,7 +147,7 @@ export default function AssistantToolResult({
           )}
           {name && <p className="break-words">{name}</p>}
           {items && (
-            <ul className="space-y-1 text-xs text-muted-foreground">
+            <ul className="space-y-0.5 text-[11px] text-muted-foreground">
               {items.slice(0, 6).map((item: unknown, index: number) => {
                 const entry =
                   item && typeof item === 'object'
@@ -168,9 +179,9 @@ export default function AssistantToolResult({
           )}
         </>
       )}
-      <details className="text-xs text-muted-foreground">
+      <details className="text-[11px] text-muted-foreground">
         <summary className="cursor-pointer">{t('assistant.details')}</summary>
-        <pre className="mt-2 max-h-48 overflow-auto whitespace-pre-wrap break-all">
+        <pre className="mt-1.5 max-h-32 overflow-auto whitespace-pre-wrap break-all">
           {tool ? JSON.stringify(tool.result, null, 2) : content}
         </pre>
       </details>
diff --git a/web/src/app/home/components/WorkspaceAssistant.tsx b/web/src/app/home/components/WorkspaceAssistant.tsx
index 58834a7b0..8457e6992 100644
--- a/web/src/app/home/components/WorkspaceAssistant.tsx
+++ b/web/src/app/home/components/WorkspaceAssistant.tsx
@@ -1,43 +1,43 @@
+import { useEffect, useRef, useState, useId, type ReactNode } from 'react';
+import { createPortal } from 'react-dom';
+import { Plus, Send, ChevronDown, LoaderCircle, Square } from 'lucide-react';
 import {
-  useEffect,
-  useRef,
-  useState,
-  type MouseEvent as ReactMouseEvent,
-  type PointerEvent as ReactPointerEvent,
-} from 'react';
-import { MessageCircle, Plus, Send, X, LoaderCircle } from 'lucide-react';
+  RiSparkling2Line,
+  RiArrowUpSLine,
+  RiCompass3Line,
+  RiMagicLine,
+  RiArrowRightUpLine,
+} from '@remixicon/react';
+import styles from './workspace-assistant.module.css';
 import { useTranslation } from 'react-i18next';
 import ReactMarkdown from 'react-markdown';
 import remarkGfm from 'remark-gfm';
-import { backendClient, useCurrentWorkspace, userInfo } from '@/app/infra/http';
+import rehypeHighlight from 'rehype-highlight';
+import '@/styles/github-markdown.css';
+import {
+  backendClient,
+  systemInfo,
+  useCurrentWorkspace,
+  userInfo,
+} from '@/app/infra/http';
+import { toast } from 'sonner';
+import { httpClient } from '@/app/infra/http/HttpClient';
+import { MODEL_CONFIGURATION_CHANGED } from '@/app/infra/http/modelConfigEvents';
+import SettingsDialog, {
+  SettingsSection,
+} from './settings-dialog/SettingsDialog';
 import { Button } from '@/components/ui/button';
+import {
+  Select,
+  SelectContent,
+  SelectItem,
+  SelectTrigger,
+  SelectValue,
+} from '@/components/ui/select';
 import DynamicFormItemComponent from './dynamic-form/DynamicFormItemComponent';
 import { DynamicFormItemType } from '@/app/infra/entities/form/dynamic';
 import AssistantToolResult, { AssistantTool } from './AssistantToolResult';
-import {
-  Popover,
-  PopoverContent,
-  PopoverTrigger,
-} from '@/components/ui/popover';
-import {
-  ASSISTANT_RAIL_WIDTH,
-  clampAssistantPosition as clampInViewport,
-  resolveAssistantEdge,
-  shouldCollapseRail,
-  shouldExpandRail,
-  type AssistantDragPosition,
-  type AssistantEdge,
-} from './assistant-dock';
-
-const ASSISTANT_LONG_PRESS_MS = 260;
-
-function viewport(): { width: number; height: number } {
-  return { width: window.innerWidth, height: window.innerHeight };
-}
-
-function clampAssistantPosition(x: number, y: number): AssistantDragPosition {
-  return clampInViewport(x, y, viewport());
-}
+import { LANGBOT_MODELS_PROVIDER_REQUESTER } from './models-dialog/types';
 
 type Conversation = {
   uuid: string;
@@ -48,6 +48,12 @@ type Conversation = {
   error: string | null;
   model_name: string | null;
   model_uuid: string | null;
+  progress?: {
+    text?: string;
+    phase?: 'thinking' | 'tool';
+    round?: number;
+    tool?: Omit<AssistantTool, 'result'>;
+  };
 };
 
 export default function WorkspaceAssistant() {
@@ -59,348 +65,437 @@ export default function WorkspaceAssistant() {
     return null;
   const identity = `${workspace.workspace.uuid}:${userInfo.account_uuid}`;
   return (
-    <AssistantPanel
+    <AssistantEntry
       key={identity}
       storageKey={`langbot-assistant:${identity}`}
     />
   );
 }
 
-function AssistantPanel({ storageKey }: { storageKey: string }) {
+function AssistantEntry({ storageKey }: { storageKey: string }) {
   const { t } = useTranslation();
-  const [open, setOpen] = useState(false);
+  const [available, setAvailable] = useState<boolean | null>(null);
+  const [localModel, setLocalModel] = useState<
+    { uuid: string; name: string } | undefined
+  >();
+  const [failed, setFailed] = useState(false);
+  const [settingsOpen, setSettingsOpen] = useState(false);
+  const [section, setSection] = useState<SettingsSection>('models');
+  const [loginBusy, setLoginBusy] = useState(false);
+  const [retry, setRetry] = useState(0);
+
+  useEffect(() => {
+    let active = true;
+    let latestCheck = 0;
+    async function check() {
+      const currentCheck = ++latestCheck;
+      try {
+        const { models } = await backendClient.getProviderLLMModels();
+        if (active && currentCheck === latestCheck) {
+          const usableModels = models.filter(
+            (model) =>
+              !!model.uuid?.trim() &&
+              model.abilities?.includes('func_call') &&
+              (model.provider?.requester !==
+                LANGBOT_MODELS_PROVIDER_REQUESTER ||
+                (userInfo?.account_type === 'space' &&
+                  !systemInfo.disable_models_service)),
+          );
+          setAvailable(usableModels.length > 0);
+          const firstLocalModel =
+            userInfo?.account_type !== 'space' ? usableModels[0] : undefined;
+          setLocalModel(
+            firstLocalModel
+              ? { uuid: firstLocalModel.uuid, name: firstLocalModel.name }
+              : undefined,
+          );
+          setFailed(false);
+        }
+      } catch {
+        if (active && currentCheck === latestCheck) setFailed(true);
+      }
+    }
+    if (!settingsOpen) void check();
+    window.addEventListener('focus', check);
+    window.addEventListener(MODEL_CONFIGURATION_CHANGED, check);
+    return () => {
+      active = false;
+      window.removeEventListener('focus', check);
+      window.removeEventListener(MODEL_CONFIGURATION_CHANGED, check);
+    };
+  }, [settingsOpen, retry]);
+
+  async function login() {
+    setLoginBusy(true);
+    try {
+      const response = await httpClient.getSpaceBindAuthorizeUrl(
+        `${window.location.origin}/auth/space/callback?mode=bind`,
+      );
+      window.location.href = response.authorize_url;
+    } catch {
+      toast.error(t('common.spaceLoginFailed'));
+      setLoginBusy(false);
+    }
+  }
+
+  return (
+    <>
+      <AssistantSessions
+        storageKey={storageKey}
+        canUseModels={available === true && !failed}
+        initialModel={localModel}
+        setup={
+          <div className="mx-3 mb-[max(0.75rem,env(safe-area-inset-bottom))] shrink-0 border-t pt-3">
+            <p
+              className="text-[13px] leading-relaxed text-muted-foreground"
+              role={failed ? 'alert' : undefined}
+            >
+              {t(
+                failed
+                  ? 'assistant.providerCheckFailed'
+                  : available === null
+                    ? 'assistant.preparing'
+                    : 'assistant.setupRequired',
+              )}
+            </p>
+            <div className="mt-2 flex flex-wrap gap-2">
+              {failed ? (
+                <Button
+                  size="sm"
+                  variant="outline"
+                  onClick={() => setRetry((value) => value + 1)}
+                >
+                  {t('assistant.retrySetup')}
+                </Button>
+              ) : available === false ? (
+                <>
+                  <Button
+                    size="sm"
+                    disabled={loginBusy}
+                    onClick={() => void login()}
+                  >
+                    {t('assistant.loginAccount')}
+                  </Button>
+                  <Button
+                    size="sm"
+                    variant="outline"
+                    onClick={() => {
+                      setSection('models');
+                      setSettingsOpen(true);
+                    }}
+                  >
+                    {t('assistant.configureModels')}
+                  </Button>
+                </>
+              ) : null}
+            </div>
+          </div>
+        }
+      />
+      <SettingsDialog
+        open={settingsOpen}
+        onOpenChange={setSettingsOpen}
+        section={section}
+        onSectionChange={setSection}
+      />
+    </>
+  );
+}
+
+type AssistantPanelProps = {
+  storageKey: string;
+  canUseModels: boolean;
+  initialModel?: { uuid: string; name: string };
+  setup: ReactNode;
+};
+
+function AssistantSessions(props: AssistantPanelProps) {
+  const { storageKey } = props;
+  const [version, setVersion] = useState(0);
+  function select(id: string | null) {
+    if (id) localStorage.setItem(storageKey, id);
+    else localStorage.removeItem(storageKey);
+    setVersion((value) => value + 1);
+  }
+  return <AssistantPanel key={version} {...props} onSelect={select} />;
+}
+
+const ASSISTANT_HEIGHT_STORAGE_KEY = 'langbot-assistant-panel-height';
+const ASSISTANT_POSITION_STORAGE_KEY = 'langbot-assistant-panel-position';
+const ASSISTANT_OPEN_STORAGE_KEY = 'langbot-assistant-panel-open';
+
+function AssistantPanel({
+  storageKey,
+  canUseModels,
+  initialModel,
+  setup,
+  onSelect,
+}: AssistantPanelProps & {
+  onSelect: (id: string | null) => void;
+}) {
+  const { t } = useTranslation();
+  const initialModelUuid = initialModel?.uuid;
+  const initialModelName = initialModel?.name;
+  const [open, setOpen] = useState(() => {
+    try {
+      return localStorage.getItem(ASSISTANT_OPEN_STORAGE_KEY) !== 'false';
+    } catch {
+      return true;
+    }
+  });
+  useEffect(() => {
+    try {
+      localStorage.setItem(ASSISTANT_OPEN_STORAGE_KEY, String(open));
+    } catch {
+      // The panel remains usable when browser storage is unavailable.
+    }
+  }, [open]);
+  const [closing, setClosing] = useState(false);
+  const [horizontalPosition, setHorizontalPosition] = useState(() => {
+    try {
+      const saved = Number(
+        localStorage.getItem(ASSISTANT_POSITION_STORAGE_KEY),
+      );
+      return Number.isFinite(saved) ? Math.max(-1, Math.min(1, saved)) : 0;
+    } catch {
+      return 0;
+    }
+  });
+  const [viewportWidth, setViewportWidth] = useState(window.innerWidth);
+  useEffect(() => {
+    const updateWidth = () => setViewportWidth(window.innerWidth);
+    window.addEventListener('resize', updateWidth);
+    return () => window.removeEventListener('resize', updateWidth);
+  }, []);
+  useEffect(() => {
+    try {
+      localStorage.setItem(
+        ASSISTANT_POSITION_STORAGE_KEY,
+        String(horizontalPosition),
+      );
+    } catch {
+      // Positioning remains available when storage is blocked.
+    }
+  }, [horizontalPosition]);
+  // Store a relative position so it follows screen changes while keeping a
+  // 12px margin. The collapsed dock remains aligned with the expanded panel.
+  const horizontalRange = Math.max(0, (viewportWidth - 480 - 24) / 2);
+  const moveStart = useRef<{ x: number; offset: number } | null>(null);
+  const dockWasDragged = useRef(false);
+  const [panelHeight, setPanelHeight] = useState<number | undefined>(() => {
+    try {
+      const saved = Number(localStorage.getItem(ASSISTANT_HEIGHT_STORAGE_KEY));
+      return Number.isFinite(saved) && saved > 0
+        ? Math.max(280, saved)
+        : undefined;
+    } catch {
+      return undefined;
+    }
+  });
+  useEffect(() => {
+    if (panelHeight === undefined) return;
+    try {
+      localStorage.setItem(ASSISTANT_HEIGHT_STORAGE_KEY, String(panelHeight));
+    } catch {
+      // Resizing still works when browser storage is unavailable.
+    }
+  }, [panelHeight]);
+  const panelRef = useRef<HTMLElement>(null);
+  const resizeStart = useRef<{ y: number; height: number } | null>(null);
+  const resizeHeight = (height: number) => {
+    const maximum = Math.max(160, window.innerHeight - 48);
+    setPanelHeight(Math.min(maximum, Math.max(Math.min(280, maximum), height)));
+  };
+  const closeTimer = useRef<number | null>(null);
   const [conversation, setConversation] = useState<Conversation | null>(null);
   const [text, setText] = useState('');
   const [sending, setBusy] = useState(false);
   const [loading, setLoading] = useState(false);
-  const busy = sending || loading;
+  const busy = sending || loading || conversation?.status === 'running';
+  const [stopping, setStopping] = useState(false);
+  const [sessions, setSessions] = useState<
+    { uuid: string; title: string; status: Conversation['status'] }[]
+  >([]);
   const [error, setError] = useState(false);
   const [modelUuid, setModelUuid] = useState('');
+  const [modelName, setModelName] = useState<string | undefined>();
+  const manualModel = useRef(false);
+  const [modelError, setModelError] = useState(false);
+  const [liveText, setLiveText] = useState('');
+  const [liveTool, setLiveTool] = useState<AssistantTool | null>(null);
+  const [phase, setPhase] = useState<'thinking' | 'tool' | null>(null);
+  const [round, setRound] = useState(1);
   const [pendingText, setPendingText] = useState<string | null>(null);
   const controller = useRef(new AbortController());
-  const end = useRef<HTMLDivElement>(null);
+  const messageList = useRef<HTMLDivElement>(null);
+  const followLatest = useRef(true);
+  const scrollPosition = useRef(0);
 
-  const [dragPosition, setDragPosition] =
-    useState<AssistantDragPosition | null>(null);
-  const [dragging, setDragging] = useState(false);
-  const [dockedEdge, setDockedEdge] = useState<AssistantEdge>(null);
-  // The rail collapses only while the pointer is away. Hovering any part of the
-  // control restores the full button, and the two states never race because
-  // every transition is derived from the same `dockedEdge` snapshot.
-  const [railExpanded, setRailExpanded] = useState(false);
-  const pointerOver = useRef(false);
-  const containerRef = useRef<HTMLDivElement>(null);
-  const hoverLocked = useRef(false);
-  const dragState = useRef<{
-    pointerId: number;
-    startX: number;
-    startY: number;
-    originX: number;
-    originY: number;
-    active: boolean;
-  } | null>(null);
-  const longPressTimer = useRef<number | null>(null);
-  const suppressClick = useRef(false);
-  const buttonRef = useRef<HTMLButtonElement>(null);
+  const panelId = useId();
+  const expandButton = useRef<HTMLButtonElement>(null);
+  const input = useRef<HTMLTextAreaElement>(null);
 
-  function clearLongPressTimer() {
-    if (longPressTimer.current !== null) {
-      window.clearTimeout(longPressTimer.current);
-      longPressTimer.current = null;
-    }
-  }
+  useEffect(
+    () => () => {
+      if (closeTimer.current !== null) window.clearTimeout(closeTimer.current);
+    },
+    [],
+  );
 
-  useEffect(() => clearLongPressTimer, []);
-
-  /*
-   * Seed the resting position from the rendered default (bottom-right) so the
-   * very first visit already docks and collapses. Without this the button would
-   * only ever dock after a manual drag, which reads as "collapse is broken".
-   */
-  useEffect(() => {
-    if (window.localStorage.getItem(`${storageKey}:button-position`)) return;
-    const rect = buttonRef.current?.getBoundingClientRect();
-    if (!rect) return;
-    const position = clampAssistantPosition(rect.left, rect.top);
-    setDragPosition(position);
-    const edge = resolveAssistantEdge(position.x, viewport());
-    setDockedEdge(edge);
-    if (edge) {
-      try {
-        window.localStorage.setItem(`${storageKey}:button-docked-edge`, edge);
-      } catch {
-        // Persisting the dock is best-effort only.
-      }
-    }
-  }, [storageKey]);
-
-  useEffect(() => {
-    const stored = window.localStorage.getItem(`${storageKey}:button-position`);
-    if (stored) {
-      try {
-        const parsed = JSON.parse(stored) as AssistantDragPosition;
-        if (typeof parsed?.x === 'number' && typeof parsed?.y === 'number') {
-          const position = clampAssistantPosition(parsed.x, parsed.y);
-          setDragPosition(position);
-          setDockedEdge(resolveAssistantEdge(position.x, viewport()));
-        }
-      } catch {
-        window.localStorage.removeItem(`${storageKey}:button-position`);
-      }
-    }
-    const storedEdge = window.localStorage.getItem(
-      `${storageKey}:button-docked-edge`,
-    );
-    if (storedEdge === 'left' || storedEdge === 'right') {
-      setDockedEdge((current) => current ?? storedEdge);
-    }
-  }, [storageKey]);
-
-  useEffect(() => {
-    const onResize = () => {
-      setDragPosition((prev) => {
-        if (!prev) return prev;
-        const next = clampAssistantPosition(prev.x, prev.y);
-        setDockedEdge(resolveAssistantEdge(next.x, viewport()));
-        return next;
-      });
+  function collapse() {
+    if (closeTimer.current !== null) return;
+    const finish = () => {
+      closeTimer.current = null;
+      setOpen(false);
+      setClosing(false);
+      window.requestAnimationFrame(() => expandButton.current?.focus());
     };
-    window.addEventListener('resize', onResize);
-    return () => window.removeEventListener('resize', onResize);
-  }, []);
-
-  function persistDragPosition(position: AssistantDragPosition) {
-    try {
-      window.localStorage.setItem(
-        `${storageKey}:button-position`,
-        JSON.stringify(position),
-      );
-    } catch {
-      // Ignore storage failures (private mode or quota); dragging still works.
+    if (window.matchMedia('(prefers-reduced-motion: reduce)').matches) {
+      finish();
+      return;
     }
+    setClosing(true);
+    closeTimer.current = window.setTimeout(finish, 180);
   }
 
-  function applyRestingPosition(position: AssistantDragPosition) {
-    const edge = resolveAssistantEdge(position.x, viewport());
-    setDragPosition(position);
-    setDockedEdge(edge);
-    // A fresh dock always collapses; the rail re-expands on the next hover.
-    setRailExpanded(false);
-    // Hold the collapse until the pointer leaves, otherwise the still-hovering
-    // cursor would fight the new state.
-    if (edge) lockHoverUntilPointerExit();
-    try {
-      if (edge) {
-        window.localStorage.setItem(`${storageKey}:button-docked-edge`, edge);
-      } else {
-        window.localStorage.removeItem(`${storageKey}:button-docked-edge`);
-      }
-    } catch {
-      // Persisting the dock is best-effort only.
-    }
+  function expand() {
+    setOpen(true);
+    window.requestAnimationFrame(() => input.current?.focus());
   }
 
-  function endDrag(commit: boolean, clientX?: number, clientY?: number) {
-    const state = dragState.current;
-    clearLongPressTimer();
-    dragState.current = null;
-    if (!state?.active) return;
-    setDragging(false);
-    const next = clampAssistantPosition(
-      state.originX + ((clientX ?? state.startX) - state.startX),
-      state.originY + ((clientY ?? state.startY) - state.startY),
-    );
-    if (commit) {
-      applyRestingPosition(next);
-      persistDragPosition(next);
-    } else {
-      setDragPosition(next);
-    }
-  }
-
-  /*
-   * Rail hover recovery. The subtle race: a drag usually ends with the pointer
-   * still sitting on the button, so the browser fires no new `pointerenter`
-   * once the button collapses. Re-expanding on `pointermove` would therefore
-   * undo the collapse immediately.
-   *
-   * Instead the drop "locks" hover until the pointer physically leaves the
-   * control. A window-level move listener watches for that exit (the element
-   * can shift under a stationary cursor, so `pointerleave` alone is not
-   * reliable) and clears the lock; only then does hovering reveal the button.
-   */
-  function handlePointerEnter() {
-    pointerOver.current = true;
-    if (hoverLocked.current) return;
-    if (shouldExpandRail({ dockedEdge, railExpanded, dragging }))
-      setRailExpanded(true);
-  }
-
-  function handlePointerLeave() {
-    pointerOver.current = false;
-    hoverLocked.current = false;
-    // Never collapse mid-drag; the drop handler owns the final state.
-    if (dragState.current || dragging) return;
-    if (shouldCollapseRail({ dockedEdge, railExpanded, dragging }))
-      setRailExpanded(false);
-  }
-
-  /*
-   * A click or an aborted swipe never arms the long press, so no drop handler
-   * runs. Re-sync the rail from the real pointer position on release, otherwise
-   * a button expanded by hover would stay expanded with the cursor gone.
-   */
-  function settleRailAfterRelease() {
-    if (pointerOver.current || dragState.current || dragging) return;
-    if (shouldCollapseRail({ dockedEdge, railExpanded, dragging }))
-      setRailExpanded(false);
-  }
-
-  function lockHoverUntilPointerExit() {
-    hoverLocked.current = true;
-    const releaseOnExit = (moveEvent: PointerEvent) => {
-      const rect = containerRef.current?.getBoundingClientRect();
-      if (!rect) return;
-      const outside =
-        moveEvent.clientX < rect.left ||
-        moveEvent.clientX > rect.right ||
-        moveEvent.clientY < rect.top ||
-        moveEvent.clientY > rect.bottom;
-      if (!outside) return;
-      hoverLocked.current = false;
-      window.removeEventListener('pointermove', releaseOnExit);
-    };
-    window.addEventListener('pointermove', releaseOnExit);
-  }
-
-  function onButtonPointerDown(event: ReactPointerEvent<HTMLButtonElement>) {
-    if (event.button !== 0) return;
-    suppressClick.current = false;
-    const rect = buttonRef.current?.getBoundingClientRect();
-    if (!rect) return;
-    const state = {
-      pointerId: event.pointerId,
-      startX: event.clientX,
-      startY: event.clientY,
-      originX: rect.left,
-      originY: rect.top,
-      active: false,
-    };
-    dragState.current = state;
-    clearLongPressTimer();
-
-    const onWindowMove = (moveEvent: PointerEvent) => {
-      if (moveEvent.pointerId !== state.pointerId) return;
-      const deltaX = moveEvent.clientX - state.startX;
-      const deltaY = moveEvent.clientY - state.startY;
-
-      if (!state.active) {
-        // Cancel the long-press when the user is clearly scrolling or swiping.
-        if (Math.hypot(deltaX, deltaY) > 8) clearLongPressTimer();
-        return;
-      }
-
-      moveEvent.preventDefault();
-      setDragPosition(
-        clampAssistantPosition(state.originX + deltaX, state.originY + deltaY),
-      );
-    };
-
-    const onWindowUp = (upEvent: PointerEvent) => {
-      if (upEvent.pointerId !== state.pointerId) return;
-      window.removeEventListener('pointermove', onWindowMove);
-      window.removeEventListener('pointerup', onWindowUp);
-      window.removeEventListener('pointercancel', onWindowUp);
-      endDrag(true, upEvent.clientX, upEvent.clientY);
-    };
-
-    window.addEventListener('pointermove', onWindowMove, { passive: false });
-    window.addEventListener('pointerup', onWindowUp);
-    window.addEventListener('pointercancel', onWindowUp);
-
-    // Release outside a drag still needs to reconcile the rail.
-    window.addEventListener('pointerup', settleRailAfterRelease, {
-      once: true,
-    });
-
-    longPressTimer.current = window.setTimeout(() => {
-      if (dragState.current !== state) return;
-      state.active = true;
-      setDragging(true);
-      suppressClick.current = true;
-      setDragPosition(clampAssistantPosition(state.originX, state.originY));
-    }, ASSISTANT_LONG_PRESS_MS);
-  }
-
-  function onButtonClick(event: ReactMouseEvent<HTMLButtonElement>) {
-    // After a drag the trailing click must not toggle the panel. Radix's
-    // trigger skips its own toggle when the event default is prevented.
-    if (suppressClick.current) {
-      suppressClick.current = false;
-      event.preventDefault();
-      event.stopPropagation();
-    }
-  }
-
-  useEffect(() => {
-    if (conversation?.model_uuid) setModelUuid(conversation.model_uuid);
-  }, [conversation?.model_uuid]);
-
   useEffect(() => {
     const abort = new AbortController();
     controller.current = abort;
+    setLoading(true);
+    async function initialize() {
+      try {
+        const id = localStorage.getItem(storageKey);
+        let saved: Conversation | null = null;
+        if (id) {
+          saved = await backendClient.request<Conversation>({
+            method: 'GET',
+            url: `/api/v1/assistant/conversations/${encodeURIComponent(id)}`,
+            signal: abort.signal,
+          });
+          if (abort.signal.aborted) return;
+          setConversation(saved);
+        }
+        if (!canUseModels) return;
+        const model = saved?.model_uuid
+          ? { uuid: saved.model_uuid, name: saved.model_name ?? undefined }
+          : initialModelUuid
+            ? { uuid: initialModelUuid, name: initialModelName }
+            : await backendClient.request<{ uuid: string; name: string }>({
+                method: 'GET',
+                url: '/api/v1/assistant/recommended-model',
+                signal: abort.signal,
+              });
+        if (!abort.signal.aborted && !manualModel.current) {
+          setModelUuid(model.uuid);
+          setModelName(model.name);
+        }
+      } catch {
+        if (!abort.signal.aborted) setModelError(true);
+      } finally {
+        if (!abort.signal.aborted) setLoading(false);
+      }
+    }
+    void initialize();
     return () => abort.abort();
-  }, []);
+  }, [storageKey, canUseModels, initialModelUuid, initialModelName]);
 
   useEffect(() => {
-    if (!open || busy || (conversation && conversation.status !== 'running'))
-      return;
-    const id = localStorage.getItem(storageKey);
-    if (!id) return;
     let active = true;
-    setLoading(true);
-    backendClient
-      .request<Conversation>({
-        method: 'GET',
-        url: `/api/v1/assistant/conversations/${encodeURIComponent(id)}`,
-        signal: controller.current.signal,
-      })
-      .then((value) => {
-        if (active) setConversation(value);
-      })
-      .catch(() => {
-        if (active) {
-          localStorage.removeItem(storageKey);
-          setError(true);
+    let timer: ReturnType<typeof setTimeout>;
+    async function refresh() {
+      try {
+        const result = await backendClient.request<{
+          conversations: typeof sessions;
+        }>({
+          method: 'GET',
+          url: '/api/v1/assistant/conversations',
+          signal: controller.current.signal,
+        });
+        if (active) setSessions(result.conversations ?? []);
+        if (!sending && conversation?.uuid) {
+          const latest = await backendClient.request<Conversation>({
+            method: 'GET',
+            url: `/api/v1/assistant/conversations/${conversation.uuid}`,
+            signal: controller.current.signal,
+          });
+          if (active) {
+            setConversation(latest);
+            setLiveText(latest.progress?.text ?? '');
+            setPhase(latest.progress?.phase ?? null);
+            setRound(latest.progress?.round ?? 1);
+            setLiveTool(
+              latest.progress?.tool
+                ? { ...latest.progress.tool, result: { status: 'running' } }
+                : null,
+            );
+          }
         }
-      })
-      .finally(() => {
-        if (active) setLoading(false);
-      });
+      } catch {
+        // Polling is read-only. A failed refresh never retries a turn.
+      } finally {
+        if (active) timer = setTimeout(refresh, 1500);
+      }
+    }
+    void refresh();
     return () => {
       active = false;
-      setLoading(false);
+      clearTimeout(timer);
     };
-    // Load only when opening; turn requests own subsequent state updates.
-    // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [open, storageKey]);
+  }, [conversation?.uuid, sending]);
+
+  async function stop() {
+    if (!conversation || stopping) return;
+    setStopping(true);
+    try {
+      const latest = await backendClient.request<Conversation>({
+        method: 'POST',
+        url: `/api/v1/assistant/conversations/${conversation.uuid}/stop`,
+        data: { revision: conversation.revision },
+        signal: controller.current.signal,
+      });
+      setConversation(latest);
+      setLiveText('');
+      setLiveTool(null);
+      setPhase(null);
+    } catch {
+      setError(true);
+    } finally {
+      setStopping(false);
+    }
+  }
+
+  // Restore the reading position when reopening; new conversations start at the bottom.
+  useEffect(() => {
+    if (open && messageList.current && !followLatest.current) {
+      messageList.current.scrollTop = scrollPosition.current;
+    }
+  }, [open]);
 
-  // `open` is a dependency so a freshly opened panel jumps to the newest turn
-  // instead of leaving the user at the oldest message. The rAF waits for the
-  // popover to lay out before measuring the sentinel.
   useEffect(() => {
     if (!open) return;
     const frame = window.requestAnimationFrame(() => {
-      end.current?.scrollIntoView({ behavior: 'auto', block: 'nearest' });
+      const list = messageList.current;
+      if (list && followLatest.current) {
+        list.scrollTop = list.scrollHeight;
+        scrollPosition.current = list.scrollTop;
+      }
     });
     return () => window.cancelAnimationFrame(frame);
-  }, [open, conversation, busy, pendingText]);
+  }, [open, conversation, busy, pendingText, liveText, liveTool, panelHeight]);
 
   async function submit(approved?: boolean) {
     if (
+      !canUseModels ||
       busy ||
       (approved === undefined &&
-        (!text.trim() || (conversation && conversation.status !== 'ready')))
+        (!text.trim() ||
+          !modelUuid ||
+          (conversation && conversation.status !== 'ready')))
     )
       return;
     const sentText = approved === undefined ? text.trim() : null;
@@ -411,6 +506,8 @@ function AssistantPanel({ storageKey }: { storageKey: string }) {
     }
     setBusy(true);
     setError(false);
+    setRound(1);
+    setPhase('thinking');
     try {
       let current = conversation;
       if (!current) {
@@ -422,24 +519,48 @@ function AssistantPanel({ storageKey }: { storageKey: string }) {
         localStorage.setItem(storageKey, current.uuid);
         setConversation(current);
       }
-      const updated = await backendClient.request<Conversation>({
-        method: 'POST',
-        url: `/api/v1/assistant/conversations/${current.uuid}/turn`,
-        data: {
+      const updated = await backendClient.streamAssistant<Conversation>(
+        current.uuid,
+        {
           revision: current.revision,
           ...(approved === undefined
-            ? {
-                text: sentText,
-                ...(modelUuid ? { model_uuid: modelUuid } : {}),
-              }
+            ? { text: sentText!, model_uuid: modelUuid }
             : { approved }),
         },
-        timeout: 130000,
-        signal: controller.current.signal,
-      });
+        (event) => {
+          if (event.kind === 'snapshot') {
+            const snapshot = event.data as Conversation;
+            setConversation(snapshot);
+            setPendingText(null);
+            setLiveText('');
+            setLiveTool(null);
+            if (snapshot.model_uuid) {
+              setModelUuid(snapshot.model_uuid);
+              setModelName(snapshot.model_name ?? undefined);
+            }
+          } else if (event.kind === 'text') {
+            setLiveText((event.data as { text: string }).text);
+          } else if (event.kind === 'phase') {
+            const state = event.data as {
+              phase: 'thinking' | 'tool';
+              round?: number;
+              tool?: Omit<AssistantTool, 'result'>;
+            };
+            setPhase(state.phase);
+            if (state.round) setRound(state.round);
+            setLiveTool(
+              state.tool
+                ? { ...state.tool, result: { status: 'running' } }
+                : null,
+            );
+          }
+        },
+        controller.current.signal,
+      );
       setConversation(updated);
       if (sentText) setPendingText(null);
     } catch {
+      if (controller.current.signal.aborted) return;
       setError(true);
       // A lost response may already have executed a write. Refresh, never replay.
       const id = localStorage.getItem(storageKey);
@@ -466,291 +587,633 @@ function AssistantPanel({ storageKey }: { storageKey: string }) {
       }
     } finally {
       setBusy(false);
+      setLiveText('');
+      setLiveTool(null);
+      setPhase(null);
     }
   }
 
   function reset() {
-    localStorage.removeItem(storageKey);
-    setConversation(null);
-    setText('');
-    setError(false);
-    setPendingText(null);
+    onSelect(null);
   }
 
-  // Collapse only when docked, idle and not hovered. The container keeps its
-  // resting box, so the hidden button and the visible strip share one anchor
-  // and cannot drift apart; only the strip is painted while collapsed.
-  const railCollapsed = !!dockedEdge && !railExpanded && !dragging;
-  const inlinePosition = dragPosition
-    ? { left: dragPosition.x, top: dragPosition.y }
-    : undefined;
+  // A user message starts a turn; its tool steps and replies share one bubble.
+  const messageGroups: {
+    key: number;
+    role: 'user' | 'assistant';
+    messages: Conversation['messages'];
+  }[] = [];
+  const visibleMessages: Conversation['messages'] = [
+    ...(conversation?.messages ?? []),
+  ];
+  if (liveText) visibleMessages.push({ role: 'assistant', content: liveText });
+  if (liveTool)
+    visibleMessages.push({ role: 'tool', content: '', tool: liveTool });
+  for (const [index, message] of visibleMessages.entries()) {
+    const role = message.role === 'user' ? 'user' : 'assistant';
+    const previous = messageGroups.at(-1);
+    if (role === 'assistant' && previous?.role === 'assistant') {
+      previous.messages.push(message);
+    } else {
+      messageGroups.push({ key: index, role, messages: [message] });
+    }
+  }
 
-  return (
-    <div
-      ref={containerRef}
-      className={dragPosition ? 'fixed z-50' : 'fixed bottom-20 right-5 z-50'}
-      style={inlinePosition}
-      onPointerEnter={handlePointerEnter}
-      onPointerLeave={handlePointerLeave}
-    >
-      <Popover open={open} onOpenChange={setOpen}>
-        <PopoverTrigger asChild>
+  return createPortal(
+    <div className="pointer-events-none fixed inset-x-0 bottom-0 z-40 flex justify-center px-3">
+      <div
+        className="flex w-full max-w-[480px] justify-center"
+        style={{
+          transform: `translateX(${horizontalPosition * horizontalRange}px)`,
+        }}
+      >
+        {!open && (
           <Button
-            ref={buttonRef}
-            size="icon"
-            aria-label={t('assistant.title')}
-            aria-expanded={railCollapsed ? false : undefined}
-            className={`size-12 rounded-full shadow-lg ${
-              dragging
-                ? 'cursor-grabbing scale-105'
-                : railCollapsed
-                  ? 'cursor-pointer'
-                  : 'cursor-grab'
-            } transition-[transform,opacity] duration-200`}
-            style={{
-              touchAction: 'none',
-              opacity: railCollapsed ? 0 : 1,
+            ref={expandButton}
+            className={`pointer-events-auto touch-none select-none active:cursor-grabbing h-[calc(21px+env(safe-area-inset-bottom))] min-h-0 w-24 gap-2 rounded-b-none rounded-t-md border-0 bg-[#2288ee] px-3 py-0 pb-[env(safe-area-inset-bottom)] text-white shadow-lg hover:bg-[#2277e0] ${styles.dock}`}
+            title={t('assistant.expand')}
+            aria-label={t('assistant.expand')}
+            aria-expanded={false}
+            aria-controls={panelId}
+            onPointerDown={(event) => {
+              if (event.button !== 0 || !event.isPrimary) return;
+              dockWasDragged.current = false;
+              moveStart.current = {
+                x: event.clientX,
+                offset: horizontalPosition * horizontalRange,
+              };
+              event.currentTarget.setPointerCapture(event.pointerId);
+            }}
+            onPointerMove={(event) => {
+              if (!moveStart.current) return;
+              const delta = event.clientX - moveStart.current.x;
+              if (Math.abs(delta) >= 5) dockWasDragged.current = true;
+              if (!dockWasDragged.current || horizontalRange === 0) return;
+              setHorizontalPosition(
+                Math.max(
+                  -1,
+                  Math.min(
+                    1,
+                    (moveStart.current.offset + delta) / horizontalRange,
+                  ),
+                ),
+              );
+            }}
+            onPointerUp={(event) => {
+              moveStart.current = null;
+              if (event.currentTarget.hasPointerCapture(event.pointerId))
+                event.currentTarget.releasePointerCapture(event.pointerId);
+            }}
+            onPointerCancel={() => {
+              moveStart.current = null;
+              dockWasDragged.current = true;
+            }}
+            onLostPointerCapture={() => {
+              moveStart.current = null;
+            }}
+            onClick={(event) => {
+              if (dockWasDragged.current && event.detail !== 0) {
+                event.preventDefault();
+                dockWasDragged.current = false;
+                return;
+              }
+              expand();
             }}
-            onClick={onButtonClick}
-            onPointerDown={onButtonPointerDown}
-            onContextMenu={(event) => event.preventDefault()}
           >
-            <MessageCircle className="size-6" />
+            <RiMagicLine
+              aria-hidden="true"
+              className={`size-3.5 ${styles.magic}`}
+            />
+            <RiSparkling2Line
+              aria-hidden="true"
+              className={`size-3 ${styles.sparkle}`}
+            />
+            <RiArrowUpSLine aria-hidden="true" className="size-3" />
           </Button>
-        </PopoverTrigger>
-        {railCollapsed && (
-          <span
-            role="presentation"
-            aria-hidden="true"
-            className={`pointer-events-none absolute top-1/2 h-14 -translate-y-1/2 rounded-full bg-[#3b82f6] shadow-md ${
-              dockedEdge === 'right' ? 'right-0' : 'left-0'
-            }`}
-            style={{ width: ASSISTANT_RAIL_WIDTH }}
-          />
         )}
-        <PopoverContent
-          align="end"
-          side="top"
-          sideOffset={12}
-          className="flex h-[min(680px,80dvh)] w-[min(440px,calc(100vw-24px))] flex-col overflow-hidden rounded-2xl p-0 shadow-xl"
-          onOpenAutoFocus={(event) => event.preventDefault()}
-        >
-          <header className="flex items-center gap-2 border-b p-3">
-            <MessageCircle className="size-5 text-primary" />
-            <div className="min-w-0 flex-1">
-              <h2 className="truncate font-semibold">{t('assistant.title')}</h2>
-              <p className="truncate text-xs text-muted-foreground">
-                {t('assistant.subtitle')}
-              </p>
-            </div>
-            {open && (
-              <div
-                className="w-28 min-w-0 shrink-0"
-                title={t('assistant.modelHint')}
-              >
-                <DynamicFormItemComponent
-                  config={{
-                    id: 'assistant-model',
-                    name: 'assistant-model',
-                    type: DynamicFormItemType.LLM_MODEL_SELECTOR,
-                    default: '',
-                    required: false,
-                    label: { en_US: 'Assistant model', zh_Hans: '助手模型' },
-                  }}
-                  field={{
-                    name: 'assistant-model',
-                    value: modelUuid,
-                    onChange: setModelUuid,
-                    onBlur: () => {},
-                    ref: () => {},
-                    disabled:
-                      busy ||
-                      (!!conversation && conversation.status !== 'ready'),
-                  }}
-                  requiredModelAbility="func_call"
-                  compactModelSelector
-                />
-              </div>
-            )}
-            <Button
-              variant="ghost"
-              size="icon"
-              className="size-7 shrink-0"
-              disabled={busy}
-              onClick={reset}
-              aria-label={t('assistant.newChat')}
-            >
-              <Plus />
-            </Button>
-            <Button
-              variant="ghost"
-              size="icon"
-              className="size-7 shrink-0"
-              onClick={() => setOpen(false)}
-              aria-label={t('assistant.close')}
-            >
-              <X />
-            </Button>
-          </header>
-          <div
-            className="flex-1 space-y-3 overflow-y-auto p-4"
-            aria-live="polite"
-          >
-            {!conversation?.messages.length && !pendingText && (
-              <>
-                <p className="text-sm text-muted-foreground">
-                  {t('assistant.welcome')}
-                </p>
-                {(['discover', 'build'] as const).map((key) => (
-                  <button
-                    key={key}
-                    type="button"
-                    disabled={busy}
-                    className="w-full rounded-lg border p-3 text-left text-sm hover:bg-muted"
-                    onClick={() => setText(t(`assistant.${key}`))}
-                  >
-                    {t(`assistant.${key}`)}
-                  </button>
-                ))}
-              </>
-            )}
-            {conversation?.messages.map((message, index) =>
-              message.role === 'tool' ? (
-                <AssistantToolResult
-                  key={index}
-                  tool={message.tool}
-                  content={message.content}
-                  defaultCollapsed={!busy && conversation.status === 'ready'}
-                />
-              ) : (
-                <div
-                  key={index}
-                  className={`rounded-xl p-3 text-sm break-words ${message.role === 'user' ? 'ml-6 bg-primary/10' : 'bg-muted'}`}
-                >
-                  <ReactMarkdown
-                    remarkPlugins={[remarkGfm]}
-                    components={{
-                      img: () => null,
-                      a: ({ href, children }) => (
-                        <a
-                          href={href}
-                          className="text-primary underline"
-                          target="_blank"
-                          rel="noopener noreferrer"
-                        >
-                          {children}
-                        </a>
-                      ),
-                    }}
-                  >
-                    {message.content}
-                  </ReactMarkdown>
-                </div>
-              ),
-            )}
-            {pendingText && (
-              <div className="ml-6 rounded-xl bg-primary/10 p-3 text-sm whitespace-pre-wrap break-words">
-                {pendingText}
-                {error && (
-                  <p className="mt-1 text-xs text-destructive">
-                    {t('assistant.sendUnconfirmed')}
-                  </p>
-                )}
-              </div>
-            )}
-            {conversation?.status === 'approval' && (
-              <div className="space-y-3 rounded-xl border border-primary/30 p-3">
-                <p className="text-sm font-medium">{t('assistant.review')}</p>
-                {conversation.pending.map((call, index) => (
-                  <div key={index}>
-                    <p className="text-sm font-medium">{call.name}</p>
-                    <pre className="max-h-48 overflow-auto whitespace-pre-wrap break-all text-xs">
-                      {JSON.stringify(call.arguments, null, 2)}
-                    </pre>
-                  </div>
-                ))}
-                <div className="flex gap-2">
-                  <Button
-                    size="sm"
-                    disabled={busy}
-                    onClick={() => submit(true)}
-                  >
-                    {t('assistant.confirm')}
-                  </Button>
-                  <Button
-                    size="sm"
-                    variant="outline"
-                    disabled={busy}
-                    onClick={() => submit(false)}
-                  >
-                    {t('assistant.decline')}
-                  </Button>
-                </div>
-              </div>
-            )}
-            {busy && (
-              <p className="flex items-center gap-2 text-sm text-muted-foreground">
-                <LoaderCircle className="size-4 animate-spin" />
-                {t('assistant.working')}
-              </p>
-            )}
-            {(error || conversation?.status === 'failed') && (
-              <p role="alert" className="text-sm text-destructive">
-                {conversation?.error === 'model_unavailable'
-                  ? t('assistant.modelUnavailable')
-                  : t('assistant.error')}
-              </p>
-            )}
-            {!busy && conversation?.status === 'running' && (
-              <p className="text-sm text-muted-foreground">
-                {t('assistant.running')}
-              </p>
-            )}
-            <div ref={end} />
-          </div>
-          <form
-            className="flex items-end gap-2 border-t p-3"
-            onSubmit={(event) => {
-              event.preventDefault();
-              void submit();
+        {open && (
+          <section
+            ref={panelRef}
+            // CSS max-height adapts to the current viewport without overwriting
+            // the saved preference, so a larger screen restores the chosen size.
+            style={
+              panelHeight === undefined ? undefined : { height: panelHeight }
+            }
+            id={panelId}
+            inert={closing}
+            role="dialog"
+            aria-modal={false}
+            aria-labelledby={`${panelId}-title`}
+            className={`pointer-events-auto flex max-h-[calc(100dvh-48px)] w-full max-w-[480px] flex-col overflow-hidden rounded-t-2xl border border-b-0 bg-background ${styles.panel} ${closing ? styles.closing : ''} ${conversation?.messages.length || pendingText ? styles.conversationPanel : ''}`}
+            onKeyDown={(event) => {
+              if (event.key === 'Escape' && !event.defaultPrevented) {
+                event.preventDefault();
+                event.stopPropagation();
+                collapse();
+              }
             }}
           >
-            <textarea
-              value={text}
-              onChange={(event) => setText(event.target.value)}
-              maxLength={8000}
-              rows={2}
-              aria-label={t('assistant.placeholder')}
-              placeholder={t(
-                busy ? 'assistant.draftPlaceholder' : 'assistant.placeholder',
-              )}
+            <div
+              role="separator"
+              tabIndex={0}
+              aria-label={t('assistant.resizeHeight')}
+              aria-orientation="horizontal"
+              aria-controls={panelId}
+              className="group flex h-5 shrink-0 touch-none select-none items-center justify-center cursor-ns-resize focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-inset focus-visible:ring-ring"
+              onPointerDown={(event) => {
+                if (event.button !== 0 || !event.isPrimary) return;
+                event.preventDefault();
+                resizeStart.current = {
+                  y: event.clientY,
+                  height: panelRef.current!.getBoundingClientRect().height,
+                };
+                event.currentTarget.setPointerCapture(event.pointerId);
+              }}
+              onPointerMove={(event) => {
+                if (resizeStart.current)
+                  resizeHeight(
+                    resizeStart.current.height +
+                      resizeStart.current.y -
+                      event.clientY,
+                  );
+              }}
+              onPointerUp={(event) => {
+                resizeStart.current = null;
+                if (event.currentTarget.hasPointerCapture(event.pointerId))
+                  event.currentTarget.releasePointerCapture(event.pointerId);
+              }}
+              onLostPointerCapture={() => {
+                resizeStart.current = null;
+              }}
+              onPointerCancel={() => {
+                resizeStart.current = null;
+              }}
               onKeyDown={(event) => {
                 if (
-                  event.key === 'Enter' &&
-                  !event.shiftKey &&
-                  !event.nativeEvent.isComposing
-                ) {
+                  !['ArrowUp', 'ArrowDown', 'Home', 'End'].includes(event.key)
+                )
+                  return;
+                event.preventDefault();
+                const height = panelRef.current!.getBoundingClientRect().height;
+                resizeHeight(
+                  event.key === 'Home'
+                    ? 280
+                    : event.key === 'End'
+                      ? window.innerHeight - 48
+                      : height + (event.key === 'ArrowUp' ? 24 : -24),
+                );
+              }}
+            >
+              <span
+                aria-hidden="true"
+                className="h-1 w-9 rounded-full bg-muted-foreground/30 transition-colors group-hover:bg-muted-foreground/60 group-focus-visible:bg-muted-foreground/60"
+              />
+            </div>
+            <header
+              className={`flex shrink-0 items-center gap-2 px-3 pb-1 pt-2 ${styles.header}`}
+            >
+              <span className={styles.avatar} aria-hidden="true">
+                <RiMagicLine className={`size-4 ${styles.magic}`} />
+              </span>
+              <div
+                className="min-w-0 flex-1 cursor-grab touch-none select-none active:cursor-grabbing focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring rounded"
+                role="slider"
+                tabIndex={0}
+                aria-label={t('assistant.movePosition')}
+                aria-orientation="horizontal"
+                aria-valuemin={-100}
+                aria-valuemax={100}
+                aria-valuenow={Math.round(horizontalPosition * 100)}
+                onPointerDown={(event) => {
+                  if (event.button !== 0 || !event.isPrimary) return;
+                  event.preventDefault();
+                  moveStart.current = {
+                    x: event.clientX,
+                    offset: horizontalPosition * horizontalRange,
+                  };
+                  event.currentTarget.setPointerCapture(event.pointerId);
+                }}
+                onPointerMove={(event) => {
+                  if (!moveStart.current || horizontalRange === 0) return;
+                  setHorizontalPosition(
+                    Math.max(
+                      -1,
+                      Math.min(
+                        1,
+                        (moveStart.current.offset +
+                          event.clientX -
+                          moveStart.current.x) /
+                          horizontalRange,
+                      ),
+                    ),
+                  );
+                }}
+                onPointerUp={(event) => {
+                  moveStart.current = null;
+                  if (event.currentTarget.hasPointerCapture(event.pointerId))
+                    event.currentTarget.releasePointerCapture(event.pointerId);
+                }}
+                onPointerCancel={() => {
+                  moveStart.current = null;
+                }}
+                onLostPointerCapture={() => {
+                  moveStart.current = null;
+                }}
+                onKeyDown={(event) => {
+                  if (
+                    !['ArrowLeft', 'ArrowRight', 'Home', 'End'].includes(
+                      event.key,
+                    )
+                  )
+                    return;
+                  event.preventDefault();
+                  setHorizontalPosition(
+                    event.key === 'Home'
+                      ? -1
+                      : event.key === 'End'
+                        ? 1
+                        : Math.max(
+                            -1,
+                            Math.min(
+                              1,
+                              horizontalPosition +
+                                (event.key === 'ArrowLeft' ? -0.1 : 0.1),
+                            ),
+                          ),
+                  );
+                }}
+              >
+                <h2
+                  id={`${panelId}-title`}
+                  className="truncate text-sm font-medium"
+                >
+                  {t('assistant.title')}
+                </h2>
+              </div>
+              <Select
+                value={conversation?.uuid ?? 'new'}
+                onValueChange={(value) =>
+                  onSelect(value === 'new' ? null : value)
+                }
+              >
+                <SelectTrigger
+                  aria-label={t('assistant.sessions')}
+                  className="w-32 min-w-0 bg-background px-2 text-xs data-[size=default]:h-7"
+                >
+                  <SelectValue />
+                </SelectTrigger>
+                <SelectContent className="max-w-[min(20rem,calc(100vw-24px))]">
+                  <SelectItem value="new" className="text-xs">
+                    {t('assistant.newChat')}
+                  </SelectItem>
+                  {conversation &&
+                    !sessions.some(
+                      (session) => session.uuid === conversation.uuid,
+                    ) && (
+                      <SelectItem value={conversation.uuid} className="text-xs">
+                        {t('assistant.currentChat')}
+                      </SelectItem>
+                    )}
+                  {sessions.map((session) => (
+                    <SelectItem
+                      key={session.uuid}
+                      value={session.uuid}
+                      className="text-xs [&_[data-slot=select-item-text]]:min-w-0"
+                    >
+                      <span className="block truncate">
+                        {session.status === 'running'
+                          ? '◌ '
+                          : session.status === 'approval'
+                            ? '… '
+                            : ''}
+                        {session.title || t('assistant.newChat')}
+                      </span>
+                    </SelectItem>
+                  ))}
+                </SelectContent>
+              </Select>
+              <Button
+                variant="ghost"
+                size="icon"
+                className="size-7 shrink-0"
+                onClick={reset}
+                aria-label={t('assistant.newChat')}
+              >
+                <Plus />
+              </Button>
+              <Button
+                variant="ghost"
+                size="icon"
+                className="size-7 shrink-0"
+                onClick={collapse}
+                aria-label={t('assistant.collapse')}
+                aria-expanded={true}
+                aria-controls={panelId}
+              >
+                <ChevronDown />
+              </Button>
+            </header>
+            <div
+              ref={messageList}
+              className="min-h-0 flex-1 space-y-3 overflow-y-auto overscroll-contain px-3 py-2"
+              onWheel={(event) => {
+                if (event.deltaY < 0) followLatest.current = false;
+              }}
+              onScroll={(event) => {
+                const list = event.currentTarget;
+                const delta = list.scrollTop - scrollPosition.current;
+                if (delta < -1) followLatest.current = false;
+                else if (delta > 1) {
+                  followLatest.current =
+                    list.scrollHeight - list.clientHeight - list.scrollTop < 32;
+                }
+                // Delayed events from our own scroll have zero delta. Content may
+                // have grown since then, so they must not disable following.
+                scrollPosition.current = list.scrollTop;
+              }}
+              aria-live="polite"
+            >
+              {!conversation?.messages.length && !pendingText && (
+                <div className={styles.welcome}>
+                  <h3 className="text-base font-medium tracking-tight">
+                    {t('assistant.subtitle')}
+                  </h3>
+                  <p className="max-w-lg text-[13px] leading-relaxed text-muted-foreground">
+                    {t('assistant.welcome')}
+                  </p>
+                  <div className="flex flex-wrap gap-2 pt-2">
+                    {(['discover', 'build'] as const).map((key) => (
+                      <button
+                        key={key}
+                        type="button"
+                        disabled={busy || !canUseModels}
+                        className={styles.prompt}
+                        onClick={() => {
+                          setText(t(`assistant.${key}`));
+                          input.current?.focus();
+                        }}
+                      >
+                        <span className={styles.promptIcon} aria-hidden="true">
+                          {key === 'discover' ? (
+                            <RiCompass3Line className="size-4" />
+                          ) : (
+                            <RiMagicLine className="size-4" />
+                          )}
+                        </span>
+                        <span className="flex-1">{t(`assistant.${key}`)}</span>
+                        <RiArrowRightUpLine
+                          className="size-4 shrink-0 opacity-60"
+                          aria-hidden="true"
+                        />
+                      </button>
+                    ))}
+                  </div>
+                </div>
+              )}
+              {messageGroups.map((group) => (
+                <div
+                  key={group.key}
+                  className={`flex min-w-0 ${group.role === 'user' ? 'justify-end' : 'justify-start'}`}
+                >
+                  <div
+                    className={`min-w-0 max-w-[94%] space-y-2 rounded-2xl px-3 py-2 text-sm ${group.role === 'user' ? styles.userMessage : styles.assistantMessage}`}
+                  >
+                    {group.messages.map((message, index) =>
+                      message.role === 'tool' ? (
+                        <AssistantToolResult
+                          key={index}
+                          tool={message.tool}
+                          content={message.content}
+                          defaultCollapsed
+                        />
+                      ) : message.role === 'user' ? (
+                        <p
+                          key={index}
+                          className="whitespace-pre-wrap break-words [overflow-wrap:anywhere] leading-relaxed"
+                        >
+                          {message.content}
+                        </p>
+                      ) : (
+                        <div
+                          key={index}
+                          className={`markdown-body ${styles.messageContent}`}
+                        >
+                          <ReactMarkdown
+                            remarkPlugins={[remarkGfm]}
+                            rehypePlugins={[rehypeHighlight]}
+                            components={{
+                              ul: ({ children }) => (
+                                <ul className="list-disc">{children}</ul>
+                              ),
+                              ol: ({ children }) => (
+                                <ol className="list-decimal">{children}</ol>
+                              ),
+                              img: () => null,
+                              a: ({ href, children }) => (
+                                <a
+                                  href={href}
+                                  target="_blank"
+                                  rel="noopener noreferrer"
+                                >
+                                  {children}
+                                </a>
+                              ),
+                            }}
+                          >
+                            {message.content}
+                          </ReactMarkdown>
+                        </div>
+                      ),
+                    )}
+                  </div>
+                </div>
+              ))}
+              {pendingText && (
+                <div
+                  className={`ml-auto w-fit max-w-[94%] rounded-2xl px-3 py-2 text-sm leading-relaxed whitespace-pre-wrap break-words [overflow-wrap:anywhere] ${styles.userMessage}`}
+                >
+                  {pendingText}
+                  {error && (
+                    <p className="mt-1 text-xs text-destructive">
+                      {t('assistant.sendUnconfirmed')}
+                    </p>
+                  )}
+                </div>
+              )}
+              {conversation?.status === 'approval' && (
+                <div className="space-y-3 rounded-xl border border-primary/30 p-3">
+                  <p className="text-sm font-medium">{t('assistant.review')}</p>
+                  {conversation.pending.map((call, index) => (
+                    <div key={index}>
+                      <p className="text-sm font-medium">{call.name}</p>
+                      <pre className="max-h-48 overflow-auto whitespace-pre-wrap break-all text-xs">
+                        {JSON.stringify(call.arguments, null, 2)}
+                      </pre>
+                    </div>
+                  ))}
+                  <div className="flex gap-2">
+                    <Button
+                      size="sm"
+                      disabled={busy}
+                      onClick={() => submit(true)}
+                    >
+                      {t('assistant.confirm')}
+                    </Button>
+                    <Button
+                      size="sm"
+                      variant="outline"
+                      disabled={busy}
+                      onClick={() => submit(false)}
+                    >
+                      {t('assistant.decline')}
+                    </Button>
+                  </div>
+                </div>
+              )}
+              {busy && (
+                <p className="flex items-center gap-2 text-sm text-muted-foreground">
+                  <LoaderCircle className="size-4 animate-spin" />
+                  {loading
+                    ? t('assistant.preparing')
+                    : t(
+                        phase === 'tool'
+                          ? 'assistant.executing'
+                          : 'assistant.thinking',
+                        { round },
+                      )}
+                </p>
+              )}
+              {canUseModels && modelError && !modelUuid && (
+                <p role="alert" className="text-xs text-destructive">
+                  {t('assistant.recommendationFailed')}
+                </p>
+              )}
+              {(error || conversation?.status === 'failed') && (
+                <p role="alert" className="text-sm text-destructive">
+                  {conversation?.error === 'stopped'
+                    ? t('assistant.stopped')
+                    : conversation?.error === 'model_unavailable'
+                      ? t('assistant.modelUnavailable')
+                      : t('assistant.error')}
+                </p>
+              )}
+              {!busy && conversation?.status === 'running' && (
+                <p className="text-sm text-muted-foreground">
+                  {t('assistant.running')}
+                </p>
+              )}
+            </div>
+            {canUseModels ? (
+              <form
+                className={`mx-2 mb-[max(0.5rem,env(safe-area-inset-bottom))] flex shrink-0 flex-col gap-1 rounded-xl p-1.5 ${styles.composer}`}
+                onSubmit={(event) => {
                   event.preventDefault();
                   void submit();
-                }
-              }}
-              className="min-w-0 flex-1 resize-none rounded-lg border bg-background p-2 text-sm focus-visible:outline-2 focus-visible:outline-primary"
-            />
-            <Button
-              type="submit"
-              size="icon"
-              aria-label={t('assistant.send')}
-              disabled={
-                busy ||
-                !text.trim() ||
-                (!!conversation && conversation.status !== 'ready')
-              }
-            >
-              <Send className="size-4" />
-            </Button>
-          </form>
-        </PopoverContent>
-      </Popover>
-    </div>
+                }}
+              >
+                <textarea
+                  ref={input}
+                  value={text}
+                  onChange={(event) => setText(event.target.value)}
+                  maxLength={8000}
+                  rows={2}
+                  aria-label={t('assistant.placeholder')}
+                  placeholder={t(
+                    busy
+                      ? 'assistant.draftPlaceholder'
+                      : 'assistant.placeholder',
+                  )}
+                  onKeyDown={(event) => {
+                    if (
+                      event.key === 'Enter' &&
+                      !event.shiftKey &&
+                      !event.nativeEvent.isComposing
+                    ) {
+                      event.preventDefault();
+                      void submit();
+                    }
+                  }}
+                  className={`w-full min-w-0 resize-none rounded-lg bg-transparent px-2 py-1.5 text-sm outline-none focus-visible:outline-none ${styles.input}`}
+                />
+                <div className="flex w-full items-center justify-between gap-3">
+                  {open && (
+                    <div
+                      className="w-28 min-w-0 shrink-0"
+                      title={t('assistant.modelHint')}
+                    >
+                      <DynamicFormItemComponent
+                        config={{
+                          id: 'assistant-model',
+                          name: 'assistant-model',
+                          type: DynamicFormItemType.LLM_MODEL_SELECTOR,
+                          default: '',
+                          required: false,
+                          label: {
+                            en_US: 'Assistant model',
+                            zh_Hans: '助手模型',
+                          },
+                        }}
+                        field={{
+                          name: 'assistant-model',
+                          value: modelUuid,
+                          onChange: (value: string) => {
+                            // Radix's native form bridge can emit an empty value while
+                            // async options mount. It is not a user model selection.
+                            if (!value) return;
+                            manualModel.current = true;
+                            setModelUuid(value);
+                            setModelName(undefined);
+                            setModelError(false);
+                          },
+                          onBlur: () => {},
+                          ref: () => {},
+                          disabled:
+                            sending ||
+                            (!!conversation && conversation.status !== 'ready'),
+                        }}
+                        requiredModelAbility="func_call"
+                        compactModelSelector
+                        selectedModelLabel={modelName}
+                      />
+                    </div>
+                  )}
+                  {conversation?.status === 'running' ? (
+                    <Button
+                      type="button"
+                      size="icon"
+                      variant="outline"
+                      className="size-7 rounded-lg"
+                      aria-label={t('assistant.stop')}
+                      disabled={stopping}
+                      onClick={() => void stop()}
+                    >
+                      <Square className="size-3 fill-current" />
+                    </Button>
+                  ) : (
+                    <Button
+                      type="submit"
+                      size="icon"
+                      className={`size-7 rounded-lg ${styles.send}`}
+                      aria-label={t('assistant.send')}
+                      disabled={
+                        busy ||
+                        !text.trim() ||
+                        !modelUuid ||
+                        (!!conversation && conversation.status !== 'ready')
+                      }
+                    >
+                      <Send className="size-4" />
+                    </Button>
+                  )}
+                </div>
+              </form>
+            ) : (
+              setup
+            )}
+          </section>
+        )}
+      </div>
+    </div>,
+    document.body,
   );
 }
diff --git a/web/src/app/home/components/account-settings-dialog/AccountSettingsPanel.tsx b/web/src/app/home/components/account-settings-dialog/AccountSettingsPanel.tsx
index ea12d95f4..1839815ce 100644
--- a/web/src/app/home/components/account-settings-dialog/AccountSettingsPanel.tsx
+++ b/web/src/app/home/components/account-settings-dialog/AccountSettingsPanel.tsx
@@ -131,9 +131,11 @@ export default function AccountSettingsPanel({
   async function loadTotpAccounts() {
     try {
       const res = await httpClient.getTotpAccounts();
-      const list = res.accounts || [];
-      setIsManager(list.length > 1);
-      setTotpRows(list);
+      // The role rule lives in the API: owners and admins receive the Workspace
+      // list, everyone else receives 403. Deriving manager state from the list
+      // length would hide oversight in a single-member Workspace.
+      setIsManager(true);
+      setTotpRows(res.accounts || []);
     } catch {
       // A non-manager receives 403 here; the panel keeps working on own state.
       setIsManager(false);
diff --git a/web/src/app/home/components/assistant-dock.ts b/web/src/app/home/components/assistant-dock.ts
deleted file mode 100644
index 3f37d04eb..000000000
--- a/web/src/app/home/components/assistant-dock.ts
+++ /dev/null
@@ -1,89 +0,0 @@
-/**
- * Pure geometry + state helpers for the floating workspace assistant button.
- *
- * The component keeps these decisions here so the dock/rail transitions stay
- * testable without a DOM. The key invariant is that the *same* edge value
- * drives the collapsed strip and the expanded button, so they can never drift
- * apart or fight each other (the race the UI is careful to avoid).
- */
-
-export const ASSISTANT_BUTTON_SIZE = 48;
-/** Distance to the viewport edge that snaps the button into the rail. */
-export const ASSISTANT_SNAP_THRESHOLD = 24;
-/** Rail strip that stays visible while collapsed, in pixels. */
-export const ASSISTANT_RAIL_WIDTH = 6;
-
-export type AssistantDragPosition = { x: number; y: number };
-export type AssistantEdge = 'left' | 'right' | null;
-
-/** Viewport box the button may occupy. Kept explicit so tests can pin a size. */
-export type AssistantViewport = { width: number; height: number };
-
-export function clampAssistantPosition(
-  x: number,
-  y: number,
-  viewport: AssistantViewport,
-): AssistantDragPosition {
-  const maxX = Math.max(0, viewport.width - ASSISTANT_BUTTON_SIZE);
-  const maxY = Math.max(0, viewport.height - ASSISTANT_BUTTON_SIZE);
-  return {
-    x: Math.min(Math.max(0, x), maxX),
-    y: Math.min(Math.max(0, y), maxY),
-  };
-}
-
-/**
- * Decide whether a resting position is docked against a vertical edge.
- * The left edge wins ties so a centred drag always resolves deterministically.
- */
-export function resolveAssistantEdge(
-  x: number,
-  viewport: AssistantViewport,
-): AssistantEdge {
-  const maxX = Math.max(0, viewport.width - ASSISTANT_BUTTON_SIZE);
-  const leftGap = x;
-  const rightGap = maxX - x;
-  if (leftGap <= ASSISTANT_SNAP_THRESHOLD && leftGap <= rightGap) return 'left';
-  if (rightGap <= ASSISTANT_SNAP_THRESHOLD) return 'right';
-  return null;
-}
-
-export type AssistantHoverInput = {
-  dockedEdge: AssistantEdge;
-  railExpanded: boolean;
-  /** A long press has armed the drag, or the pointer is actively moving. */
-  dragging: boolean;
-};
-
-/** Distance to the viewport edge that snaps the button into the rail. */
-
-/**
- * Hovering any part of the control must reveal the full button again.
- * Only a docked *and still collapsed* button needs to expand, so a pointer that
- * merely jitters inside an already expanded button does no state churn.
- */
-export function shouldExpandRail(input: AssistantHoverInput): boolean {
-  if (input.dragging) return false;
-  return !!input.dockedEdge && !input.railExpanded;
-}
-
-/**
- * A docked button collapses to the strip when the pointer leaves. This is the
- * mirror of `shouldExpandRail` and is keyed on `railExpanded` being *true* —
- * the earlier `!railExpanded` form made a revealed button impossible to
- * re-collapse, which is the bug this guards against. The panel being open pins
- * it, because the popover stays anchored to the button.
- */
-export function shouldCollapseRail(input: AssistantHoverInput): boolean {
-  if (input.dragging) return false;
-  return !!input.dockedEdge && input.railExpanded;
-}
-
-/**
- * After a drop the button always starts collapsed, even though the pointer is
- * still over it. The rail is then held collapsed until the pointer exits, which
- * avoids the "drops under the cursor and never comes back" race.
- */
-export function restingRailExpanded(): boolean {
-  return false;
-}
diff --git a/web/src/app/home/components/dynamic-form/DynamicFormComponent.tsx b/web/src/app/home/components/dynamic-form/DynamicFormComponent.tsx
index 5ec6eff88..f5a4541e1 100644
--- a/web/src/app/home/components/dynamic-form/DynamicFormComponent.tsx
+++ b/web/src/app/home/components/dynamic-form/DynamicFormComponent.tsx
@@ -3,7 +3,7 @@ import {
   SYSTEM_FIELD_PREFIX,
   DynamicFormItemType,
 } from '@/app/infra/entities/form/dynamic';
-import { useForm } from 'react-hook-form';
+import { useForm, useWatch, type UseFormReturn } from 'react-hook-form';
 import type { ControllerRenderProps } from 'react-hook-form';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { isJsonValue, isPromptValue } from './StructuredFieldValue';
@@ -24,7 +24,8 @@ import {
 import QrCodeLoginDialog, {
   QrLoginPlatform,
 } from '@/app/home/components/qrcode-login/QrCodeLoginDialog';
-import { useCallback, useEffect, useMemo, useRef, useState } from 'react';
+import SlackSetupDialog from '@/app/home/components/qrcode-login/SlackSetupDialog';
+import { memo, useCallback, useEffect, useMemo, useRef, useState } from 'react';
 import type { ReactNode } from 'react';
 import { extractI18nObject } from '@/i18n/I18nProvider';
 import { useTranslation } from 'react-i18next';
@@ -47,7 +48,7 @@ import {
   TooltipProvider,
   TooltipTrigger,
 } from '@/components/ui/tooltip';
-import { systemInfo } from '@/app/infra/http';
+import { backendClient, systemInfo } from '@/app/infra/http';
 import { getAdapterDocUrl } from '@/app/infra/entities/adapter-docs';
 import { parseDynamicFormItemType } from './DynamicFormItemConfig';
 import {
@@ -304,18 +305,56 @@ function DownloadLinkField({
   helpUrl?: string | null;
   helpLabel: string;
 }) {
-  const baseUrl = import.meta.env.VITE_API_BASE_URL || window.location.origin;
-  const downloadUrl = url.startsWith('http') ? url : `${baseUrl}${url}`;
+  const { t } = useTranslation();
+  const [downloading, setDownloading] = useState(false);
+  const [failed, setFailed] = useState(false);
+
+  const download = async () => {
+    if (downloading) return;
+    setDownloading(true);
+    setFailed(false);
+    try {
+      let blob: Blob;
+      if (/^https?:\/\//i.test(url)) {
+        // External templates must not receive backend credentials.
+        const response = await fetch(url);
+        if (!response.ok)
+          throw new Error(`Download failed: ${response.status}`);
+        blob = await response.blob();
+      } else {
+        const response = await backendClient.downloadFile(url);
+        blob = response.data;
+      }
+      const objectUrl = URL.createObjectURL(blob);
+      const link = document.createElement('a');
+      link.href = objectUrl;
+      link.download =
+        filename || url.split('/').pop()?.split('?')[0] || 'download';
+      document.body.appendChild(link);
+      link.click();
+      link.remove();
+      window.setTimeout(() => URL.revokeObjectURL(objectUrl), 1000);
+    } catch {
+      setFailed(true);
+    } finally {
+      setDownloading(false);
+    }
+  };
 
   return (
     <FormItem className="min-w-0">
       <FormLabel className="break-words">{label}</FormLabel>
       <div className="flex min-w-0 flex-wrap items-center gap-2">
-        <Button asChild variant="outline" size="sm">
-          <a href={downloadUrl} download={filename}>
-            <Download className="h-4 w-4" />
-            {label}
-          </a>
+        <Button
+          type="button"
+          variant="outline"
+          size="sm"
+          onClick={download}
+          disabled={downloading}
+          aria-busy={downloading}
+        >
+          <Download className="h-4 w-4" />
+          {label}
         </Button>
         {helpUrl && (
           <Button asChild variant="ghost" size="sm">
@@ -326,6 +365,11 @@ function DownloadLinkField({
           </Button>
         )}
       </div>
+      {failed && (
+        <p role="alert" className="text-sm text-destructive">
+          {t('common.operationFailed')}
+        </p>
+      )}
       {description && (
         <p className="max-w-2xl text-sm break-words text-muted-foreground">
           {description}
@@ -441,6 +485,7 @@ export default function DynamicFormComponent({
   systemContext,
   onValidate,
   renderItem,
+  hiddenItemNames,
 }: {
   itemConfigList: IDynamicFormItemSchema[];
   onSubmit?: (val: object) => unknown;
@@ -463,10 +508,12 @@ export default function DynamicFormComponent({
     formValues: Record<string, unknown>;
     setFormValue: (name: string, value: unknown) => void;
   }) => ReactNode | undefined;
+  /** Fields retained in form state but omitted from the rendered form. */
+  hiddenItemNames?: readonly string[];
 }) {
   const isInitialMount = useRef(true);
   const previousInitialValues = useRef(initialValues);
-  const { t, i18n } = useTranslation();
+  const { t } = useTranslation();
 
   // Filter out display-only fields (webhook-url/embed-code/qr-code-login types
   // and `__system.*`-named fields) that should not participate in form state,
@@ -490,28 +537,32 @@ export default function DynamicFormComponent({
   );
 
   // 根据 itemConfigList 动态生成 zod schema
-  const formSchema = z.object(
-    editableValueSpecs.reduce(
-      (acc, item) => {
-        let fieldSchema = getValueSchema(item);
+  const formSchema = useMemo(
+    () =>
+      z.object(
+        editableValueSpecs.reduce(
+          (acc, item) => {
+            let fieldSchema = getValueSchema(item);
 
-        if (
-          item.required &&
-          (fieldSchema instanceof z.ZodString ||
-            fieldSchema instanceof z.ZodArray)
-        ) {
-          fieldSchema = fieldSchema.min(1, {
-            message: t('common.fieldRequired'),
-          });
-        }
+            if (
+              item.required &&
+              (fieldSchema instanceof z.ZodString ||
+                fieldSchema instanceof z.ZodArray)
+            ) {
+              fieldSchema = fieldSchema.min(1, {
+                message: t('common.fieldRequired'),
+              });
+            }
 
-        return {
-          ...acc,
-          [item.name]: fieldSchema,
-        };
-      },
-      {} as Record<string, z.ZodTypeAny>,
-    ),
+            return {
+              ...acc,
+              [item.name]: fieldSchema,
+            };
+          },
+          {} as Record<string, z.ZodTypeAny>,
+        ),
+      ),
+    [editableValueSpecs, t],
   );
 
   type FormValues = z.infer<typeof formSchema>;
@@ -577,21 +628,12 @@ export default function DynamicFormComponent({
     }
   }, [initialValues, form, editableValueSpecs]);
 
-  // Get reactive form values for conditional rendering
-  const watchedValues = form.watch();
-  const setFormValue = (name: string, value: unknown) => {
-    form.setValue(name as keyof FormValues, value as never, {
-      shouldDirty: true,
-      shouldValidate: true,
-    });
-  };
-
   // Stable ref for onSubmit to avoid re-triggering the effect when the
   // parent passes a new closure on every render.
   const onSubmitRef = useRef(onSubmit);
   onSubmitRef.current = onSubmit;
 
-  // 监听表单值变化
+  // Keep parent drafts synchronous with field changes.
   useEffect(() => {
     // Emit initial form values immediately so the parent always has a valid snapshot,
     // even if the user saves without modifying any field.
@@ -618,21 +660,120 @@ export default function DynamicFormComponent({
         editableValueSpecs,
         formValues as Record<string, unknown>,
       );
-      onSubmitRef.current?.(finalValues);
       previousInitialValues.current = finalValues as Record<string, object>;
+      onSubmitRef.current?.(finalValues);
     });
     return () => subscription.unsubscribe();
   }, [form, editableValueSpecs]);
 
+  return (
+    <DynamicFormFields
+      form={form}
+      itemConfigList={itemConfigList}
+      onFileUploaded={onFileUploaded}
+      isEditing={isEditing}
+      externalDependentValues={externalDependentValues}
+      systemContext={systemContext}
+      renderItem={renderItem}
+      hiddenItemNames={hiddenItemNames}
+    />
+  );
+}
+
+type DynamicFormProps = Parameters<typeof DynamicFormComponent>[0];
+
+type DynamicFormFieldsProps = Omit<
+  DynamicFormProps,
+  'initialValues' | 'onSubmit' | 'onValidate'
+> & {
+  form: UseFormReturn<Record<string, any>>;
+};
+
+// The stable provider also isolates useWatch/useController from an outer form's
+// context updates. The field list itself must subscribe inside this provider.
+const DynamicFormFields = memo(function DynamicFormFields(
+  props: DynamicFormFieldsProps,
+) {
+  return (
+    <Form {...props.form}>
+      <DynamicFormFieldList {...props} />
+    </Form>
+  );
+});
+
+function DynamicFormFieldList({
+  form,
+  itemConfigList,
+  onFileUploaded,
+  isEditing,
+  externalDependentValues,
+  systemContext,
+  renderItem,
+  hiddenItemNames,
+}: DynamicFormFieldsProps) {
+  type FormValues = Record<string, any>;
+  const { t, i18n } = useTranslation();
+  const dependencies = useMemo(
+    () => [
+      ...new Set(
+        itemConfigList
+          .flatMap((item) => [
+            item.show_if?.field,
+            item.disable_if?.field,
+            ...(item.disabled_tooltip_overrides ?? []).map(
+              (rule) => rule.when.field,
+            ),
+            ...(item.type === DynamicFormItemType.RICH_TOOLS_SELECTOR ||
+            item.type === DynamicFormItemType.RESOURCES_SELECTOR
+              ? getValueSpecs(item).map((spec) => spec.name)
+              : []),
+          ])
+          .filter(
+            (name): name is string =>
+              !!name && !name.startsWith(SYSTEM_FIELD_PREFIX),
+          ),
+      ),
+    ],
+    [itemConfigList],
+  );
+  const watchedValues = useWatch({
+    control: form.control,
+    // Custom controls can read arbitrary fields; regular controls only need
+    // their declared conditions and composite-field dependencies here.
+    compute: (values: Record<string, any>) =>
+      renderItem
+        ? values
+        : Object.fromEntries(dependencies.map((name) => [name, values[name]])),
+  });
+  const setFormValue = useCallback(
+    (name: string, value: unknown) => {
+      form.setValue(name, value, { shouldDirty: true, shouldValidate: true });
+    },
+    [form],
+  );
+
   // State for QR code login dialog
   const [qrDialogOpen, setQrDialogOpen] = useState(false);
-  const [qrDialogPlatform, setQrDialogPlatform] =
-    useState<QrLoginPlatform>('feishu');
+  const [qrDialogPlatform, setQrDialogPlatform] = useState<
+    QrLoginPlatform | 'slack'
+  >('feishu');
 
   return (
-    <Form {...form}>
-      <div className="min-w-0 max-w-full space-y-4 overflow-x-hidden">
-        {/* QR code login dialog */}
+    <div className="min-w-0 max-w-full space-y-4 overflow-x-hidden">
+      {/* QR code login dialog */}
+      {qrDialogPlatform === 'slack' ? (
+        <SlackSetupDialog
+          open={qrDialogOpen}
+          onOpenChange={setQrDialogOpen}
+          botId={systemContext?.bot_uuid as string | undefined}
+          webhookUrl={(systemContext?.webhook_url as string) || ''}
+          socketMode={!!form.getValues('socket_mode')}
+          onSuccess={(credentials) => {
+            for (const [key, value] of Object.entries(credentials))
+              setFormValue(key, value);
+          }}
+        />
+      ) : (
         <QrCodeLoginDialog
           open={qrDialogOpen}
           onOpenChange={setQrDialogOpen}
@@ -645,296 +786,354 @@ export default function DynamicFormComponent({
             }
           }}
         />
+      )}
 
-        {itemConfigList.map((config, index) => {
-          // Create a normalized config with type converted to frontend format
-          const normalizedConfig = {
-            ...config,
-            type: normalizeItemType(config.type),
-          };
-          const fieldKey = config.id || config.name || `field-${index}`;
+      {itemConfigList.map((config, index) => {
+        if (hiddenItemNames?.includes(config.name)) return null;
 
-          let isHiddenByCondition = false;
-          if (config.show_if) {
-            const dependValue = resolveShowIfValue(
-              config.show_if.field,
-              watchedValues as Record<string, unknown>,
-              externalDependentValues,
-              systemContext,
-            );
+        // Create a normalized config with type converted to frontend format
+        const normalizedConfig = {
+          ...config,
+          type: normalizeItemType(config.type),
+        };
+        const fieldKey = config.id || config.name || `field-${index}`;
 
-            if (
-              config.show_if.operator === 'eq' &&
-              dependValue !== config.show_if.value
-            ) {
-              isHiddenByCondition = true;
-            }
-            if (
-              config.show_if.operator === 'neq' &&
-              dependValue === config.show_if.value
-            ) {
-              isHiddenByCondition = true;
-            }
-            if (
-              config.show_if.operator === 'in' &&
-              Array.isArray(config.show_if.value) &&
-              !config.show_if.value.includes(dependValue)
-            ) {
-              isHiddenByCondition = true;
-            }
-          }
+        let isHiddenByCondition = false;
+        if (config.show_if) {
+          const dependValue = resolveShowIfValue(
+            config.show_if.field,
+            watchedValues as Record<string, unknown>,
+            externalDependentValues,
+            systemContext,
+          );
 
-          // Keep structured drafts mounted across conditional hiding so invalid
-          // JSON cannot disappear from validation when Advanced Settings closes.
           if (
-            isHiddenByCondition &&
-            normalizedConfig.type !== DynamicFormItemType.JSON &&
-            normalizedConfig.type !== DynamicFormItemType.PROMPT_EDITOR
-          )
-            return null;
-
-          // Keep locked fields visible and resolve only the applicable reason.
-          const { isDisabledByCondition, disabledTooltip: tooltip } =
-            resolveDisabledState(
-              config,
-              watchedValues as Record<string, unknown>,
-              externalDependentValues,
-              systemContext,
-            );
-
-          // All fields are disabled when editing (creation_settings are
-          // immutable) or when ``disable_if`` matches.
-          const isFieldDisabled = !!isEditing || isDisabledByCondition;
-          const disabledTooltip = tooltip ? extractI18nObject(tooltip) : '';
-          const renderDisabledTooltipIcon = () =>
-            disabledTooltip ? (
-              <DisabledTooltipIcon text={disabledTooltip} />
-            ) : null;
-
-          // `__system.*` fields are display-only; their value is resolved
-          // from systemContext (same namespace as show_if), not user input.
-          // Hidden entirely when the deployment doesn't provide the value.
-          if (config.name.startsWith(SYSTEM_FIELD_PREFIX)) {
-            const rawValue =
-              systemContext?.[config.name.slice(SYSTEM_FIELD_PREFIX.length)];
-            const values = (Array.isArray(rawValue) ? rawValue : [rawValue])
-              .filter((v) => v !== undefined && v !== null && v !== '')
-              .map(String);
-            if (values.length === 0) return null;
-
-            return (
-              <SystemInfoField
-                key={config.id}
-                label={extractI18nObject(config.label)}
-                description={
-                  config.description
-                    ? extractI18nObject(config.description)
-                    : undefined
-                }
-                values={values}
-              />
-            );
+            config.show_if.operator === 'eq' &&
+            dependValue !== config.show_if.value
+          ) {
+            isHiddenByCondition = true;
           }
-
-          // Webhook URL fields are display-only; render outside of form binding
-          if (normalizedConfig.type === 'webhook-url') {
-            const webhookUrl = (systemContext?.webhook_url as string) || '';
-            const extraWebhookUrl =
-              (systemContext?.extra_webhook_url as string) || '';
-
-            if (!webhookUrl) return null;
-
-            return (
-              <WebhookUrlField
-                key={fieldKey}
-                label={extractI18nObject(config.label)}
-                description={
-                  config.description
-                    ? extractI18nObject(config.description)
-                    : undefined
-                }
-                url={webhookUrl}
-                extraUrl={extraWebhookUrl || undefined}
-              />
-            );
+          if (
+            config.show_if.operator === 'neq' &&
+            dependValue === config.show_if.value
+          ) {
+            isHiddenByCondition = true;
           }
-
-          if (normalizedConfig.type === 'embed-code') {
-            const botUuid = (systemContext?.bot_uuid as string) || '';
-            if (!botUuid) return null;
-
-            const baseUrl =
-              import.meta.env.VITE_API_BASE_URL || window.location.origin;
-            const widgetTitle =
-              ((systemContext?.adapter_config as Record<string, unknown>)
-                ?.title as string) || 'LangBot';
-            const safeTitle = widgetTitle
-              .replace(/&/g, '&amp;')
-              .replace(/"/g, '&quot;')
-              .replace(/</g, '&lt;')
-              .replace(/>/g, '&gt;');
-            const embedSnippet = `<script data-title="${safeTitle}" src="${baseUrl}/api/v1/embed/${botUuid}/widget.js"><\/script>`;
-
-            return (
-              <EmbedCodeField
-                key={fieldKey}
-                label={extractI18nObject(config.label)}
-                description={
-                  config.description
-                    ? extractI18nObject(config.description)
-                    : undefined
-                }
-                snippet={embedSnippet}
-              />
-            );
+          if (
+            config.show_if.operator === 'in' &&
+            Array.isArray(config.show_if.value) &&
+            !config.show_if.value.includes(dependValue)
+          ) {
+            isHiddenByCondition = true;
           }
+        }
 
-          if (config.type === 'download-link') {
-            if (!config.url) return null;
+        // Keep structured drafts mounted across conditional hiding so invalid
+        // JSON cannot disappear from validation when Advanced Settings closes.
+        if (
+          isHiddenByCondition &&
+          normalizedConfig.type !== DynamicFormItemType.JSON &&
+          normalizedConfig.type !== DynamicFormItemType.PROMPT_EDITOR
+        )
+          return null;
 
-            return (
-              <DownloadLinkField
-                key={config.id}
-                label={extractI18nObject(config.label)}
-                description={
-                  config.description
-                    ? extractI18nObject(config.description)
-                    : undefined
-                }
-                url={config.url}
-                filename={config.download_filename}
-                helpUrl={getAdapterDocUrl(config.help_links, i18n.language)}
-                helpLabel={
-                  config.help_label
-                    ? extractI18nObject(config.help_label)
-                    : t('bots.viewAdapterDocs')
-                }
-              />
-            );
-          }
+        // Keep locked fields visible and resolve only the applicable reason.
+        const { isDisabledByCondition, disabledTooltip: tooltip } =
+          resolveDisabledState(
+            config,
+            watchedValues as Record<string, unknown>,
+            externalDependentValues,
+            systemContext,
+          );
 
-          // QR code login button (e.g. Feishu one-click create, WeChat scan login)
-          if (config.type === 'qr-code-login') {
-            return (
-              <FormItem key={fieldKey}>
-                <div
-                  className="relative flex items-center gap-4 p-4 rounded-xl border-2 border-dashed cursor-pointer transition-all hover:border-solid hover:shadow-md group"
-                  style={{
-                    borderColor:
-                      'color-mix(in srgb, var(--primary) 25%, transparent)',
-                    background:
-                      'color-mix(in srgb, var(--primary) 3%, transparent)',
-                  }}
-                  onClick={() => {
-                    if (!isEditing) {
-                      setQrDialogPlatform(
-                        (config.login_platform as QrLoginPlatform) || 'feishu',
-                      );
-                      setQrDialogOpen(true);
-                    }
+        // All fields are disabled when editing (creation_settings are
+        // immutable) or when ``disable_if`` matches.
+        const isFieldDisabled = !!isEditing || isDisabledByCondition;
+        const disabledTooltip = tooltip ? extractI18nObject(tooltip) : '';
+        const renderDisabledTooltipIcon = () =>
+          disabledTooltip ? (
+            <DisabledTooltipIcon text={disabledTooltip} />
+          ) : null;
+
+        // `__system.*` fields are display-only; their value is resolved
+        // from systemContext (same namespace as show_if), not user input.
+        // Hidden entirely when the deployment doesn't provide the value.
+        if (config.name.startsWith(SYSTEM_FIELD_PREFIX)) {
+          const rawValue =
+            systemContext?.[config.name.slice(SYSTEM_FIELD_PREFIX.length)];
+          const values = (Array.isArray(rawValue) ? rawValue : [rawValue])
+            .filter((v) => v !== undefined && v !== null && v !== '')
+            .map(String);
+          if (values.length === 0) return null;
+
+          return (
+            <SystemInfoField
+              key={config.id}
+              label={extractI18nObject(config.label)}
+              description={
+                config.description
+                  ? extractI18nObject(config.description)
+                  : undefined
+              }
+              values={values}
+            />
+          );
+        }
+
+        // Webhook URL fields are display-only; render outside of form binding
+        if (normalizedConfig.type === 'webhook-url') {
+          const webhookUrl = (systemContext?.webhook_url as string) || '';
+          const extraWebhookUrl =
+            (systemContext?.extra_webhook_url as string) || '';
+
+          if (!webhookUrl) return null;
+
+          return (
+            <WebhookUrlField
+              key={fieldKey}
+              label={extractI18nObject(config.label)}
+              description={
+                config.description
+                  ? extractI18nObject(config.description)
+                  : undefined
+              }
+              url={webhookUrl}
+              extraUrl={extraWebhookUrl || undefined}
+            />
+          );
+        }
+
+        if (normalizedConfig.type === 'embed-code') {
+          const botUuid = (systemContext?.bot_uuid as string) || '';
+          if (!botUuid) return null;
+
+          const baseUrl =
+            import.meta.env.VITE_API_BASE_URL || window.location.origin;
+          const widgetTitle =
+            ((systemContext?.adapter_config as Record<string, unknown>)
+              ?.title as string) || 'LangBot';
+          const safeTitle = widgetTitle
+            .replace(/&/g, '&amp;')
+            .replace(/"/g, '&quot;')
+            .replace(/</g, '&lt;')
+            .replace(/>/g, '&gt;');
+          const embedSnippet = `<script data-title="${safeTitle}" src="${baseUrl}/api/v1/embed/${botUuid}/widget.js"><\/script>`;
+
+          return (
+            <EmbedCodeField
+              key={fieldKey}
+              label={extractI18nObject(config.label)}
+              description={
+                config.description
+                  ? extractI18nObject(config.description)
+                  : undefined
+              }
+              snippet={embedSnippet}
+            />
+          );
+        }
+
+        if (config.type === 'download-link') {
+          if (!config.url) return null;
+
+          return (
+            <DownloadLinkField
+              key={config.id}
+              label={extractI18nObject(config.label)}
+              description={
+                config.description
+                  ? extractI18nObject(config.description)
+                  : undefined
+              }
+              url={config.url}
+              filename={config.download_filename}
+              helpUrl={getAdapterDocUrl(config.help_links, i18n.language)}
+              helpLabel={
+                config.help_label
+                  ? extractI18nObject(config.help_label)
+                  : t('bots.viewAdapterDocs')
+              }
+            />
+          );
+        }
+
+        // QR code login button (e.g. Feishu one-click create, WeChat scan login)
+        if (config.type === 'qr-code-login') {
+          return (
+            <FormItem key={fieldKey}>
+              <div
+                className="relative flex items-center gap-4 p-4 rounded-xl border-2 border-dashed cursor-pointer transition-all hover:border-solid hover:shadow-md group"
+                style={{
+                  borderColor:
+                    'color-mix(in srgb, var(--primary) 25%, transparent)',
+                  background:
+                    'color-mix(in srgb, var(--primary) 3%, transparent)',
+                }}
+                onClick={() => {
+                  if (!isEditing) {
+                    setQrDialogPlatform(
+                      (config.login_platform as QrLoginPlatform | 'slack') ||
+                        'feishu',
+                    );
+                    setQrDialogOpen(true);
+                  }
+                }}
+              >
+                <div className="flex items-center justify-center h-12 w-12 rounded-lg bg-primary/10 shrink-0">
+                  <QrCode className="h-6 w-6 text-primary" />
+                </div>
+                <div className="flex-1 min-w-0">
+                  <div className="flex items-center gap-2">
+                    <span className="text-sm font-semibold text-foreground">
+                      {extractI18nObject(config.label)}
+                    </span>
+                    <span className="px-1.5 py-0.5 text-[10px] font-bold rounded bg-primary text-primary-foreground">
+                      {t('common.recommend')}
+                    </span>
+                  </div>
+                  {config.description && (
+                    <p className="text-xs text-muted-foreground mt-1 leading-relaxed">
+                      {extractI18nObject(config.description)}
+                    </p>
+                  )}
+                </div>
+                <Button
+                  type="button"
+                  size="sm"
+                  disabled={!!isEditing}
+                  className="shrink-0"
+                  onClick={(e) => {
+                    e.stopPropagation();
+                    setQrDialogPlatform(
+                      (config.login_platform as QrLoginPlatform | 'slack') ||
+                        'feishu',
+                    );
+                    setQrDialogOpen(true);
                   }}
                 >
-                  <div className="flex items-center justify-center h-12 w-12 rounded-lg bg-primary/10 shrink-0">
-                    <QrCode className="h-6 w-6 text-primary" />
-                  </div>
-                  <div className="flex-1 min-w-0">
-                    <div className="flex items-center gap-2">
-                      <span className="text-sm font-semibold text-foreground">
-                        {extractI18nObject(config.label)}
-                      </span>
-                      <span className="px-1.5 py-0.5 text-[10px] font-bold rounded bg-primary text-primary-foreground">
-                        {t('common.recommend')}
-                      </span>
-                    </div>
-                    {config.description && (
-                      <p className="text-xs text-muted-foreground mt-1 leading-relaxed">
-                        {extractI18nObject(config.description)}
-                      </p>
-                    )}
-                  </div>
-                  <Button
-                    type="button"
-                    size="sm"
-                    disabled={!!isEditing}
-                    className="shrink-0"
-                    onClick={(e) => {
-                      e.stopPropagation();
-                      setQrDialogPlatform(
-                        (config.login_platform as QrLoginPlatform) || 'feishu',
-                      );
-                      setQrDialogOpen(true);
-                    }}
-                  >
-                    <QrCode className="h-3.5 w-3.5 mr-1" />
-                    {t('common.start')}
-                  </Button>
-                </div>
-              </FormItem>
-            );
-          }
+                  <QrCode className="h-3.5 w-3.5 mr-1" />
+                  {t('common.start')}
+                </Button>
+              </div>
+            </FormItem>
+          );
+        }
 
-          if (
-            config.type === DynamicFormItemType.RICH_TOOLS_SELECTOR ||
-            config.type === DynamicFormItemType.RESOURCES_SELECTOR
-          ) {
-            return (
-              <FormField
-                key={config.id}
-                control={form.control}
-                name={config.name as keyof FormValues}
-                render={({ field }) => (
-                  <FormItem className="min-w-0">
-                    <FormControl>
-                      <div
-                        className={cn(
-                          'min-w-0 max-w-full overflow-x-hidden',
-                          isFieldDisabled && 'pointer-events-none opacity-60',
-                        )}
-                      >
-                        <DynamicFormItemComponent
-                          config={config}
-                          field={field}
-                          formValues={watchedValues as Record<string, unknown>}
-                          onFileUploaded={onFileUploaded}
-                          setFormValue={setFormValue}
-                          systemContext={systemContext}
-                        />
-                      </div>
-                    </FormControl>
-                    <FormMessage />
-                  </FormItem>
-                )}
-              />
-            );
-          }
-
-          // Boolean fields use a special inline layout
-          if (normalizedConfig.type === 'boolean') {
-            return (
-              <FormField
-                key={fieldKey}
-                control={form.control}
-                name={config.name as keyof FormValues}
-                render={({ field }) => (
-                  <FormItem className="min-w-0">
+        if (
+          config.type === DynamicFormItemType.RICH_TOOLS_SELECTOR ||
+          config.type === DynamicFormItemType.RESOURCES_SELECTOR
+        ) {
+          return (
+            <FormField
+              key={config.id}
+              control={form.control}
+              name={config.name as keyof FormValues}
+              render={({ field }) => (
+                <FormItem className="min-w-0">
+                  <FormControl>
                     <div
                       className={cn(
-                        'flex w-full min-w-0 max-w-full flex-row items-center justify-between rounded-lg border p-4',
+                        'min-w-0 max-w-full overflow-x-hidden',
                         isFieldDisabled && 'pointer-events-none opacity-60',
                       )}
                     >
-                      <div className="min-w-0 space-y-0.5">
-                        <FormLabel className="flex min-w-0 items-center gap-1.5 text-base">
-                          {extractI18nObject(config.label)}
-                          {renderDisabledTooltipIcon()}
-                        </FormLabel>
-                        {config.description && (
-                          <p className="text-sm break-words text-muted-foreground">
-                            {extractI18nObject(config.description)}
-                          </p>
-                        )}
-                      </div>
-                      <FormControl>
+                      <DynamicFormItemComponent
+                        config={config}
+                        field={field}
+                        formValues={watchedValues as Record<string, unknown>}
+                        onFileUploaded={onFileUploaded}
+                        setFormValue={setFormValue}
+                        systemContext={systemContext}
+                      />
+                    </div>
+                  </FormControl>
+                  <FormMessage />
+                </FormItem>
+              )}
+            />
+          );
+        }
+
+        // Boolean fields use a special inline layout
+        if (normalizedConfig.type === 'boolean') {
+          return (
+            <FormField
+              key={fieldKey}
+              control={form.control}
+              name={config.name as keyof FormValues}
+              render={({ field }) => (
+                <FormItem className="min-w-0">
+                  <div
+                    className={cn(
+                      'flex w-full min-w-0 max-w-full flex-row items-center justify-between rounded-lg border p-4',
+                      isFieldDisabled && 'pointer-events-none opacity-60',
+                    )}
+                  >
+                    <div className="min-w-0 space-y-0.5">
+                      <FormLabel className="flex min-w-0 items-center gap-1.5 text-base">
+                        {extractI18nObject(config.label)}
+                        {renderDisabledTooltipIcon()}
+                      </FormLabel>
+                      {config.description && (
+                        <p className="text-sm break-words text-muted-foreground">
+                          {extractI18nObject(config.description)}
+                        </p>
+                      )}
+                    </div>
+                    <FormControl>
+                      <DynamicFormItemComponent
+                        config={normalizedConfig}
+                        field={field}
+                        formValues={watchedValues as Record<string, unknown>}
+                        onFileUploaded={onFileUploaded}
+                        setFormValue={setFormValue}
+                        systemContext={systemContext}
+                      />
+                    </FormControl>
+                  </div>
+                  <FormMessage />
+                </FormItem>
+              )}
+            />
+          );
+        }
+
+        return (
+          <FormField
+            key={fieldKey}
+            control={form.control}
+            name={config.name as keyof FormValues}
+            render={({ field }) => {
+              const customItem = renderItem?.({
+                config: normalizedConfig,
+                field,
+                formValues: watchedValues as Record<string, unknown>,
+                setFormValue,
+              });
+              return (
+                <FormItem
+                  className={cn('min-w-0', isHiddenByCondition && 'hidden')}
+                  hidden={isHiddenByCondition}
+                >
+                  <FormLabel className="flex min-w-0 items-center gap-1.5">
+                    <span className="min-w-0 break-words">
+                      {extractI18nObject(config.label)}{' '}
+                      {config.required && (
+                        <span className="text-red-500">*</span>
+                      )}
+                    </span>
+                    {renderDisabledTooltipIcon()}
+                  </FormLabel>
+                  <FormControl>
+                    <div
+                      className={cn(
+                        'min-w-0 max-w-full overflow-x-hidden',
+                        isFieldDisabled && 'pointer-events-none opacity-60',
+                      )}
+                    >
+                      {customItem !== undefined ? (
+                        customItem
+                      ) : (
                         <DynamicFormItemComponent
                           config={normalizedConfig}
                           field={field}
@@ -943,77 +1142,21 @@ export default function DynamicFormComponent({
                           setFormValue={setFormValue}
                           systemContext={systemContext}
                         />
-                      </FormControl>
+                      )}
                     </div>
-                    <FormMessage />
-                  </FormItem>
-                )}
-              />
-            );
-          }
-
-          return (
-            <FormField
-              key={fieldKey}
-              control={form.control}
-              name={config.name as keyof FormValues}
-              render={({ field }) => {
-                const customItem = renderItem?.({
-                  config: normalizedConfig,
-                  field,
-                  formValues: watchedValues as Record<string, unknown>,
-                  setFormValue,
-                });
-                return (
-                  <FormItem
-                    className={cn('min-w-0', isHiddenByCondition && 'hidden')}
-                    hidden={isHiddenByCondition}
-                  >
-                    <FormLabel className="flex min-w-0 items-center gap-1.5">
-                      <span className="min-w-0 break-words">
-                        {extractI18nObject(config.label)}{' '}
-                        {config.required && (
-                          <span className="text-red-500">*</span>
-                        )}
-                      </span>
-                      {renderDisabledTooltipIcon()}
-                    </FormLabel>
-                    <FormControl>
-                      <div
-                        className={cn(
-                          'min-w-0 max-w-full overflow-x-hidden',
-                          isFieldDisabled && 'pointer-events-none opacity-60',
-                        )}
-                      >
-                        {customItem !== undefined ? (
-                          customItem
-                        ) : (
-                          <DynamicFormItemComponent
-                            config={normalizedConfig}
-                            field={field}
-                            formValues={
-                              watchedValues as Record<string, unknown>
-                            }
-                            onFileUploaded={onFileUploaded}
-                            setFormValue={setFormValue}
-                            systemContext={systemContext}
-                          />
-                        )}
-                      </div>
-                    </FormControl>
-                    {config.description && (
-                      <p className="text-sm break-words text-muted-foreground">
-                        {extractI18nObject(config.description)}
-                      </p>
-                    )}
-                    <FormMessage />
-                  </FormItem>
-                );
-              }}
-            />
-          );
-        })}
-      </div>
-    </Form>
+                  </FormControl>
+                  {config.description && (
+                    <p className="text-sm break-words text-muted-foreground">
+                      {extractI18nObject(config.description)}
+                    </p>
+                  )}
+                  <FormMessage />
+                </FormItem>
+              );
+            }}
+          />
+        );
+      })}
+    </div>
   );
 }
diff --git a/web/src/app/home/components/dynamic-form/DynamicFormItemComponent.tsx b/web/src/app/home/components/dynamic-form/DynamicFormItemComponent.tsx
index fd7811a02..3f14128c8 100644
--- a/web/src/app/home/components/dynamic-form/DynamicFormItemComponent.tsx
+++ b/web/src/app/home/components/dynamic-form/DynamicFormItemComponent.tsx
@@ -22,6 +22,8 @@ import { ControllerRenderProps } from 'react-hook-form';
 import { Button } from '@/components/ui/button';
 import { useEffect, useState } from 'react';
 import { httpClient, systemInfo, userInfo } from '@/app/infra/http';
+import { MODEL_CONFIGURATION_CHANGED } from '@/app/infra/http/modelConfigEvents';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 import {
   LLMModel,
   Bot,
@@ -94,19 +96,6 @@ function hasUsableOptionName(option: { name?: string | null }): boolean {
   return typeof option.name === 'string' && option.name.trim().length > 0;
 }
 
-function getPluginComponentIconURL(value?: string): string | null {
-  if (!value?.startsWith('plugin:')) {
-    return null;
-  }
-
-  const match = value.match(/^plugin:([^/]+)\/([^/]+)(?:\/|$)/);
-  if (!match) {
-    return null;
-  }
-
-  return httpClient.getPluginIconURL(match[1], match[2]);
-}
-
 function SelectOptionContent({
   label,
   value,
@@ -114,7 +103,10 @@ function SelectOptionContent({
   label: string;
   value: string;
 }) {
-  const iconURL = getPluginComponentIconURL(value);
+  const match = value?.startsWith('plugin:')
+    ? value.match(/^plugin:([^/]+)\/([^/]+)(?:\/|$)/)
+    : null;
+  const iconURL = useInstalledPluginIcon(match?.[1], match?.[2]);
 
   return (
     <div className="flex min-w-0 items-center gap-2">
@@ -141,6 +133,7 @@ export default function DynamicFormItemComponent({
   systemContext,
   requiredModelAbility,
   compactModelSelector = false,
+  selectedModelLabel,
 }: {
   config: IDynamicFormItemSchema;
   field: ControllerRenderProps<any, any>;
@@ -150,6 +143,7 @@ export default function DynamicFormItemComponent({
   systemContext?: Record<string, unknown>;
   requiredModelAbility?: string;
   compactModelSelector?: boolean;
+  selectedModelLabel?: string;
 }) {
   const [llmModels, setLlmModels] = useState<LLMModel[]>([]);
   const [embeddingModels, setEmbeddingModels] = useState<EmbeddingModel[]>([]);
@@ -313,6 +307,10 @@ export default function DynamicFormItemComponent({
   useEffect(() => {
     if (config.type === DynamicFormItemType.LLM_MODEL_SELECTOR) {
       fetchLlmModels();
+      window.addEventListener(MODEL_CONFIGURATION_CHANGED, fetchLlmModels);
+      return () => {
+        window.removeEventListener(MODEL_CONFIGURATION_CHANGED, fetchLlmModels);
+      };
     }
   }, [config.type]);
 
@@ -648,11 +646,25 @@ export default function DynamicFormItemComponent({
                 aria-label={t('models.selectModel')}
                 className={
                   compactModelSelector
-                    ? 'w-full min-w-0 gap-1 border-0 bg-transparent px-1 text-xs text-muted-foreground shadow-none hover:bg-muted data-[size=default]:h-7 [&_[data-slot=select-value]_svg]:hidden'
+                    ? 'w-full min-w-0 gap-1 rounded-lg border border-border bg-background px-2 text-xs text-foreground shadow-[0_1px_2px_rgba(0,0,0,0.04)] hover:border-muted-foreground/40 hover:bg-background dark:bg-background dark:hover:bg-background data-[state=open]:border-muted-foreground/50 focus-visible:border-muted-foreground/50 focus-visible:ring-1 focus-visible:ring-muted-foreground/20 data-[size=default]:h-7 [&_[data-slot=select-value]_svg]:hidden'
                     : MODEL_SELECT_TRIGGER_CLASS
                 }
               >
-                <SelectValue placeholder={t('models.selectModel')} />
+                <SelectValue placeholder={t('models.selectModel')}>
+                  {/* Keep compact labels explicitly rendered even when a stream
+                      omits the model name. Switching to Radix's ItemText portal
+                      would give two renderers ownership of the same DOM node. */}
+                  {compactModelSelector ? (
+                    <span className="truncate">
+                      {selectedModelLabel ||
+                        selectableModels.find(
+                          (model) => model.uuid === field.value,
+                        )?.name ||
+                        field.value ||
+                        t('models.selectModel')}
+                    </span>
+                  ) : undefined}
+                </SelectValue>
               </SelectTrigger>
               <SelectContent>
                 {Object.entries(groupedModels).map(([providerName, models]) => (
diff --git a/web/src/app/home/components/dynamic-form/StructuredFieldValue.ts b/web/src/app/home/components/dynamic-form/StructuredFieldValue.ts
index 09a4c34e0..6e4415452 100644
--- a/web/src/app/home/components/dynamic-form/StructuredFieldValue.ts
+++ b/web/src/app/home/components/dynamic-form/StructuredFieldValue.ts
@@ -1,10 +1,5 @@
 export type JsonValue =
-  | null
-  | boolean
-  | number
-  | string
-  | JsonValue[]
-  | { [key: string]: JsonValue };
+  null | boolean | number | string | JsonValue[] | { [key: string]: JsonValue };
 export type PromptMessage = {
   role: string;
   content?: string | null | JsonValue[];
diff --git a/web/src/app/home/components/dynamic-form/ToolResourceSelectors.tsx b/web/src/app/home/components/dynamic-form/ToolResourceSelectors.tsx
index df729d634..4b06b991c 100644
--- a/web/src/app/home/components/dynamic-form/ToolResourceSelectors.tsx
+++ b/web/src/app/home/components/dynamic-form/ToolResourceSelectors.tsx
@@ -204,19 +204,17 @@ function buildToolSourceGroups(
         (rightIndex === -1 ? TOOL_SOURCE_ORDER.length : rightIndex)
       );
     })
-    .map(
-      ([source, providerGroups]): ToolSourceGroup => ({
-        key: source,
-        label: sourceLabels[source] || source,
-        groups: Array.from(providerGroups.values()).sort((left, right) =>
-          left.name.localeCompare(right.name),
-        ),
-        total: Array.from(providerGroups.values()).reduce(
-          (count, group) => count + group.tools.length,
-          0,
-        ),
-      }),
-    );
+    .map(([source, providerGroups]): ToolSourceGroup => ({
+      key: source,
+      label: sourceLabels[source] || source,
+      groups: Array.from(providerGroups.values()).sort((left, right) =>
+        left.name.localeCompare(right.name),
+      ),
+      total: Array.from(providerGroups.values()).reduce(
+        (count, group) => count + group.tools.length,
+        0,
+      ),
+    }));
 }
 
 function buildMCPToolOwners(servers: MCPServer[]) {
diff --git a/web/src/app/home/components/guided-tour/GuidedTour.tsx b/web/src/app/home/components/guided-tour/GuidedTour.tsx
index 7ffdc6ab0..436d61aed 100644
--- a/web/src/app/home/components/guided-tour/GuidedTour.tsx
+++ b/web/src/app/home/components/guided-tour/GuidedTour.tsx
@@ -57,6 +57,7 @@ export default function GuidedTour({
     useState<PopoverPosition | null>(null);
   const popoverRef = useRef<HTMLDivElement | null>(null);
   const previousStorageKeyRef = useRef(storageKey);
+  const enteredStepRef = useRef<string | null>(null);
 
   const activeIndex = steps.findIndex((step) => step.id === activeStepId);
   const activeStep = activeIndex >= 0 ? steps[activeIndex] : undefined;
@@ -86,8 +87,17 @@ export default function GuidedTour({
   }, [activeStepId, enabled, finished, steps, storageKey]);
 
   useEffect(() => {
-    if (enabled && !finished) activeStep?.onEnter?.();
-  }, [activeStep, enabled, finished]);
+    if (!enabled || finished || !activeStep) {
+      enteredStepRef.current = null;
+      return;
+    }
+    // Callback identities can change during navigation. Enter each step once,
+    // so a URL update cannot replay its navigation and override the user's tab.
+    const stepKey = JSON.stringify([storageKey, activeStep.id]);
+    if (enteredStepRef.current === stepKey) return;
+    enteredStepRef.current = stepKey;
+    activeStep.onEnter?.();
+  }, [activeStep, enabled, finished, storageKey]);
 
   const measure = useCallback(() => {
     if (!activeStep) return;
diff --git a/web/src/app/home/components/guided-tour/TourOverlay.tsx b/web/src/app/home/components/guided-tour/TourOverlay.tsx
index e1a05476f..a52b542a9 100644
--- a/web/src/app/home/components/guided-tour/TourOverlay.tsx
+++ b/web/src/app/home/components/guided-tour/TourOverlay.tsx
@@ -45,6 +45,12 @@ export function getTourPosition(
         ? rect.bottom + VIEWPORT_GAP
         : rect.top - height - VIEWPORT_GAP;
   }
+  // The mobile sidebar can still be sliding in when its target is measured.
+  // Clamp the card independently of that temporarily off-screen target.
+  left = Math.max(
+    VIEWPORT_GAP,
+    Math.min(left, viewportWidth - width - VIEWPORT_GAP),
+  );
   top = Math.max(
     VIEWPORT_GAP,
     Math.min(top, viewportHeight - height - VIEWPORT_GAP),
diff --git a/web/src/app/home/components/home-sidebar/HomeSidebar.tsx b/web/src/app/home/components/home-sidebar/HomeSidebar.tsx
index 230d13c2b..55d3326d6 100644
--- a/web/src/app/home/components/home-sidebar/HomeSidebar.tsx
+++ b/web/src/app/home/components/home-sidebar/HomeSidebar.tsx
@@ -655,6 +655,16 @@ function NavItems({
           if (item.extensionType === 'skill') {
             return `/home/skills?id=${encodeURIComponent(item.id)}`;
           }
+          const processorTab = searchParams.get('tab');
+          if (
+            hasDetailPages &&
+            (routePrefix === '/home/agents' ||
+              routePrefix === '/home/pipelines') &&
+            (pathname === '/home/agents' || pathname === '/home/pipelines') &&
+            (processorTab === 'logs' || processorTab === 'config')
+          ) {
+            return `${routePrefix}?id=${encodeURIComponent(item.id)}&tab=${processorTab}`;
+          }
           return hasDetailPages
             ? `${routePrefix}?id=${encodeURIComponent(item.id)}`
             : routePrefix;
@@ -794,7 +804,7 @@ function NavItems({
                       {(isBot || isMCP) && (
                         <span
                           className={cn(
-                            'absolute -bottom-0.5 -right-0.5 size-2 rounded-full border-2 border-popover',
+                            'absolute -bottom-1 -right-1 size-3 rounded-full border-2 border-popover',
                             isMCP
                               ? mcpStatusColor(item)
                               : item.enabled === false
@@ -862,7 +872,7 @@ function NavItems({
                             {(isBot || isMCP) && (
                               <span
                                 className={cn(
-                                  'absolute -bottom-0.5 -right-0.5 size-2 rounded-full border-2 border-sidebar',
+                                  'absolute -bottom-1 -right-1 size-3 rounded-full border-2 border-sidebar',
                                   isMCP
                                     ? mcpStatusColor(item)
                                     : item.enabled === false
@@ -1871,6 +1881,9 @@ export default function HomeSidebar({
   const canViewStorageAnalysis =
     currentWorkspace?.workspace.source !== 'cloud_projection' &&
     currentWorkspace?.permissions.includes('audit.view');
+  // The operation log is an audit surface: owner / admin only.
+  const canViewOperationTrace =
+    currentWorkspace?.permissions.includes('audit.view') ?? false;
   const [settingsOpen, setSettingsOpen] = useState(false);
   const [settingsSection, setSettingsSection] =
     useState<SettingsSection>('models');
@@ -2221,19 +2234,8 @@ export default function HomeSidebar({
                 <span>{t('models.title')}</span>
               </SidebarMenuButton>
             </SidebarMenuItem>
-            {/* Operation traceability is an audit surface: only roles holding
-                audit.view (owner / admin) get a direct sidebar entry. */}
-            {currentWorkspace?.permissions.includes('audit.view') && (
-              <SidebarMenuItem data-sidebar-guide="operation-trace">
-                <SidebarMenuButton
-                  onClick={() => openSettings('operationTrace')}
-                  tooltip={t('operationTrace.title')}
-                >
-                  <History className="size-4 text-blue-500" />
-                  <span>{t('operationTrace.title')}</span>
-                </SidebarMenuButton>
-              </SidebarMenuItem>
-            )}
+            {/* Operation log entry moved to the account menu: it is an audit
+                surface, not a primary navigation target. */}
           </SidebarMenu>
 
           {/* API-key management is available only to authorized Workspace roles. */}
@@ -2358,6 +2360,17 @@ export default function HomeSidebar({
                         {t('storageAnalysis.title')}
                       </DropdownMenuItem>
                     )}
+                    {canViewOperationTrace && (
+                      <DropdownMenuItem
+                        onClick={() => {
+                          setUserMenuOpen(false);
+                          openSettings('operationTrace');
+                        }}
+                      >
+                        <History />
+                        {t('operationTrace.title')}
+                      </DropdownMenuItem>
+                    )}
                     <DropdownMenuItem
                       onClick={() => {
                         setUserMenuOpen(false);
diff --git a/web/src/app/home/components/home-sidebar/SidebarDataContext.tsx b/web/src/app/home/components/home-sidebar/SidebarDataContext.tsx
index 871d4d7b4..aeef4895d 100644
--- a/web/src/app/home/components/home-sidebar/SidebarDataContext.tsx
+++ b/web/src/app/home/components/home-sidebar/SidebarDataContext.tsx
@@ -1,81 +1,20 @@
-import React, {
-  createContext,
-  useContext,
-  useState,
-  useEffect,
-  useCallback,
-  useRef,
-} from 'react';
+import React, { useState, useEffect, useCallback, useRef } from 'react';
 import { httpClient, getCloudServiceClientSync } from '@/app/infra/http';
 import { extractI18nObject } from '@/i18n/I18nProvider';
 import { isNewerVersion } from '@/app/utils/versionCompare';
 
-// Lightweight entity item for sidebar display
-export interface SidebarEntityItem {
-  id: string;
-  name: string;
-  description?: string;
-  emoji?: string;
-  iconURL?: string;
-  updatedAt?: string; // ISO timestamp for sorting by most recently edited
-  // Bot-specific fields
-  enabled?: boolean;
-  legacyAdapter?: boolean;
-  // MCP-specific fields
-  runtimeStatus?: 'connecting' | 'connected' | 'error';
-  // Plugin-specific fields
-  installSource?: string;
-  installInfo?: Record<string, unknown>;
-  hasUpdate?: boolean;
-  debug?: boolean;
-  // Set when this item appears in the unified extensions list
-  extensionType?: 'plugin' | 'mcp' | 'skill';
-  // Agent-specific: distinguishes Agent processors from Pipelines
-  kind?: 'agent' | 'pipeline' | 'event_processor';
-}
-
-// Plugin page registered by a plugin
-export interface PluginPageItem {
-  id: string; // "author/name/pageId"
-  name: string; // display label
-  pluginAuthor: string;
-  pluginName: string;
-  pluginLabel: string; // human-readable plugin display name
-  pluginIconURL: string; // plugin icon URL
-  pageId: string;
-  path: string; // asset path (HTML file)
-}
-
-// Entity lists and refresh functions exposed via context
-export interface SidebarDataContextValue {
-  bots: SidebarEntityItem[];
-  pipelines: SidebarEntityItem[];
-  knowledgeBases: SidebarEntityItem[];
-  plugins: SidebarEntityItem[];
-  pluginCount: number;
-  mcpServers: SidebarEntityItem[];
-  skills: SidebarEntityItem[];
-  pluginPages: PluginPageItem[];
-  quotaDataLoaded: boolean;
-  refreshBots: () => Promise<void>;
-  refreshPipelines: () => Promise<void>;
-  refreshKnowledgeBases: () => Promise<void>;
-  refreshPlugins: () => Promise<void>;
-  refreshMCPServers: () => Promise<void>;
-  refreshSkills: () => Promise<void>;
-  refreshAll: () => Promise<void>;
-  // Breadcrumb: entity name shown when viewing a detail page
-  detailEntityName: string | null;
-  setDetailEntityName: (name: string | null) => void;
-  // Whether the extensions list is grouped by type (shared between page and sidebar)
-  extensionsGroupByType: boolean;
-  setExtensionsGroupByType: (enabled: boolean) => void;
-  // Whether the Agent list is grouped by kind (Agent vs Pipeline)
-  agentsGroupByKind: boolean;
-  setAgentsGroupByKind: (enabled: boolean) => void;
-}
-
-const SidebarDataContext = createContext<SidebarDataContextValue | null>(null);
+// Keep the context identity independent of provider dependencies during HMR.
+import {
+  SidebarDataContext,
+  type SidebarEntityItem,
+  type PluginPageItem,
+} from './sidebarDataStore';
+export { useSidebarData } from './sidebarDataStore';
+export type {
+  SidebarEntityItem,
+  PluginPageItem,
+  SidebarDataContextValue,
+} from './sidebarDataStore';
 
 export function SidebarDataProvider({
   children,
@@ -432,11 +371,3 @@ export function SidebarDataProvider({
     </SidebarDataContext.Provider>
   );
 }
-
-export function useSidebarData(): SidebarDataContextValue {
-  const ctx = useContext(SidebarDataContext);
-  if (!ctx) {
-    throw new Error('useSidebarData must be used within a SidebarDataProvider');
-  }
-  return ctx;
-}
diff --git a/web/src/app/home/components/home-sidebar/sidebarDataStore.ts b/web/src/app/home/components/home-sidebar/sidebarDataStore.ts
new file mode 100644
index 000000000..bba5e02f7
--- /dev/null
+++ b/web/src/app/home/components/home-sidebar/sidebarDataStore.ts
@@ -0,0 +1,78 @@
+import { createContext, useContext } from 'react';
+
+// Lightweight entity item for sidebar display
+export interface SidebarEntityItem {
+  id: string;
+  name: string;
+  description?: string;
+  emoji?: string;
+  iconURL?: string;
+  updatedAt?: string; // ISO timestamp for sorting by most recently edited
+  // Bot-specific fields
+  enabled?: boolean;
+  legacyAdapter?: boolean;
+  // MCP-specific fields
+  runtimeStatus?: 'connecting' | 'connected' | 'error';
+  // Plugin-specific fields
+  installSource?: string;
+  installInfo?: Record<string, unknown>;
+  hasUpdate?: boolean;
+  debug?: boolean;
+  // Set when this item appears in the unified extensions list
+  extensionType?: 'plugin' | 'mcp' | 'skill';
+  // Agent-specific: distinguishes Agent processors from Pipelines
+  kind?: 'agent' | 'pipeline' | 'event_processor';
+}
+
+// Plugin page registered by a plugin
+export interface PluginPageItem {
+  id: string; // "author/name/pageId"
+  name: string; // display label
+  pluginAuthor: string;
+  pluginName: string;
+  pluginLabel: string; // human-readable plugin display name
+  pluginIconURL: string; // plugin icon URL
+  pageId: string;
+  path: string; // asset path (HTML file)
+}
+
+// Entity lists and refresh functions exposed via context
+export interface SidebarDataContextValue {
+  bots: SidebarEntityItem[];
+  pipelines: SidebarEntityItem[];
+  knowledgeBases: SidebarEntityItem[];
+  plugins: SidebarEntityItem[];
+  pluginCount: number;
+  mcpServers: SidebarEntityItem[];
+  skills: SidebarEntityItem[];
+  pluginPages: PluginPageItem[];
+  quotaDataLoaded: boolean;
+  refreshBots: () => Promise<void>;
+  refreshPipelines: () => Promise<void>;
+  refreshKnowledgeBases: () => Promise<void>;
+  refreshPlugins: () => Promise<void>;
+  refreshMCPServers: () => Promise<void>;
+  refreshSkills: () => Promise<void>;
+  refreshAll: () => Promise<void>;
+  // Breadcrumb: entity name shown when viewing a detail page
+  detailEntityName: string | null;
+  setDetailEntityName: (name: string | null) => void;
+  // Whether the extensions list is grouped by type (shared between page and sidebar)
+  extensionsGroupByType: boolean;
+  setExtensionsGroupByType: (enabled: boolean) => void;
+  // Whether the Agent list is grouped by kind (Agent vs Pipeline)
+  agentsGroupByKind: boolean;
+  setAgentsGroupByKind: (enabled: boolean) => void;
+}
+
+export const SidebarDataContext = createContext<SidebarDataContextValue | null>(
+  null,
+);
+
+export function useSidebarData(): SidebarDataContextValue {
+  const ctx = useContext(SidebarDataContext);
+  if (!ctx) {
+    throw new Error('useSidebarData must be used within a SidebarDataProvider');
+  }
+  return ctx;
+}
diff --git a/web/src/app/home/components/inflight/DockIcon.tsx b/web/src/app/home/components/inflight/DockIcon.tsx
new file mode 100644
index 000000000..b58acbaa7
--- /dev/null
+++ b/web/src/app/home/components/inflight/DockIcon.tsx
@@ -0,0 +1,32 @@
+export default function DockIcon({ kind }: { kind: string }) {
+  const paths: Record<string, string> = {
+    event: 'M13 2 4 14h7l-1 8 10-12h-7l0-8z',
+    activity: 'M3 12h4l3-8 4 16 3-8h4',
+    bot: 'M8 4h8v3H8z M5 7h14v13H5z M9 12h.01 M15 12h.01 M9 16h6',
+    pipeline: 'M4 3h6v6H4z M14 15h6v6h-6z M7 9v9h7',
+    agent: 'M12 3a4 4 0 1 0 0 8 4 4 0 0 0 0-8 M4 21v-2a8 8 0 0 1 16 0v2',
+    processor: 'M9 3h6v6h6v6h-6v6H9v-6H3V9h6z',
+    completed: 'm5 12 4 4L19 6',
+    failed: 'm6 6 12 12 M6 18 18 6',
+    cancelled: 'm6 6 12 12 M6 18 18 6',
+    queued: 'M8 5v14 M16 5v14',
+    waiting: 'M8 5v14 M16 5v14',
+    move: 'm8 6 4-4 4 4 M12 2v20 m-4-4 4 4 4-4',
+    collapse: 'm9 5 7 7-7 7',
+  };
+  return (
+    <svg
+      width="16"
+      height="16"
+      viewBox="0 0 24 24"
+      fill="none"
+      stroke="currentColor"
+      strokeWidth="1.7"
+      strokeLinecap="round"
+      strokeLinejoin="round"
+      aria-hidden="true"
+    >
+      <path d={paths[kind] || paths.activity} />
+    </svg>
+  );
+}
diff --git a/web/src/app/home/components/inflight/ExecutionNodes.tsx b/web/src/app/home/components/inflight/ExecutionNodes.tsx
new file mode 100644
index 000000000..d37ba8bf5
--- /dev/null
+++ b/web/src/app/home/components/inflight/ExecutionNodes.tsx
@@ -0,0 +1,84 @@
+import { Link } from 'react-router-dom';
+import { useTranslation } from 'react-i18next';
+import { Avatar, AvatarFallback, AvatarImage } from '@/components/ui/avatar';
+import { useSidebarData } from '../home-sidebar/SidebarDataContext';
+import type { ExecutionRow } from '@/app/infra/entities/api/monitoring-executions';
+import { executionProcessorLabel } from '../../monitoring/components/executions/ExecutionTable';
+import DockIcon from './DockIcon';
+import styles from './inflight.module.css';
+
+export default function ExecutionNodes({
+  row,
+  processorUrl,
+  running,
+}: {
+  row: ExecutionRow;
+  processorUrl: string | null;
+  running: boolean;
+}) {
+  const { t } = useTranslation();
+  const { bots, pipelines } = useSidebarData();
+  const bot = bots.find((item) => item.id === row.bot_id);
+  const processor = pipelines.find((item) => item.id === row.target_id);
+  const nodes = [
+    {
+      name: bot?.name || row.bot_name || t('inflight.debug'),
+      iconURL: bot?.iconURL,
+      emoji: bot?.emoji,
+      kind: 'bot',
+      url: row.bot_id
+        ? `/home/bots?id=${encodeURIComponent(row.bot_id)}&tab=logs`
+        : null,
+    },
+    {
+      name: processor?.name || executionProcessorLabel(row),
+      iconURL: processor?.iconURL,
+      emoji: processor?.emoji,
+      kind:
+        row.target_kind === 'event_processor' ? 'processor' : row.target_kind,
+      url: processorUrl,
+    },
+  ];
+  const node = (index: number) => {
+    const item = nodes[index];
+    const content = (
+      <>
+        <Avatar className="size-5 shrink-0 rounded">
+          {item.iconURL && (
+            <AvatarImage src={item.iconURL} alt="" className="object-contain" />
+          )}
+          <AvatarFallback className="rounded bg-transparent text-base [&_svg]:size-4">
+            {item.emoji || <DockIcon kind={item.kind} />}
+          </AvatarFallback>
+        </Avatar>
+        <span className="truncate text-[11px]">{item.name}</span>
+      </>
+    );
+    const className =
+      'flex h-8 min-w-0 max-w-[125px] items-center gap-1.5 rounded-md bg-muted/40 px-1.5';
+    return item.url ? (
+      <Link
+        to={item.url}
+        title={`${index === 0 ? t('inflight.botLogs') : t('inflight.processorLogs')}: ${item.name}`}
+        aria-label={`${index === 0 ? t('inflight.botLogs') : t('inflight.processorLogs')}: ${item.name}`}
+        className={`${className} hover:bg-muted hover:text-primary focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring`}
+      >
+        {content}
+      </Link>
+    ) : (
+      <div className={className}>{content}</div>
+    );
+  };
+  return (
+    <div className="mt-1.5 flex items-center">
+      {node(0)}
+      <span
+        className={`w-5 shrink-0 ${styles.connection} ${running ? styles.flowing : ''}`}
+        aria-hidden="true"
+      >
+        <i />
+      </span>
+      {node(1)}
+    </div>
+  );
+}
diff --git a/web/src/app/home/components/inflight/InFlightMonitor.tsx b/web/src/app/home/components/inflight/InFlightMonitor.tsx
new file mode 100644
index 000000000..3e831ee23
--- /dev/null
+++ b/web/src/app/home/components/inflight/InFlightMonitor.tsx
@@ -0,0 +1,363 @@
+import { useEffect, useRef, useState } from 'react';
+import { useTranslation } from 'react-i18next';
+import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
+import type { InflightSnapshot } from '@/app/infra/entities/api/monitoring-executions';
+import { Button } from '@/components/ui/button';
+import ExecutionDetailSheet, {
+  type ExecutionSelection,
+} from '@/app/home/monitoring/components/executions/ExecutionDetailSheet';
+import { executionStatusLabel } from '@/app/home/monitoring/components/executions/ExecutionTable';
+import styles from './inflight.module.css';
+import ExecutionNodes from './ExecutionNodes';
+import { useEdgeDock } from './useEdgeDock';
+import DockIcon from './DockIcon';
+
+type Row = InflightSnapshot['items'][number] & { settledAt?: number };
+const keyOf = (row: Row) => `${row.source}:${row.id}`;
+const active = (row: Row) => ['running', 'queued'].includes(row.status_group);
+
+function stage(row: Row): string {
+  if (!active(row)) return row.status_group;
+  if (row.status_group === 'queued') return 'queued';
+  const event = row.progress_event || '';
+  if (event.startsWith('interaction.')) return 'waiting';
+  if (event.startsWith('tool.') && !event.endsWith('completed')) return 'tool';
+  if (event.includes('reasoning')) return 'thinking';
+  if (event === 'message.completed') return 'replying';
+  if (event.startsWith('message.')) return 'generating';
+  return 'running';
+}
+
+function WorkspaceMonitor() {
+  const { t } = useTranslation();
+  const [rows, setRows] = useState<Row[]>([]);
+  const dock = useEdgeDock(true);
+  const { collapsed, setCollapsed } = dock;
+  const [connected, setConnected] = useState(true);
+  const [total, setTotal] = useState(0);
+  const [truncated, setTruncated] = useState(false);
+  const [selected, setSelected] = useState<ExecutionSelection | null>(null);
+  const [now, setNow] = useState(Date.now());
+  const known = useRef(new Map<string, Row>());
+  const [autoCollapse, setAutoCollapse] = useState(() => {
+    try {
+      return localStorage.getItem('langbot.inflight.autoCollapse') !== 'false';
+    } catch {
+      return true;
+    }
+  });
+  // Keep terminal results visible until their exit animation has finished.
+  const hasWork = total > 0 || rows.length > 0;
+  useEffect(() => {
+    try {
+      localStorage.setItem(
+        'langbot.inflight.autoCollapse',
+        String(autoCollapse),
+      );
+    } catch {
+      /* Optional preference. */
+    }
+  }, [autoCollapse]);
+  useEffect(() => {
+    if (autoCollapse) setCollapsed(!hasWork);
+  }, [autoCollapse, hasWork, setCollapsed]);
+
+  useEffect(() => {
+    let disposed = false;
+    let first = true;
+    let controller: AbortController | undefined;
+    let retry: ReturnType<typeof setTimeout> | undefined;
+    let failures = 0;
+    const seen = new Set<string>();
+    const accept = (snapshot: InflightSnapshot) => {
+      const time = Date.now();
+      const present = new Set(snapshot.items.map(keyOf));
+      // A reconnect is a fresh authoritative snapshot. Do not leave vanished
+      // executions spinning forever or infer successful completion from absence.
+      if (!snapshot.truncated) {
+        for (const [key, row] of known.current) {
+          if (active(row) && !present.has(key)) known.current.delete(key);
+        }
+      }
+      for (const row of snapshot.items) {
+        const key = keyOf(row);
+        const previous = known.current.get(key);
+        if (active(row) || previous || (!first && !seen.has(key))) {
+          known.current.set(key, {
+            ...row,
+            settledAt: active(row) ? undefined : (previous?.settledAt ?? time),
+          });
+        }
+        seen.add(key);
+      }
+      // Retain terminal results briefly, including runs completed between frames.
+      for (const [key, row] of known.current) {
+        if (row.settledAt && time - row.settledAt > 8000)
+          known.current.delete(key);
+      }
+      while (seen.size > 2000) seen.delete(seen.values().next().value!);
+      while (known.current.size > 200)
+        known.current.delete(known.current.keys().next().value!);
+      first = false;
+      setRows([...known.current.values()]);
+      setTotal(snapshot.active_total);
+      setTruncated(snapshot.truncated);
+    };
+    const connect = async () => {
+      if (disposed || document.hidden) return;
+      controller = new AbortController();
+      const connection = controller;
+      try {
+        await backendClient.streamInflight((frame) => {
+          if (disposed || connection.signal.aborted) return;
+          failures = 0;
+          if (frame.kind !== 'heartbeat')
+            setConnected(frame.kind !== 'unavailable');
+          if (frame.kind === 'snapshot') accept(frame.data);
+        }, connection.signal);
+      } catch {
+        if (!disposed && !connection.signal.aborted) {
+          setConnected(false);
+          failures += 1;
+        }
+      }
+      if (!disposed && !connection.signal.aborted && !document.hidden) {
+        retry = setTimeout(
+          connect,
+          Math.min(30000, 1000 * 2 ** failures) + Math.random() * 500,
+        );
+      }
+    };
+    const visibility = () => {
+      clearTimeout(retry);
+      controller?.abort();
+      if (!document.hidden) retry = setTimeout(connect, 100);
+    };
+    void connect();
+    document.addEventListener('visibilitychange', visibility);
+    const tick = setInterval(() => {
+      const time = Date.now();
+      let changed = false;
+      for (const [key, row] of known.current) {
+        if (row.settledAt && time - row.settledAt > 8000) {
+          known.current.delete(key);
+          changed = true;
+        }
+      }
+      if (changed) setRows([...known.current.values()]);
+      if (known.current.size) setNow(time);
+    }, 1000);
+    return () => {
+      disposed = true;
+      controller?.abort();
+      clearTimeout(retry);
+      clearInterval(tick);
+      document.removeEventListener('visibilitychange', visibility);
+    };
+  }, []);
+
+  const ordered = [...rows].sort(
+    (a, b) =>
+      Number(active(b)) - Number(active(a)) ||
+      (b.created_at_ms ?? 0) - (a.created_at_ms ?? 0),
+  );
+  return (
+    <>
+      <aside
+        ref={dock.panel}
+        style={{ width: collapsed ? 26 : 'min(320px, calc(100vw - 24px))' }}
+        className={`fixed top-0 right-0 z-40 max-h-[calc(100dvh-24px)] ${collapsed ? 'overflow-hidden' : 'overflow-y-auto'} rounded-l-xl border border-r-0 bg-popover text-popover-foreground shadow-lg ${styles.panel}`}
+        aria-label={t('inflight.title')}
+      >
+        <div
+          key={collapsed ? 'collapsed' : 'expanded'}
+          className={styles.dockContent}
+        >
+          {collapsed ? (
+            <button
+              type="button"
+              {...dock.handle}
+              className="flex h-12 w-full touch-none select-none flex-col items-center justify-center gap-0.5 bg-[#2288ee] text-white hover:bg-[#2277e0] focus-visible:ring-2 focus-visible:ring-inset focus-visible:ring-ring active:cursor-grabbing"
+              title={`${t('inflight.expand')} · ${t('inflight.move')}`}
+              aria-label={`${t('inflight.title')}: ${total}. ${t('inflight.expand')}. ${t('inflight.move')}`}
+              aria-expanded={false}
+              onClick={(event) => {
+                if (event.detail === 0 || !dock.suppressClick.current)
+                  setCollapsed(false);
+              }}
+            >
+              <DockIcon kind="activity" />
+              <span
+                className={`text-[10px] font-mono ${connected ? '' : 'text-amber-100'}`}
+              >
+                {total}
+              </span>
+            </button>
+          ) : (
+            <div className="flex items-center justify-between gap-2 px-3 py-2">
+              <button
+                type="button"
+                {...dock.handle}
+                aria-label={t('inflight.move')}
+                title={t('inflight.move')}
+                className="flex min-w-0 flex-1 touch-none select-none items-center gap-2 rounded py-1 text-xs font-medium cursor-grab active:cursor-grabbing focus-visible:ring-2 focus-visible:ring-ring"
+              >
+                <span className={connected ? 'text-primary' : 'text-amber-600'}>
+                  <DockIcon kind="activity" />
+                </span>
+                {t('inflight.title')}{' '}
+                <span className="font-mono text-muted-foreground">{total}</span>
+                <span className="ml-auto text-muted-foreground">
+                  <DockIcon kind="move" />
+                </span>
+              </button>
+              <Button
+                variant="ghost"
+                size="sm"
+                className={`h-7 shrink-0 gap-1 px-2 text-[10px] ${autoCollapse ? 'bg-primary/10 text-primary hover:bg-primary/15' : 'text-muted-foreground'}`}
+                aria-pressed={autoCollapse}
+                title={t('inflight.autoCollapse')}
+                onClick={() => setAutoCollapse((enabled) => !enabled)}
+              >
+                {autoCollapse && <DockIcon kind="completed" />}
+                {t('inflight.autoCollapse')}
+              </Button>
+              <Button
+                variant="ghost"
+                size="icon"
+                className="size-8"
+                aria-label={t('inflight.collapse')}
+                title={t('inflight.collapse')}
+                aria-expanded={true}
+                onClick={() => setCollapsed(true)}
+              >
+                <DockIcon kind="collapse" />
+              </Button>
+            </div>
+          )}
+          {!collapsed && (
+            <>
+              {!connected && (
+                <p
+                  className="border-t px-4 py-2 text-xs text-amber-600"
+                  role="status"
+                >
+                  {t('inflight.reconnecting')}
+                </p>
+              )}
+              <div className="max-h-[min(420px,55dvh)] overflow-y-auto border-t">
+                {rows.length === 0 && connected && (
+                  <div
+                    className="flex flex-col items-center gap-2 px-4 py-8 text-muted-foreground"
+                    role="status"
+                  >
+                    <DockIcon kind="completed" />
+                    <p className="text-xs">{t('inflight.empty')}</p>
+                  </div>
+                )}
+                {ordered.slice(0, 20).map((row) => {
+                  const phase = stage(row);
+                  const stageKey = `inflight.${phase}`;
+                  const label = t(stageKey, {
+                    defaultValue: executionStatusLabel(row, t),
+                  });
+                  const elapsed = Math.max(
+                    0,
+                    Math.floor(
+                      ((row.finished_at_ms ?? row.settledAt ?? now) -
+                        (row.started_at_ms ?? row.created_at_ms ?? now)) /
+                        1000,
+                    ),
+                  );
+                  const eventKey = `bots.eventNames.${(row.event_type || row.title).replaceAll('.', '_')}`;
+                  const processorUrl =
+                    row.target_id &&
+                    [
+                      'pipeline',
+                      'agent',
+                      'processor',
+                      'event_processor',
+                    ].includes(row.target_kind)
+                      ? `/home/${row.target_kind === 'pipeline' ? 'pipelines' : 'agents'}?id=${encodeURIComponent(row.target_id)}&tab=logs`
+                      : null;
+                  const leaving =
+                    !!row.settledAt && now - row.settledAt >= 7000;
+                  return (
+                    <div
+                      key={keyOf(row)}
+                      className={`${styles.task} ${leaving ? styles.leaving : ''} border-b px-3 py-2 last:border-b-0`}
+                    >
+                      <div className="flex items-center gap-2">
+                        <span
+                          className={
+                            row.has_error ? 'text-destructive' : 'text-primary'
+                          }
+                          title={label}
+                          role="img"
+                          aria-label={label}
+                        >
+                          <DockIcon kind={active(row) ? 'event' : phase} />
+                        </span>
+                        <button
+                          type="button"
+                          onClick={() => setSelected(row)}
+                          className="min-w-0 flex-1 truncate rounded py-1 text-left text-xs hover:underline focus-visible:ring-2 focus-visible:ring-ring"
+                          title={`${t('inflight.details')}: ${row.input_preview || row.event_type || row.title}`}
+                          aria-label={`${t('inflight.details')}: ${row.event_type || row.title}`}
+                        >
+                          {t(eventKey, {
+                            defaultValue: row.event_type || row.title,
+                          })}
+                        </button>
+                        <span className="shrink-0 font-mono text-[10px] tabular-nums text-muted-foreground">
+                          {elapsed}s
+                        </span>
+                      </div>
+                      <ExecutionNodes
+                        row={row}
+                        processorUrl={processorUrl}
+                        running={active(row) && connected}
+                      />
+                      <div
+                        className="mt-1.5 h-1 overflow-hidden rounded-full bg-muted"
+                        role="progressbar"
+                        aria-label={label}
+                        {...(!active(row)
+                          ? {
+                              'aria-valuenow': 100,
+                              'aria-valuemin': 0,
+                              'aria-valuemax': 100,
+                            }
+                          : {})}
+                      >
+                        <div
+                          className={`h-full rounded-full ${row.has_error ? 'bg-destructive' : 'bg-primary'} ${active(row) && connected ? styles.active : 'w-full'}`}
+                        />
+                      </div>
+                    </div>
+                  );
+                })}
+              </div>
+              {(truncated || rows.length > 20) && (
+                <p className="border-t px-4 py-2 text-xs text-muted-foreground">
+                  {t('inflight.limited')}
+                </p>
+              )}
+            </>
+          )}
+        </div>
+      </aside>
+      <ExecutionDetailSheet
+        row={selected}
+        onSelect={setSelected}
+        onClose={() => setSelected(null)}
+      />
+    </>
+  );
+}
+
+export default function InFlightMonitor() {
+  const workspace = useCurrentWorkspace();
+  if (!workspace?.permissions.includes('resource.view')) return null;
+  return <WorkspaceMonitor key={workspace.workspace.uuid} />;
+}
diff --git a/web/src/app/home/components/inflight/inflight.module.css b/web/src/app/home/components/inflight/inflight.module.css
new file mode 100644
index 000000000..8698f57b0
--- /dev/null
+++ b/web/src/app/home/components/inflight/inflight.module.css
@@ -0,0 +1,125 @@
+.active {
+  width: 40%;
+  animation: travel 1.6s ease-in-out infinite;
+}
+@keyframes travel {
+  from {
+    transform: translateX(-100%);
+  }
+  to {
+    transform: translateX(350%);
+  }
+}
+@media (prefers-reduced-motion: reduce) {
+  .active {
+    animation: none;
+    width: 40%;
+  }
+}
+.panel {
+  contain: layout;
+  overflow-x: hidden;
+  transition:
+    width 180ms ease-out,
+    transform 180ms ease-out;
+}
+.task {
+  animation: arrive 180ms ease-out both;
+}
+.leaving {
+  animation: depart 180ms ease-in both;
+  pointer-events: none;
+}
+@keyframes arrive {
+  from {
+    opacity: 0;
+    transform: translateX(8px);
+  }
+  to {
+    opacity: 1;
+    transform: translateX(0);
+  }
+}
+@keyframes depart {
+  from {
+    opacity: 1;
+    transform: translateX(0);
+  }
+  to {
+    opacity: 0;
+    transform: translateX(8px);
+  }
+}
+@media (prefers-reduced-motion: reduce) {
+  .task,
+  .leaving {
+    animation: none;
+  }
+}
+
+.panel[data-dragging='true'] {
+  transition: none;
+}
+.dockContent {
+  animation: dockReveal 160ms ease-out both;
+}
+@keyframes dockReveal {
+  from {
+    opacity: 0.35;
+    transform: translateX(5px);
+  }
+  to {
+    opacity: 1;
+    transform: translateX(0);
+  }
+}
+@media (prefers-reduced-motion: reduce) {
+  .panel {
+    transition: none;
+  }
+  .dockContent {
+    animation: none;
+  }
+}
+
+.connection {
+  position: relative;
+  height: 2px;
+  background: var(--border);
+}
+.connection::after {
+  content: '';
+  position: absolute;
+  right: 1px;
+  top: -2px;
+  width: 6px;
+  height: 6px;
+  border-top: 1.5px solid var(--muted-foreground);
+  border-right: 1.5px solid var(--muted-foreground);
+  transform: rotate(45deg);
+}
+.flowing {
+  overflow: hidden;
+  background: color-mix(in srgb, var(--primary) 20%, transparent);
+}
+.flowing i {
+  display: block;
+  width: 8px;
+  height: 2px;
+  background: var(--primary);
+  animation: nodeFlow 1.4s ease-in-out infinite;
+}
+@keyframes nodeFlow {
+  from {
+    transform: translateX(-8px);
+  }
+  to {
+    transform: translateX(24px);
+  }
+}
+@media (prefers-reduced-motion: reduce) {
+  .flowing i {
+    animation: none;
+    transform: translateX(8px);
+  }
+}
diff --git a/web/src/app/home/components/inflight/useEdgeDock.ts b/web/src/app/home/components/inflight/useEdgeDock.ts
new file mode 100644
index 000000000..4a7e4ddb0
--- /dev/null
+++ b/web/src/app/home/components/inflight/useEdgeDock.ts
@@ -0,0 +1,128 @@
+import {
+  useEffect,
+  useLayoutEffect,
+  useRef,
+  useState,
+  type PointerEvent,
+  type KeyboardEvent,
+} from 'react';
+
+const KEY = 'langbot.inflight.dock';
+function saved() {
+  try {
+    return JSON.parse(localStorage.getItem(KEY) || '{}');
+  } catch {
+    return {};
+  }
+}
+
+export function useEdgeDock(visible: boolean) {
+  const [preference] = useState(saved);
+  const [collapsed, setCollapsed] = useState(preference?.collapsed === true);
+  const panel = useRef<HTMLElement>(null);
+  const position = useRef(
+    Number.isFinite(preference?.position)
+      ? Math.min(1, Math.max(0, preference.position))
+      : 0.55,
+  );
+  const drag = useRef<{
+    id: number;
+    y: number;
+    top: number;
+    moved: boolean;
+  } | null>(null);
+  const suppressClick = useRef(false);
+  // Anchor the top edge to the viewport, never to the changing content height.
+  const range = () => Math.max(0, window.innerHeight - 48 - 24);
+  const place = (top = 12 + position.current * range()) => {
+    const bounded = Math.max(12, Math.min(top, 12 + range()));
+    if (panel.current) {
+      panel.current.style.transform = `translateY(${bounded}px)`;
+      panel.current.style.maxHeight = `${Math.max(48, window.innerHeight - bounded - 12)}px`;
+    }
+    return bounded;
+  };
+  const persist = () => {
+    try {
+      localStorage.setItem(
+        KEY,
+        JSON.stringify({ collapsed, position: position.current }),
+      );
+    } catch {
+      /* Optional preference. */
+    }
+  };
+  useEffect(() => {
+    persist();
+  }, [collapsed]);
+  useLayoutEffect(() => {
+    if (!visible || !panel.current) return;
+    const update = () => {
+      place();
+    };
+    window.addEventListener('resize', update);
+    update();
+    return () => {
+      window.removeEventListener('resize', update);
+    };
+  }, [visible, collapsed]);
+  const onPointerDown = (event: PointerEvent<HTMLButtonElement>) => {
+    if (event.button !== 0) return;
+    if (panel.current) panel.current.dataset.dragging = 'true';
+    suppressClick.current = false;
+    drag.current = {
+      id: event.pointerId,
+      y: event.clientY,
+      top: panel.current?.getBoundingClientRect().top || 12,
+      moved: false,
+    };
+    event.currentTarget.setPointerCapture(event.pointerId);
+  };
+  const onPointerMove = (event: PointerEvent<HTMLButtonElement>) => {
+    const current = drag.current;
+    if (!current || current.id !== event.pointerId) return;
+    const delta = event.clientY - current.y;
+    if (Math.abs(delta) > 5) current.moved = true;
+    if (current.moved) {
+      const top = place(current.top + delta);
+      position.current = range() ? (top - 12) / range() : 0;
+    }
+  };
+  const finish = (event: PointerEvent<HTMLButtonElement>) => {
+    if (!drag.current || drag.current.id !== event.pointerId) return;
+    suppressClick.current = drag.current.moved;
+    drag.current = null;
+    if (panel.current) delete panel.current.dataset.dragging;
+    if (event.currentTarget.hasPointerCapture(event.pointerId))
+      event.currentTarget.releasePointerCapture(event.pointerId);
+    persist();
+  };
+  const onKeyDown = (event: KeyboardEvent<HTMLButtonElement>) => {
+    if (!['ArrowUp', 'ArrowDown', 'Home', 'End'].includes(event.key)) return;
+    event.preventDefault();
+    const top =
+      event.key === 'Home'
+        ? 12
+        : event.key === 'End'
+          ? 12 + range()
+          : (panel.current?.getBoundingClientRect().top || 12) +
+            (event.key === 'ArrowUp' ? -24 : 24);
+    const bounded = place(top);
+    position.current = range() ? (bounded - 12) / range() : 0;
+    persist();
+  };
+  return {
+    panel,
+    collapsed,
+    setCollapsed,
+    suppressClick,
+    handle: {
+      onPointerDown,
+      onPointerMove,
+      onPointerUp: finish,
+      onPointerCancel: finish,
+      onLostPointerCapture: finish,
+      onKeyDown,
+    },
+  };
+}
diff --git a/web/src/app/home/components/model-availability/LangBotModelMetadata.tsx b/web/src/app/home/components/model-availability/LangBotModelMetadata.tsx
index bd5d58c16..e44e5b83c 100644
--- a/web/src/app/home/components/model-availability/LangBotModelMetadata.tsx
+++ b/web/src/app/home/components/model-availability/LangBotModelMetadata.tsx
@@ -47,7 +47,7 @@ export default function LangBotModelMetadata({
           <Tooltip>
             <TooltipTrigger asChild>
               <span
-                className="inline-flex items-center gap-1 text-xs tabular-nums text-muted-foreground"
+                className="inline-flex items-center gap-1 whitespace-nowrap text-xs tabular-nums text-muted-foreground"
                 onMouseDown={(event) => event.preventDefault()}
               >
                 <Coins className="size-3" />
@@ -59,12 +59,12 @@ export default function LangBotModelMetadata({
             <TooltipContent side="top" className="max-w-64">
               <div className="space-y-0.5">
                 <p>{t('models.pricing.title')}</p>
-                <p className="text-xs text-muted-foreground">
+                <p className="text-xs">
                   {t('models.pricing.input', {
                     credits: input,
                   })}
                 </p>
-                <p className="text-xs text-muted-foreground">
+                <p className="text-xs">
                   {t('models.pricing.output', {
                     credits: output,
                   })}
diff --git a/web/src/app/home/components/models-dialog/component/provider-form/ProviderForm.tsx b/web/src/app/home/components/models-dialog/component/provider-form/ProviderForm.tsx
index 0b07a6966..19cccc884 100644
--- a/web/src/app/home/components/models-dialog/component/provider-form/ProviderForm.tsx
+++ b/web/src/app/home/components/models-dialog/component/provider-form/ProviderForm.tsx
@@ -1,3 +1,4 @@
+import LoadErrorState from '@/components/LoadErrorState';
 import { useEffect, useState, useRef } from 'react';
 import { httpClient } from '@/app/infra/http/HttpClient';
 
@@ -270,9 +271,7 @@ export default function ProviderForm({
             <LoadingSpinner text={t('common.loading')} />
           </div>
         ) : (
-          <p role="alert" className="py-8 text-sm text-destructive">
-            {t('models.loadError')}
-          </p>
+          <LoadErrorState compact title={t('models.loadError')} />
         )}
         <DialogFooter>
           {loadState === 'error' && (
diff --git a/web/src/app/home/components/models-dialog/components/ModelItem.tsx b/web/src/app/home/components/models-dialog/components/ModelItem.tsx
index c8c25bde1..aa2f47041 100644
--- a/web/src/app/home/components/models-dialog/components/ModelItem.tsx
+++ b/web/src/app/home/components/models-dialog/components/ModelItem.tsx
@@ -25,6 +25,7 @@ import {
   TestResult,
 } from '../types';
 import ExtraArgsEditor from './ExtraArgsEditor';
+import ModelStar from './ModelStar';
 import { userInfo } from '@/app/infra/http';
 import LangBotModelMetadata from '../../model-availability/LangBotModelMetadata';
 
@@ -197,14 +198,14 @@ export default function ModelItem({
     >
       <PopoverTrigger asChild>
         <div
-          className={`flex items-center justify-between py-2 px-3 rounded-md border bg-background ${
+          className={`flex items-center gap-3 py-2 px-3 rounded-md border bg-background ${
             isPopoverDisabled
               ? 'cursor-not-allowed opacity-60'
               : 'hover:bg-accent cursor-pointer'
           }`}
         >
-          <div className="flex min-w-0 items-center gap-2 flex-wrap">
-            <span className="text-sm font-medium">{model.name}</span>
+          <div className="flex min-w-0 flex-1 items-center gap-2 flex-wrap">
+            <span className="break-all text-sm font-medium">{model.name}</span>
             <Badge variant="secondary" className="text-xs">
               {modelType === 'llm'
                 ? t('models.chat')
@@ -240,6 +241,12 @@ export default function ModelItem({
               loaded={availabilityLoaded}
             />
           )}
+          {modelType === 'llm' && (
+            <ModelStar
+              uuid={model.uuid}
+              disabled={!canManage || isPopoverDisabled}
+            />
+          )}
           {canManage && !isLangBotModels && (
             <Popover
               open={isDeleteOpen}
diff --git a/web/src/app/home/components/models-dialog/components/ModelStar.tsx b/web/src/app/home/components/models-dialog/components/ModelStar.tsx
new file mode 100644
index 000000000..7714676cb
--- /dev/null
+++ b/web/src/app/home/components/models-dialog/components/ModelStar.tsx
@@ -0,0 +1,83 @@
+import { useEffect, useSyncExternalStore } from 'react';
+import { Star } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+import { toast } from 'sonner';
+import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
+import { Button } from '@/components/ui/button';
+
+const records = new Map<
+  string,
+  { uuid: string | null; busy: boolean; loaded: boolean }
+>();
+const listeners = new Set<() => void>();
+const pending = new Map<string, Promise<void>>();
+const empty = { uuid: null, busy: false, loaded: false };
+const subscribe = (listener: () => void) => {
+  listeners.add(listener);
+  return () => {
+    listeners.delete(listener);
+  };
+};
+function update(
+  workspace: string,
+  value: typeof empty | { uuid: string | null; busy: boolean; loaded: boolean },
+) {
+  records.set(workspace, value);
+  listeners.forEach((listener) => listener());
+}
+
+export default function ModelStar({
+  uuid,
+  disabled,
+}: {
+  uuid: string;
+  disabled: boolean;
+}) {
+  const workspace = useCurrentWorkspace()?.workspace.uuid ?? '';
+  const { t } = useTranslation();
+  const state = useSyncExternalStore(
+    subscribe,
+    () => records.get(workspace) ?? empty,
+  );
+  useEffect(() => {
+    if (!workspace || pending.has(workspace)) return;
+    const request = backendClient
+      .getStarredModel()
+      .then((result) => {
+        update(workspace, { uuid: result.uuid, busy: false, loaded: true });
+      })
+      .catch(() => {
+        update(workspace, { uuid: null, busy: false, loaded: false });
+      })
+      .finally(() => pending.delete(workspace));
+    pending.set(workspace, request);
+  }, [workspace]);
+  const starred = state.uuid === uuid;
+  return (
+    <Button
+      type="button"
+      variant="ghost"
+      size="icon"
+      className={`size-7 shrink-0 ${starred ? 'text-amber-500 hover:text-amber-600' : 'text-muted-foreground'}`}
+      disabled={disabled || state.busy || !state.loaded}
+      aria-pressed={starred}
+      aria-label={t(starred ? 'models.unstarModel' : 'models.starModel')}
+      title={t(starred ? 'models.unstarModel' : 'models.starModel')}
+      onPointerDown={(event) => event.stopPropagation()}
+      onClick={async (event) => {
+        event.stopPropagation();
+        const next = starred ? null : uuid;
+        update(workspace, { ...state, busy: true });
+        try {
+          await backendClient.setStarredModel(next);
+          update(workspace, { uuid: next, busy: false, loaded: true });
+        } catch {
+          update(workspace, { ...state, busy: false });
+          toast.error(t('models.starModelFailed'));
+        }
+      }}
+    >
+      <Star className={`size-3.5 ${starred ? 'fill-current' : ''}`} />
+    </Button>
+  );
+}
diff --git a/web/src/app/home/components/models-dialog/components/ProviderCard.tsx b/web/src/app/home/components/models-dialog/components/ProviderCard.tsx
index ac522f258..a0b49612e 100644
--- a/web/src/app/home/components/models-dialog/components/ProviderCard.tsx
+++ b/web/src/app/home/components/models-dialog/components/ProviderCard.tsx
@@ -195,9 +195,18 @@ export default function ProviderCard({
                     {t('models.modelsCount', { count: totalModels })}
                   </Badge>
                 </div>
-                <p className="text-xs text-muted-foreground truncate">
+                <p
+                  className={`text-xs text-muted-foreground ${isLangBotModels ? 'leading-relaxed' : 'truncate'}`}
+                >
                   {isLangBotModels ? (
-                    t('models.langbotModelsDescription')
+                    <>
+                      {t('models.langbotModelsDescription')}
+                      {!ownerSpaceBound && (
+                        <span className="mt-1 block">
+                          {t('models.ownerMustBindSpace')}
+                        </span>
+                      )}
+                    </>
                   ) : (
                     <>
                       {provider.base_url}
@@ -222,7 +231,7 @@ export default function ProviderCard({
                   }}
                 >
                   <LogIn className="h-4 w-4 mr-1" />
-                  {t('models.ownerMustBindSpace')}
+                  {t('models.loginWithSpace')}
                 </Button>
               )}
               {isLangBotModels && ownerSpaceBound && spaceCredits !== null && (
@@ -253,11 +262,6 @@ export default function ProviderCard({
                   {t('models.usesOwnerSpaceBilling')}
                 </span>
               )}
-              {isLangBotModels && !isWorkspaceOwner && !ownerSpaceBound && (
-                <span className="text-xs text-muted-foreground">
-                  {t('models.ownerMustBindSpace')}
-                </span>
-              )}
               {canManage && !isLangBotModels && (
                 <>
                   <Button
@@ -363,7 +367,7 @@ export default function ProviderCard({
                       }}
                     >
                       <Plus className="h-3 w-3 mr-1" />
-                      {t('models.addModel')}
+                      {t('common.add')}
                     </Button>
                   }
                   onOpen={() => {
@@ -390,14 +394,15 @@ export default function ProviderCard({
                   trigger={
                     <Button
                       variant="ghost"
-                      size="icon"
-                      className="h-6 w-6"
+                      size="sm"
+                      className="h-6 text-xs"
                       onClick={(e) => {
                         e.stopPropagation();
                         setAddModelMode('scan');
                       }}
                     >
-                      <Radar className="h-3 w-3" />
+                      <Radar className="h-3 w-3 mr-1" />
+                      {t('models.scanModels')}
                     </Button>
                   }
                   onOpen={() => {
diff --git a/web/src/app/home/components/models-dialog/types.ts b/web/src/app/home/components/models-dialog/types.ts
index f8f72e0b1..6b4eaf614 100644
--- a/web/src/app/home/components/models-dialog/types.ts
+++ b/web/src/app/home/components/models-dialog/types.ts
@@ -28,8 +28,7 @@ export interface ProviderModels {
 }
 
 export type TestResult =
-  | { success: true; duration: number }
-  | { success: false; message: string };
+  { success: true; duration: number } | { success: false; message: string };
 
 export type SelectedScannedModel = {
   model: ScannedProviderModel;
diff --git a/web/src/app/home/components/processor-detail/ProcessorDetailWorkbench.tsx b/web/src/app/home/components/processor-detail/ProcessorDetailWorkbench.tsx
index 3c608fa40..c8211885a 100644
--- a/web/src/app/home/components/processor-detail/ProcessorDetailWorkbench.tsx
+++ b/web/src/app/home/components/processor-detail/ProcessorDetailWorkbench.tsx
@@ -1,4 +1,5 @@
-import { ReactNode, useState } from 'react';
+import { ReactNode } from 'react';
+import { useSearchParams } from 'react-router-dom';
 import { BarChart3, Bug, Info, Settings } from 'lucide-react';
 import { Badge } from '@/components/ui/badge';
 import { Button } from '@/components/ui/button';
@@ -56,6 +57,7 @@ interface ProcessorDetailWorkbenchProps {
   debugDisconnectedLabel?: string;
   unsavedLabel?: string;
   monitoring?: ProcessorMonitoringView;
+  view?: 'workbench' | 'monitoring';
   onViewChange?: (view: 'workbench' | 'monitoring') => void;
 }
 
@@ -83,11 +85,12 @@ export default function ProcessorDetailWorkbench({
   debugDisconnectedLabel,
   unsavedLabel,
   monitoring,
+  view,
   onViewChange,
 }: ProcessorDetailWorkbenchProps) {
-  const [activeView, setActiveView] = useState<'workbench' | 'monitoring'>(
-    'workbench',
-  );
+  const [searchParams, setSearchParams] = useSearchParams();
+  const activeView =
+    searchParams.get('tab') === 'logs' ? 'monitoring' : 'workbench';
   const hasDebug = Boolean(debugTitle && debugContent);
 
   const configPanel = (
@@ -143,10 +146,17 @@ export default function ProcessorDetailWorkbench({
 
   return (
     <Tabs
-      value={activeView}
+      value={view ?? activeView}
       onValueChange={(value) => {
         const view = value as 'workbench' | 'monitoring';
-        setActiveView(view);
+        setSearchParams(
+          (current) => {
+            const next = new URLSearchParams(current);
+            next.set('tab', view === 'monitoring' ? 'logs' : 'config');
+            return next;
+          },
+          { preventScrollReset: true },
+        );
         onViewChange?.(view);
       }}
       className="flex h-full min-h-0 min-w-0 flex-col gap-0"
@@ -226,7 +236,7 @@ export default function ProcessorDetailWorkbench({
           )}
         </div>
         <div className="flex items-center gap-2">
-          {canSave && activeView === 'workbench' && (
+          {canSave && (view ?? activeView) === 'workbench' && (
             <Button
               type="submit"
               form={saveFormId}
@@ -236,7 +246,7 @@ export default function ProcessorDetailWorkbench({
               {saveLabel}
             </Button>
           )}
-          {activeView === 'workbench' && headerActions}
+          {(view ?? activeView) === 'workbench' && headerActions}
         </div>
       </div>
 
diff --git a/web/src/app/home/components/qrcode-login/QrCodeLoginDialog.tsx b/web/src/app/home/components/qrcode-login/QrCodeLoginDialog.tsx
index 29459d416..0be94ce32 100644
--- a/web/src/app/home/components/qrcode-login/QrCodeLoginDialog.tsx
+++ b/web/src/app/home/components/qrcode-login/QrCodeLoginDialog.tsx
@@ -19,11 +19,7 @@ import { getActiveWorkspaceUuid } from '@/app/infra/http/workspaceContext';
 import { getBackendBaseUrl } from '@/app/infra/http/backendUrl';
 
 export type QrLoginPlatform =
-  | 'feishu'
-  | 'weixin'
-  | 'dingtalk'
-  | 'wecombot'
-  | 'qqofficial';
+  'feishu' | 'weixin' | 'dingtalk' | 'wecombot' | 'qqofficial';
 
 interface PlatformConfig {
   titleKey: string;
@@ -83,7 +79,6 @@ const PLATFORM_CONFIGS: Record<QrLoginPlatform, PlatformConfig> = {
       client_id: data.client_id,
       client_secret: data.client_secret,
     }),
-    successNoteKey: 'dingtalk.robotCodeNote',
   },
   wecombot: {
     titleKey: 'wecombot.createBot',
diff --git a/web/src/app/home/components/qrcode-login/SlackSetupDialog.tsx b/web/src/app/home/components/qrcode-login/SlackSetupDialog.tsx
new file mode 100644
index 000000000..3bbf97a69
--- /dev/null
+++ b/web/src/app/home/components/qrcode-login/SlackSetupDialog.tsx
@@ -0,0 +1,331 @@
+import { useEffect, useRef, useState } from 'react';
+import { useTranslation } from 'react-i18next';
+import { Loader2, ExternalLink } from 'lucide-react';
+import {
+  Dialog,
+  DialogContent,
+  DialogHeader,
+  DialogTitle,
+  DialogDescription,
+} from '@/components/ui/dialog';
+import { Button } from '@/components/ui/button';
+import { Input } from '@/components/ui/input';
+import { Label } from '@/components/ui/label';
+import { getBackendBaseUrl } from '@/app/infra/http/backendUrl';
+import { getActiveWorkspaceUuid } from '@/app/infra/http/workspaceContext';
+
+type Config = Record<string, string | boolean>;
+
+export default function SlackSetupDialog({
+  open,
+  onOpenChange,
+  botId,
+  webhookUrl,
+  socketMode,
+  onSuccess,
+}: {
+  open: boolean;
+  onOpenChange: (open: boolean) => void;
+  botId?: string;
+  webhookUrl: string;
+  socketMode: boolean;
+  onSuccess: (config: Config) => void;
+}) {
+  const { t } = useTranslation();
+  const [access, setAccess] = useState('');
+  const [refresh, setRefresh] = useState('');
+  const [name, setName] = useState('LangBot');
+  const [url, setUrl] = useState(webhookUrl);
+  const [redirect, setRedirect] = useState('');
+  const [appToken, setAppToken] = useState('');
+  const [status, setStatus] = useState('idle');
+  const [error, setError] = useState('');
+  const [authorizeUrl, setAuthorizeUrl] = useState('');
+  const [appId, setAppId] = useState('');
+  const [config, setConfig] = useState<Config | null>(null);
+  const generation = useRef(0);
+  const session = useRef<{
+    id: string;
+    base: string;
+    headers: Record<string, string>;
+  } | null>(null);
+  const onSuccessRef = useRef(onSuccess);
+  onSuccessRef.current = onSuccess;
+
+  const cleanup = () => {
+    generation.current++;
+    const current = session.current;
+    session.current = null;
+    if (current)
+      void fetch(`${current.base}/${current.id}`, {
+        method: 'DELETE',
+        headers: current.headers,
+        keepalive: true,
+      }).catch(() => {});
+  };
+
+  useEffect(() => {
+    if (open) {
+      setUrl(webhookUrl);
+      const backend = new URL(
+        getBackendBaseUrl() || window.location.origin,
+        window.location.origin,
+      );
+      if (backend.hostname === '127.0.0.1') backend.hostname = 'localhost';
+      setRedirect(
+        `${backend.href.replace(/\/$/, '')}/api/v1/platform/adapters/slack/setup/callback`,
+      );
+      setStatus('idle');
+      setError('');
+      setConfig(null);
+      setAppId('');
+      setAuthorizeUrl('');
+    }
+    return () => {
+      cleanup();
+      setAccess('');
+      setRefresh('');
+      setAppToken('');
+    };
+  }, [open, webhookUrl]);
+
+  const start = async () => {
+    cleanup();
+    const version = generation.current;
+    setStatus('creating');
+    setError('');
+    const workspace = getActiveWorkspaceUuid();
+    const headers: Record<string, string> = {
+      'Content-Type': 'application/json',
+      Authorization: `Bearer ${localStorage.getItem('token')}`,
+      ...(workspace ? { 'X-Workspace-Id': workspace } : {}),
+    };
+    const base = `${getBackendBaseUrl()}/api/v1/platform/adapters/slack/setup`;
+    try {
+      const response = await fetch(base, {
+        method: 'POST',
+        headers,
+        body: JSON.stringify({
+          bot_uuid: botId,
+          access_token: access,
+          refresh_token: refresh,
+          name,
+          webhook_url: url,
+          socket_mode: socketMode,
+          redirect_url: redirect,
+        }),
+      });
+      const result = await response.json();
+      if (!response.ok || result.code !== 0)
+        throw new Error(result.msg || `HTTP ${response.status}`);
+      const current = { id: result.data.session_id as string, base, headers };
+      if (generation.current !== version) {
+        void fetch(`${base}/${current.id}`, { method: 'DELETE', headers });
+        return;
+      }
+      session.current = current;
+      setAccess('');
+      setRefresh('');
+      const deadline = Date.now() + 15 * 60 * 1000;
+      while (generation.current === version && Date.now() < deadline) {
+        const poll = await fetch(`${base}/${current.id}`, {
+          headers,
+          cache: 'no-store',
+        });
+        const json = await poll.json();
+        if (generation.current !== version) return;
+        if (!poll.ok || json.code !== 0)
+          throw new Error(json.msg || `HTTP ${poll.status}`);
+        const data = json.data;
+        setStatus(data.status);
+        if (data.app_id) setAppId(data.app_id);
+        if (data.authorize_url) setAuthorizeUrl(data.authorize_url);
+        if (data.status === 'error') throw new Error(data.error);
+        if (data.status === 'success') {
+          setConfig(data.config);
+          return;
+        }
+        await new Promise((resolve) => setTimeout(resolve, 2000));
+      }
+      if (generation.current === version) throw new Error('session_expired');
+    } catch (cause) {
+      if (generation.current !== version) return;
+      setError(cause instanceof Error ? cause.message : 'setup_failed');
+      setStatus('error');
+    }
+  };
+
+  const apply = () => {
+    if (!config) return;
+    onSuccessRef.current({
+      ...config,
+      ...(socketMode ? { app_token: appToken.trim() } : {}),
+    });
+    onOpenChange(false);
+  };
+
+  return (
+    <Dialog open={open} onOpenChange={onOpenChange}>
+      <DialogContent className="sm:max-w-lg max-h-[85vh] overflow-y-auto">
+        <DialogHeader>
+          <DialogTitle>{t('slackSetup.title')}</DialogTitle>
+          <DialogDescription>{t('slackSetup.description')}</DialogDescription>
+        </DialogHeader>
+        {!botId ? (
+          <p>{t('slackSetup.saveFirst')}</p>
+        ) : (
+          <div className="space-y-4">
+            {status === 'idle' && (
+              <>
+                <p className="text-sm text-muted-foreground">
+                  {t('slackSetup.permissionsHint')}
+                </p>
+                <a
+                  className="text-sm text-primary underline"
+                  href="https://api.slack.com/apps"
+                  target="_blank"
+                  rel="noreferrer"
+                >
+                  {t('slackSetup.generate')}
+                </a>
+                <div className="space-y-2">
+                  <Label htmlFor="slack-name">{t('slackSetup.name')}</Label>
+                  <Input
+                    id="slack-name"
+                    value={name}
+                    maxLength={35}
+                    onChange={(e) => setName(e.target.value)}
+                  />
+                </div>
+                <div className="space-y-2">
+                  <Label htmlFor="slack-access">Access Token</Label>
+                  <Input
+                    id="slack-access"
+                    type="password"
+                    autoComplete="off"
+                    value={access}
+                    onChange={(e) => setAccess(e.target.value)}
+                  />
+                </div>
+                <div className="space-y-2">
+                  <Label htmlFor="slack-refresh">Refresh Token</Label>
+                  <Input
+                    id="slack-refresh"
+                    type="password"
+                    autoComplete="off"
+                    value={refresh}
+                    onChange={(e) => setRefresh(e.target.value)}
+                  />
+                </div>
+                <div className="space-y-2">
+                  <Label htmlFor="slack-url">
+                    {t(
+                      socketMode ? 'slackSetup.redirect' : 'slackSetup.webhook',
+                    )}
+                  </Label>
+                  <Input
+                    id="slack-url"
+                    value={socketMode ? redirect : url}
+                    onChange={(e) =>
+                      socketMode
+                        ? setRedirect(e.target.value)
+                        : setUrl(e.target.value)
+                    }
+                  />
+                  <p className="text-xs text-muted-foreground">
+                    {t(
+                      socketMode
+                        ? 'slackSetup.redirectHint'
+                        : 'slackSetup.webhookHint',
+                    )}
+                  </p>
+                </div>
+                <Button
+                  type="button"
+                  onClick={() => void start()}
+                  disabled={!access.trim() || !refresh.trim() || !name.trim()}
+                >
+                  {t('slackSetup.create')}
+                </Button>
+              </>
+            )}
+            {(status === 'creating' || status === 'authorizing') && (
+              <p className="flex items-center gap-2 text-sm">
+                <Loader2 className="size-4 animate-spin" />
+                {t(`slackSetup.${status}`)}
+              </p>
+            )}
+            {status === 'waiting' && (
+              <div className="space-y-3">
+                <p className="text-sm">{t('slackSetup.waiting')}</p>
+                <Button asChild>
+                  <a href={authorizeUrl} target="_blank" rel="noreferrer">
+                    {t('slackSetup.authorize')}
+                    <ExternalLink className="ml-2 size-4" />
+                  </a>
+                </Button>
+              </div>
+            )}
+            {socketMode && appId && (
+              <div className="space-y-2 rounded-lg border p-3">
+                <p className="text-sm">{t('slackSetup.appTokenHint')}</p>
+                <a
+                  className="text-sm text-primary underline"
+                  href={`https://api.slack.com/apps/${appId}/general`}
+                  target="_blank"
+                  rel="noreferrer"
+                >
+                  {t('slackSetup.openApp')}
+                </a>
+                <Label htmlFor="slack-app-token" className="block">
+                  App-Level Token
+                </Label>
+                <Input
+                  id="slack-app-token"
+                  type="password"
+                  autoComplete="off"
+                  placeholder="xapp-…"
+                  value={appToken}
+                  onChange={(e) => setAppToken(e.target.value)}
+                />
+              </div>
+            )}
+            {status === 'error' && (
+              <div className="space-y-2">
+                <p role="alert" className="text-sm text-destructive">
+                  {t('slackSetup.failed')} {error}
+                </p>
+                {appId && (
+                  <p className="text-xs text-muted-foreground">
+                    {t('slackSetup.appCreated')} {appId}
+                  </p>
+                )}
+                <Button
+                  type="button"
+                  onClick={() => {
+                    cleanup();
+                    setStatus('idle');
+                  }}
+                >
+                  {t('slackSetup.retry')}
+                </Button>
+              </div>
+            )}
+            {status === 'success' && (
+              <div className="space-y-3">
+                <p className="text-sm">{t('slackSetup.success')}</p>
+                <Button
+                  type="button"
+                  onClick={apply}
+                  disabled={socketMode && !appToken.trim().startsWith('xapp-')}
+                >
+                  {t('slackSetup.apply')}
+                </Button>
+              </div>
+            )}
+          </div>
+        )}
+      </DialogContent>
+    </Dialog>
+  );
+}
diff --git a/web/src/app/home/components/settings-dialog/SettingsDialog.tsx b/web/src/app/home/components/settings-dialog/SettingsDialog.tsx
index f727442bc..e642337fc 100644
--- a/web/src/app/home/components/settings-dialog/SettingsDialog.tsx
+++ b/web/src/app/home/components/settings-dialog/SettingsDialog.tsx
@@ -105,13 +105,6 @@ export default function SettingsDialog({
       description: t('workspace.description'),
       icon: <UsersRound className="size-4" />,
     },
-    {
-      id: 'operationTrace',
-      label: t('settingsDialog.nav.operationTrace'),
-      title: t('operationTrace.title'),
-      description: t('operationTrace.description'),
-      icon: <History className="size-4" />,
-    },
     {
       id: 'models',
       label: t('settingsDialog.nav.models'),
@@ -140,6 +133,13 @@ export default function SettingsDialog({
       description: t('account.settingsDescription'),
       icon: <Settings className="size-4" />,
     },
+    {
+      id: 'operationTrace',
+      label: t('settingsDialog.nav.operationTrace'),
+      title: t('operationTrace.title'),
+      description: t('operationTrace.description'),
+      icon: <History className="size-4" />,
+    },
   ];
   const permissions = currentWorkspace?.permissions ?? [];
   const canManageApiKeys = permissions.includes('api_key.manage');
@@ -191,7 +191,7 @@ export default function SettingsDialog({
       }}
     >
       <DialogContent
-        className="h-[80vh] max-h-[800px] overflow-hidden p-0 sm:max-w-[52rem] [&>button:last-child]:z-20"
+        className="flex h-[80dvh] max-h-[800px] min-h-0 flex-col gap-0 overflow-hidden p-0 sm:max-w-[52rem] [&>button:last-child]:z-20"
         // Fixed height so switching sections never resizes the dialog; each
         // panel scrolls its own content internally.
       >
@@ -202,7 +202,7 @@ export default function SettingsDialog({
 
         {/* Override the SidebarProvider wrapper's default h-svh so the two
             columns fill the dialog's fixed height instead of the viewport. */}
-        <SidebarProvider className="!min-h-0 h-full">
+        <SidebarProvider className="!min-h-0 flex-1 overflow-hidden">
           <Sidebar
             collapsible="none"
             className="hidden h-full w-44 shrink-0 border-r md:flex"
@@ -231,7 +231,7 @@ export default function SettingsDialog({
             </SidebarContent>
           </Sidebar>
 
-          <main className="flex h-full min-w-0 flex-1 flex-col overflow-hidden">
+          <main className="flex min-h-0 min-w-0 flex-1 flex-col overflow-hidden">
             {/* Mobile section switcher (sidebar is hidden on small screens) */}
             <div className="flex shrink-0 items-center gap-1 overflow-x-auto border-b px-3 py-2 md:hidden">
               {navItems.map((item) => (
diff --git a/web/src/app/home/components/workspace-assistant.module.css b/web/src/app/home/components/workspace-assistant.module.css
new file mode 100644
index 000000000..8546573b7
--- /dev/null
+++ b/web/src/app/home/components/workspace-assistant.module.css
@@ -0,0 +1,238 @@
+/* Brief gestures separated by long rests keep the collapsed dock quiet. */
+.magic {
+  animation: assistant-wave 9s ease-in-out infinite;
+  transform-origin: bottom center;
+}
+
+.sparkle {
+  animation: assistant-sparkle 9s ease-in-out infinite;
+}
+
+.dock:hover .magic,
+.dock:hover .sparkle,
+.dock:focus-visible .magic,
+.dock:focus-visible .sparkle {
+  animation-play-state: paused;
+}
+
+@keyframes assistant-wave {
+  0%,
+  84%,
+  100% {
+    transform: translateY(0) rotate(0);
+  }
+  87% {
+    transform: translateY(-2px) rotate(-10deg);
+  }
+  90% {
+    transform: translateY(-2px) rotate(8deg);
+  }
+  93% {
+    transform: translateY(-1px) rotate(-5deg);
+  }
+  96% {
+    transform: translateY(0) rotate(0);
+  }
+}
+
+@keyframes assistant-sparkle {
+  0%,
+  86%,
+  100% {
+    transform: scale(1);
+  }
+  90% {
+    transform: scale(1.2) rotate(12deg);
+  }
+  96% {
+    transform: scale(1);
+  }
+}
+
+@media (prefers-reduced-motion: reduce) {
+  .magic,
+  .sparkle {
+    animation: none;
+  }
+}
+
+.panel {
+  border-color: var(--border);
+  background: var(--background);
+  box-shadow: 0 -4px 20px #0000000d;
+  transform-origin: bottom center;
+  animation: assistant-arrive 220ms cubic-bezier(0.2, 0.8, 0.2, 1) both;
+}
+.closing {
+  animation: assistant-leave 180ms ease-in both;
+  pointer-events: none;
+}
+.dock {
+  animation: assistant-arrive 160ms ease-out both;
+}
+.conversationPanel {
+  height: min(420px, calc(100dvh - 48px));
+}
+.header {
+  background: var(--background);
+}
+.avatar {
+  display: grid;
+  place-items: center;
+  flex-shrink: 0;
+  width: 26px;
+  height: 26px;
+  border-radius: 10px;
+  color: #2288ee;
+  background: color-mix(in srgb, #2288ee 8%, var(--background));
+}
+.welcome {
+  padding: 6px 0;
+  display: grid;
+  gap: 8px;
+}
+.prompt {
+  display: inline-flex;
+  align-items: center;
+  gap: 8px;
+  padding: 6px 8px;
+  border-radius: 10px;
+  text-align: left;
+  font-size: 12px;
+  line-height: 1.5;
+  background: var(--muted);
+  border: 1px solid transparent;
+  cursor: pointer;
+  transition:
+    transform 160ms ease,
+    border-color 160ms ease;
+}
+.promptIcon {
+  display: flex;
+  flex-shrink: 0;
+  color: #2288ee;
+}
+.prompt:hover:not(:disabled) {
+  transform: translateY(-1px);
+  border-color: var(--border);
+}
+.prompt:focus-visible {
+  outline: 2px solid #2288ee;
+  outline-offset: 2px;
+}
+.prompt:disabled {
+  opacity: 0.5;
+  cursor: default;
+}
+.userMessage {
+  background: color-mix(in srgb, #2288ee 8%, var(--background));
+  border-bottom-right-radius: 0;
+}
+.assistantMessage {
+  background: var(--muted);
+  border-bottom-left-radius: 0;
+}
+
+/* Use the debug dialog's Markdown styling, scaled to the compact assistant. */
+.messageContent:global(.markdown-body) {
+  color: var(--foreground);
+  font-size: 13px;
+  line-height: 1.7;
+  overflow-wrap: anywhere;
+}
+.messageContent:global(.markdown-body) :is(p, ul, ol, blockquote, pre, table) {
+  margin-top: 0;
+  margin-bottom: 10px;
+}
+.messageContent:global(.markdown-body) :is(h1, h2, h3, h4, h5, h6) {
+  margin-top: 16px;
+  margin-bottom: 8px;
+  padding: 0;
+  border: 0;
+  font-size: 14px;
+  line-height: 1.5;
+}
+.messageContent:global(.markdown-body) > :first-child {
+  margin-top: 0;
+}
+.messageContent:global(.markdown-body) > :last-child {
+  margin-bottom: 0;
+}
+.messageContent:global(.markdown-body) :is(ul, ol) {
+  padding-left: 1.5em;
+}
+.messageContent:global(.markdown-body) li + li {
+  margin-top: 4px;
+}
+.messageContent:global(.markdown-body) li > p {
+  margin-bottom: 4px;
+}
+.messageContent:global(.markdown-body) pre {
+  max-width: 100%;
+  overflow-x: auto;
+  padding: 10px;
+  border-radius: 8px;
+  background: var(--background);
+  font-size: 12px;
+}
+.messageContent:global(.markdown-body) table {
+  display: block;
+  max-width: 100%;
+  overflow-x: auto;
+  font-size: 12px;
+}
+.messageContent:global(.markdown-body) :is(th, td) {
+  padding: 6px 8px;
+}
+
+.composer {
+  background: var(--muted);
+  border: 1px solid transparent;
+}
+.input {
+  border: 0;
+  box-shadow: none;
+}
+.send,
+.dock {
+  background: #2288ee;
+  box-shadow: none;
+}
+.send:hover:not(:disabled),
+.dock:hover {
+  background: #1878d8;
+}
+@keyframes assistant-arrive {
+  from {
+    opacity: 0;
+    transform: translateY(14px) scale(0.98);
+  }
+  to {
+    opacity: 1;
+    transform: translateY(0) scale(1);
+  }
+}
+@keyframes assistant-leave {
+  from {
+    opacity: 1;
+    transform: translateY(0) scale(1);
+  }
+  to {
+    opacity: 0;
+    transform: translateY(18px) scale(0.98);
+  }
+}
+@media (prefers-reduced-motion: reduce) {
+  .panel,
+  .closing,
+  .dock {
+    animation: none;
+  }
+  .prompt,
+  .composer {
+    transition: none;
+  }
+  .prompt:hover:not(:disabled) {
+    transform: none;
+  }
+}
diff --git a/web/src/app/home/components/workspace-quota/useWorkspaceQuotaStatus.ts b/web/src/app/home/components/workspace-quota/useWorkspaceQuotaStatus.ts
index 4788b7477..94d021b4d 100644
--- a/web/src/app/home/components/workspace-quota/useWorkspaceQuotaStatus.ts
+++ b/web/src/app/home/components/workspace-quota/useWorkspaceQuotaStatus.ts
@@ -1,3 +1,4 @@
+import { useCurrentWorkspace } from '@/app/infra/http';
 import { systemInfo } from '@/app/infra/http/HttpClient';
 import { useSidebarData } from '@/app/home/components/home-sidebar/SidebarDataContext';
 
@@ -46,27 +47,28 @@ export function useWorkspaceQuotaStatus(): WorkspaceQuotaStatus {
     skills,
     quotaDataLoaded,
   } = useSidebarData();
-  const limitation = systemInfo.limitation;
+  const currentWorkspace = useCurrentWorkspace();
+  const cloud = currentWorkspace?.workspace.source === 'cloud_projection';
+  const limitation = cloud
+    ? currentWorkspace.limitation
+    : systemInfo.limitation;
+  const loaded = quotaDataLoaded && (!cloud || limitation != null);
 
-  const botQuota = quotaItem(
-    bots.length,
-    limitation?.max_bots,
-    quotaDataLoaded,
-  );
+  const botQuota = quotaItem(bots.length, limitation?.max_bots, loaded);
   const pipelineQuota = quotaItem(
     pipelines.length,
     limitation?.max_pipelines,
-    quotaDataLoaded,
+    loaded,
   );
   const knowledgeBaseQuota = quotaItem(
     knowledgeBases.length,
     limitation?.max_knowledge_bases,
-    quotaDataLoaded,
+    loaded,
   );
   const extensionQuota = quotaItem(
     pluginCount + mcpServers.length + skills.length,
     limitation?.max_extensions,
-    quotaDataLoaded,
+    loaded,
   );
 
   return {
diff --git a/web/src/app/home/components/workspace-settings/CloudPlansDialog.tsx b/web/src/app/home/components/workspace-settings/CloudPlansDialog.tsx
new file mode 100644
index 000000000..8e753b18b
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/CloudPlansDialog.tsx
@@ -0,0 +1,687 @@
+import { useEffect, useState, type ReactNode, type ComponentType } from 'react';
+import { useTranslation } from 'react-i18next';
+import {
+  ArrowRight,
+  Check,
+  Info,
+  Layers,
+  Loader2,
+  RefreshCw,
+} from 'lucide-react';
+import * as RemixIcons from '@remixicon/react';
+import {
+  Dialog,
+  DialogContent,
+  DialogDescription,
+  DialogHeader,
+  DialogTitle,
+} from '@/components/ui/dialog';
+import { Button } from '@/components/ui/button';
+import { Badge } from '@/components/ui/badge';
+import {
+  Tooltip,
+  TooltipContent,
+  TooltipProvider,
+  TooltipTrigger,
+} from '@/components/ui/tooltip';
+import { cn } from '@/lib/utils';
+import type { CurrentWorkspace } from '@/app/infra/entities/workspace';
+import {
+  cloudPlanCheckoutURL,
+  cloudPlanPlatformIconURL,
+  cloudSubscriptionManagementURL,
+  cloudPlanPrice,
+  cloudPlanCurrencyForLanguage,
+  fetchCloudPlanRevision,
+  fetchPublicCloudPlans,
+  localizedPlanText,
+  type CloudPlanBillingCycle,
+  type PublicCloudPlan,
+} from './cloud-plans';
+import { revisionRows } from './plan-revisions';
+import { supportedPlanPlatforms } from './cloud-plan-platforms';
+import PlanRevisionStack from './PlanRevisionStack';
+
+function PlanIcon({ name }: { name?: string }) {
+  const key = name
+    ?.replace(/^ri-/, '')
+    .split('-')
+    .map((part) => part.charAt(0).toUpperCase() + part.slice(1))
+    .join('');
+  const Icon =
+    name && /^ri-[a-z0-9-]+-(line|fill)$/.test(name)
+      ? (
+          RemixIcons as unknown as Record<
+            string,
+            ComponentType<{ className?: string }>
+          >
+        )[`Ri${key}`]
+      : undefined;
+  return Icon ? (
+    <Icon className="size-4 shrink-0 text-blue-500" />
+  ) : (
+    <Check className="size-4 shrink-0 text-blue-500" />
+  );
+}
+
+function Help({ label, children }: { label: string; children: ReactNode }) {
+  return (
+    <Tooltip>
+      <TooltipTrigger asChild>
+        <button
+          type="button"
+          aria-label={label}
+          className="flex h-5 shrink-0 items-center rounded text-muted-foreground hover:text-foreground focus-visible:outline-2 focus-visible:outline-primary"
+        >
+          <Info className="size-3.5" aria-hidden />
+        </button>
+      </TooltipTrigger>
+      <TooltipContent className="max-w-80 whitespace-pre-line">
+        {children}
+      </TooltipContent>
+    </Tooltip>
+  );
+}
+
+function Benefits({
+  plan,
+  spaceURL,
+}: {
+  plan: Pick<
+    PublicCloudPlan,
+    | 'limits'
+    | 'features'
+    | 'disabled_adapters'
+    | 'display_config'
+    | 'chat_history_retention_days'
+    | 'monthly_bonus_credits'
+  >;
+  spaceURL: string;
+}) {
+  const { t, i18n } = useTranslation();
+  const text = (values: Parameters<typeof localizedPlanText>[0]) =>
+    localizedPlanText(values, i18n.language);
+  const config = plan.display_config;
+  const platforms = supportedPlanPlatforms(plan.disabled_adapters);
+  const quotaLabel = text(config?.quota_summary) || t('cloudPlans.quotas');
+  const legacyFeatures: Record<string, string> = {
+    'Multi-user workspace': 'multiUserWorkspace',
+    'Remote E2B sandbox configuration': 'remoteE2BSandbox',
+    'Community support': 'communitySupport',
+    'Everything in Free': 'everythingInFree',
+    'One always-on global sandbox': 'alwaysOnSandbox',
+    'Higher workspace limits': 'higherWorkspaceLimits',
+  };
+  const items =
+    config?.items ??
+    (plan.features ?? []).map((feature) => ({
+      label: {
+        en_US: legacyFeatures[feature]
+          ? t(`cloudPlans.features.${legacyFeatures[feature]}`)
+          : feature,
+      },
+      icon: 'ri-check-line',
+      tooltip: {},
+    }));
+  const rows: Record<string, ReactNode> = {
+    quota_summary: (
+      <>
+        <Layers className="size-4 shrink-0 text-blue-500" />
+        <span className="min-w-0 break-words">{quotaLabel}</span>
+        <Help label={quotaLabel}>
+          {text(config?.tooltips?.quota_summary) && (
+            <p className="mb-2">{text(config?.tooltips?.quota_summary)}</p>
+          )}
+          <ul className="space-y-2">
+            {Object.entries(plan.limits)
+              .filter(([key]) => key !== 'chat_history.retention_days')
+              .map(([key, limit]) => {
+                const resource = key.replace(/\.max$/, '');
+                const label =
+                  text(config?.quotas?.[key]?.label) ||
+                  t(`cloudPlans.limits.${resource}`, {
+                    defaultValue: resource.replace(/_/g, ' '),
+                  });
+                return (
+                  <li key={key}>
+                    <div className="flex justify-between gap-5">
+                      <span>{label}</span>
+                      <span>
+                        {limit < 0
+                          ? t('cloudPlans.unlimited')
+                          : limit.toLocaleString()}
+                      </span>
+                    </div>
+                    {text(config?.quotas?.[key]?.tooltip) && (
+                      <p className="mt-1 opacity-75">
+                        {text(config?.quotas?.[key]?.tooltip)}
+                      </p>
+                    )}
+                  </li>
+                );
+              })}
+          </ul>
+        </Help>
+      </>
+    ),
+    retention: (
+      <>
+        <PlanIcon name="ri-history-line" />
+        <span className="min-w-0 break-words">
+          {t('cloudPlans.retention', {
+            count: plan.chat_history_retention_days,
+          })}
+        </span>
+        <Help label={t('cloudPlans.retentionLabel')}>
+          {text(config?.tooltips?.retention) || t('cloudPlans.retentionHelp')}
+        </Help>
+      </>
+    ),
+    monthly_credits: (
+      <>
+        <PlanIcon name="ri-coin-line" />
+        <span className="min-w-0 break-words">
+          {t('cloudPlans.credits', { count: plan.monthly_bonus_credits })}
+        </span>
+        <Help label={t('cloudPlans.creditsLabel')}>
+          {text(config?.tooltips?.monthly_credits) ||
+            t('cloudPlans.creditsHelp')}
+        </Help>
+      </>
+    ),
+    supported_platforms: (
+      <>
+        <PlanIcon name="ri-apps-2-line" />
+        <span className="min-w-0 break-words">{t('cloudPlans.platforms')}</span>
+        <Tooltip>
+          <TooltipTrigger
+            type="button"
+            aria-label={t('cloudPlans.platforms')}
+            className="flex h-5 shrink-0 items-center rounded text-muted-foreground hover:text-foreground focus-visible:outline-2 focus-visible:outline-primary"
+            onPointerDown={(event) => {
+              if (event.pointerType === 'touch') event.preventDefault();
+            }}
+            onClick={(event) => {
+              event.preventDefault();
+              event.stopPropagation();
+            }}
+          >
+            <Info className="size-3.5" aria-hidden />
+          </TooltipTrigger>
+          <TooltipContent
+            className="max-w-[min(360px,calc(100vw-2rem))] bg-foreground p-3 text-background"
+            arrowClassName="bg-foreground fill-foreground"
+          >
+            {text(config?.tooltips?.supported_platforms) && (
+              <p className="mb-2 whitespace-pre-line break-words">
+                {text(config?.tooltips?.supported_platforms)}
+              </p>
+            )}
+            <div className="flex flex-wrap gap-1.5" data-supported-platforms>
+              {platforms.map((platform) => (
+                <span
+                  key={platform.name}
+                  className="inline-flex items-center gap-1.5 rounded border border-background/30 px-1.5 py-1 text-xs"
+                >
+                  <img
+                    src={cloudPlanPlatformIconURL(spaceURL, platform.icon)}
+                    alt=""
+                    referrerPolicy="no-referrer"
+                    className="size-4 shrink-0 rounded-sm bg-white p-0.5 object-contain"
+                  />
+                  {t(
+                    `cloudPlans.platformNames.${platform.name.replace('LINE', 'line').replace('openclaw-weixin', 'openclaw_weixin')}`,
+                    { defaultValue: platform.label },
+                  )}
+                </span>
+              ))}
+            </div>
+            {platforms.length === 0 && <p>{t('common.none')}</p>}
+          </TooltipContent>
+        </Tooltip>
+      </>
+    ),
+  };
+  items.forEach((item, index) => {
+    const label = text(item.label);
+    rows[`item:${index}`] = (
+      <>
+        <PlanIcon name={item.icon} />
+        <span className="min-w-0 whitespace-pre-line break-words">{label}</span>
+        {text(item.tooltip) && <Help label={label}>{text(item.tooltip)}</Help>}
+      </>
+    );
+  });
+  const order = [
+    ...new Set([
+      ...(config?.order ?? []),
+      'quota_summary',
+      'supported_platforms',
+      'retention',
+      'monthly_credits',
+      ...items.map((_, index) => `item:${index}`),
+    ]),
+  ].filter((key) => key in rows);
+  return (
+    <ul className="space-y-3 text-sm">
+      {order.map((key) => (
+        <li key={key} className="flex items-start gap-2 break-words">
+          {rows[key]}
+        </li>
+      ))}
+    </ul>
+  );
+}
+
+export default function CloudPlansDialog({
+  open,
+  onOpenChange,
+  workspace,
+  spaceURL,
+}: {
+  open: boolean;
+  onOpenChange: (open: boolean) => void;
+  workspace: CurrentWorkspace;
+  spaceURL: string;
+}) {
+  const { t, i18n } = useTranslation();
+  const [plans, setPlans] = useState<PublicCloudPlan[]>([]);
+  const [loading, setLoading] = useState(true);
+  const [error, setError] = useState(false);
+  const [attempt, setAttempt] = useState(0);
+  const [cycle, setCycle] = useState<CloudPlanBillingCycle>('annual');
+  const currency = cloudPlanCurrencyForLanguage(i18n.language);
+  const canManage = ['owner', 'admin'].includes(workspace.membership.role);
+  useEffect(() => {
+    if (!open) return;
+    const controller = new AbortController();
+    let active = true;
+    const timeout = window.setTimeout(() => controller.abort(), 15000);
+    setLoading(true);
+    setError(false);
+    fetchPublicCloudPlans(spaceURL, controller.signal)
+      .then((value) => {
+        if (active) setPlans(value);
+      })
+      .catch(() => {
+        if (active) setError(true);
+      })
+      .finally(() => {
+        window.clearTimeout(timeout);
+        if (active) setLoading(false);
+      });
+    return () => {
+      active = false;
+      controller.abort();
+      window.clearTimeout(timeout);
+    };
+  }, [open, spaceURL, attempt]);
+  const owned = workspace.current_plan;
+  const latest = plans.find((plan) => plan.name === owned?.name);
+  // Catalog copy of the purchased revision, so its card renders exactly like
+  // the latest one. Benefits still come from the entitlement snapshot below.
+  const [ownedRevision, setOwnedRevision] = useState<PublicCloudPlan | null>(
+    null,
+  );
+  const ownedName = owned?.name;
+  const ownedRevisionNumber = owned?.revision;
+  const needsOwnedRevision =
+    !!latest && !!ownedName && latest.revision !== ownedRevisionNumber;
+  useEffect(() => {
+    setOwnedRevision(null);
+    if (!open || !needsOwnedRevision || !ownedName || !ownedRevisionNumber)
+      return;
+    const controller = new AbortController();
+    fetchCloudPlanRevision(
+      spaceURL,
+      ownedName,
+      ownedRevisionNumber,
+      controller.signal,
+    )
+      .then(setOwnedRevision)
+      // Fall back to the snapshot-only card when the revision is unavailable.
+      .catch(() => {});
+    return () => controller.abort();
+  }, [open, spaceURL, needsOwnedRevision, ownedName, ownedRevisionNumber]);
+  const rows = owned ? revisionRows(owned, latest) : [];
+  const distinct =
+    !!owned &&
+    (!latest ||
+      (owned.revision !== latest.revision && rows.some((row) => row.changed)));
+  const entries: (PublicCloudPlan | 'owned')[] = [...plans];
+  if (distinct) {
+    const index = entries.findIndex(
+      (plan) => plan !== 'owned' && plan.name === owned.name,
+    );
+    entries.splice(index < 0 ? 0 : index, 0, 'owned');
+  }
+  // Copy, prices and display items come from the purchased revision itself;
+  // limits, platforms, retention and credits come only from the workspace
+  // entitlement snapshot, never from the latest revision.
+  const ownedCatalog =
+    ownedRevision?.revision === owned?.revision ? ownedRevision : null;
+  const ownedPlan: PublicCloudPlan | null =
+    distinct && owned
+      ? {
+          uuid: `owned-${owned.name}-${owned.revision}`,
+          name: owned.name,
+          revision: owned.revision,
+          display_name:
+            ownedCatalog?.display_name ?? latest?.display_name ?? {},
+          description: ownedCatalog?.description ?? {},
+          price_cny: ownedCatalog?.price_cny ?? '0',
+          price_usd: ownedCatalog?.price_usd ?? '0',
+          price_jpy: ownedCatalog?.price_jpy ?? '0',
+          limits: owned.limits,
+          features: ownedCatalog?.features ?? [],
+          is_active: true,
+          sort_order: latest?.sort_order ?? 0,
+          highlight: ownedCatalog?.highlight ?? false,
+          chat_history_retention_days:
+            owned.limits['chat_history.retention_days'] ?? 30,
+          monthly_bonus_credits: owned.monthly_bonus_credits,
+          disabled_adapters: Object.entries(owned.entitlement_features)
+            .filter(([key, enabled]) => key.startsWith('platform.') && !enabled)
+            .map(([key]) => key.slice(9)),
+          display_config: ownedCatalog?.display_config
+            ? ownedCatalog.display_config
+            : latest?.display_config
+              ? {
+                  icon: latest.display_config.icon,
+                  order: latest.display_config.order,
+                  tooltips: latest.display_config.tooltips,
+                  quotas: latest.display_config.quotas,
+                  items: [],
+                }
+              : { items: [] },
+        }
+      : null;
+  const subscribedOn = owned?.subscription_started_at
+    ? t('planRevision.subscribedOn', {
+        interpolation: { escapeValue: false },
+        date: new Date(owned.subscription_started_at * 1000).toLocaleDateString(
+          i18n.language.replace('_', '-'),
+        ),
+      })
+    : null;
+  const formatPrice = (price: number) =>
+    new Intl.NumberFormat(i18n.language.replace('_', '-'), {
+      style: 'currency',
+      currency,
+      maximumFractionDigits: currency === 'JPY' ? 0 : 2,
+    }).format(price);
+  return (
+    <Dialog open={open} onOpenChange={onOpenChange}>
+      <DialogContent
+        className="flex max-h-[90dvh] flex-col gap-5 overflow-hidden sm:max-w-5xl"
+        data-cloud-plans
+      >
+        <DialogHeader>
+          <DialogTitle>{t('cloudPlans.title')}</DialogTitle>
+          <DialogDescription>
+            {t('cloudPlans.description', {
+              workspace: workspace.workspace.name,
+            })}
+          </DialogDescription>
+        </DialogHeader>
+        <div className="flex flex-wrap items-center justify-between gap-3">
+          <div
+            className="inline-flex rounded-lg bg-muted p-1"
+            role="group"
+            aria-label={t('cloudPlans.billing')}
+          >
+            {(['monthly', 'annual'] as const).map((value) => (
+              <Button
+                key={value}
+                size="sm"
+                variant="ghost"
+                aria-pressed={cycle === value}
+                className={cn(cycle === value && 'bg-background shadow-sm')}
+                onClick={() => setCycle(value)}
+              >
+                {t(`cloudPlans.${value}`)}
+              </Button>
+            ))}
+          </div>
+        </div>
+        <div className="min-h-0 overflow-y-auto [scrollbar-gutter:stable]">
+          {!loading && (
+            <>
+              {!workspace.current_plan && (
+                <p className="mb-4 text-sm" role="status">
+                  {t('planRevision.unknown')}
+                </p>
+              )}
+            </>
+          )}
+          {loading ? (
+            <div
+              className="flex min-h-48 items-center justify-center"
+              role="status"
+            >
+              <Loader2 className="size-6 animate-spin" />
+              <span className="sr-only">{t('cloudPlans.loading')}</span>
+            </div>
+          ) : error ? (
+            <div className="space-y-3 py-10 text-center" role="alert">
+              <p>{t('cloudPlans.error')}</p>
+              <Button
+                variant="outline"
+                onClick={() => setAttempt((value) => value + 1)}
+              >
+                <RefreshCw />
+                {t('cloudPlans.retry')}
+              </Button>
+            </div>
+          ) : !entries.length ? (
+            <p className="py-10 text-center text-muted-foreground">
+              {t('cloudPlans.empty')}
+            </p>
+          ) : (
+            <TooltipProvider delayDuration={200}>
+              <div
+                className={cn(
+                  'flex gap-4 overflow-x-auto p-1 [&>*]:w-72 [&>*]:shrink-0 [&>*]:grow',
+                )}
+              >
+                {entries.map((plan) => {
+                  const renderCard = (
+                    plan: PublicCloudPlan,
+                    ownedCard = false,
+                  ) => {
+                    const samePlan = workspace.plan_name === plan.name;
+                    const current =
+                      ownedCard || (samePlan && !distinct && !!owned);
+                    // Both stacked revisions share one card style; only the
+                    // badge text tells them apart.
+                    const stacked = distinct && plan.name === owned?.name;
+                    const downgradeBlocked =
+                      Boolean(
+                        workspace.plan_name && workspace.plan_name !== 'free',
+                      ) && plan.name === 'free';
+                    const price = cloudPlanPrice(plan, cycle, currency);
+                    const name =
+                      localizedPlanText(plan.display_name, i18n.language) ||
+                      plan.name.toUpperCase();
+                    return (
+                      <article
+                        key={plan.uuid}
+                        data-owned-plan={ownedCard || undefined}
+                        className={cn(
+                          'flex min-w-0 flex-col rounded-xl border p-5',
+                          current &&
+                            !stacked &&
+                            'border-blue-500/50 bg-blue-500/5',
+                        )}
+                        aria-label={name}
+                      >
+                        <div className="flex flex-wrap items-center justify-between gap-2">
+                          <h3 className="flex items-center gap-2 font-semibold">
+                            <PlanIcon name={plan.display_config?.icon} />
+                            {name}
+                          </h3>
+                          {current ? (
+                            <Badge variant={stacked ? 'default' : 'secondary'}>
+                              {t('cloudPlans.current')}
+                            </Badge>
+                          ) : plan.highlight ? (
+                            <Badge>{t('cloudPlans.recommended')}</Badge>
+                          ) : null}
+                        </div>
+                        <p className="mt-2 text-sm text-muted-foreground">
+                          {localizedPlanText(plan.description, i18n.language)}
+                          {distinct &&
+                            !ownedCard &&
+                            plan.name === owned?.name && (
+                              <span className="block" data-latest-quota>
+                                {t('planRevision.latestQuota')}
+                              </span>
+                            )}
+                          {ownedCard && subscribedOn && (
+                            <span className="block" data-subscribed-on>
+                              {subscribedOn}
+                            </span>
+                          )}
+                        </p>
+                        {ownedCard && !ownedCatalog ? (
+                          <div className="mb-5" />
+                        ) : (
+                          <>
+                            <p className="mt-5 text-2xl font-semibold">
+                              {price === 0
+                                ? t('cloudPlans.free')
+                                : formatPrice(price)}
+                              {price > 0 && (
+                                <span className="ml-1 text-xs font-normal text-muted-foreground">
+                                  / {t(`cloudPlans.${cycle}`)}
+                                </span>
+                              )}
+                            </p>
+                            <p className="mb-5 mt-1 min-h-4 text-xs text-muted-foreground">
+                              {cycle === 'annual' && price > 0
+                                ? t('cloudPlans.annualNote', {
+                                    price: formatPrice(price / 12),
+                                  })
+                                : ''}
+                            </p>
+                          </>
+                        )}
+                        <div className="flex-1">
+                          <Benefits plan={plan} spaceURL={spaceURL} />
+                        </div>
+                        {plan.subscription_disabled && (
+                          <p className="mt-4 text-xs text-muted-foreground">
+                            {t('cloudPlans.subscriptionUnavailableHint')}
+                          </p>
+                        )}
+                        {plan.subscription_disabled && !current ? (
+                          <Button
+                            className="mt-6 w-full"
+                            variant="outline"
+                            disabled
+                          >
+                            {t('cloudPlans.subscriptionUnavailable')}
+                          </Button>
+                        ) : downgradeBlocked ? null : !canManage ? (
+                          <Button
+                            className="mt-6 w-full"
+                            variant="outline"
+                            disabled
+                          >
+                            {t('cloudPlans.ownerOnly')}
+                          </Button>
+                        ) : distinct &&
+                          !ownedCard &&
+                          plan.name === owned?.name ? (
+                          <Tooltip>
+                            <TooltipTrigger asChild>
+                              <Button
+                                asChild
+                                className="mt-6 w-full"
+                                variant={plan.highlight ? 'default' : 'outline'}
+                              >
+                                <a
+                                  href={cloudSubscriptionManagementURL(
+                                    spaceURL,
+                                    workspace.workspace.uuid,
+                                  )}
+                                  target="_blank"
+                                  rel="noopener noreferrer"
+                                >
+                                  {t('cloudPlans.manageSubscription')}
+                                  <ArrowRight className="size-4" />
+                                </a>
+                              </Button>
+                            </TooltipTrigger>
+                            <TooltipContent>
+                              {t('planRevision.cancelToSubscribe')}
+                            </TooltipContent>
+                          </Tooltip>
+                        ) : (
+                          <Button
+                            asChild
+                            className="mt-6 w-full"
+                            variant={plan.highlight ? 'default' : 'outline'}
+                          >
+                            <a
+                              href={
+                                workspace.plan_name &&
+                                workspace.plan_name !== 'free'
+                                  ? cloudSubscriptionManagementURL(
+                                      spaceURL,
+                                      workspace.workspace.uuid,
+                                    )
+                                  : cloudPlanCheckoutURL(
+                                      spaceURL,
+                                      workspace.workspace.uuid,
+                                      plan.name,
+                                      cycle,
+                                      currency,
+                                    )
+                              }
+                              target="_blank"
+                              rel="noopener noreferrer"
+                            >
+                              {t(
+                                workspace.plan_name &&
+                                  workspace.plan_name !== 'free'
+                                  ? 'cloudPlans.manageSubscription'
+                                  : 'cloudPlans.choose',
+                              )}
+                              <ArrowRight className="size-4" />
+                            </a>
+                          </Button>
+                        )}
+                        {ownedCard && !latest && (
+                          <p className="mt-4 text-sm">
+                            {t('planRevision.unavailable')}
+                          </p>
+                        )}
+                      </article>
+                    );
+                  };
+                  if (plan === 'owned')
+                    return ownedPlan && !latest
+                      ? renderCard(ownedPlan, true)
+                      : null;
+                  return ownedPlan && latest === plan ? (
+                    <PlanRevisionStack
+                      key={plan.uuid}
+                      current={renderCard(ownedPlan, true)}
+                      latest={renderCard(plan)}
+                    />
+                  ) : (
+                    renderCard(plan)
+                  );
+                })}
+              </div>
+            </TooltipProvider>
+          )}
+        </div>
+        <p className="text-xs text-muted-foreground">
+          {t('cloudPlans.handoff')}
+        </p>
+      </DialogContent>
+    </Dialog>
+  );
+}
diff --git a/web/src/app/home/components/workspace-settings/OperationTraceCollectionDialog.tsx b/web/src/app/home/components/workspace-settings/OperationTraceCollectionDialog.tsx
new file mode 100644
index 000000000..75414dd1f
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/OperationTraceCollectionDialog.tsx
@@ -0,0 +1,216 @@
+import { useEffect, useState } from 'react';
+import { Loader2, SlidersHorizontal, Timer } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+
+import { Badge } from '@/components/ui/badge';
+import { Button } from '@/components/ui/button';
+import {
+  Dialog,
+  DialogContent,
+  DialogDescription,
+  DialogFooter,
+  DialogHeader,
+  DialogTitle,
+} from '@/components/ui/dialog';
+import { Input } from '@/components/ui/input';
+import type {
+  OperationGovernance,
+  OperationLevel,
+} from '@/app/infra/entities/operation-log';
+
+export interface OperationTraceCollectionSettings {
+  level: OperationLevel;
+  retentionDays: number;
+  maxRows: number;
+  dedupeSeconds: number;
+}
+
+export interface OperationTraceCollectionDialogProps {
+  open: boolean;
+  onOpenChange: (open: boolean) => void;
+  governance: OperationGovernance | null;
+  canConfigure: boolean;
+  saving: boolean;
+  onSave: (settings: OperationTraceCollectionSettings) => void;
+}
+
+/**
+ * Collection settings for the operation log: capture level and retention.
+ *
+ * Edits stay local until Save is pressed, so a click on a level card never
+ * changes what the instance collects. The Save button stays disabled while the
+ * draft matches the persisted policy.
+ */
+export default function OperationTraceCollectionDialog({
+  open,
+  onOpenChange,
+  governance,
+  canConfigure,
+  saving,
+  onSave,
+}: OperationTraceCollectionDialogProps) {
+  const { t } = useTranslation();
+  const [level, setLevel] = useState<OperationLevel>(0);
+  const [retentionDays, setRetentionDays] = useState<number>(30);
+  const [maxRows, setMaxRows] = useState<number>(20000);
+  const [dedupeSeconds, setDedupeSeconds] = useState<number>(60);
+
+  // Every open starts from the persisted policy, so a dialog dismissed with
+  // unsaved edits never leaks them into a later session.
+  useEffect(() => {
+    if (!open || !governance) {
+      return;
+    }
+    setLevel(governance.configured_level);
+    setRetentionDays(governance.retention_days);
+    setMaxRows(governance.max_rows);
+    setDedupeSeconds(governance.dedupe_window_seconds);
+  }, [open, governance]);
+
+  const levelOptions = governance?.supported_levels ?? [];
+  const dirty =
+    governance !== null &&
+    (level !== governance.configured_level ||
+      retentionDays !== governance.retention_days ||
+      maxRows !== governance.max_rows ||
+      dedupeSeconds !== governance.dedupe_window_seconds);
+
+  function resetDraft() {
+    if (!governance) {
+      return;
+    }
+    setLevel(governance.configured_level);
+    setRetentionDays(governance.retention_days);
+    setMaxRows(governance.max_rows);
+    setDedupeSeconds(governance.dedupe_window_seconds);
+  }
+
+  return (
+    <Dialog open={open} onOpenChange={onOpenChange}>
+      <DialogContent className="sm:max-w-lg">
+        <DialogHeader>
+          <DialogTitle>{t('operationTrace.collectionSettings')}</DialogTitle>
+          <DialogDescription>
+            {t('operationTrace.description')}
+          </DialogDescription>
+        </DialogHeader>
+
+        <div className="space-y-6">
+          <section className="space-y-3">
+            <div className="flex items-center gap-2">
+              <SlidersHorizontal className="size-4" />
+              <h3 className="text-sm font-semibold">
+                {t('operationTrace.captureLevel')}
+              </h3>
+            </div>
+            <div className="grid gap-2 sm:grid-cols-3">
+              {levelOptions.map((option) => {
+                const selected = level === option.level;
+                return (
+                  <button
+                    key={option.level}
+                    type="button"
+                    disabled={!canConfigure || saving}
+                    onClick={() => setLevel(option.level)}
+                    className={`rounded-lg border p-3 text-left transition-colors disabled:opacity-60 ${
+                      selected
+                        ? 'border-primary bg-primary/5'
+                        : 'border-border hover:bg-muted/50'
+                    }`}
+                  >
+                    <div className="flex items-center justify-between">
+                      <span className="text-sm font-medium">
+                        {t(`${option.i18n_key}.label`)}
+                      </span>
+                      <Badge variant={selected ? 'default' : 'outline'}>
+                        L{option.level}
+                      </Badge>
+                    </div>
+                  </button>
+                );
+              })}
+            </div>
+          </section>
+
+          <section className="space-y-3">
+            <div className="flex items-center gap-2">
+              <Timer className="size-4" />
+              <h3 className="text-sm font-semibold">
+                {t('operationTrace.retention')}
+              </h3>
+            </div>
+            <div className="flex flex-wrap items-end gap-2">
+              <label className="flex flex-col gap-1 text-xs">
+                <span className="text-muted-foreground">
+                  {t('operationTrace.retentionDays')}
+                </span>
+                <Input
+                  type="number"
+                  className="w-28"
+                  min={governance?.limits.min_retention_days ?? 1}
+                  max={governance?.limits.max_retention_days ?? 3650}
+                  value={retentionDays}
+                  disabled={!canConfigure}
+                  onChange={(event) =>
+                    setRetentionDays(Number(event.target.value))
+                  }
+                />
+              </label>
+              <label className="flex flex-col gap-1 text-xs">
+                <span className="text-muted-foreground">
+                  {t('operationTrace.maxRows')}
+                </span>
+                <Input
+                  type="number"
+                  className="w-32"
+                  min={governance?.limits.min_max_rows ?? 100}
+                  max={governance?.limits.max_max_rows ?? 500000}
+                  value={maxRows}
+                  disabled={!canConfigure}
+                  onChange={(event) => setMaxRows(Number(event.target.value))}
+                />
+              </label>
+              <label className="flex flex-col gap-1 text-xs">
+                <span className="text-muted-foreground">
+                  {t('operationTrace.dedupeWindow')}
+                </span>
+                <Input
+                  type="number"
+                  className="w-28"
+                  min={governance?.limits.min_dedupe_window_seconds ?? 0}
+                  max={governance?.limits.max_dedupe_window_seconds ?? 3600}
+                  value={dedupeSeconds}
+                  disabled={!canConfigure}
+                  onChange={(event) =>
+                    setDedupeSeconds(Number(event.target.value))
+                  }
+                />
+              </label>
+            </div>
+          </section>
+        </div>
+
+        <DialogFooter className="gap-2 sm:justify-end">
+          <Button
+            variant="outline"
+            size="sm"
+            onClick={resetDraft}
+            disabled={!dirty || saving}
+          >
+            {t('operationTrace.reset')}
+          </Button>
+          <Button
+            size="sm"
+            onClick={() =>
+              onSave({ level, retentionDays, maxRows, dedupeSeconds })
+            }
+            disabled={!dirty || saving || !canConfigure}
+          >
+            {saving ? <Loader2 className="size-3.5 animate-spin" /> : null}
+            {t('operationTrace.save')}
+          </Button>
+        </DialogFooter>
+      </DialogContent>
+    </Dialog>
+  );
+}
diff --git a/web/src/app/home/components/workspace-settings/OperationTracePanel.tsx b/web/src/app/home/components/workspace-settings/OperationTracePanel.tsx
index 31ed3d04d..7940d4873 100644
--- a/web/src/app/home/components/workspace-settings/OperationTracePanel.tsx
+++ b/web/src/app/home/components/workspace-settings/OperationTracePanel.tsx
@@ -15,12 +15,6 @@ import { toast } from 'sonner';
 
 import { Badge } from '@/components/ui/badge';
 import { Button } from '@/components/ui/button';
-import {
-  Collapsible,
-  CollapsibleContent,
-  CollapsibleTrigger,
-} from '@/components/ui/collapsible';
-import { Input } from '@/components/ui/input';
 import { Item, ItemContent, ItemMedia, ItemTitle } from '@/components/ui/item';
 import {
   Select,
@@ -41,6 +35,7 @@ import type {
 } from '@/app/infra/entities/operation-log';
 import { Download } from 'lucide-react';
 import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
+import OperationTraceCollectionDialog from './OperationTraceCollectionDialog';
 import {
   PanelBody,
   PanelToolbar,
@@ -343,12 +338,7 @@ export default function OperationTracePanel({
   const [saving, setSaving] = useState(false);
   const [exporting, setExporting] = useState(false);
   const [query, setQuery] = useState<OperationLogQuery>({});
-  const [retentionDays, setRetentionDays] = useState<number>(30);
-  const [maxRows, setMaxRows] = useState<number>(20000);
-  const [dedupeSeconds, setDedupeSeconds] = useState<number>(60);
-  // Retention is an advanced, rarely-touched policy: it collapses by default so
-  // the capture level and the records list stay the focus of the panel.
-  const [retentionOpen, setRetentionOpen] = useState(false);
+  const [settingsOpen, setSettingsOpen] = useState(false);
   // Only one record shows its full change detail at a time, so the list stays
   // scannable: the payload diff is long and is opt-in per row.
   const [expandedId, setExpandedId] = useState<number | null>(null);
@@ -395,9 +385,6 @@ export default function OperationTracePanel({
         setGovernance(governanceResponse);
         setFilters(filtersResponse);
         setPage(pageResponse);
-        setRetentionDays(governanceResponse.retention_days);
-        setMaxRows(governanceResponse.max_rows);
-        setDedupeSeconds(governanceResponse.dedupe_window_seconds);
       } catch {
         toast.error(t('operationTrace.loadFailed'));
       } finally {
@@ -463,16 +450,28 @@ export default function OperationTracePanel({
     }));
   }
 
-  async function changeLevel(level: OperationLevel) {
+  async function saveCollectionSettings(settings: {
+    level: OperationLevel;
+    retentionDays: number;
+    maxRows: number;
+    dedupeSeconds: number;
+  }) {
     if (!canConfigure) return;
     setSaving(true);
     try {
-      const updated = await backendClient.updateOperationGovernance({ level });
+      // One governed write per dialog save: the level and the retention policy
+      // are applied together, and the maintenance loop enforces them after.
+      const updated = await backendClient.updateOperationGovernance({
+        level: settings.level,
+        retention_days: settings.retentionDays,
+        max_rows: settings.maxRows,
+        dedupe_window_seconds: settings.dedupeSeconds,
+      });
       setGovernance(updated);
-      toast.success(t('operationTrace.levelUpdated'));
+      toast.success(t('operationTrace.settingsUpdated'));
       await load({ ...query, offset: 0 });
     } catch {
-      toast.error(t('operationTrace.levelUpdateFailed'));
+      toast.error(t('operationTrace.settingsUpdateFailed'));
     } finally {
       setSaving(false);
     }
@@ -496,8 +495,7 @@ export default function OperationTracePanel({
       });
       const response = await backendClient.downloadFile(url);
       const disposition = response.headers['content-disposition'] as
-        | string
-        | undefined;
+        string | undefined;
       const match = disposition?.match(/filename="?([^";\n]+)"?/);
       const filename = match?.[1] ?? `operation-logs-${Date.now()}.csv`;
       const blob = new Blob([response.data], {
@@ -518,33 +516,6 @@ export default function OperationTracePanel({
     }
   }
 
-  async function saveRetention() {
-    if (!canConfigure || !governance) return;
-    setSaving(true);
-    try {
-      // Saving the retention policy applies it immediately and keeps the
-      // maintenance loop enforcing it afterwards; there is no manual prune.
-      const updated = await backendClient.updateOperationGovernance({
-        level: governance.configured_level,
-        retention_days: retentionDays,
-        max_rows: maxRows,
-        dedupe_window_seconds: dedupeSeconds,
-      });
-      setGovernance(updated);
-      toast.success(t('operationTrace.retentionUpdated'));
-      await load({ ...query, offset: 0 });
-    } catch {
-      toast.error(t('operationTrace.retentionUpdateFailed'));
-    } finally {
-      setSaving(false);
-    }
-  }
-
-  const levelOptions = useMemo(
-    () => governance?.supported_levels ?? [],
-    [governance],
-  );
-
   if (loading && !governance) {
     return (
       <div className="flex flex-1 items-center justify-center">
@@ -556,23 +527,33 @@ export default function OperationTracePanel({
   return (
     <>
       <PanelToolbar>
-        <div className="min-w-0">
+        <div className="flex min-w-0 items-center gap-2">
           <p className="truncate text-sm font-medium">
             {t('operationTrace.title')}
           </p>
-        </div>
-        {/* Keep the actions in one right-aligned cluster: the panel toolbar
-            spreads direct children apart, so they are grouped instead. */}
-        <div className="ml-auto flex items-center gap-2">
           {governance && (
             <Badge
               variant={governance.configured_level > 0 ? 'default' : 'outline'}
             >
-              {t(
-                `operationTrace.levelNames.${governance.configured_level_name}`,
-              )}
+              {t('operationTrace.levelBadge', {
+                level: t(
+                  `operationTrace.levelNames.${governance.configured_level_name}`,
+                ),
+              })}
             </Badge>
           )}
+        </div>
+        {/* Keep the actions in one right-aligned cluster: the panel toolbar
+            spreads direct children apart, so they are grouped instead. */}
+        <div className="ml-auto flex items-center gap-2">
+          <Button
+            size="sm"
+            variant="outline"
+            onClick={() => setSettingsOpen(true)}
+          >
+            <SlidersHorizontal className="size-3.5" />
+            {t('operationTrace.collectionSettings')}
+          </Button>
           <Button
             size="sm"
             variant="outline"
@@ -603,121 +584,14 @@ export default function OperationTracePanel({
       </PanelToolbar>
 
       <PanelBody className="space-y-6">
-        <section className="space-y-3">
-          <div className="flex items-center gap-2">
-            <SlidersHorizontal className="size-4" />
-            <h3 className="text-sm font-semibold">
-              {t('operationTrace.captureLevel')}
-            </h3>
-          </div>
-          <div className="grid gap-2 sm:grid-cols-3">
-            {levelOptions.map((option) => {
-              const selected = governance?.configured_level === option.level;
-              return (
-                <button
-                  key={option.level}
-                  type="button"
-                  disabled={!canConfigure || saving}
-                  onClick={() => void changeLevel(option.level)}
-                  className={`rounded-lg border p-3 text-left transition-colors disabled:opacity-60 ${
-                    selected
-                      ? 'border-primary bg-primary/5'
-                      : 'border-border hover:bg-muted/50'
-                  }`}
-                >
-                  <div className="flex items-center justify-between">
-                    <span className="text-sm font-medium">
-                      {t(`${option.i18n_key}.label`)}
-                    </span>
-                    <Badge variant={selected ? 'default' : 'outline'}>
-                      L{option.level}
-                    </Badge>
-                  </div>
-                </button>
-              );
-            })}
-          </div>
-        </section>
-
-        <Collapsible open={retentionOpen} onOpenChange={setRetentionOpen}>
-          <section className="space-y-3">
-            <CollapsibleTrigger className="flex w-full items-center gap-2 text-left">
-              <ChevronDown
-                className={`size-4 shrink-0 text-muted-foreground transition-transform ${
-                  retentionOpen ? '' : '-rotate-90'
-                }`}
-              />
-              <h3 className="text-sm font-semibold">
-                {t('operationTrace.retention')}
-              </h3>
-              {!retentionOpen && (
-                <span className="ml-auto truncate text-xs text-muted-foreground">
-                  {t('operationTrace.retentionSummary', {
-                    days: retentionDays,
-                    rows: maxRows,
-                  })}
-                </span>
-              )}
-            </CollapsibleTrigger>
-            <CollapsibleContent>
-              <div className="flex flex-wrap items-end gap-2 pt-3">
-                <label className="flex flex-col gap-1 text-xs">
-                  <span className="text-muted-foreground">
-                    {t('operationTrace.retentionDays')}
-                  </span>
-                  <Input
-                    type="number"
-                    className="w-28"
-                    min={governance?.limits.min_retention_days ?? 1}
-                    max={governance?.limits.max_retention_days ?? 3650}
-                    value={retentionDays}
-                    disabled={!canConfigure}
-                    onChange={(event) =>
-                      setRetentionDays(Number(event.target.value))
-                    }
-                  />
-                </label>
-                <label className="flex flex-col gap-1 text-xs">
-                  <span className="text-muted-foreground">
-                    {t('operationTrace.maxRows')}
-                  </span>
-                  <Input
-                    type="number"
-                    className="w-32"
-                    min={governance?.limits.min_max_rows ?? 100}
-                    max={governance?.limits.max_max_rows ?? 500000}
-                    value={maxRows}
-                    disabled={!canConfigure}
-                    onChange={(event) => setMaxRows(Number(event.target.value))}
-                  />
-                </label>
-                <label className="flex flex-col gap-1 text-xs">
-                  <span className="text-muted-foreground">
-                    {t('operationTrace.dedupeWindow')}
-                  </span>
-                  <Input
-                    type="number"
-                    className="w-28"
-                    min={governance?.limits.min_dedupe_window_seconds ?? 0}
-                    max={governance?.limits.max_dedupe_window_seconds ?? 3600}
-                    value={dedupeSeconds}
-                    disabled={!canConfigure}
-                    onChange={(event) =>
-                      setDedupeSeconds(Number(event.target.value))
-                    }
-                  />
-                </label>
-                <Button
-                  size="sm"
-                  onClick={() => void saveRetention()}
-                  disabled={!canConfigure || saving}
-                >
-                  {t('operationTrace.save')}
-                </Button>
-              </div>
-            </CollapsibleContent>
-          </section>
-        </Collapsible>
+        <OperationTraceCollectionDialog
+          open={settingsOpen}
+          onOpenChange={setSettingsOpen}
+          governance={governance}
+          canConfigure={canConfigure}
+          saving={saving}
+          onSave={(settings) => void saveCollectionSettings(settings)}
+        />
 
         <section className="space-y-3">
           <div className="flex flex-wrap items-center justify-between gap-2">
diff --git a/web/src/app/home/components/workspace-settings/PlanRevisionStack.tsx b/web/src/app/home/components/workspace-settings/PlanRevisionStack.tsx
new file mode 100644
index 000000000..22e5130a2
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/PlanRevisionStack.tsx
@@ -0,0 +1,84 @@
+import { useState, type ReactNode } from 'react';
+import { ChevronLeft } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+import { cn } from '@/lib/utils';
+
+type Layer = 'current' | 'latest';
+
+/**
+ * Stacks the purchased revision in front of the plan's latest revision. The
+ * toggle only changes which card is shown; it never changes the subscription.
+ * Layers stay opaque; the back card is dimmed with an overlay instead of
+ * opacity so the two cards never show through each other while swapping.
+ */
+export default function PlanRevisionStack({
+  current,
+  latest,
+}: {
+  current: ReactNode;
+  latest: ReactNode;
+}) {
+  const { t } = useTranslation();
+  const [front, setFront] = useState<Layer>('current');
+  const back: Layer = front === 'current' ? 'latest' : 'current';
+  const label = t(
+    back === 'latest' ? 'planRevision.showLatest' : 'planRevision.showCurrent',
+  );
+  const layer = (key: Layer, node: ReactNode) => {
+    const isFront = front === key;
+    return (
+      <div
+        key={key}
+        data-revision-layer={key}
+        data-front={isFront || undefined}
+        aria-hidden={!isFront || undefined}
+        inert={!isFront}
+        className={cn(
+          'relative mr-8 flex min-w-0 origin-right rounded-xl bg-background transition-transform duration-500 ease-[cubic-bezier(0.22,1,0.36,1)] will-change-transform after:pointer-events-none after:absolute after:inset-0 after:rounded-xl after:bg-background after:transition-opacity after:duration-500 [grid-area:stack] motion-reduce:transition-none [&>article]:flex-1',
+          isFront
+            ? 'z-10 after:opacity-0'
+            : 'pointer-events-none z-0 translate-x-8 scale-[0.95] after:opacity-50 [&>article]:shadow-none',
+        )}
+      >
+        {node}
+      </div>
+    );
+  };
+  return (
+    <div
+      className='relative grid min-w-0 [grid-template-areas:"stack"]'
+      data-plan-revision-stack
+      data-front-layer={front}
+    >
+      {layer('current', current)}
+      {layer('latest', latest)}
+      <button
+        type="button"
+        aria-label={label}
+        title={label}
+        data-revision-toggle
+        onClick={(event) => {
+          setFront(back);
+          // Pointer clicks should not leave the strip highlighted.
+          if (event.detail > 0) event.currentTarget.blur();
+        }}
+        className="group absolute inset-y-2 right-0 z-20 flex w-8 flex-col items-center justify-center gap-2 rounded-r-xl text-muted-foreground outline-none transition-colors hover:bg-primary/5 hover:text-primary focus-visible:bg-primary/10 focus-visible:text-primary"
+      >
+        <ChevronLeft
+          className="size-4 shrink-0 transition-transform group-hover:-translate-x-0.5"
+          aria-hidden
+        />
+        <span
+          className="text-xs font-medium [writing-mode:vertical-rl]"
+          aria-hidden
+        >
+          {t(
+            back === 'latest'
+              ? 'planRevision.latestTab'
+              : 'planRevision.currentTab',
+          )}
+        </span>
+      </button>
+    </div>
+  );
+}
diff --git a/web/src/app/home/components/workspace-settings/WorkspacePlanBadge.tsx b/web/src/app/home/components/workspace-settings/WorkspacePlanBadge.tsx
new file mode 100644
index 000000000..b5eb7471b
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/WorkspacePlanBadge.tsx
@@ -0,0 +1,28 @@
+import { useTranslation } from 'react-i18next';
+
+import { Badge } from '@/components/ui/badge';
+import { cn } from '@/lib/utils';
+
+export default function WorkspacePlanBadge({
+  planName,
+}: {
+  planName?: string | null;
+}) {
+  const { t } = useTranslation();
+  const isPaidPlan =
+    Boolean(planName?.trim()) && planName?.trim().toLowerCase() !== 'free';
+
+  return (
+    <Badge
+      variant="outline"
+      className={cn(
+        'text-[11px] uppercase',
+        isPaidPlan
+          ? 'border-blue-200 bg-blue-50 font-semibold text-blue-700 dark:border-blue-400/30 dark:bg-blue-500/15 dark:text-blue-300'
+          : 'bg-muted text-muted-foreground',
+      )}
+    >
+      {planName || t('workspace.planUnavailable')}
+    </Badge>
+  );
+}
diff --git a/web/src/app/home/components/workspace-settings/WorkspaceSettingsPanel.tsx b/web/src/app/home/components/workspace-settings/WorkspaceSettingsPanel.tsx
index ecd3e08cb..9e614c5d0 100644
--- a/web/src/app/home/components/workspace-settings/WorkspaceSettingsPanel.tsx
+++ b/web/src/app/home/components/workspace-settings/WorkspaceSettingsPanel.tsx
@@ -1,7 +1,14 @@
-import { useCallback, useEffect, useMemo, useState } from 'react';
+import {
+  lazy,
+  Suspense,
+  useCallback,
+  useEffect,
+  useMemo,
+  useState,
+} from 'react';
 import {
   Copy,
-  ExternalLink,
+  ArrowRight,
   Loader2,
   Trash2,
   UserPlus,
@@ -39,6 +46,9 @@ import {
   PanelBody,
   PanelToolbar,
 } from '@/app/home/components/settings-dialog/panel-layout';
+import WorkspacePlanBadge from './WorkspacePlanBadge';
+
+const CloudPlansDialog = lazy(() => import('./CloudPlansDialog'));
 
 interface WorkspaceSettingsPanelProps {
   active: boolean;
@@ -51,16 +61,31 @@ const ASSIGNABLE_ROLES: Exclude<WorkspaceRole, 'owner'>[] = [
   'viewer',
 ];
 
+const MESSAGE_LANGUAGES: Record<string, string> = {
+  en_US: 'English',
+  zh_Hans: '简体中文',
+  zh_Hant: '繁體中文',
+  ja_JP: '日本語',
+  es_ES: 'Español',
+  ru_RU: 'Русский',
+  th_TH: 'ไทย',
+  vi_VN: 'Tiếng Việt',
+};
+
 export default function WorkspaceSettingsPanel({
   active,
 }: WorkspaceSettingsPanelProps) {
   const { t } = useTranslation();
+  const [plansOpen, setPlansOpen] = useState(false);
   const [workspaceInfo, setWorkspaceInfo] = useState<CurrentWorkspace | null>(
     null,
   );
   const [members, setMembers] = useState<WorkspaceMembership[]>([]);
   const [invitations, setInvitations] = useState<WorkspaceInvitation[]>([]);
   const [loading, setLoading] = useState(false);
+  const [messageLanguage, setMessageLanguage] = useState('en_US');
+  const [supportedLanguages, setSupportedLanguages] = useState<string[]>([]);
+  const [savingLanguage, setSavingLanguage] = useState(false);
   const [inviteEmail, setInviteEmail] = useState('');
   const [inviteRole, setInviteRole] =
     useState<Exclude<WorkspaceRole, 'owner'>>('viewer');
@@ -79,15 +104,17 @@ export default function WorkspaceSettingsPanel({
   const canInvite = permissions.has('member.invite');
   const canUpdateMembers = permissions.has('member.update_role');
   const canRemoveMembers = permissions.has('member.remove');
-  const cloudPortalURL = workspaceInfo
-    ? `${systemInfo.cloud_service_url.replace(/\/$/, '')}/cloud?workspace=${encodeURIComponent(workspaceInfo.workspace.uuid)}&step=plan`
-    : '';
 
   const loadWorkspace = useCallback(async () => {
     setLoading(true);
     try {
-      const current = await backendClient.getCurrentWorkspace();
+      const [current, languageSettings] = await Promise.all([
+        backendClient.getCurrentWorkspace(),
+        backendClient.getWorkspaceLanguage(),
+      ]);
       setWorkspaceInfo(current);
+      setMessageLanguage(languageSettings.language);
+      setSupportedLanguages(languageSettings.supported_languages);
 
       const [memberResponse, invitationResponse] = await Promise.all([
         current.permissions.includes('member.view')
@@ -106,6 +133,19 @@ export default function WorkspaceSettingsPanel({
     }
   }, [t]);
 
+  async function saveLanguage(locale: string) {
+    setSavingLanguage(true);
+    try {
+      const result = await backendClient.setWorkspaceLanguage(locale);
+      setMessageLanguage(result.language);
+      toast.success(t('workspaceLanguage.saved'));
+    } catch {
+      toast.error(t('workspaceLanguage.saveFailed'));
+    } finally {
+      setSavingLanguage(false);
+    }
+  }
+
   useEffect(() => {
     if (active) void loadWorkspace();
   }, [active, loadWorkspace]);
@@ -193,11 +233,21 @@ export default function WorkspaceSettingsPanel({
 
   return (
     <>
-      <PanelToolbar>
-        <div className="min-w-0">
-          <p className="truncate text-sm font-medium">
-            {workspaceInfo?.workspace.name ?? t('workspace.title')}
-          </p>
+      <PanelToolbar className="gap-4 py-4">
+        <div className="min-w-0 flex-1 basis-56 space-y-1.5">
+          <div className="flex min-w-0 items-center gap-2">
+            <p className="truncate text-sm font-semibold">
+              {workspaceInfo?.workspace.name ?? t('workspace.title')}
+            </p>
+            {workspaceInfo && (
+              <Badge
+                variant="secondary"
+                className="shrink-0 text-[11px] font-normal"
+              >
+                {t(`workspace.roles.${workspaceInfo.membership.role}`)}
+              </Badge>
+            )}
+          </div>
           <p className="text-xs text-muted-foreground">
             {t(
               isCloudProjection
@@ -206,22 +256,62 @@ export default function WorkspaceSettingsPanel({
             )}
           </p>
         </div>
-        {workspaceInfo && (
-          <Badge variant="secondary">
-            {t(`workspace.roles.${workspaceInfo.membership.role}`)}
-          </Badge>
-        )}
         {isCloudProjection && workspaceInfo && (
-          <Button asChild size="sm">
-            <a href={cloudPortalURL} target="_blank" rel="noopener noreferrer">
-              {t('workspace.upgradePlan')}
-              <ExternalLink className="size-3.5" />
-            </a>
-          </Button>
+          <div className="ml-auto flex max-w-full flex-wrap items-center justify-end gap-3">
+            <div className="space-y-1 text-right">
+              <p className="text-[11px] text-muted-foreground">
+                {t('workspace.currentPlan')}
+              </p>
+              <WorkspacePlanBadge planName={workspaceInfo.plan_name} />
+            </div>
+            <Button
+              size="sm"
+              className="shrink-0"
+              onClick={() => setPlansOpen(true)}
+            >
+              {t('cloudPlans.title')}
+              <ArrowRight className="size-3.5" />
+            </Button>
+          </div>
         )}
       </PanelToolbar>
 
       <PanelBody className="space-y-6">
+        <section className="space-y-3">
+          <div>
+            <h3 id="workspace-language-label" className="text-sm font-semibold">
+              {t('workspaceLanguage.label')}
+            </h3>
+            <p
+              id="workspace-language-description"
+              className="text-xs text-muted-foreground"
+            >
+              {t('workspaceLanguage.description')}
+            </p>
+          </div>
+          <Select
+            value={messageLanguage}
+            onValueChange={saveLanguage}
+            disabled={
+              loading || savingLanguage || !permissions.has('workspace.update')
+            }
+          >
+            <SelectTrigger
+              className="w-full sm:w-64"
+              aria-labelledby="workspace-language-label"
+              aria-describedby="workspace-language-description"
+            >
+              <SelectValue />
+            </SelectTrigger>
+            <SelectContent>
+              {supportedLanguages.map((locale) => (
+                <SelectItem key={locale} value={locale}>
+                  {MESSAGE_LANGUAGES[locale] || locale}
+                </SelectItem>
+              ))}
+            </SelectContent>
+          </Select>
+        </section>
         {canInvite && (
           <section className="space-y-3">
             <div>
@@ -404,6 +494,17 @@ export default function WorkspaceSettingsPanel({
           </section>
         )}
       </PanelBody>
+      {isCloudProjection && workspaceInfo && plansOpen && (
+        <Suspense fallback={null}>
+          <CloudPlansDialog
+            key={workspaceInfo.workspace.uuid}
+            open={plansOpen}
+            onOpenChange={setPlansOpen}
+            workspace={workspaceInfo}
+            spaceURL={systemInfo.cloud_service_url}
+          />
+        </Suspense>
+      )}
     </>
   );
 }
diff --git a/web/src/app/home/components/workspace-settings/WorkspaceSwitcher.tsx b/web/src/app/home/components/workspace-settings/WorkspaceSwitcher.tsx
index c8fc53924..5c5ca2b81 100644
--- a/web/src/app/home/components/workspace-settings/WorkspaceSwitcher.tsx
+++ b/web/src/app/home/components/workspace-settings/WorkspaceSwitcher.tsx
@@ -1,4 +1,4 @@
-import { Building2, Check, Settings } from 'lucide-react';
+import { Building2, Settings } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
 
 import {
@@ -15,6 +15,7 @@ import {
   DropdownMenuTrigger,
 } from '@/components/ui/dropdown-menu';
 import { cn } from '@/lib/utils';
+import WorkspacePlanBadge from './WorkspacePlanBadge';
 
 export const OPEN_WORKSPACE_SETTINGS_EVENT = 'langbot:open-workspace-settings';
 
@@ -56,7 +57,10 @@ export default function WorkspaceSwitcher({
           return (
             <DropdownMenuItem
               key={entry.workspace.uuid}
-              className="min-h-11 gap-2 px-2 py-1.5"
+              className={cn(
+                'min-h-11 gap-2 px-2 py-1.5',
+                selected && 'bg-muted',
+              )}
               onClick={() => {
                 if (!selected)
                   void switchWorkspaceAndReload(entry.workspace.uuid);
@@ -67,11 +71,8 @@ export default function WorkspaceSwitcher({
                 {entry.workspace.name}
               </span>
               {entry.workspace.source === 'cloud_projection' && (
-                <span className="rounded-md border bg-muted px-2 py-0.5 text-[11px] font-medium uppercase text-muted-foreground">
-                  {entry.plan_name || t('workspace.planUnavailable')}
-                </span>
+                <WorkspacePlanBadge planName={entry.plan_name} />
               )}
-              {selected && <Check className="size-4" />}
               {selected && (
                 <Button
                   variant="ghost"
diff --git a/web/src/app/home/components/workspace-settings/cloud-plan-platforms.ts b/web/src/app/home/components/workspace-settings/cloud-plan-platforms.ts
new file mode 100644
index 000000000..e4b9b74a9
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/cloud-plan-platforms.ts
@@ -0,0 +1,201 @@
+// Keep this public catalog aligned with Space web/lib/platform-adapters.ts.
+// Do not use the workspace-filtered adapter API to describe other plans.
+export const KNOWN_ADAPTERS = [
+  {
+    name: 'LINE',
+    label: 'LINE',
+    icon: '/platforms/line.png',
+  },
+  {
+    name: 'aiocqhttp',
+    label: 'OneBot v11',
+    icon: '/platforms/aiocqhttp.svg',
+  },
+  {
+    name: 'aiocqhttp-omni',
+    label: 'OneBot v11 (Omni)',
+    icon: '/platforms/aiocqhttp.svg',
+  },
+  {
+    name: 'dingtalk',
+    label: 'DingTalk',
+    icon: '/platforms/dingtalk.svg',
+  },
+  {
+    name: 'dingtalk-omni',
+    label: 'DingTalk (Omni)',
+    icon: '/platforms/dingtalk.svg',
+  },
+  {
+    name: 'discord',
+    label: 'Discord',
+    icon: '/platforms/discord.svg',
+  },
+  {
+    name: 'discord-omni',
+    label: 'Discord (Omni)',
+    icon: '/platforms/discord.svg',
+  },
+  {
+    name: 'http_bot',
+    label: 'HTTP Bot',
+    icon: '/platforms/http_bot.svg',
+  },
+  {
+    name: 'kook',
+    label: 'KOOK',
+    icon: '/platforms/kook.png',
+  },
+  {
+    name: 'kook-omni',
+    label: 'KOOK (Omni)',
+    icon: '/platforms/kook.png',
+  },
+  {
+    name: 'lark',
+    label: 'Lark',
+    icon: '/platforms/lark.svg',
+  },
+  {
+    name: 'lark-omni',
+    label: 'Lark / Feishu (Omni)',
+    icon: '/platforms/lark.svg',
+  },
+  {
+    name: 'matrix',
+    label: 'Matrix',
+    icon: '/platforms/matrix.png',
+  },
+  {
+    name: 'mattermost',
+    label: 'Mattermost',
+    icon: '/platforms/mattermost.svg',
+  },
+  {
+    name: 'officialaccount',
+    label: 'Official Account',
+    icon: '/platforms/officialaccount.png',
+  },
+  {
+    name: 'officialaccount-omni',
+    label: 'Official Account (Omni)',
+    icon: '/platforms/officialaccount.png',
+  },
+  {
+    name: 'openclaw-weixin',
+    label: 'OpenClaw WeChat',
+    icon: '/platforms/openclaw_weixin.png',
+  },
+  {
+    name: 'qqofficial',
+    label: 'QQ Official API',
+    icon: '/platforms/qqofficial.svg',
+  },
+  {
+    name: 'qqofficial-omni',
+    label: 'QQ Official API (Omni)',
+    icon: '/platforms/qqofficial.svg',
+  },
+  {
+    name: 'satori',
+    label: 'Satori',
+    icon: '/platforms/satori.png',
+  },
+  {
+    name: 'slack',
+    label: 'Slack',
+    icon: '/platforms/slack.png',
+  },
+  {
+    name: 'slack-omni',
+    label: 'Slack (Omni)',
+    icon: '/platforms/slack.png',
+  },
+  {
+    name: 'telegram',
+    label: 'Telegram',
+    icon: '/platforms/telegram.svg',
+  },
+  {
+    name: 'telegram-omni',
+    label: 'Telegram (Omni)',
+    icon: '/platforms/telegram.svg',
+  },
+  {
+    name: 'web_page_bot',
+    label: 'Page Bot',
+    icon: '/platforms/webpage.webp',
+  },
+  {
+    name: 'wechatpad',
+    label: 'WeChatPad',
+    icon: '/platforms/wechatpad.png',
+  },
+  {
+    name: 'wecom',
+    label: 'WeCom',
+    icon: '/platforms/wecom.png',
+  },
+  {
+    name: 'wecom-omni',
+    label: 'WeCom (Omni)',
+    icon: '/platforms/wecom.png',
+  },
+  {
+    name: 'wecombot',
+    label: 'WeComBot',
+    icon: '/platforms/wecombot.png',
+  },
+  {
+    name: 'wecombot-omni',
+    label: 'WeComBot (Omni)',
+    icon: '/platforms/wecombot.png',
+  },
+  {
+    name: 'wecomcs',
+    label: 'WeComCustomerService',
+    icon: '/platforms/wecomcs.png',
+  },
+  {
+    name: 'wecomcs-omni',
+    label: 'WeCom Customer Service (Omni)',
+    icon: '/platforms/wecomcs.png',
+  },
+];
+
+export function supportedPlanAdapters(disabled: string[] = []) {
+  const blocked = new Set(
+    disabled.map((name) =>
+      name === 'line'
+        ? 'LINE'
+        : name.endsWith('-eba')
+          ? name.slice(0, -4) + '-omni'
+          : name,
+    ),
+  );
+  return KNOWN_ADAPTERS.filter(
+    (adapter) =>
+      !blocked.has(adapter.name) &&
+      !(
+        adapter.name.endsWith('-omni') && blocked.has(adapter.name.slice(0, -5))
+      ),
+  );
+}
+
+/** Collapse transport variants into one platform logo after applying restrictions. */
+export function supportedPlanPlatforms(disabled: string[] = []) {
+  const platforms = new Map<
+    string,
+    { name: string; label: string; icon: string }
+  >();
+  for (const adapter of supportedPlanAdapters(disabled)) {
+    const name = adapter.name.replace(/-omni$/, '');
+    if (!platforms.has(name))
+      platforms.set(name, {
+        ...adapter,
+        name,
+        label: adapter.label.replace(/ \(Omni\)$/, ''),
+      });
+  }
+  return [...platforms.values()];
+}
diff --git a/web/src/app/home/components/workspace-settings/cloud-plans.ts b/web/src/app/home/components/workspace-settings/cloud-plans.ts
new file mode 100644
index 000000000..1ad7a392a
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/cloud-plans.ts
@@ -0,0 +1,183 @@
+export type CloudPlanBillingCycle = 'monthly' | 'annual';
+export type CloudPlanCurrency = 'CNY' | 'USD' | 'JPY';
+export type PlanLocalizedText = Partial<Record<string, string>>;
+export interface PublicCloudPlan {
+  entitlement_features?: Record<string, boolean>;
+  uuid: string;
+  name: string;
+  revision: number;
+  display_name: PlanLocalizedText;
+  description: PlanLocalizedText;
+  price_cny: string;
+  price_usd: string;
+  price_jpy: string;
+  limits: Record<string, number>;
+  features: string[];
+  is_active: boolean;
+  subscription_disabled?: boolean;
+  sort_order: number;
+  highlight: boolean;
+  chat_history_retention_days: number;
+  monthly_bonus_credits: number;
+  disabled_adapters: string[];
+  display_config?: {
+    icon?: string;
+    order?: string[];
+    quota_summary?: PlanLocalizedText;
+    tooltips?: Partial<Record<string, PlanLocalizedText>>;
+    quotas?: Record<
+      string,
+      { label?: PlanLocalizedText; tooltip?: PlanLocalizedText }
+    >;
+    items?: Array<{
+      label: PlanLocalizedText;
+      tooltip?: PlanLocalizedText;
+      icon: string;
+    }>;
+  } | null;
+}
+
+export function localizedPlanText(
+  values: PlanLocalizedText | undefined,
+  language: string,
+): string {
+  if (!values) return '';
+  const locale = language.replace(/-/g, '_');
+  const key = Object.keys(values).find(
+    (key) => key.toLowerCase() === locale.toLowerCase(),
+  );
+  return (
+    (key && values[key]) ||
+    values.en_US ||
+    values.zh_Hans ||
+    Object.values(values).find(Boolean) ||
+    ''
+  );
+}
+
+export function cloudPlanCurrencyForLanguage(
+  language: string,
+): CloudPlanCurrency {
+  const locale = language.replace(/_/g, '-').toLowerCase();
+  if (locale === 'zh-hans' || locale.startsWith('zh-hans-')) return 'CNY';
+  if (locale === 'ja' || locale.startsWith('ja-')) return 'JPY';
+  return 'USD';
+}
+
+export function cloudPlanPrice(
+  plan: PublicCloudPlan,
+  cycle: CloudPlanBillingCycle,
+  currency: CloudPlanCurrency,
+): number {
+  const amount = Number(
+    plan[
+      currency === 'CNY'
+        ? 'price_cny'
+        : currency === 'JPY'
+          ? 'price_jpy'
+          : 'price_usd'
+    ],
+  );
+  if (!Number.isFinite(amount) || amount < 0)
+    throw new Error('Invalid public plan price');
+  return amount * (cycle === 'annual' ? 10 : 1);
+}
+
+function spaceURL(base: string, path: string): URL {
+  const url = new URL(base);
+  if (
+    !['http:', 'https:'].includes(url.protocol) ||
+    url.username ||
+    url.password
+  )
+    throw new Error('Invalid Space URL');
+  url.pathname = `${url.pathname.replace(/\/$/, '')}${path}`;
+  url.search = '';
+  url.hash = '';
+  return url;
+}
+
+export function cloudPlanCheckoutURL(
+  base: string,
+  workspace: string,
+  plan: string,
+  billing: CloudPlanBillingCycle,
+  currency: CloudPlanCurrency,
+): string {
+  const url = spaceURL(base, '/cloud');
+  url.search = new URLSearchParams({
+    environment: 'new',
+    step: 'plan',
+    workspace,
+    plan,
+    billing,
+    currency,
+  }).toString();
+  return url.toString();
+}
+
+/** Workspace navigation only; Space authenticates and authorizes independently. */
+export function cloudSubscriptionManagementURL(
+  base: string,
+  workspace: string,
+): string {
+  const url = spaceURL(base, '/cloud');
+  url.search = new URLSearchParams({
+    environment: 'new',
+    step: 'subscription',
+    workspace,
+  }).toString();
+  return url.toString();
+}
+
+export function cloudPlanPlatformIconURL(base: string, icon: string): string {
+  return spaceURL(base, icon).toString();
+}
+
+/** This request is deliberately independent of Core and Space authenticated clients. */
+export async function fetchPublicCloudPlans(
+  base: string,
+  signal: AbortSignal,
+): Promise<PublicCloudPlan[]> {
+  const response = await fetch(spaceURL(base, '/api/v1/cloud/v2/plans'), {
+    credentials: 'omit',
+    referrerPolicy: 'no-referrer',
+    signal,
+  });
+  if (!response.ok) throw new Error('Could not load public plans');
+  const result = await response.json();
+  if (result.code !== 0 || !Array.isArray(result.data))
+    throw new Error('Invalid plan catalog');
+  const plans = (result.data as PublicCloudPlan[]).filter(
+    (plan) => plan.is_active,
+  );
+  for (const plan of plans) {
+    if (!plan.uuid || !plan.name || !plan.limits)
+      throw new Error('Invalid public plan');
+    for (const currency of ['CNY', 'USD', 'JPY'] as const)
+      cloudPlanPrice(plan, 'monthly', currency);
+  }
+  return plans.sort((a, b) => a.sort_order - b.sort_order);
+}
+
+// Catalog metadata (copy, prices, display items) of one historical revision.
+export async function fetchCloudPlanRevision(
+  base: string,
+  name: string,
+  revision: number,
+  signal: AbortSignal,
+): Promise<PublicCloudPlan> {
+  const response = await fetch(
+    spaceURL(
+      base,
+      `/api/v1/cloud/v2/plans/${encodeURIComponent(name)}/revisions/${revision}`,
+    ),
+    { credentials: 'omit', referrerPolicy: 'no-referrer', signal },
+  );
+  if (!response.ok) throw new Error('Could not load plan revision');
+  const result = await response.json();
+  const plan = result?.data as PublicCloudPlan | undefined;
+  if (result?.code !== 0 || !plan?.name || plan.name !== name)
+    throw new Error('Invalid plan revision');
+  return plan;
+}
diff --git a/web/src/app/home/components/workspace-settings/plan-revisions.ts b/web/src/app/home/components/workspace-settings/plan-revisions.ts
new file mode 100644
index 000000000..0c5cf31ec
--- /dev/null
+++ b/web/src/app/home/components/workspace-settings/plan-revisions.ts
@@ -0,0 +1,48 @@
+/** Only trusted subscription data may populate the current side of a comparison. */
+export interface PlanRevisionBenefits {
+  name: string;
+  revision: number;
+  limits: Record<string, number>;
+  entitlement_features?: Record<string, boolean>;
+  disabled_adapters?: string[];
+  chat_history_retention_days?: number;
+  monthly_bonus_credits?: number;
+}
+
+export function benefitValues(plan: PlanRevisionBenefits) {
+  const values: Record<string, number | boolean> = {};
+  for (const [key, value] of Object.entries(plan.limits))
+    values[`limit:${key}`] = value;
+  values['limit:chat_history.retention_days'] =
+    plan.limits['chat_history.retention_days'] ??
+    plan.chat_history_retention_days ??
+    30;
+  values.monthly_credits = plan.monthly_bonus_credits ?? 0;
+  for (const [key, value] of Object.entries(plan.entitlement_features ?? {}))
+    values[`feature:${key}`] = value;
+  for (const adapter of plan.disabled_adapters ?? [])
+    values[`feature:platform.${adapter}`] = false;
+  return values;
+}
+
+export function revisionRows(
+  current: PlanRevisionBenefits,
+  latest?: PlanRevisionBenefits,
+) {
+  const left = benefitValues(current);
+  const right = latest ? benefitValues(latest) : {};
+  return [...new Set([...Object.keys(left), ...Object.keys(right)])]
+    .sort()
+    .map((key) => {
+      // Platform adapters are allowed unless explicitly disabled. Other feature
+      // switches must be explicitly granted, matching Core's admission semantics.
+      const fallback = key.startsWith('feature:platform.')
+        ? true
+        : key.startsWith('feature:')
+          ? false
+          : undefined;
+      const before = left[key] ?? fallback;
+      const after = right[key] ?? fallback;
+      return { key, before, after, changed: !!latest && before !== after };
+    });
+}
diff --git a/web/src/app/home/knowledge/components/kb-form/KBForm.tsx b/web/src/app/home/knowledge/components/kb-form/KBForm.tsx
index 9ea13e488..79b3f9b4d 100644
--- a/web/src/app/home/knowledge/components/kb-form/KBForm.tsx
+++ b/web/src/app/home/knowledge/components/kb-form/KBForm.tsx
@@ -112,6 +112,8 @@ export default function KBForm({
   // Dirty tracking: snapshot of saved state for comparison
   const savedSnapshotRef = useRef<string>('');
   const isInitializing = useRef(true);
+  const loadVersion = useRef(0);
+  const initialSettings = useRef(new Set<'config' | 'retrieval'>());
   const suppressNextAutoSelectRef = useRef(false);
 
   // Refs to store validation functions from dynamic forms
@@ -135,36 +137,65 @@ export default function KBForm({
     (e) => e.plugin_id === selectedEngineId,
   );
 
-  // Dirty tracking: compare current form + dynamic settings against saved snapshot
-  const watchedFormValues = form.watch();
-  useEffect(() => {
+  const draftRef = useRef({
+    config: configSettings,
+    retrieval: retrievalSettings,
+  });
+  draftRef.current = { config: configSettings, retrieval: retrievalSettings };
+  const lastDirtyRef = useRef(false);
+  const refreshDirty = useCallback(() => {
     if (!savedSnapshotRef.current || isInitializing.current) return;
-    const currentSnapshot = JSON.stringify({
-      form: watchedFormValues,
-      config: configSettings,
-      retrieval: retrievalSettings,
-    });
-    const dirty = currentSnapshot !== savedSnapshotRef.current;
-    onDirtyChange?.(dirty);
-  }, [watchedFormValues, configSettings, retrievalSettings, onDirtyChange]);
+    const next =
+      !engineInitialized ||
+      JSON.stringify({
+        form: form.getValues(),
+        ...draftRef.current,
+      }) !== savedSnapshotRef.current;
+    if (next !== lastDirtyRef.current) {
+      lastDirtyRef.current = next;
+      onDirtyChange?.(next);
+    }
+  }, [engineInitialized, form, onDirtyChange]);
+  useEffect(() => {
+    refreshDirty();
+    const subscription = form.watch(refreshDirty);
+    return () => subscription.unsubscribe();
+  }, [form, refreshDirty, configSettings, retrievalSettings]);
 
-  const captureSnapshot = () => {
-    savedSnapshotRef.current = JSON.stringify({
-      form: form.getValues(),
-      config: configSettings,
-      retrieval: retrievalSettings,
-    });
-  };
+  function updateSettings(
+    slot: 'config' | 'retrieval',
+    values: Record<string, unknown>,
+  ) {
+    // Normalize only this form's initial emission. Never absorb another field's
+    // edits or depend on a timer to decide when the user may start typing.
+    if (initialSettings.current.delete(slot) && savedSnapshotRef.current) {
+      const baseline = JSON.parse(savedSnapshotRef.current);
+      baseline[slot] = values;
+      savedSnapshotRef.current = JSON.stringify(baseline);
+    }
+    draftRef.current = { ...draftRef.current, [slot]: values };
+    if (slot === 'config') setConfigSettings(values);
+    else setRetrievalSettings(values);
+  }
 
   useEffect(() => {
+    const version = ++loadVersion.current;
     setInitialDataLoaded(false);
     setLoadFailed(false);
-    loadRagEngines()
+    loadRagEngines(version)
       .then(() => {
-        if (initKbId) return loadKbConfig(initKbId);
+        if (version === loadVersion.current && initKbId)
+          return loadKbConfig(initKbId, version);
       })
-      .catch(() => setLoadFailed(true))
-      .finally(() => setInitialDataLoaded(true));
+      .catch(() => {
+        if (version === loadVersion.current) setLoadFailed(true);
+      })
+      .finally(() => {
+        if (version === loadVersion.current) setInitialDataLoaded(true);
+      });
+    return () => {
+      loadVersion.current += 1;
+    };
   }, [initKbId, loadAttempt]);
 
   // Auto-select first engine when engines are loaded and no selection
@@ -188,54 +219,63 @@ export default function KBForm({
     }
   }, [ragEngines, selectedEngineId, isEditing]);
 
-  const loadRagEngines = async () => {
+  const loadRagEngines = async (version: number) => {
     setLoading(true);
     try {
       const resp = await httpClient.getKnowledgeEngines();
-      setRagEngines(resp.engines);
+      if (version === loadVersion.current) setRagEngines(resp.engines);
     } catch (err) {
       throw err;
     } finally {
-      setLoading(false);
+      if (version === loadVersion.current) setLoading(false);
     }
   };
 
-  const loadKbConfig = async (kbId: string) => {
+  const loadKbConfig = async (kbId: string, version: number) => {
     try {
       isInitializing.current = true;
       setIsEditing(true);
 
       const res = await httpClient.getKnowledgeBase(kbId);
+      if (version !== loadVersion.current) return;
       const kb = res.base;
 
       const engineId = kb.knowledge_engine_plugin_id || '';
       setSelectedEngineId(engineId);
 
-      form.reset({
+      const loadedForm = {
         name: kb.name,
         description: kb.description,
         emoji: kb.emoji || '📚',
         ragEngineId: engineId,
-      });
+      };
+      form.reset(loadedForm);
 
       setConfigSettings(kb.creation_settings || {});
       setRetrievalSettings(kb.retrieval_settings || {});
       setEngineInitialized(kb.initialized !== false);
 
-      // Capture snapshot after a tick so dynamic forms have emitted initial values
-      setTimeout(() => {
-        captureSnapshot();
-        isInitializing.current = false;
-        onDirtyChange?.(kb.initialized === false);
-      }, 500);
-    } catch (err) {
+      draftRef.current = {
+        config: kb.creation_settings || {},
+        retrieval: kb.retrieval_settings || {},
+      };
+      savedSnapshotRef.current = JSON.stringify({
+        form: loadedForm,
+        ...draftRef.current,
+      });
+      initialSettings.current = new Set(['config', 'retrieval']);
       isInitializing.current = false;
+      lastDirtyRef.current = kb.initialized === false;
+      onDirtyChange?.(lastDirtyRef.current);
+    } catch (err) {
+      if (version === loadVersion.current) isInitializing.current = false;
       throw err;
     }
   };
 
   const handleEngineChange = useCallback(
     (engineId: string, installedEngine?: KnowledgeEngine) => {
+      initialSettings.current.clear();
       setSelectedEngineId(engineId);
       form.setValue('ragEngineId', engineId, {
         shouldDirty: true,
@@ -300,13 +340,22 @@ export default function KBForm({
         : { defer_initialization: true }),
     };
 
+    const submittedSnapshot = JSON.stringify({
+      form: data,
+      config: configSettings,
+      retrieval: retrievalSettings,
+    });
     if (initKbId) {
       httpClient
         .updateKnowledgeBase(initKbId, kbData)
         .then((res) => {
           setEngineInitialized(true);
-          captureSnapshot();
-          onDirtyChange?.(false);
+          savedSnapshotRef.current = submittedSnapshot;
+          const dirty =
+            JSON.stringify({ form: form.getValues(), ...draftRef.current }) !==
+            submittedSnapshot;
+          lastDirtyRef.current = dirty;
+          onDirtyChange?.(dirty);
           onKbUpdated(res.uuid);
           toast.success(t('knowledge.updateKnowledgeBaseSuccess'));
         })
@@ -515,7 +564,7 @@ export default function KBForm({
                   itemConfigList={configFormItems}
                   initialValues={configSettings as Record<string, object>}
                   onSubmit={(val) =>
-                    setConfigSettings(val as Record<string, unknown>)
+                    updateSettings('config', val as Record<string, unknown>)
                   }
                   isEditing={engineInitialized}
                   externalDependentValues={retrievalSettings}
@@ -542,7 +591,7 @@ export default function KBForm({
                 itemConfigList={retrievalFormItems}
                 initialValues={retrievalSettings as Record<string, object>}
                 onSubmit={(val) =>
-                  setRetrievalSettings(val as Record<string, unknown>)
+                  updateSettings('retrieval', val as Record<string, unknown>)
                 }
                 externalDependentValues={configSettings}
                 onValidate={(validateFn) =>
diff --git a/web/src/app/home/knowledge/components/kb-form/knowledge-engine-marketplace.ts b/web/src/app/home/knowledge/components/kb-form/knowledge-engine-marketplace.ts
index 2abd6ee3f..7850026a2 100644
--- a/web/src/app/home/knowledge/components/kb-form/knowledge-engine-marketplace.ts
+++ b/web/src/app/home/knowledge/components/kb-form/knowledge-engine-marketplace.ts
@@ -13,9 +13,7 @@ const INSTALL_INTENT_KEY_PREFIX = 'langbot-knowledge-engine-install';
 const INSTALL_INTENT_EVENT = 'langbot-knowledge-engine-install-change';
 
 export type KnowledgeEngineMarketplaceErrorCode =
-  | 'version-unavailable'
-  | 'install-timeout'
-  | 'registration-timeout';
+  'version-unavailable' | 'install-timeout' | 'registration-timeout';
 
 export class KnowledgeEngineMarketplaceError extends Error {
   constructor(public readonly code: KnowledgeEngineMarketplaceErrorCode) {
diff --git a/web/src/app/home/layout.tsx b/web/src/app/home/layout.tsx
index 237f56761..056b715f9 100644
--- a/web/src/app/home/layout.tsx
+++ b/web/src/app/home/layout.tsx
@@ -1,5 +1,6 @@
 import HomeSidebar from '@/app/home/components/home-sidebar/HomeSidebar';
 import SurveyWidget from '@/app/home/components/survey/SurveyWidget';
+import InFlightMonitor from '@/app/home/components/inflight/InFlightMonitor';
 import WorkspaceAssistant from '@/app/home/components/WorkspaceAssistant';
 import React, {
   useState,
@@ -280,6 +281,7 @@ function HomeLayoutInner({ children }: { children: React.ReactNode }) {
 
         <SurveyWidget />
         <WorkspaceAssistant />
+        <InFlightMonitor />
       </SidebarInset>
     </SidebarProvider>
   );
diff --git a/web/src/app/home/mcp/MCPDetailContent.tsx b/web/src/app/home/mcp/MCPDetailContent.tsx
index 9403febf2..66cb0607f 100644
--- a/web/src/app/home/mcp/MCPDetailContent.tsx
+++ b/web/src/app/home/mcp/MCPDetailContent.tsx
@@ -31,11 +31,7 @@ import { useCurrentWorkspace } from '@/app/infra/http';
 
 type MCPRuntimeState = 'connected' | 'connecting' | 'error';
 type MCPConnectionState =
-  | 'connected'
-  | 'connecting'
-  | 'error'
-  | 'disabled'
-  | 'disconnected';
+  'connected' | 'connecting' | 'error' | 'disabled' | 'disconnected';
 
 export default function MCPDetailContent({ id }: { id: string }) {
   const isCreateMode = id === 'new';
diff --git a/web/src/app/home/monitoring/components/ConversationTurnList.tsx b/web/src/app/home/monitoring/components/ConversationTurnList.tsx
index afcaa1388..52fae2228 100644
--- a/web/src/app/home/monitoring/components/ConversationTurnList.tsx
+++ b/web/src/app/home/monitoring/components/ConversationTurnList.tsx
@@ -1,23 +1,12 @@
 import React from 'react';
 import { useTranslation } from 'react-i18next';
-import {
-  AlertCircle,
-  Bot,
-  ChevronDown,
-  ChevronRight,
-  Clock,
-  Cpu,
-  Hash,
-  User,
-  Wrench,
-} from 'lucide-react';
+import { AlertCircle, Cpu, MessageCircle, Send, Wrench } from 'lucide-react';
 import { cn } from '@/lib/utils';
+import { Badge } from '@/components/ui/badge';
+import { ProcessorPayload } from '@/app/home/agents/components/PluginProcessorTrace';
+import { toolSummary } from '@/app/home/agents/components/tool-summary';
 import { MessageContentRenderer } from './MessageContentRenderer';
-import {
-  ConversationTurn,
-  hasRenderableMessageContent,
-} from '../utils/conversationTurns';
-import { MonitoringMessage } from '../types/monitoring';
+import { ConversationTurn } from '../utils/conversationTurns';
 
 interface ConversationTurnListProps {
   turns: ConversationTurn[];
@@ -25,124 +14,23 @@ interface ConversationTurnListProps {
   onToggleTurn: (turnId: string) => void;
 }
 
-function shortId(id?: string) {
-  if (!id) return '-';
-  if (id.length <= 12) return id;
-  return `${id.slice(0, 8)}...${id.slice(-4)}`;
-}
-
 function formatDuration(ms: number) {
-  if (!ms) return '0ms';
-  if (ms < 1000) return `${ms}ms`;
-  return `${(ms / 1000).toFixed(2)}s`;
+  return ms < 1000 ? `${ms} ms` : `${(ms / 1000).toFixed(2)} s`;
 }
 
-function truncateDetail(value?: string) {
-  if (!value) return '';
-  return value.length > 1200 ? `${value.slice(0, 1200)}...` : value;
-}
-
-function roleLabel(message: MonitoringMessage | undefined) {
-  const role = message?.role?.toLowerCase();
-  if (role === 'assistant') return 'assistant';
-  if (role === 'user') return 'user';
-  return 'message';
-}
-
-function statusClass(level: ConversationTurn['level']) {
-  if (level === 'error') {
-    return 'border-red-200 bg-red-50 text-red-700 dark:border-red-900 dark:bg-red-950/40 dark:text-red-300';
+function parsePayload(value?: string): unknown {
+  if (!value) return null;
+  try {
+    return JSON.parse(value);
+  } catch {
+    return value;
   }
-  if (level === 'warning') {
-    return 'border-yellow-200 bg-yellow-50 text-yellow-700 dark:border-yellow-900 dark:bg-yellow-950/40 dark:text-yellow-300';
-  }
-  return 'border-green-200 bg-green-50 text-green-700 dark:border-green-900 dark:bg-green-950/40 dark:text-green-300';
 }
 
-function Metric({
-  icon,
-  label,
-  tone = 'default',
-}: {
-  icon: React.ReactNode;
-  label: string;
-  tone?: 'default' | 'error';
-}) {
+function MessageBody({ content }: { content: string }) {
   return (
-    <span
-      className={cn(
-        'inline-flex h-7 items-center gap-1.5 rounded-md border px-2 text-xs font-medium',
-        tone === 'error'
-          ? 'border-red-200 bg-red-50 text-red-700 dark:border-red-900 dark:bg-red-950/40 dark:text-red-300'
-          : 'border-border bg-background text-muted-foreground',
-      )}
-    >
-      {icon}
-      {label}
-    </span>
-  );
-}
-
-function MetaItem({ label, value }: { label: string; value?: string }) {
-  return (
-    <div className="min-w-0 rounded-md bg-background px-3 py-2">
-      <div className="text-xs text-muted-foreground">{label}</div>
-      <div className="truncate text-sm font-medium text-foreground">
-        {value || '-'}
-      </div>
-    </div>
-  );
-}
-
-function MessageLane({
-  label,
-  icon,
-  content,
-  empty,
-  maxLines,
-}: {
-  label: string;
-  icon: React.ReactNode;
-  content?: string;
-  empty: string;
-  maxLines: number;
-}) {
-  return (
-    <div className="grid grid-cols-[5.25rem_minmax(0,1fr)] items-start gap-3 text-sm sm:grid-cols-[6rem_minmax(0,1fr)]">
-      <div className="flex h-7 items-center gap-1.5 text-xs font-medium text-muted-foreground">
-        {icon}
-        <span>{label}</span>
-      </div>
-      <div className="min-w-0 rounded-md bg-muted/45 px-3 py-2 text-foreground">
-        {content && hasRenderableMessageContent(content) ? (
-          <MessageContentRenderer content={content} maxLines={maxLines} />
-        ) : (
-          <span className="italic text-muted-foreground">{empty}</span>
-        )}
-      </div>
-    </div>
-  );
-}
-
-function ExpandedMessage({
-  message,
-  label,
-}: {
-  message: MonitoringMessage;
-  label: string;
-}) {
-  return (
-    <div className="border-t border-border/70 py-3 first:border-t-0 first:pt-0 last:pb-0">
-      <div className="mb-2 flex flex-wrap items-center gap-2 text-xs text-muted-foreground">
-        <span className="rounded-md bg-muted px-2 py-1 font-medium text-foreground">
-          {label}
-        </span>
-        <span>{message.timestamp.toLocaleString()}</span>
-        <span className="font-mono">ID: {shortId(message.id)}</span>
-      </div>
-      <div className="text-sm leading-6 text-foreground">
-        <MessageContentRenderer content={message.messageContent} maxLines={4} />
-      </div>
+    <div className="min-w-0 whitespace-pre-wrap break-words text-sm leading-7 [overflow-wrap:anywhere]">
+      <MessageContentRenderer content={content} maxLines={0} />
     </div>
   );
 }
@@ -153,497 +41,276 @@ export function ConversationTurnList({
   onToggleTurn,
 }: ConversationTurnListProps) {
   const { t } = useTranslation();
-  const [expandedToolCallIds, setExpandedToolCallIds] = React.useState<
-    Record<string, boolean>
-  >({});
+  const turn = turns.find((item) => item.id === expandedTurnId) ?? turns[0];
+  if (!turn) return null;
 
-  const toggleToolCallDetails = (toolCallKey: string) => {
-    setExpandedToolCallIds((previous) => ({
-      ...previous,
-      [toolCallKey]: !previous[toolCallKey],
-    }));
-  };
+  const statusBadge = (status: ConversationTurn['status']) => (
+    <Badge status={status} className="shrink-0">
+      {t(`monitoring.pipelineTrace.${status}`)}
+    </Badge>
+  );
+  const steps = [
+    ...turn.llmCalls.map((call) => ({ kind: 'model' as const, call })),
+    ...turn.toolCalls.map((call) => ({ kind: 'tool' as const, call })),
+    ...turn.errors.map((call) => ({ kind: 'error' as const, call })),
+  ].sort((a, b) => a.call.timestamp.getTime() - b.call.timestamp.getTime());
 
   return (
-    <div className="space-y-4">
-      <div className="flex items-center justify-between text-sm text-muted-foreground">
-        <span className="font-medium text-foreground">
-          {t('monitoring.messageList.turns', {
-            defaultValue: '{{count}} 轮对话',
-            count: turns.length,
-          })}
-        </span>
-      </div>
-
-      {turns.map((turn) => {
-        const expanded = expandedTurnId === turn.id;
-        const firstAssistant = turn.assistantMessages[0];
-        const assistantOverflow = Math.max(
-          turn.assistantMessages.length - 1,
-          0,
-        );
-
-        return (
-          <div
-            key={turn.id}
-            className={cn(
-              'overflow-hidden rounded-xl border bg-card transition-colors',
-              turn.level === 'error' && 'border-red-200 dark:border-red-900',
-            )}
-          >
+    <div className="grid min-h-0 gap-5 lg:h-full lg:grid-cols-[18rem_minmax(0,1fr)]">
+      <aside className="min-h-0 lg:overflow-y-auto lg:pr-1">
+        <div className="mb-3 text-sm font-medium">
+          {t('monitoring.messageList.turns', { count: turns.length })}
+        </div>
+        <div className="max-h-72 overflow-y-auto rounded-xl border lg:max-h-none">
+          {turns.map((item) => (
             <div
+              key={item.id}
               role="button"
               tabIndex={0}
-              className="cursor-pointer p-3 outline-none transition-colors hover:bg-accent/60 focus-visible:ring-2 focus-visible:ring-ring sm:p-5"
-              onClick={() => onToggleTurn(turn.id)}
+              aria-pressed={turn.id === item.id}
+              onClick={() => onToggleTurn(item.id)}
               onKeyDown={(event) => {
-                if (event.key === 'Enter' || event.key === ' ') {
+                if (
+                  event.target === event.currentTarget &&
+                  (event.key === 'Enter' || event.key === ' ')
+                ) {
                   event.preventDefault();
-                  onToggleTurn(turn.id);
+                  onToggleTurn(item.id);
                 }
               }}
+              className={cn(
+                'cursor-pointer border-b border-l-2 border-l-transparent p-3 text-left last:border-b-0 hover:bg-muted/50 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-inset focus-visible:ring-ring',
+                turn.id === item.id && 'border-l-primary bg-primary/5',
+              )}
             >
-              <div className="flex flex-col gap-4 lg:flex-row lg:items-start lg:justify-between">
-                <div className="min-w-0 flex-1">
-                  <div className="mb-2 flex min-w-0 items-center gap-2">
-                    {expanded ? (
-                      <ChevronDown className="h-5 w-5 shrink-0 text-muted-foreground" />
-                    ) : (
-                      <ChevronRight className="h-5 w-5 shrink-0 text-muted-foreground" />
-                    )}
-                    <span className="truncate font-mono text-xs text-muted-foreground">
-                      Turn: {shortId(turn.id)}
-                    </span>
-                  </div>
-
-                  <div className="mb-3 flex min-w-0 flex-wrap items-center gap-2">
-                    <span className="truncate text-sm font-medium text-foreground">
-                      {turn.botName}
-                    </span>
-                    <span className="text-muted-foreground">→</span>
-                    <span className="truncate text-sm text-muted-foreground">
-                      {turn.pipelineName}
-                    </span>
-                    {turn.runnerName && (
-                      <>
-                        <span className="text-muted-foreground">→</span>
-                        <span className="truncate text-sm text-muted-foreground">
-                          {turn.runnerName}
-                        </span>
-                      </>
-                    )}
-                  </div>
-
-                  <div className="space-y-2">
-                    <MessageLane
-                      label={t('monitoring.messageList.userMessage', {
-                        defaultValue: '用户',
-                      })}
-                      icon={<User className="h-3.5 w-3.5" />}
-                      content={turn.userMessage?.messageContent}
-                      empty={t('monitoring.messageList.noUserMessage', {
-                        defaultValue: '未记录用户输入',
-                      })}
-                      maxLines={2}
-                    />
-                    <MessageLane
-                      label={
-                        assistantOverflow > 0
-                          ? t('monitoring.messageList.assistantMessageCount', {
-                              defaultValue: '助手 +{{count}}',
-                              count: assistantOverflow,
-                            })
-                          : t('monitoring.messageList.assistantMessage', {
-                              defaultValue: '助手',
-                            })
-                      }
-                      icon={<Bot className="h-3.5 w-3.5" />}
-                      content={firstAssistant?.messageContent}
-                      empty={t('monitoring.messageList.noAssistantMessage', {
-                        defaultValue: '未记录助手回复',
-                      })}
-                      maxLines={2}
-                    />
-                  </div>
-                </div>
-
-                <div className="flex shrink-0 flex-col gap-2 lg:items-end">
-                  <div className="text-xs text-muted-foreground">
-                    {turn.lastActivityAt.toLocaleString()}
-                  </div>
-                  <div
-                    className={cn(
-                      'inline-flex h-7 items-center rounded-md border px-2 text-xs font-medium',
-                      statusClass(turn.level),
-                    )}
-                  >
-                    {turn.level}
-                  </div>
-                  <div className="flex flex-wrap gap-2 lg:justify-end">
-                    <Metric
-                      icon={<Cpu className="h-3.5 w-3.5" />}
-                      label={`${turn.llmCalls.length} LLM`}
-                    />
-                    {turn.toolCalls.length > 0 && (
-                      <Metric
-                        icon={<Wrench className="h-3.5 w-3.5" />}
-                        label={`${turn.toolCalls.length} tools`}
-                      />
-                    )}
-                    <Metric
-                      icon={<Hash className="h-3.5 w-3.5" />}
-                      label={`${turn.totalTokens.toLocaleString()} tokens`}
-                    />
-                    <Metric
-                      icon={<Clock className="h-3.5 w-3.5" />}
-                      label={formatDuration(turn.totalDuration)}
-                    />
-                    {turn.errors.length > 0 && (
-                      <Metric
-                        icon={<AlertCircle className="h-3.5 w-3.5" />}
-                        label={`${turn.errors.length} errors`}
-                        tone="error"
-                      />
-                    )}
-                  </div>
-                </div>
+              <div className="mb-2 flex items-center justify-between gap-2 text-xs">
+                <span className="truncate font-medium">{item.botName}</span>
+                {statusBadge(item.status)}
+              </div>
+              <div className="pointer-events-none max-h-12 overflow-hidden break-words text-sm">
+                {item.userMessage ? (
+                  <MessageContentRenderer
+                    content={item.userMessage.messageContent}
+                    maxLines={2}
+                  />
+                ) : (
+                  t('monitoring.messageList.noUserMessage')
+                )}
+              </div>
+              <div className="mt-2 text-xs tabular-nums text-muted-foreground">
+                {item.startedAt.toLocaleString()}
               </div>
             </div>
+          ))}
+        </div>
+      </aside>
 
-            {expanded && (
-              <div className="border-t bg-muted/40 p-3 sm:p-5">
-                <div className="space-y-5 border-l-2 border-border pl-4 sm:pl-6">
-                  <div className="grid grid-cols-2 gap-2 lg:grid-cols-5">
-                    <MetaItem
-                      label={t('monitoring.messageList.platform', {
-                        defaultValue: '平台',
-                      })}
-                      value={turn.platform}
-                    />
-                    <MetaItem
-                      label={t('monitoring.messageList.user', {
-                        defaultValue: '用户',
-                      })}
-                      value={turn.userName || turn.userId}
-                    />
-                    <MetaItem
-                      label={t('monitoring.messageList.runner', {
-                        defaultValue: '执行器',
-                      })}
-                      value={turn.runnerName}
-                    />
-                    <MetaItem
-                      label={t('monitoring.sessions.sessionId', {
-                        defaultValue: '会话 ID',
-                      })}
-                      value={turn.sessionId}
-                    />
-                    <MetaItem
-                      label={t('monitoring.messageList.messageCount', {
-                        defaultValue: '消息数',
-                      })}
-                      value={String(turn.messages.length)}
-                    />
-                  </div>
-
-                  <section>
-                    <h4 className="mb-3 flex items-center gap-2 text-sm font-semibold text-foreground">
-                      <Bot className="h-4 w-4" />
-                      {t('monitoring.messageList.conversationTrace', {
-                        defaultValue: '消息链路',
-                      })}
-                    </h4>
-                    <div className="rounded-lg bg-background px-3 py-3">
-                      {turn.messages.map((message) => (
-                        <ExpandedMessage
-                          key={message.id}
-                          message={message}
-                          label={t(
-                            `monitoring.messageList.roles.${roleLabel(message)}`,
-                            {
-                              defaultValue:
-                                roleLabel(message) === 'assistant'
-                                  ? '助手'
-                                  : roleLabel(message) === 'user'
-                                    ? '用户'
-                                    : '消息',
-                            },
-                          )}
-                        />
-                      ))}
-                    </div>
-                  </section>
-
-                  <section>
-                    <h4 className="mb-3 flex items-center gap-2 text-sm font-semibold text-foreground">
-                      <Cpu className="h-4 w-4" />
-                      {t('monitoring.llmCalls.title', {
-                        defaultValue: 'LLM 调用',
-                      })}{' '}
-                      ({turn.llmCalls.length})
-                    </h4>
-
-                    <div className="grid grid-cols-3 gap-2">
-                      <MetaItem
-                        label={t('monitoring.llmCalls.totalTokens', {
-                          defaultValue: '总 Token',
-                        })}
-                        value={turn.totalTokens.toLocaleString()}
-                      />
-                      <MetaItem
-                        label={t('monitoring.llmCalls.inputTokens', {
-                          defaultValue: '输入 Token',
-                        })}
-                        value={turn.inputTokens.toLocaleString()}
-                      />
-                      <MetaItem
-                        label={t('monitoring.llmCalls.duration', {
-                          defaultValue: '耗时',
-                        })}
-                        value={formatDuration(turn.totalDuration)}
-                      />
-                    </div>
-
-                    <div className="mt-3 rounded-lg bg-background px-3 py-3">
-                      {turn.llmCalls.length > 0 ? (
-                        turn.llmCalls.map((call, index) => (
-                          <div
-                            key={call.id}
-                            className="border-t border-border/70 py-3 first:border-t-0 first:pt-0 last:pb-0"
-                          >
-                            <div className="mb-2 flex flex-wrap items-center justify-between gap-2">
-                              <div className="flex min-w-0 flex-wrap items-center gap-2">
-                                <span className="text-sm font-medium text-foreground">
-                                  #{index + 1} {call.modelName}
-                                </span>
-                                <span
-                                  className={cn(
-                                    'rounded-md px-2 py-1 text-xs font-medium',
-                                    call.status === 'success'
-                                      ? 'bg-green-100 text-green-700 dark:bg-green-950 dark:text-green-300'
-                                      : 'bg-red-100 text-red-700 dark:bg-red-950 dark:text-red-300',
-                                  )}
-                                >
-                                  {call.status}
-                                </span>
-                              </div>
-                              <span className="text-xs text-muted-foreground">
-                                {formatDuration(call.duration)}
-                              </span>
-                            </div>
-                            <div className="flex flex-wrap gap-x-8 gap-y-1 text-xs text-muted-foreground">
-                              <span>In: {call.tokens.input}</span>
-                              <span>Out: {call.tokens.output}</span>
-                              <span>Total: {call.tokens.total}</span>
-                              <span className="font-mono">
-                                ID: {shortId(call.id)}
-                              </span>
-                            </div>
-                            {call.errorMessage && (
-                              <div className="mt-2 whitespace-pre-wrap break-words text-xs text-red-600 dark:text-red-400">
-                                {call.errorMessage}
-                              </div>
-                            )}
-                          </div>
-                        ))
-                      ) : (
-                        <div className="py-4 text-center text-sm text-muted-foreground">
-                          {t('monitoring.messageList.noLlmCalls', {
-                            defaultValue: '未记录模型调用',
-                          })}
-                        </div>
-                      )}
-                    </div>
-                  </section>
-
-                  <section>
-                    <h4 className="mb-3 flex items-center gap-2 text-sm font-semibold text-foreground">
-                      <Wrench className="h-4 w-4" />
-                      {t('monitoring.toolCalls.title', {
-                        defaultValue: '工具调用',
-                      })}{' '}
-                      ({turn.toolCalls.length})
-                    </h4>
-
-                    <div className="grid grid-cols-2 gap-2 lg:grid-cols-3">
-                      <MetaItem
-                        label={t('monitoring.toolCalls.totalCalls', {
-                          defaultValue: '调用次数',
-                        })}
-                        value={String(turn.toolCalls.length)}
-                      />
-                      <MetaItem
-                        label={t('monitoring.toolCalls.duration', {
-                          defaultValue: '工具耗时',
-                        })}
-                        value={formatDuration(turn.totalToolDuration)}
-                      />
-                      <MetaItem
-                        label={t('monitoring.toolCalls.errorCalls', {
-                          defaultValue: '失败次数',
-                        })}
-                        value={String(
-                          turn.toolCalls.filter(
-                            (call) => call.status === 'error',
-                          ).length,
-                        )}
-                      />
-                    </div>
-
-                    <div className="mt-3 rounded-lg bg-background px-3 py-3">
-                      {turn.toolCalls.length > 0 ? (
-                        turn.toolCalls.map((call, index) => {
-                          const toolCallKey = `${turn.id}:${call.id}`;
-                          const hasToolDetails = Boolean(
-                            call.arguments || call.result || call.errorMessage,
-                          );
-                          const expandedToolCall = Boolean(
-                            expandedToolCallIds[toolCallKey],
-                          );
-                          const detailsId = `monitoring-tool-call-details-${call.id}`;
-
-                          return (
-                            <div
-                              key={call.id}
-                              className="border-t border-border/70 py-2 first:border-t-0 first:pt-0 last:pb-0"
-                            >
-                              <button
-                                type="button"
-                                className={cn(
-                                  'flex w-full items-start justify-between gap-3 rounded-md px-2 py-2 text-left outline-none transition-colors',
-                                  hasToolDetails &&
-                                    'cursor-pointer hover:bg-muted/60 focus-visible:ring-2 focus-visible:ring-ring',
-                                )}
-                                aria-expanded={
-                                  hasToolDetails ? expandedToolCall : undefined
-                                }
-                                aria-controls={
-                                  hasToolDetails ? detailsId : undefined
-                                }
-                                aria-disabled={!hasToolDetails}
-                                onClick={() =>
-                                  hasToolDetails &&
-                                  toggleToolCallDetails(toolCallKey)
-                                }
-                              >
-                                <div className="flex min-w-0 flex-wrap items-center gap-2">
-                                  {hasToolDetails &&
-                                    (expandedToolCall ? (
-                                      <ChevronDown className="mt-0.5 h-4 w-4 shrink-0 text-muted-foreground" />
-                                    ) : (
-                                      <ChevronRight className="mt-0.5 h-4 w-4 shrink-0 text-muted-foreground" />
-                                    ))}
-                                  <span className="min-w-0 truncate text-sm font-medium text-foreground">
-                                    #{index + 1} {call.toolName}
-                                  </span>
-                                  <span className="rounded-md bg-muted px-2 py-1 text-xs font-medium text-muted-foreground">
-                                    {call.toolSource}
-                                  </span>
-                                  <span
-                                    className={cn(
-                                      'rounded-md px-2 py-1 text-xs font-medium',
-                                      call.status === 'success'
-                                        ? 'bg-green-100 text-green-700 dark:bg-green-950 dark:text-green-300'
-                                        : 'bg-red-100 text-red-700 dark:bg-red-950 dark:text-red-300',
-                                    )}
-                                  >
-                                    {call.status}
-                                  </span>
-                                  <span className="font-mono text-xs text-muted-foreground">
-                                    ID: {shortId(call.id)}
-                                  </span>
-                                </div>
-                                <span className="shrink-0 text-xs text-muted-foreground">
-                                  {formatDuration(call.duration)}
-                                </span>
-                              </button>
-
-                              {hasToolDetails && expandedToolCall && (
-                                <div
-                                  id={detailsId}
-                                  className="mt-1 grid gap-2 px-2 pb-2 text-xs lg:grid-cols-2"
-                                >
-                                  {call.arguments && (
-                                    <div className="min-w-0 rounded-md bg-muted/50 p-2">
-                                      <div className="mb-1 font-medium text-foreground">
-                                        {t('monitoring.toolCalls.arguments', {
-                                          defaultValue: '参数',
-                                        })}
-                                      </div>
-                                      <pre className="whitespace-pre-wrap break-words font-mono text-muted-foreground">
-                                        {truncateDetail(call.arguments)}
-                                      </pre>
-                                    </div>
-                                  )}
-                                  {call.result && (
-                                    <div className="min-w-0 rounded-md bg-muted/50 p-2">
-                                      <div className="mb-1 font-medium text-foreground">
-                                        {t('monitoring.toolCalls.result', {
-                                          defaultValue: '结果',
-                                        })}
-                                      </div>
-                                      <pre className="whitespace-pre-wrap break-words font-mono text-muted-foreground">
-                                        {truncateDetail(call.result)}
-                                      </pre>
-                                    </div>
-                                  )}
-                                  {call.errorMessage && (
-                                    <div className="min-w-0 whitespace-pre-wrap break-words rounded-md bg-red-50 p-2 text-red-600 dark:bg-red-950/40 dark:text-red-400 lg:col-span-2">
-                                      {call.errorMessage}
-                                    </div>
-                                  )}
-                                </div>
-                              )}
-                            </div>
-                          );
-                        })
-                      ) : (
-                        <div className="py-4 text-center text-sm text-muted-foreground">
-                          {t('monitoring.toolCalls.noToolCalls', {
-                            defaultValue: '未记录工具调用',
-                          })}
-                        </div>
-                      )}
-                    </div>
-                  </section>
-
-                  {turn.errors.length > 0 && (
-                    <section>
-                      <h4 className="mb-3 flex items-center gap-2 text-sm font-semibold text-red-700 dark:text-red-300">
-                        <AlertCircle className="h-4 w-4" />
-                        {t('monitoring.errors.title', {
-                          defaultValue: '错误日志',
-                        })}{' '}
-                        ({turn.errors.length})
-                      </h4>
-                      <div className="rounded-lg bg-background px-3 py-3">
-                        {turn.errors.map((error) => (
-                          <div
-                            key={error.id}
-                            className="border-t border-red-200/80 py-3 first:border-t-0 first:pt-0 last:pb-0 dark:border-red-900"
-                          >
-                            <div className="mb-2 flex flex-wrap items-center justify-between gap-2">
-                              <span className="text-sm font-medium text-red-700 dark:text-red-300">
-                                {error.errorType}
-                              </span>
-                              <span className="text-xs text-muted-foreground">
-                                {error.timestamp.toLocaleString()}
-                              </span>
-                            </div>
-                            <div className="whitespace-pre-wrap break-words text-sm text-red-600 dark:text-red-400">
-                              {error.errorMessage}
-                            </div>
-                          </div>
-                        ))}
-                      </div>
-                    </section>
-                  )}
-                </div>
-              </div>
-            )}
+      <article
+        key={turn.id}
+        className="min-w-0 space-y-5 lg:overflow-y-auto lg:pr-2"
+      >
+        <header className="space-y-3 border-b pb-4">
+          <div className="flex flex-wrap items-center gap-2">
+            <h3 className="text-sm font-semibold">{turn.botName}</h3>
+            {statusBadge(turn.status)}
+            <span className="text-sm text-muted-foreground">
+              {turn.userName || turn.userId}
+            </span>
           </div>
-        );
-      })}
+          <div className="flex flex-wrap gap-x-5 gap-y-2 text-xs tabular-nums text-muted-foreground">
+            <span>{turn.startedAt.toLocaleString()}</span>
+            <span>
+              {t('monitoring.llmCalls.title')} {turn.llmCalls.length}
+            </span>
+            <span>
+              {t('monitoring.toolCalls.title')} {turn.toolCalls.length}
+            </span>
+            <span>
+              {t('monitoring.llmCalls.totalTokens')}{' '}
+              {turn.totalTokens.toLocaleString()}
+            </span>
+          </div>
+        </header>
+
+        <section className="rounded-xl border p-4">
+          <h4 className="mb-3 flex items-center gap-2 text-sm font-medium">
+            <MessageCircle className="size-4 text-muted-foreground" />
+            {t('monitoring.pipelineTrace.received')}
+          </h4>
+          {turn.userMessage ? (
+            <MessageBody content={turn.userMessage.messageContent} />
+          ) : (
+            <p className="text-sm text-muted-foreground">
+              {t('monitoring.messageList.noUserMessage')}
+            </p>
+          )}
+        </section>
+
+        <section>
+          <h4 className="mb-3 text-sm font-medium">
+            {t('monitoring.pipelineTrace.processing')}
+            <span className="ml-2 font-normal text-muted-foreground">
+              {steps.length}
+            </span>
+          </h4>
+          {steps.length === 0 ? (
+            <p className="text-sm text-muted-foreground">
+              {t('monitoring.pipelineTrace.noSteps')}
+            </p>
+          ) : (
+            <ol className="space-y-3">
+              {steps.map((step, index) => {
+                const { call } = step;
+                const parameters =
+                  step.kind === 'tool'
+                    ? parsePayload(step.call.arguments)
+                    : null;
+                const summary =
+                  step.kind === 'tool' &&
+                  parameters &&
+                  typeof parameters === 'object' &&
+                  !Array.isArray(parameters)
+                    ? toolSummary(
+                        step.call.toolName,
+                        parameters as Record<string, unknown>,
+                      ).text
+                    : '';
+                return (
+                  <li
+                    key={`${step.kind}:${call.id}`}
+                    className="relative flex gap-3"
+                  >
+                    {index < steps.length - 1 && (
+                      <div className="absolute bottom-[-12px] left-4 top-8 w-px bg-border" />
+                    )}
+                    <span className="relative flex size-8 shrink-0 items-center justify-center rounded-full border bg-background text-xs tabular-nums text-muted-foreground">
+                      {index + 1}
+                    </span>
+                    <div className="min-w-0 flex-1 overflow-hidden rounded-xl border">
+                      <div className="flex flex-wrap items-center gap-2 border-b bg-muted/20 px-4 py-3 text-sm">
+                        {step.kind === 'model' ? (
+                          <Cpu className="size-4" />
+                        ) : step.kind === 'tool' ? (
+                          <Wrench className="size-4" />
+                        ) : (
+                          <AlertCircle className="size-4 text-destructive" />
+                        )}
+                        <span className="min-w-0 flex-1 break-words font-medium">
+                          {step.kind === 'model'
+                            ? step.call.modelName
+                            : step.kind === 'tool'
+                              ? step.call.toolName
+                              : step.call.errorType}
+                        </span>
+                        {statusBadge(
+                          step.kind === 'error' ? 'error' : step.call.status,
+                        )}
+                      </div>
+                      <div className="space-y-3 p-4">
+                        <div className="flex flex-wrap gap-x-4 gap-y-1 text-xs tabular-nums text-muted-foreground">
+                          <span>{call.timestamp.toLocaleTimeString()}</span>
+                          {step.kind !== 'error' && (
+                            <span>{formatDuration(step.call.duration)}</span>
+                          )}
+                          {step.kind === 'tool' && (
+                            <span>{step.call.toolSource}</span>
+                          )}
+                          {step.kind === 'model' && (
+                            <>
+                              <span>
+                                {t('monitoring.llmCalls.inputTokens')}{' '}
+                                {step.call.tokens.input.toLocaleString()}
+                              </span>
+                              <span>
+                                {t('monitoring.llmCalls.outputTokens')}{' '}
+                                {step.call.tokens.output.toLocaleString()}
+                              </span>
+                            </>
+                          )}
+                        </div>
+                        {summary && (
+                          <p className="whitespace-pre-wrap break-words text-sm [overflow-wrap:anywhere]">
+                            {summary}
+                          </p>
+                        )}
+                        {call.errorMessage && (
+                          <p className="whitespace-pre-wrap break-words text-sm text-destructive [overflow-wrap:anywhere]">
+                            {call.errorMessage}
+                          </p>
+                        )}
+                        {step.kind === 'tool' && (
+                          <div>
+                            {step.call.arguments && (
+                              <ProcessorPayload
+                                title={t('monitoring.toolCalls.arguments')}
+                                value={parameters}
+                              />
+                            )}
+                            {step.call.result &&
+                              step.call.result !== 'null' && (
+                                <ProcessorPayload
+                                  title={t('monitoring.toolCalls.result')}
+                                  value={parsePayload(step.call.result)}
+                                />
+                              )}
+                          </div>
+                        )}
+                        {step.kind === 'error' && step.call.stackTrace && (
+                          <ProcessorPayload
+                            title={t('monitoring.unified.metadata')}
+                            value={step.call.stackTrace}
+                          />
+                        )}
+                      </div>
+                    </div>
+                  </li>
+                );
+              })}
+            </ol>
+          )}
+        </section>
+
+        <section className="rounded-xl border p-4">
+          <h4 className="mb-3 flex items-center gap-2 text-sm font-medium">
+            <Send className="size-4 text-muted-foreground" />
+            {t('monitoring.pipelineTrace.replies')}
+            <span className="font-normal text-muted-foreground">
+              {turn.assistantMessages.length}
+            </span>
+          </h4>
+          {turn.assistantMessages.length ? (
+            turn.assistantMessages.map((message) => (
+              <div
+                key={message.id}
+                className="space-y-2 border-t py-3 first:border-t-0 first:pt-0 last:pb-0"
+              >
+                <div className="flex items-center justify-between gap-2 text-xs text-muted-foreground">
+                  <span>{message.timestamp.toLocaleTimeString()}</span>
+                  {statusBadge(message.status)}
+                </div>
+                <MessageBody content={message.messageContent} />
+              </div>
+            ))
+          ) : (
+            <p className="text-sm text-muted-foreground">
+              {t('monitoring.messageList.noAssistantMessage')}
+            </p>
+          )}
+        </section>
+        <ProcessorPayload
+          title={t('monitoring.unified.metadata')}
+          value={{
+            turn_id: turn.id,
+            session_id: turn.sessionId,
+            bot: turn.botName,
+            pipeline: turn.pipelineName,
+            runner: turn.runnerName,
+            platform: turn.platform,
+            user_id: turn.userId,
+            messages: turn.messages,
+          }}
+        />
+      </article>
     </div>
   );
 }
diff --git a/web/src/app/home/monitoring/components/ExportDropdown.tsx b/web/src/app/home/monitoring/components/ExportDropdown.tsx
index 29dde3fd1..be60d909e 100644
--- a/web/src/app/home/monitoring/components/ExportDropdown.tsx
+++ b/web/src/app/home/monitoring/components/ExportDropdown.tsx
@@ -8,6 +8,7 @@ import {
   Users,
   Layers,
   ThumbsUp,
+  Loader2,
 } from 'lucide-react';
 import { Button } from '@/components/ui/button';
 import {
@@ -19,6 +20,7 @@ import {
   DropdownMenuTrigger,
 } from '@/components/ui/dropdown-menu';
 import { backendClient } from '@/app/infra/http';
+import { resolveMonitoringWindow } from '../utils/dateUtils';
 import { FilterState } from '../types/monitoring';
 
 export type ExportType =
@@ -37,44 +39,8 @@ export function ExportDropdown({ filterState }: ExportDropdownProps) {
   const { t } = useTranslation();
   const [exporting, setExporting] = useState<ExportType | null>(null);
 
-  const getDateRangeParams = (): { startTime: string; endTime: string } => {
-    const now = new Date();
-    let startTime: Date;
-    let endTime: Date = now;
-
-    switch (filterState.timeRange) {
-      case 'lastHour':
-        startTime = new Date(now.getTime() - 60 * 60 * 1000);
-        break;
-      case 'last6Hours':
-        startTime = new Date(now.getTime() - 6 * 60 * 60 * 1000);
-        break;
-      case 'last24Hours':
-        startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
-        break;
-      case 'last7Days':
-        startTime = new Date(now.getTime() - 7 * 24 * 60 * 60 * 1000);
-        break;
-      case 'last30Days':
-        startTime = new Date(now.getTime() - 30 * 24 * 60 * 60 * 1000);
-        break;
-      case 'custom':
-        if (filterState.customDateRange) {
-          startTime = filterState.customDateRange.from;
-          endTime = filterState.customDateRange.to;
-        } else {
-          startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
-        }
-        break;
-      default:
-        startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
-    }
-
-    return {
-      startTime: startTime.toISOString(),
-      endTime: endTime.toISOString(),
-    };
-  };
+  const getDateRangeParams = () =>
+    resolveMonitoringWindow(filterState.timeRange, filterState.customDateRange);
 
   const handleExport = async (type: ExportType) => {
     setExporting(type);
@@ -182,26 +148,7 @@ export function ExportDropdown({ filterState }: ExportDropdownProps) {
         >
           {exporting ? (
             <>
-              <svg
-                className="w-4 h-4 mr-2 animate-spin"
-                xmlns="http://www.w3.org/2000/svg"
-                fill="none"
-                viewBox="0 0 24 24"
-              >
-                <circle
-                  className="opacity-25"
-                  cx="12"
-                  cy="12"
-                  r="10"
-                  stroke="currentColor"
-                  strokeWidth="4"
-                />
-                <path
-                  className="opacity-75"
-                  fill="currentColor"
-                  d="M4 12a8 8 0 018-8V0C5.373 0 0 5.373 0 12h4zm2 5.291A7.962 7.962 0 014 12H0c0 3.042 1.135 5.824 3 7.938l3-2.647z"
-                />
-              </svg>
+              <Loader2 className="w-4 h-4 mr-2 animate-spin" />
               {t('monitoring.export.exporting')}
             </>
           ) : (
diff --git a/web/src/app/home/monitoring/components/FeedbackCard.tsx b/web/src/app/home/monitoring/components/FeedbackCard.tsx
index c2e93e52b..b5882b7f1 100644
--- a/web/src/app/home/monitoring/components/FeedbackCard.tsx
+++ b/web/src/app/home/monitoring/components/FeedbackCard.tsx
@@ -9,6 +9,7 @@ import {
   Heart,
   Smile,
 } from 'lucide-react';
+import { MetricCard } from './overview-cards/MetricCard';
 
 interface FeedbackCardProps {
   title: string;
@@ -23,6 +24,18 @@ interface FeedbackCardProps {
   loading?: boolean;
 }
 
+// The variant only tints the icon tile; the card surface stays neutral so the
+// feedback row matches the token-monitoring cards.
+const variantAccents: Record<
+  NonNullable<FeedbackCardProps['variant']>,
+  string
+> = {
+  default: '#64748b',
+  success: '#10b981',
+  warning: '#f59e0b',
+  danger: '#ef4444',
+};
+
 export function FeedbackCard({
   title,
   value,
@@ -32,89 +45,37 @@ export function FeedbackCard({
   variant = 'default',
   loading = false,
 }: FeedbackCardProps) {
-  const variantStyles = {
-    default: 'bg-white dark:bg-[#2a2a2e] border-gray-200 dark:border-gray-700',
-    success:
-      'bg-green-50 dark:bg-green-900/20 border-green-200 dark:border-green-800',
-    warning:
-      'bg-yellow-50 dark:bg-yellow-900/20 border-yellow-200 dark:border-yellow-800',
-    danger: 'bg-red-50 dark:bg-red-900/20 border-red-200 dark:border-red-800',
-  };
-
-  const iconStyles = {
-    default: 'text-gray-500 dark:text-gray-400',
-    success: 'text-green-500 dark:text-green-400',
-    warning: 'text-yellow-500 dark:text-yellow-400',
-    danger: 'text-red-500 dark:text-red-400',
-  };
-
   const trendStyles = {
     up: 'text-green-500',
-    down: 'text-red-500',
-    neutral: 'text-gray-500',
+    down: 'text-destructive',
+    neutral: 'text-muted-foreground',
   };
 
-  if (loading) {
-    return (
-      <div
-        className={`p-6 rounded-xl border shadow-sm ${variantStyles.default} animate-pulse`}
-      >
-        <div className="flex items-start justify-between">
-          <div className="flex-1">
-            <div className="h-4 bg-gray-200 dark:bg-gray-700 rounded w-20 mb-2" />
-            <div className="h-8 bg-gray-200 dark:bg-gray-700 rounded w-16 mb-1" />
-            <div className="h-3 bg-gray-200 dark:bg-gray-700 rounded w-24" />
-          </div>
-          <div className="w-10 h-10 bg-gray-200 dark:bg-gray-700 rounded-lg" />
-        </div>
-      </div>
-    );
-  }
-
   return (
-    <div
-      className={`p-6 rounded-xl border shadow-sm ${variantStyles[variant]}`}
+    <MetricCard
+      label={title}
+      value={value}
+      hint={subtitle}
+      icon={icon}
+      accent={variantAccents[variant]}
+      loading={loading}
     >
-      <div className="flex items-start justify-between">
-        <div className="flex-1">
-          <p className="text-sm font-medium text-gray-500 dark:text-gray-400 mb-1">
-            {title}
-          </p>
-          <p className="text-2xl font-bold text-gray-900 dark:text-white">
-            {value}
-          </p>
-          {subtitle && (
-            <p className="text-xs text-gray-400 dark:text-gray-500 mt-1">
-              {subtitle}
-            </p>
-          )}
-          {trend && (
-            <div
-              className={`flex items-center mt-2 text-sm ${trendStyles[trend.direction]}`}
-            >
-              {trend.direction === 'up' && (
-                <TrendingUp className="w-4 h-4 mr-1" />
-              )}
-              {trend.direction === 'down' && (
-                <TrendingDown className="w-4 h-4 mr-1" />
-              )}
-              {trend.direction === 'neutral' && (
-                <Minus className="w-4 h-4 mr-1" />
-              )}
-              <span>
-                {trend.value > 0 ? '+' : ''}
-                {trend.value}%
-              </span>
-            </div>
-          )}
-        </div>
+      {trend && (
         <div
-          className={`p-3 rounded-lg bg-gray-100 dark:bg-gray-800 ${iconStyles[variant]}`}
+          className={`flex items-center text-sm ${trendStyles[trend.direction]}`}
         >
-          {icon}
+          {trend.direction === 'up' && <TrendingUp className="mr-1 h-4 w-4" />}
+          {trend.direction === 'down' && (
+            <TrendingDown className="mr-1 h-4 w-4" />
+          )}
+          {trend.direction === 'neutral' && <Minus className="mr-1 h-4 w-4" />}
+          <span>
+            {trend.value > 0 ? '+' : ''}
+            {trend.value}%
+          </span>
         </div>
-      </div>
-    </div>
+      )}
+    </MetricCard>
   );
 }
 
@@ -135,35 +96,37 @@ export function FeedbackStatsCards({ stats, loading }: FeedbackStatsProps) {
     {
       title: t('monitoring.feedback.totalFeedback'),
       value: stats?.totalFeedback ?? 0,
-      icon: <Heart className="w-6 h-6" />,
+      icon: <Heart className="h-4 w-4" />,
       variant: 'default' as const,
     },
     {
       title: t('monitoring.feedback.totalLikes'),
       value: stats?.totalLikes ?? 0,
-      icon: <ThumbsUp className="w-6 h-6" />,
+      icon: <ThumbsUp className="h-4 w-4" />,
       variant: 'success' as const,
     },
     {
       title: t('monitoring.feedback.totalDislikes'),
       value: stats?.totalDislikes ?? 0,
-      icon: <ThumbsDown className="w-6 h-6" />,
+      icon: <ThumbsDown className="h-4 w-4" />,
       variant: 'danger' as const,
     },
     {
       title: t('monitoring.feedback.satisfactionRate'),
       value: stats ? `${stats.satisfactionRate}%` : '0%',
-      icon: <Smile className="w-6 h-6" />,
+      icon: <Smile className="h-4 w-4" />,
+      // The rate is the one feedback metric that is good or bad on its own, so
+      // its tile keeps encoding the verdict (as it did before the restyle).
       variant: (stats && stats.satisfactionRate >= 80
         ? 'success'
         : stats && stats.satisfactionRate >= 50
           ? 'warning'
-          : 'danger') as 'default' | 'success' | 'warning' | 'danger',
+          : 'danger') as 'success' | 'warning' | 'danger',
     },
   ];
 
   return (
-    <div className="grid grid-cols-1 md:grid-cols-2 xl:grid-cols-4 gap-6">
+    <div className="grid grid-cols-1 gap-6 md:grid-cols-2 xl:grid-cols-4">
       {cards.map((card, index) => (
         <FeedbackCard
           key={index}
diff --git a/web/src/app/home/monitoring/components/FeedbackList.tsx b/web/src/app/home/monitoring/components/FeedbackList.tsx
index 10f590f2c..3e42ca4f1 100644
--- a/web/src/app/home/monitoring/components/FeedbackList.tsx
+++ b/web/src/app/home/monitoring/components/FeedbackList.tsx
@@ -10,7 +10,9 @@ import {
 } from 'lucide-react';
 import { FeedbackRecord } from '../types/monitoring';
 import { Button } from '@/components/ui/button';
-import { LoadingSpinner } from '@/components/ui/loading-spinner';
+import { Badge } from '@/components/ui/badge';
+import { Card } from '@/components/ui/card';
+import { TabState } from './TabState';
 
 interface FeedbackListProps {
   feedback: FeedbackRecord[];
@@ -31,33 +33,25 @@ export function FeedbackList({
   };
 
   if (loading) {
-    return (
-      <div className="py-12 flex justify-center">
-        <LoadingSpinner text={t('common.loading')} />
-      </div>
-    );
+    return <TabState loading rows={6} />;
   }
 
   if (!feedback || feedback.length === 0) {
     return (
-      <div className="text-center text-gray-500 dark:text-gray-400 py-16">
-        <Heart className="w-16 h-16 mx-auto mb-4 text-gray-300 dark:text-gray-600" />
-        <p className="text-base font-medium mb-2">
-          {t('monitoring.feedback.noFeedback')}
-        </p>
-        <p className="text-sm">
-          {t('monitoring.feedback.noFeedbackDescription')}
-        </p>
-      </div>
+      <TabState
+        icon={<Heart className="w-16 h-16 text-muted-foreground/50" />}
+        title={t('monitoring.feedback.noFeedback')}
+        hint={t('monitoring.feedback.noFeedbackDescription')}
+      />
     );
   }
 
   return (
     <div className="space-y-4">
       {feedback.map((item) => (
-        <div
+        <Card
           key={item.id}
-          className={`border rounded-xl overflow-hidden hover:shadow-md transition-all duration-200 ${
+          className={`gap-0 py-0 overflow-hidden hover:shadow-md transition-all duration-200 ${
             item.feedbackType === 'like'
               ? 'border-green-200 dark:border-green-900'
               : 'border-red-200 dark:border-red-900'
@@ -105,16 +99,19 @@ export function FeedbackList({
                     </span>
                     {item.botName && (
                       <>
-                        <span className="text-gray-400">→</span>
-                        <span className="text-sm text-gray-600 dark:text-gray-400">
+                        <span className="text-muted-foreground">→</span>
+                        <span className="text-sm text-muted-foreground">
                           {item.botName}
                         </span>
                       </>
                     )}
                     {item.platform && (
-                      <span className="text-xs px-2 py-0.5 rounded bg-gray-100 dark:bg-gray-800 text-gray-500 dark:text-gray-400">
+                      <Badge
+                        variant="outline"
+                        className="bg-muted text-muted-foreground"
+                      >
                         {item.platform}
-                      </span>
+                      </Badge>
                     )}
                     {item.streamId && onViewMessage && (
                       <Button
@@ -133,7 +130,7 @@ export function FeedbackList({
                   </div>
 
                   {item.feedbackContent && (
-                    <p className="text-sm text-gray-600 dark:text-gray-400 line-clamp-2">
+                    <p className="text-sm text-muted-foreground line-clamp-2">
                       {item.feedbackContent}
                     </p>
                   )}
@@ -142,12 +139,13 @@ export function FeedbackList({
                     item.inaccurateReasons.length > 0 && (
                       <div className="flex flex-wrap gap-1 mt-2">
                         {item.inaccurateReasons.map((reason, idx) => (
-                          <span
+                          <Badge
                             key={idx}
-                            className="text-xs px-2 py-0.5 rounded bg-red-100 dark:bg-red-900/30 text-red-600 dark:text-red-400"
+                            variant="outline"
+                            className="bg-red-100 dark:bg-red-900/30 text-red-600 dark:text-red-400"
                           >
                             {reason}
-                          </span>
+                          </Badge>
                         ))}
                       </div>
                     )}
@@ -156,7 +154,7 @@ export function FeedbackList({
 
               {/* Timestamp */}
               <div className="flex flex-col items-end gap-2 ml-4">
-                <span className="text-xs text-gray-500 dark:text-gray-400 whitespace-nowrap">
+                <span className="text-xs text-muted-foreground whitespace-nowrap">
                   {item.timestamp.toLocaleString()}
                 </span>
               </div>
@@ -166,75 +164,75 @@ export function FeedbackList({
           {/* Expanded Details */}
           {expandedId === item.id && (
             <div
-              className={`border-t p-5 bg-white dark:bg-gray-900 ${
+              className={`border-t p-5 bg-background ${
                 item.feedbackType === 'like'
                   ? 'border-green-200 dark:border-green-900'
                   : 'border-red-200 dark:border-red-900'
               }`}
             >
-              <div className="space-y-4 pl-8 border-l-2 border-gray-200 dark:border-gray-700 ml-4">
+              <div className="space-y-4 pl-8 border-l-2 border-border ml-4">
                 {/* Context Info */}
-                <div className="bg-gray-50 dark:bg-gray-800 rounded-lg p-3">
-                  <h4 className="text-sm font-semibold text-gray-700 dark:text-gray-300 mb-3">
+                <div className="bg-muted rounded-lg p-3">
+                  <h4 className="text-sm font-semibold text-foreground mb-3">
                     {t('monitoring.feedback.contextInfo')}
                   </h4>
                   <div className="grid grid-cols-2 md:grid-cols-3 gap-2 text-xs">
                     {item.botName && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.messageList.bot')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.botName}
                         </div>
                       </div>
                     )}
                     {item.pipelineName && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.messageList.pipeline')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.pipelineName}
                         </div>
                       </div>
                     )}
                     {item.sessionId && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.sessions.sessionId')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.sessionId}
                         </div>
                       </div>
                     )}
                     {item.userId && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.feedback.userId')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.userId}
                         </div>
                       </div>
                     )}
                     {item.messageId && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.feedback.messageId')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.messageId}
                         </div>
                       </div>
                     )}
                     {item.streamId && (
-                      <div className="bg-white dark:bg-gray-900 rounded p-2">
-                        <div className="text-gray-500 dark:text-gray-400">
+                      <div className="bg-background rounded p-2">
+                        <div className="text-muted-foreground">
                           {t('monitoring.feedback.streamId')}
                         </div>
-                        <div className="font-medium text-gray-900 dark:text-white truncate">
+                        <div className="font-medium text-foreground truncate">
                           {item.streamId}
                         </div>
                       </div>
@@ -244,11 +242,11 @@ export function FeedbackList({
 
                 {/* Feedback Content */}
                 {item.feedbackContent && (
-                  <div className="bg-gray-50 dark:bg-gray-800 rounded-lg p-3">
-                    <h4 className="text-sm font-semibold text-gray-700 dark:text-gray-300 mb-3">
+                  <div className="bg-muted rounded-lg p-3">
+                    <h4 className="text-sm font-semibold text-foreground mb-3">
                       {t('monitoring.feedback.feedbackContent')}
                     </h4>
-                    <p className="text-sm text-gray-600 dark:text-gray-400 whitespace-pre-wrap">
+                    <p className="text-sm text-muted-foreground whitespace-pre-wrap">
                       {item.feedbackContent}
                     </p>
                   </div>
@@ -256,7 +254,7 @@ export function FeedbackList({
               </div>
             </div>
           )}
-        </div>
+        </Card>
       ))}
     </div>
   );
diff --git a/web/src/app/home/monitoring/components/MessageContentRenderer.tsx b/web/src/app/home/monitoring/components/MessageContentRenderer.tsx
index cf11f5f1c..0a29102e6 100644
--- a/web/src/app/home/monitoring/components/MessageContentRenderer.tsx
+++ b/web/src/app/home/monitoring/components/MessageContentRenderer.tsx
@@ -8,6 +8,7 @@ import {
   Voice,
   Quote,
 } from '@/app/infra/entities/message';
+import { Badge } from '@/components/ui/badge';
 import ImagePreviewDialog from '@/app/home/pipelines/components/debug-dialog/ImagePreviewDialog';
 
 interface MessageContentRendererProps {
@@ -48,23 +49,23 @@ export function MessageContentRenderer({
         const displayName =
           atComponent.display || atComponent.target?.toString() || '';
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-blue-100 dark:bg-blue-900 text-blue-700 dark:text-blue-300 text-sm"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-blue-100 dark:bg-blue-900 text-blue-700 dark:text-blue-300 text-sm"
           >
             @{displayName}
-          </span>
+          </Badge>
         );
       }
 
       case 'AtAll':
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-blue-100 dark:bg-blue-900 text-blue-700 dark:text-blue-300 text-sm"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-blue-100 dark:bg-blue-900 text-blue-700 dark:text-blue-300 text-sm"
           >
             @All
-          </span>
+          </Badge>
         );
 
       case 'Image': {
@@ -73,12 +74,12 @@ export function MessageContentRenderer({
 
         if (!imageUrl) {
           return (
-            <span
+            <Badge
               key={index}
-              className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-muted text-muted-foreground text-sm"
+              className="mx-0.5 px-1.5 py-0.5 rounded bg-muted text-muted-foreground text-sm"
             >
               [Image]
-            </span>
+            </Badge>
           );
         }
 
@@ -101,41 +102,41 @@ export function MessageContentRenderer({
       case 'File': {
         const file = component as MessageChainComponent & { name?: string };
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-muted text-muted-foreground text-sm"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-muted text-muted-foreground text-sm [&>svg]:size-3.5"
           >
-            <Paperclip className="w-3.5 h-3.5 mr-1" />
+            <Paperclip className="mr-1" />
             {file.name || 'File'}
-          </span>
+          </Badge>
         );
       }
 
       case 'Voice': {
         const voice = component as Voice;
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-muted text-muted-foreground text-sm"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-muted text-muted-foreground text-sm [&>svg]:size-3.5"
           >
-            <AudioLines className="w-3.5 h-3.5 mr-1" />
+            <AudioLines className="mr-1" />
             Voice{voice.length ? ` ${voice.length}s` : ''}
-          </span>
+          </Badge>
         );
       }
 
       case 'Quote': {
         const quote = component as Quote;
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-muted text-muted-foreground text-sm border-l-2 border-muted-foreground/50"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-muted text-muted-foreground text-sm border-l-2 border-l-muted-foreground/50"
           >
             {quote.origin
               ?.filter((c) => (c as MessageChainComponent).type === 'Plain')
               .map((c) => (c as MessageChainComponent as Plain).text)
               .join('') || '[Quote]'}
-          </span>
+          </Badge>
         );
       }
 
@@ -144,12 +145,12 @@ export function MessageContentRenderer({
 
       default:
         return (
-          <span
+          <Badge
             key={index}
-            className="inline-flex items-center px-1.5 py-0.5 mx-0.5 rounded bg-muted text-muted-foreground text-sm"
+            className="mx-0.5 px-1.5 py-0.5 rounded bg-muted text-muted-foreground text-sm"
           >
             [{component.type}]
-          </span>
+          </Badge>
         );
     }
   };
diff --git a/web/src/app/home/monitoring/components/MessageDetailsCard.tsx b/web/src/app/home/monitoring/components/MessageDetailsCard.tsx
index 9e78c394c..d9ce79fe9 100644
--- a/web/src/app/home/monitoring/components/MessageDetailsCard.tsx
+++ b/web/src/app/home/monitoring/components/MessageDetailsCard.tsx
@@ -1,6 +1,9 @@
 import React, { useMemo } from 'react';
 import { useTranslation } from 'react-i18next';
 import { Info, Clock, AlertCircle, Braces } from 'lucide-react';
+import { Card } from '@/components/ui/card';
+import { Badge } from '@/components/ui/badge';
+import { cn } from '@/lib/utils';
 import { MessageDetails } from '../types/monitoring';
 
 interface MessageDetailsCardProps {
@@ -24,7 +27,7 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
     <div className="space-y-4 pl-8 border-l-2 border-border ml-4">
       {/* Context Info Section */}
       {details.message && (
-        <div className="bg-muted rounded-lg p-3">
+        <Card className="bg-muted rounded-lg border-0 p-3 gap-0">
           <h4 className="text-sm font-semibold text-foreground mb-3 flex items-center">
             <Info className="w-4 h-4 mr-2" />
             {t('monitoring.messageList.viewDetails')}
@@ -79,12 +82,12 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
               </div>
             </div>
           </div>
-        </div>
+        </Card>
       )}
 
       {/* LLM Calls Section */}
       {details.llmCalls && details.llmCalls.length > 0 && (
-        <div className="bg-muted rounded-lg p-3">
+        <Card className="bg-muted rounded-lg border-0 p-3 gap-0">
           <h4 className="text-sm font-semibold text-foreground mb-3 flex items-center">
             <Clock className="w-4 h-4 mr-2" />
             {t('monitoring.llmCalls.title')} ({details.llmCalls.length})
@@ -127,15 +130,16 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
                     <span className="font-medium text-foreground">
                       #{index + 1} {call.modelName}
                     </span>
-                    <span
-                      className={`ml-2 text-xs px-2 py-0.5 rounded ${
+                    <Badge
+                      className={cn(
+                        'ml-2',
                         call.status === 'success'
                           ? 'bg-green-100 text-green-800 dark:bg-green-900 dark:text-green-200'
-                          : 'bg-red-100 text-red-800 dark:bg-red-900 dark:text-red-200'
-                      }`}
+                          : 'bg-red-100 text-red-800 dark:bg-red-900 dark:text-red-200',
+                      )}
                     >
                       {call.status}
-                    </span>
+                    </Badge>
                   </div>
                   <span className="text-xs text-muted-foreground">
                     {call.duration}ms
@@ -163,12 +167,12 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
               </div>
             ))}
           </div>
-        </div>
+        </Card>
       )}
 
       {/* Errors Section */}
       {details.errors && details.errors.length > 0 && (
-        <div className="bg-muted rounded-lg p-3">
+        <Card className="bg-muted rounded-lg border-0 p-3 gap-0">
           <h4 className="text-sm font-semibold text-red-700 dark:text-red-400 mb-3 flex items-center">
             <AlertCircle className="w-4 h-4 mr-2" />
             {t('monitoring.errors.title')} ({details.errors.length})
@@ -198,12 +202,12 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
               </div>
             ))}
           </div>
-        </div>
+        </Card>
       )}
 
       {/* Query Variables Section */}
       {queryVariables && Object.keys(queryVariables).length > 0 && (
-        <div className="bg-muted rounded-lg p-3">
+        <Card className="bg-muted rounded-lg border-0 p-3 gap-0">
           <h4 className="text-sm font-semibold text-foreground mb-3 flex items-center">
             <Braces className="w-4 h-4 mr-2" />
             {t('monitoring.queryVariables.title')}
@@ -233,7 +237,7 @@ export function MessageDetailsCard({ details }: MessageDetailsCardProps) {
               </div>
             ))}
           </div>
-        </div>
+        </Card>
       )}
 
       {/* No data message */}
diff --git a/web/src/app/home/monitoring/components/TabState.tsx b/web/src/app/home/monitoring/components/TabState.tsx
new file mode 100644
index 000000000..09bd93678
--- /dev/null
+++ b/web/src/app/home/monitoring/components/TabState.tsx
@@ -0,0 +1,50 @@
+'use client';
+
+import React from 'react';
+import { Skeleton } from '@/components/ui/skeleton';
+
+interface TabStateProps {
+  /** Skeleton rows while the tab's data is in flight. */
+  loading?: boolean;
+  /** Number of skeleton rows; the rhythm matches the loaded content. */
+  rows?: number;
+  /** Icon for the empty state, in the same muted tone as the copy. */
+  icon?: React.ReactNode;
+  /** Empty-state headline. */
+  title?: string;
+  /** Optional second line explaining what would fill the tab. */
+  hint?: string;
+}
+
+/**
+ * The one loading/empty treatment every detail tab uses.
+ *
+ * Tabs are peers on one fixed-height surface, so they must not each invent their
+ * own placeholder shape: a reader switching tabs sees the same skeleton rhythm
+ * and the same centred empty card regardless of which record type is empty.
+ */
+export function TabState({
+  loading,
+  rows = 6,
+  icon,
+  title,
+  hint,
+}: TabStateProps) {
+  if (loading) {
+    return (
+      <div className="space-y-3" role="status" aria-busy="true">
+        {Array.from({ length: rows }).map((_, index) => (
+          <Skeleton key={index} className="h-14 w-full rounded-xl" />
+        ))}
+      </div>
+    );
+  }
+
+  return (
+    <div className="flex flex-col items-center justify-center gap-2 py-16 text-center text-muted-foreground">
+      {icon}
+      {title && <div className="text-sm">{title}</div>}
+      {hint && <div className="text-xs">{hint}</div>}
+    </div>
+  );
+}
diff --git a/web/src/app/home/monitoring/components/TokenMonitoring.tsx b/web/src/app/home/monitoring/components/TokenMonitoring.tsx
index 83b552ef3..fc9666065 100644
--- a/web/src/app/home/monitoring/components/TokenMonitoring.tsx
+++ b/web/src/app/home/monitoring/components/TokenMonitoring.tsx
@@ -19,6 +19,19 @@ import {
   AlertTriangle,
   TrendingUp,
 } from 'lucide-react';
+import { Tabs, TabsList, TabsTrigger } from '@/components/ui/tabs';
+import {
+  Table,
+  TableBody,
+  TableCell,
+  TableHead,
+  TableHeader,
+  TableRow,
+} from '@/components/ui/table';
+import { Card } from '@/components/ui/card';
+import { MetricCard } from './overview-cards/MetricCard';
+import LoadErrorState from '@/components/LoadErrorState';
+import { TabState } from './TabState';
 import { httpClient } from '@/app/infra/http/HttpClient';
 import { getErrorMessage } from '../utils';
 
@@ -64,6 +77,8 @@ interface TokenStatistics {
 }
 
 interface TokenMonitoringProps {
+  mode?: 'all' | 'real' | 'debug';
+  statusGroup?: string;
   botIds?: string[];
   pipelineIds?: string[];
   startTime?: string;
@@ -89,42 +104,9 @@ const TOOLTIP_STYLE: React.CSSProperties = {
   color: 'var(--foreground)',
 };
 
-function MetricTile({
-  icon,
-  label,
-  value,
-  sub,
-  accent,
-}: {
-  icon: React.ReactNode;
-  label: string;
-  value: string;
-  sub?: string;
-  accent?: string;
-}) {
-  return (
-    <div className="bg-card rounded-xl border p-4 flex flex-col gap-2">
-      <div className="flex items-center gap-2 text-muted-foreground text-sm">
-        <span
-          className="flex items-center justify-center h-7 w-7 rounded-lg"
-          style={{
-            backgroundColor: accent ? `${accent}1a` : 'var(--muted)',
-            color: accent || 'var(--foreground)',
-          }}
-        >
-          {icon}
-        </span>
-        {label}
-      </div>
-      <div className="text-2xl font-semibold text-foreground tabular-nums">
-        {value}
-      </div>
-      {sub && <div className="text-xs text-muted-foreground">{sub}</div>}
-    </div>
-  );
-}
-
 export default function TokenMonitoring({
+  mode,
+  statusGroup,
   botIds,
   pipelineIds,
   startTime,
@@ -145,6 +127,9 @@ export default function TokenMonitoring({
     setError(null);
     try {
       const result = await httpClient.getTokenStatistics({
+        mode,
+        status:
+          statusGroup && statusGroup !== 'all' ? [statusGroup] : undefined,
         botId: botIds,
         pipelineId: pipelineIds,
         startTime,
@@ -158,7 +143,16 @@ export default function TokenMonitoring({
       setLoading(false);
     }
     // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [botIdsKey, pipelineIdsKey, startTime, endTime, bucket, refreshKey]);
+  }, [
+    mode,
+    statusGroup,
+    botIdsKey,
+    pipelineIdsKey,
+    startTime,
+    endTime,
+    bucket,
+    refreshKey,
+  ]);
 
   useEffect(() => {
     fetchStats();
@@ -174,39 +168,27 @@ export default function TokenMonitoring({
     }));
   }, [stats]);
 
-  if (loading) {
-    return (
-      <div className="space-y-4">
-        <div className="grid grid-cols-2 md:grid-cols-3 lg:grid-cols-6 gap-4">
-          {Array.from({ length: 6 }).map((_, i) => (
-            <div
-              key={i}
-              className="bg-card rounded-xl border p-4 h-24 animate-pulse"
-            />
-          ))}
-        </div>
-        <div className="bg-card rounded-xl border p-6 h-[320px] animate-pulse" />
-      </div>
-    );
+  if (loading && !stats) {
+    return <TabState loading rows={6} />;
   }
 
   if (error) {
     return (
-      <div className="bg-card rounded-xl border p-6 text-sm text-destructive flex items-center gap-2">
-        <AlertTriangle className="h-4 w-4" />
-        {t('monitoring.tokens.loadError', { error })}
-      </div>
+      <LoadErrorState
+        title={t('monitoring.loadError')}
+        description={error}
+        onRetry={fetchStats}
+        busy={loading}
+      />
     );
   }
 
   if (!stats || !stats.summary || stats.summary.total_calls === 0) {
     return (
-      <div className="bg-card rounded-xl border p-6">
-        <div className="h-[260px] flex flex-col items-center justify-center text-muted-foreground gap-2">
-          <Coins className="h-[3rem] w-[3rem]" />
-          <div className="text-sm">{t('monitoring.tokens.noData')}</div>
-        </div>
-      </div>
+      <TabState
+        icon={<Coins className="h-12 w-12" />}
+        title={t('monitoring.tokens.noData')}
+      />
     );
   }
 
@@ -229,70 +211,69 @@ export default function TokenMonitoring({
 
       {/* Summary tiles */}
       <div className="grid grid-cols-2 md:grid-cols-3 lg:grid-cols-6 gap-4">
-        <MetricTile
+        <MetricCard
           icon={<Coins className="h-4 w-4" />}
           label={t('monitoring.tokens.totalTokens')}
           value={formatNumber(summary.total_tokens)}
-          sub={t('monitoring.tokens.acrossCalls', {
+          hint={t('monitoring.tokens.acrossCalls', {
             count: summary.total_calls,
           })}
           accent="#8b5cf6"
         />
-        <MetricTile
+        <MetricCard
           icon={<ArrowDownToLine className="h-4 w-4" />}
           label={t('monitoring.tokens.inputTokens')}
           value={formatNumber(summary.total_input_tokens)}
           accent="#3b82f6"
         />
-        <MetricTile
+        <MetricCard
           icon={<ArrowUpFromLine className="h-4 w-4" />}
           label={t('monitoring.tokens.outputTokens')}
           value={formatNumber(summary.total_output_tokens)}
           accent="#10b981"
         />
-        <MetricTile
+        <MetricCard
           icon={<TrendingUp className="h-4 w-4" />}
           label={t('monitoring.tokens.avgPerCall')}
           value={formatNumber(summary.avg_tokens_per_call)}
           accent="#f59e0b"
         />
-        <MetricTile
+        <MetricCard
           icon={<Gauge className="h-4 w-4" />}
           label={t('monitoring.tokens.throughput')}
           value={`${summary.avg_tokens_per_second}`}
-          sub={t('monitoring.tokens.tokensPerSec')}
+          hint={t('monitoring.tokens.tokensPerSec')}
           accent="#06b6d4"
         />
-        <MetricTile
+        <MetricCard
           icon={<AlertTriangle className="h-4 w-4" />}
           label={t('monitoring.tokens.errorCalls')}
           value={`${summary.error_calls}`}
-          sub={t('monitoring.tokens.ofTotal', { count: summary.total_calls })}
+          hint={t('monitoring.tokens.ofTotal', { count: summary.total_calls })}
           accent="#ef4444"
         />
       </div>
 
       {/* Token usage over time */}
-      <div className="bg-card rounded-xl border p-6">
+      <Card className="p-6">
         <div className="flex items-center justify-between mb-6">
           <h3 className="text-base font-semibold text-foreground">
             {t('monitoring.tokens.usageOverTime')}
           </h3>
-          <div className="inline-flex rounded-lg border p-0.5 text-sm">
-            {(['hour', 'day'] as const).map((b) => (
-              <button
-                key={b}
-                onClick={() => setBucket(b)}
-                className={`px-3 py-1 rounded-md transition-colors ${
-                  bucket === b
-                    ? 'bg-primary text-primary-foreground'
-                    : 'text-muted-foreground hover:text-foreground'
-                }`}
-              >
-                {t(`monitoring.tokens.bucket.${b}`)}
-              </button>
-            ))}
-          </div>
+          <Tabs
+            value={bucket}
+            onValueChange={(value) => {
+              if (value === 'hour' || value === 'day') setBucket(value);
+            }}
+          >
+            <TabsList className="h-8">
+              {(['hour', 'day'] as const).map((b) => (
+                <TabsTrigger key={b} value={b} className="text-xs">
+                  {t(`monitoring.tokens.bucket.${b}`)}
+                </TabsTrigger>
+              ))}
+            </TabsList>
+          </Tabs>
         </div>
         <div className="h-[320px]">
           <ResponsiveContainer width="100%" height="100%">
@@ -372,93 +353,86 @@ export default function TokenMonitoring({
             </ComposedChart>
           </ResponsiveContainer>
         </div>
-      </div>
+      </Card>
 
       {/* Per-model breakdown */}
-      <div className="bg-card rounded-xl border p-6">
+      <Card className="p-6">
         <h3 className="text-base font-semibold text-foreground mb-4">
           {t('monitoring.tokens.byModel')}
         </h3>
-        <div className="overflow-x-auto">
-          <table className="w-full text-sm">
-            <thead>
-              <tr className="text-left text-muted-foreground border-b">
-                <th className="py-2 pr-4 font-medium">
-                  {t('monitoring.tokens.model')}
-                </th>
-                <th className="py-2 px-4 font-medium text-right">
-                  {t('monitoring.tokens.calls')}
-                </th>
-                <th className="py-2 px-4 font-medium text-right">
-                  {t('monitoring.tokens.inputTokens')}
-                </th>
-                <th className="py-2 px-4 font-medium text-right">
-                  {t('monitoring.tokens.outputTokens')}
-                </th>
-                <th className="py-2 px-4 font-medium text-right">
-                  {t('monitoring.tokens.totalTokens')}
-                </th>
-                <th className="py-2 px-4 font-medium text-right">
-                  {t('monitoring.tokens.avgPerCall')}
-                </th>
-                <th className="py-2 pl-4 font-medium text-right">
-                  {t('monitoring.tokens.avgLatency')}
-                </th>
-              </tr>
-            </thead>
-            <tbody>
-              {by_model.map((m) => {
-                const share =
-                  summary.total_tokens > 0
-                    ? (m.total_tokens / summary.total_tokens) * 100
-                    : 0;
-                return (
-                  <tr
-                    key={m.model_name}
-                    className="border-b last:border-0 hover:bg-muted/40 transition-colors"
-                  >
-                    <td className="py-2.5 pr-4">
-                      <div className="font-medium text-foreground">
-                        {m.model_name}
-                      </div>
-                      <div className="mt-1 h-1.5 w-32 rounded-full bg-muted overflow-hidden">
-                        <div
-                          className="h-full rounded-full bg-violet-500"
-                          style={{ width: `${share}%` }}
-                        />
-                      </div>
-                    </td>
-                    <td className="py-2.5 px-4 text-right tabular-nums">
-                      {m.calls}
-                      {m.error_calls > 0 && (
-                        <span className="text-destructive">
-                          {' '}
-                          ({m.error_calls}✕)
-                        </span>
-                      )}
-                    </td>
-                    <td className="py-2.5 px-4 text-right tabular-nums">
-                      {formatNumber(m.input_tokens)}
-                    </td>
-                    <td className="py-2.5 px-4 text-right tabular-nums">
-                      {formatNumber(m.output_tokens)}
-                    </td>
-                    <td className="py-2.5 px-4 text-right tabular-nums font-medium">
-                      {formatNumber(m.total_tokens)}
-                    </td>
-                    <td className="py-2.5 px-4 text-right tabular-nums">
-                      {formatNumber(m.avg_tokens_per_call)}
-                    </td>
-                    <td className="py-2.5 pl-4 text-right tabular-nums">
-                      {m.avg_duration_ms}ms
-                    </td>
-                  </tr>
-                );
-              })}
-            </tbody>
-          </table>
-        </div>
-      </div>
+        <Table>
+          <TableHeader>
+            <TableRow>
+              <TableHead>{t('monitoring.tokens.model')}</TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.calls')}
+              </TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.inputTokens')}
+              </TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.outputTokens')}
+              </TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.totalTokens')}
+              </TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.avgPerCall')}
+              </TableHead>
+              <TableHead className="text-right">
+                {t('monitoring.tokens.avgLatency')}
+              </TableHead>
+            </TableRow>
+          </TableHeader>
+          <TableBody>
+            {by_model.map((m) => {
+              const share =
+                summary.total_tokens > 0
+                  ? (m.total_tokens / summary.total_tokens) * 100
+                  : 0;
+              return (
+                <TableRow key={m.model_name}>
+                  <TableCell>
+                    <div className="font-medium text-foreground">
+                      {m.model_name}
+                    </div>
+                    <div className="mt-1 h-1.5 w-32 rounded-full bg-muted overflow-hidden">
+                      <div
+                        className="h-full rounded-full bg-violet-500"
+                        style={{ width: `${share}%` }}
+                      />
+                    </div>
+                  </TableCell>
+                  <TableCell className="text-right tabular-nums">
+                    {m.calls}
+                    {m.error_calls > 0 && (
+                      <span className="text-destructive">
+                        {' '}
+                        ({m.error_calls}✕)
+                      </span>
+                    )}
+                  </TableCell>
+                  <TableCell className="text-right tabular-nums">
+                    {formatNumber(m.input_tokens)}
+                  </TableCell>
+                  <TableCell className="text-right tabular-nums">
+                    {formatNumber(m.output_tokens)}
+                  </TableCell>
+                  <TableCell className="text-right font-medium tabular-nums">
+                    {formatNumber(m.total_tokens)}
+                  </TableCell>
+                  <TableCell className="text-right tabular-nums">
+                    {formatNumber(m.avg_tokens_per_call)}
+                  </TableCell>
+                  <TableCell className="text-right tabular-nums">
+                    {m.avg_duration_ms}ms
+                  </TableCell>
+                </TableRow>
+              );
+            })}
+          </TableBody>
+        </Table>
+      </Card>
     </div>
   );
 }
diff --git a/web/src/app/home/monitoring/components/executions/ExecutionDetailSheet.tsx b/web/src/app/home/monitoring/components/executions/ExecutionDetailSheet.tsx
new file mode 100644
index 000000000..70d226487
--- /dev/null
+++ b/web/src/app/home/monitoring/components/executions/ExecutionDetailSheet.tsx
@@ -0,0 +1,543 @@
+import LoadErrorState from '@/components/LoadErrorState';
+import { useEffect, useRef, useState } from 'react';
+import { useTranslation } from 'react-i18next';
+import { Link } from 'react-router-dom';
+import { RefreshCw } from 'lucide-react';
+import {
+  Sheet,
+  SheetContent,
+  SheetDescription,
+  SheetHeader,
+  SheetTitle,
+} from '@/components/ui/sheet';
+import { Badge } from '@/components/ui/badge';
+import { Button } from '@/components/ui/button';
+import { Skeleton } from '@/components/ui/skeleton';
+import { ScrollArea } from '@/components/ui/scroll-area';
+import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
+import type {
+  ExecutionDetail,
+  ExecutionDetailItem,
+  ExecutionRow,
+  ExecutionSection,
+} from '@/app/infra/entities/api/monitoring-executions';
+import { ExecutionValue as Value, hasExecutionValue } from './ExecutionValue';
+import { formatRunDuration } from '@/app/home/agents/components/processor-run-timing';
+import { formatDateTime } from '../../utils/dateUtils';
+import {
+  executionKindLabel,
+  ExecutionKindBadge,
+  executionProcessorLabel,
+  executionStatusLabel,
+} from './ExecutionTable';
+
+export type ExecutionSelection = Pick<ExecutionRow, 'id'> & {
+  source: ExecutionRow['source'] | 'auto';
+};
+
+const mainSections: ExecutionSection[] = ['inputs', 'outputs', 'deliveries'];
+
+// Keep transport metadata in the payload disclosure, not in the message body.
+function messageBody(value: unknown): unknown {
+  if (typeof value === 'string') {
+    if (/^\s*[\[{]/.test(value)) {
+      try {
+        return messageBody(JSON.parse(value));
+      } catch {
+        /* Plain text. */
+      }
+    }
+    return value;
+  }
+  if (Array.isArray(value))
+    return value.map(messageBody).filter((item) => item != null && item !== '');
+  if (!value || typeof value !== 'object') return null;
+  const data = value as Record<string, unknown>;
+  if (
+    ['Image', 'File', 'Voice', 'At', 'AtAll', 'Quote'].includes(
+      String(data.type),
+    )
+  )
+    return value;
+  for (const key of ['text', 'content', 'message_chain', 'message', 'root']) {
+    if (data[key] != null) return messageBody(data[key]);
+  }
+  return null;
+}
+const secondarySections: ExecutionSection[] = [
+  'related',
+  'conversation',
+  'llm_calls',
+  'tool_calls',
+  'errors',
+  'events',
+];
+
+export default function ExecutionDetailSheet({
+  row,
+  onClose,
+  onSelect,
+}: {
+  row: ExecutionSelection | null;
+  onClose: () => void;
+  onSelect: (row: ExecutionSelection) => void;
+}) {
+  const { t } = useTranslation();
+  const workspace = useCurrentWorkspace()?.workspace.uuid;
+  const [detail, setDetail] = useState<ExecutionDetail | null>(null);
+  const [loading, setLoading] = useState(false);
+  const [error, setError] = useState(false);
+  const [sectionError, setSectionError] = useState<ExecutionSection | null>(
+    null,
+  );
+  const [busy, setBusy] = useState<ExecutionSection | null>(null);
+  const [refresh, setRefresh] = useState(0);
+  const request = useRef(0);
+  const identity = `${workspace}:${row?.source}:${row?.id}`;
+  const requestedIdentity = useRef('');
+  const [loadedIdentity, setLoadedIdentity] = useState('');
+  const visible = loadedIdentity === identity ? detail : null;
+  const id = row?.id;
+  const source = row?.source;
+  useEffect(() => {
+    const version = ++request.current;
+    // Refresh the current trace in place so its scroll position and expanded
+    // disclosures survive polling. Only a different selection starts empty.
+    if (requestedIdentity.current !== identity) {
+      setDetail(null);
+      requestedIdentity.current = identity;
+    }
+    setError(false);
+    setSectionError(null);
+    setBusy(null);
+    if (!id || !source) {
+      setLoading(false);
+      return;
+    }
+    setLoading(true);
+    backendClient
+      .getExecutionDetail(source, id)
+      .then((result) => {
+        if (version !== request.current) return;
+        setDetail(result);
+        setLoadedIdentity(identity);
+      })
+      .catch(() => {
+        if (version === request.current) setError(true);
+      })
+      .finally(() => {
+        if (version === request.current) setLoading(false);
+      });
+    return () => {
+      request.current += 1;
+    };
+  }, [id, source, workspace, identity, refresh]);
+  const status = visible?.row.status_group;
+  useEffect(() => {
+    if (!['running', 'queued'].includes(status ?? '')) return;
+    // Do not discard paged history with a background refresh.
+    if (
+      Object.values(visible?.pages ?? {}).some((page) => page.next_offset > 100)
+    )
+      return;
+    const timer = setInterval(() => setRefresh((value) => value + 1), 5000);
+    return () => clearInterval(timer);
+  }, [status, visible]);
+  const loadMore = async (section: ExecutionSection) => {
+    if (!visible || !row || busy) return;
+    const version = request.current;
+    setBusy(section);
+    setSectionError(null);
+    try {
+      const result = await backendClient.getExecutionDetail(
+        row.source,
+        row.id,
+        { section, offset: visible.pages[section]?.next_offset ?? 0 },
+      );
+      if (version !== request.current) return;
+      const next = result.pages[section];
+      if (next)
+        setDetail(
+          (previous) =>
+            previous && {
+              ...previous,
+              pages: {
+                ...previous.pages,
+                [section]: {
+                  ...next,
+                  items: [
+                    ...(previous.pages[section]?.items ?? []),
+                    ...next.items,
+                  ],
+                },
+              },
+            },
+        );
+    } catch {
+      if (version === request.current) setSectionError(section);
+    } finally {
+      if (version === request.current) setBusy(null);
+    }
+  };
+  const renderItem = (section: ExecutionSection, item: ExecutionDetailItem) => {
+    if (section === 'related') {
+      const related = item as unknown as ExecutionRow;
+      return (
+        <Button
+          variant="outline"
+          className="h-auto w-full justify-between whitespace-normal py-3 text-left"
+          onClick={() => onSelect(related)}
+        >
+          <span>
+            {executionProcessorLabel(related)} ·{' '}
+            {executionKindLabel(related, t)}
+          </span>
+          <Badge status={related.status_group}>
+            {executionStatusLabel(related, t)}
+          </Badge>
+        </Button>
+      );
+    }
+    const content = [
+      'inputs',
+      'outputs',
+      'deliveries',
+      'conversation',
+    ].includes(section);
+    const body = messageBody(item.content);
+    const title = content
+      ? [
+          item.actor_name || item.actor_id,
+          item.event_type,
+          item.role &&
+            t(`monitoring.unified.roles.${item.role}`, {
+              defaultValue: item.role,
+            }),
+        ]
+          .filter(Boolean)
+          .join(' · ')
+      : String(
+          item.model_name ||
+            item.tool_name ||
+            item.error_type ||
+            item.type ||
+            item.id,
+        );
+    return (
+      <div className="min-w-0 space-y-3 rounded-lg border bg-background p-4">
+        <div className="flex flex-wrap items-center justify-between gap-2 text-xs text-muted-foreground">
+          <span className="break-all">{title}</span>
+          <span>
+            {item.timestamp_ms
+              ? formatDateTime(new Date(item.timestamp_ms))
+              : typeof item.timestamp === 'string'
+                ? item.timestamp
+                : ''}
+          </span>
+          {item.status && (
+            <Badge status={item.status}>
+              {t(
+                `monitoring.execution.status.${['success', 'delivered'].includes(item.status) ? 'completed' : item.status}`,
+                { defaultValue: item.status },
+              )}
+            </Badge>
+          )}
+        </div>
+        {content ? (
+          <>
+            <div className="text-sm leading-7">
+              {hasExecutionValue(body) ? (
+                <Value value={body} />
+              ) : (
+                <p className="text-muted-foreground">
+                  {t('monitoring.unified.noRecords')}
+                </p>
+              )}
+            </div>
+            <Value value={item.attachments} />
+            {item.origin === 'generated' && section === 'conversation' && (
+              <Badge variant="outline">{t('monitoring.unified.outputs')}</Badge>
+            )}
+            {item.delivery?.delivery_error ? (
+              <p className="text-sm text-destructive">
+                {String(item.delivery.delivery_error)}
+              </p>
+            ) : null}
+            <details className="border-t pt-2">
+              <summary className="cursor-pointer text-xs text-muted-foreground">
+                {t('monitoring.unified.payload')}
+              </summary>
+              <div className="mt-3 max-h-80 overflow-auto rounded-md bg-muted/40 p-3">
+                <Value value={item} />
+              </div>
+            </details>
+          </>
+        ) : (
+          <>
+            <div className="flex flex-wrap gap-3 text-xs text-muted-foreground">
+              {typeof item.duration === 'number' && (
+                <span>{formatRunDuration(item.duration)}</span>
+              )}
+              {typeof item.total_tokens === 'number' && (
+                <span>{item.total_tokens.toLocaleString()} Token</span>
+              )}
+            </div>
+            <Value value={item.error_message} />
+            <details>
+              <summary className="cursor-pointer text-xs text-muted-foreground">
+                {t('monitoring.unified.payload')}
+              </summary>
+              <div className="mt-2">
+                <Value value={section === 'events' ? item.data : item} />
+              </div>
+            </details>
+          </>
+        )}
+      </div>
+    );
+  };
+  const renderSection = (section: ExecutionSection, primary = false) => {
+    const page = visible?.pages[section];
+    if (
+      !page ||
+      (!primary && page.items.length === 0 && section !== 'conversation')
+    )
+      return null;
+    const body = (
+      <div className="mt-3 space-y-3">
+        {section === 'conversation' && (
+          <p className="text-xs text-muted-foreground">
+            {t('monitoring.unified.contextHint')}
+          </p>
+        )}
+        {page.items.length === 0 && (
+          <p className="text-sm text-muted-foreground">
+            {t('monitoring.unified.noRecords')}
+          </p>
+        )}
+        {page.items.map((item, i) => (
+          <div key={`${item.id}-${i}`}>{renderItem(section, item)}</div>
+        ))}
+        {sectionError === section && (
+          <LoadErrorState
+            compact
+            title={t('monitoring.execution.detail.loadError')}
+            onRetry={() => loadMore(section)}
+          />
+        )}
+        {page.has_more && (
+          <Button
+            variant="outline"
+            size="sm"
+            disabled={busy !== null}
+            onClick={() => loadMore(section)}
+          >
+            {t('monitoring.unified.loadMore')}
+          </Button>
+        )}
+      </div>
+    );
+    const title = `${t(`monitoring.unified.${section}`)} (${page.items.length}${page.has_more ? '+' : ''})`;
+    return primary ? (
+      <section key={section} className="space-y-3">
+        <h3 className="flex items-center gap-2 text-sm font-semibold">
+          <span className="flex size-6 items-center justify-center rounded-full bg-muted text-xs text-muted-foreground">
+            {mainSections.indexOf(section) + 1}
+          </span>
+          {title}
+        </h3>
+        {body}
+      </section>
+    ) : (
+      <details
+        key={section}
+        open={section === 'errors' ? true : undefined}
+        className={`rounded-lg border p-3 ${section === 'errors' ? 'border-red-500/25 bg-red-500/5' : ''}`}
+      >
+        <summary className="cursor-pointer text-sm font-medium">
+          {title}
+        </summary>
+        {body}
+      </details>
+    );
+  };
+  const record = visible?.row;
+  const href = record?.target_id
+    ? record.target_kind === 'pipeline'
+      ? `/home/pipelines?id=${encodeURIComponent(record.target_id)}`
+      : ['agent', 'event_processor', 'processor'].includes(record.target_kind)
+        ? `/home/agents?id=${encodeURIComponent(record.target_id)}`
+        : null
+    : null;
+  return (
+    <Sheet
+      open={row !== null}
+      onOpenChange={(open) => {
+        if (!open) onClose();
+      }}
+    >
+      <SheetContent side="right" className="w-full p-0 sm:max-w-3xl">
+        <SheetHeader className="sr-only">
+          <SheetTitle>{t('monitoring.execution.detail.title')}</SheetTitle>
+          <SheetDescription>
+            {t('monitoring.execution.detail.a11yDescription')}
+          </SheetDescription>
+        </SheetHeader>
+        <ScrollArea className="h-full">
+          <div className="space-y-5 p-5 sm:p-6">
+            <div className="flex items-start justify-between gap-3 pr-8">
+              <h2 className="text-lg font-semibold">
+                {record
+                  ? record.source === 'event'
+                    ? t('monitoring.unified.unhandled')
+                    : executionProcessorLabel(record)
+                  : t('monitoring.execution.detail.title')}
+              </h2>
+              <Button
+                variant="ghost"
+                size="icon"
+                disabled={loading}
+                aria-label={t('monitoring.refreshData')}
+                title={
+                  error
+                    ? t('monitoring.execution.detail.loadError')
+                    : t('monitoring.refreshData')
+                }
+                onClick={() => setRefresh((value) => value + 1)}
+              >
+                <RefreshCw
+                  className={`h-4 w-4 ${loading ? 'animate-spin' : ''} ${error ? 'text-destructive' : ''}`}
+                />
+              </Button>
+            </div>
+            {loading && !visible && <Skeleton className="h-36 w-full" />}
+            {error && !visible && (
+              <LoadErrorState
+                title={t('monitoring.execution.detail.loadError')}
+                onRetry={() => setRefresh((value) => value + 1)}
+              />
+            )}
+            {record && (
+              <>
+                <div className="flex flex-wrap items-center gap-2">
+                  <Badge status={record.status_group}>
+                    {executionStatusLabel(record, t)}
+                  </Badge>
+                  <ExecutionKindBadge row={record} />
+                  {record.debug && (
+                    <Badge status="debug">
+                      {t('monitoring.execution.debug')}
+                    </Badge>
+                  )}
+                  {href && (
+                    <Button asChild variant="outline" size="sm">
+                      <Link to={href}>
+                        {t('monitoring.execution.detail.openObject')}
+                      </Link>
+                    </Button>
+                  )}
+                </div>
+                <p className="text-sm text-muted-foreground">
+                  {[
+                    record.bot_name,
+                    record.platform,
+                    record.user_name || record.user_id,
+                  ]
+                    .filter(Boolean)
+                    .join(' · ')}
+                </p>
+                <div className="grid grid-cols-3 divide-x rounded-lg border bg-muted/20">
+                  <div className="space-y-1 p-3">
+                    <p className="text-xs text-muted-foreground">
+                      {t('monitoring.execution.detail.duration')}
+                    </p>
+                    <p className="text-sm font-semibold tabular-nums">
+                      {record.duration_ms != null
+                        ? formatRunDuration(record.duration_ms)
+                        : '—'}
+                    </p>
+                  </div>
+                  <div className="space-y-1 p-3">
+                    <p className="text-xs text-muted-foreground">
+                      {t('monitoring.llmCalls.totalTokens')}
+                    </p>
+                    <p className="text-sm font-semibold tabular-nums">
+                      {record.usage?.total_tokens?.toLocaleString() ?? '—'}
+                    </p>
+                  </div>
+                  <div className="space-y-1 p-3">
+                    <p className="text-xs text-muted-foreground">
+                      {t('monitoring.unified.deliveries')}
+                    </p>
+                    <p className="text-sm font-semibold tabular-nums">
+                      {visible?.pages.deliveries?.items.length ?? 0}
+                      {visible?.pages.deliveries?.has_more ? '+' : ''}
+                    </p>
+                  </div>
+                </div>
+                {record.status_reason &&
+                  !['stop', 'completed', 'success'].includes(
+                    record.status_reason,
+                  ) && (
+                    <p className="rounded-md bg-muted p-3 text-sm">
+                      {record.status_reason
+                        .split('; ')
+                        .map((reason) =>
+                          reason === 'failure_reason_not_recorded'
+                            ? t('common.failureReasonNotRecorded')
+                            : reason,
+                        )
+                        .join('; ')}
+                    </p>
+                  )}
+                {mainSections.map((section) => renderSection(section, true))}
+                {secondarySections.map((section) => renderSection(section))}
+                <details className="rounded-lg border p-3">
+                  <summary className="cursor-pointer text-sm">
+                    {t('monitoring.unified.metadata')}
+                  </summary>
+                  <dl className="mt-3 space-y-2 text-xs">
+                    {[
+                      [t('monitoring.execution.detail.executionId'), record.id],
+                      [
+                        t('monitoring.execution.detail.eventId'),
+                        record.event_id,
+                      ],
+                      [
+                        t('monitoring.unified.conversation'),
+                        record.conversation_id || record.session_id,
+                      ],
+                      [
+                        t('monitoring.execution.detail.runner'),
+                        record.runner_id,
+                      ],
+                      [
+                        t('monitoring.execution.detail.created'),
+                        record.created_at_ms
+                          ? formatDateTime(new Date(record.created_at_ms))
+                          : null,
+                      ],
+                      [
+                        t('monitoring.execution.detail.duration'),
+                        record.duration_ms != null
+                          ? formatRunDuration(record.duration_ms)
+                          : null,
+                      ],
+                    ]
+                      .filter(([, value]) => value)
+                      .map(([label, value]) => (
+                        <div key={label}>
+                          <dt className="text-muted-foreground">{label}</dt>
+                          <dd className="break-all">{value}</dd>
+                        </div>
+                      ))}
+                  </dl>
+                </details>
+              </>
+            )}
+          </div>
+        </ScrollArea>
+      </SheetContent>
+    </Sheet>
+  );
+}
diff --git a/web/src/app/home/monitoring/components/executions/ExecutionTable.tsx b/web/src/app/home/monitoring/components/executions/ExecutionTable.tsx
new file mode 100644
index 000000000..82f4ca935
--- /dev/null
+++ b/web/src/app/home/monitoring/components/executions/ExecutionTable.tsx
@@ -0,0 +1,216 @@
+import React from 'react';
+import { Bot, Workflow, Puzzle, CircleDashed } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+import type { TFunction } from 'i18next';
+import {
+  Table,
+  TableBody,
+  TableCell,
+  TableHead,
+  TableHeader,
+  TableRow,
+} from '@/components/ui/table';
+import { Badge } from '@/components/ui/badge';
+import type { ExecutionRow } from '@/app/infra/entities/api/monitoring-executions';
+import { eventPatternLabel } from '@/app/home/components/event-patterns/event-pattern-groups';
+import { formatRunDuration } from '@/app/home/agents/components/processor-run-timing';
+import { formatDateTime } from '../../utils/dateUtils';
+
+interface ExecutionTableProps {
+  rows: ExecutionRow[];
+  selectedId?: string | null;
+  onSelect: (row: ExecutionRow) => void;
+}
+
+const STATUS_GROUPS = [
+  'completed',
+  'failed',
+  'running',
+  'queued',
+  'cancelled',
+  'ignored',
+];
+
+export function executionStatusLabel(row: ExecutionRow, t: TFunction): string {
+  if (STATUS_GROUPS.includes(row.status_group)) {
+    return t(`monitoring.execution.status.${row.status_group}`);
+  }
+  return row.status;
+}
+
+export function executionStatusVariant(
+  statusGroup: string,
+): 'default' | 'secondary' | 'destructive' | 'outline' {
+  if (statusGroup === 'failed') return 'destructive';
+  if (statusGroup === 'ignored' || statusGroup === 'queued') return 'secondary';
+  return 'outline';
+}
+
+const KIND_GROUPS = [
+  'agent',
+  'pipeline',
+  'processor',
+  'event_processor',
+  'event',
+];
+
+export function executionKindLabel(row: ExecutionRow, t: TFunction): string {
+  if (KIND_GROUPS.includes(row.target_kind)) {
+    if (row.target_kind === 'event_processor')
+      return t('monitoring.execution.kind.processor');
+    if (row.target_kind === 'event') return t('monitoring.unified.unhandled');
+    return t(`monitoring.execution.kind.${row.target_kind}`);
+  }
+  return row.target_kind;
+}
+
+export function ExecutionKindBadge({ row }: { row: ExecutionRow }) {
+  const { t } = useTranslation();
+  const Icon =
+    row.target_kind === 'pipeline'
+      ? Workflow
+      : row.target_kind === 'agent'
+        ? Bot
+        : ['processor', 'event_processor'].includes(row.target_kind)
+          ? Puzzle
+          : CircleDashed;
+  const color =
+    row.target_kind === 'pipeline'
+      ? 'border-blue-200/70 bg-blue-50/70 text-blue-700 dark:border-blue-800/50 dark:bg-blue-950/30 dark:text-blue-300'
+      : row.target_kind === 'agent'
+        ? 'border-emerald-200/70 bg-emerald-50/70 text-emerald-700 dark:border-emerald-800/50 dark:bg-emerald-950/30 dark:text-emerald-300'
+        : ['processor', 'event_processor'].includes(row.target_kind)
+          ? 'border-amber-200/70 bg-amber-50/70 text-amber-700 dark:border-amber-800/50 dark:bg-amber-950/30 dark:text-amber-300'
+          : 'border-border bg-muted/40 text-muted-foreground';
+  return (
+    <Badge variant="outline" className={`gap-1 whitespace-nowrap ${color}`}>
+      <Icon className="size-3 shrink-0" aria-hidden="true" />
+      {executionKindLabel(row, t)}
+    </Badge>
+  );
+}
+
+export function executionTargetLabel(row: ExecutionRow, t: TFunction): string {
+  if (row.source !== 'pipeline') {
+    return eventPatternLabel(row.title, t);
+  }
+  const title = row.title ?? '';
+  return title.length > 48 ? `${title.slice(0, 48)}…` : title;
+}
+
+// The processor the row ran: the Agent or Pipeline that handled it. A run whose
+// Agent was deleted still names the Runner that executed it.
+export function executionProcessorLabel(row: ExecutionRow): string {
+  return row.target_name || row.runner_id || row.target_id || '—';
+}
+
+export default function ExecutionTable({
+  rows,
+  selectedId,
+  onSelect,
+}: ExecutionTableProps) {
+  const { t } = useTranslation();
+
+  return (
+    // The page hands this table a fixed-height region: the table's own scroll
+    // container takes that height so the header stays sticky while rows scroll
+    // under it.
+    <div className="flex h-full min-h-0 flex-col overflow-hidden rounded-xl border">
+      <Table containerClassName="min-h-0 flex-1 overflow-y-auto">
+        <TableHeader className="bg-card sticky top-0 z-10">
+          <TableRow>
+            <TableHead>{t('monitoring.execution.columns.time')}</TableHead>
+            <TableHead>{t('monitoring.execution.columns.target')}</TableHead>
+            <TableHead>{t('monitoring.unified.inputs')}</TableHead>
+            <TableHead>{t('monitoring.execution.columns.status')}</TableHead>
+            <TableHead className="text-right">
+              {t('monitoring.execution.columns.duration')}
+            </TableHead>
+            <TableHead className="text-right">
+              {t('monitoring.execution.columns.tokens')}
+            </TableHead>
+          </TableRow>
+        </TableHeader>
+        <TableBody>
+          {rows.map((row) => {
+            const selected = selectedId === row.id;
+            return (
+              <TableRow
+                key={`${row.source}-${row.id}`}
+                role="button"
+                tabIndex={0}
+                aria-pressed={selected}
+                data-state={selected ? 'selected' : undefined}
+                className="cursor-pointer data-[state=selected]:bg-muted"
+                onClick={() => onSelect(row)}
+                onKeyDown={(
+                  event: React.KeyboardEvent<HTMLTableRowElement>,
+                ) => {
+                  if (event.key === 'Enter' || event.key === ' ') {
+                    event.preventDefault();
+                    onSelect(row);
+                  }
+                }}
+              >
+                <TableCell className="whitespace-nowrap text-muted-foreground">
+                  {row.created_at_ms == null
+                    ? '—'
+                    : formatDateTime(new Date(row.created_at_ms))}
+                </TableCell>
+                <TableCell>
+                  <div className="flex flex-col gap-1">
+                    <div className="flex items-center gap-2">
+                      <ExecutionKindBadge row={row} />
+                      <span className="truncate font-medium">
+                        {row.source === 'event'
+                          ? t('monitoring.unified.unhandled')
+                          : executionProcessorLabel(row)}
+                      </span>
+                    </div>
+                    <span className="truncate text-xs text-muted-foreground">
+                      {t('monitoring.execution.triggeredBy', {
+                        trigger: executionTargetLabel(row, t),
+                      })}
+                    </span>
+                  </div>
+                </TableCell>
+                <TableCell className="max-w-xs">
+                  <div className="line-clamp-2 break-words text-sm">
+                    {row.input_preview || executionTargetLabel(row, t) || '—'}
+                  </div>
+                  <div className="mt-1 truncate text-xs text-muted-foreground">
+                    {[row.bot_name, row.user_name || row.user_id]
+                      .filter(Boolean)
+                      .join(' · ')}
+                  </div>
+                </TableCell>
+                <TableCell>
+                  <div className="flex flex-wrap items-center gap-1.5">
+                    <Badge status={row.status_group}>
+                      {executionStatusLabel(row, t)}
+                    </Badge>
+                    {row.debug && (
+                      <Badge status="debug">
+                        {t('monitoring.execution.debug')}
+                      </Badge>
+                    )}
+                  </div>
+                </TableCell>
+                <TableCell className="text-right tabular-nums">
+                  {row.duration_ms == null
+                    ? '—'
+                    : formatRunDuration(row.duration_ms)}
+                </TableCell>
+                <TableCell className="text-right tabular-nums">
+                  {row.usage?.total_tokens != null
+                    ? row.usage.total_tokens.toLocaleString()
+                    : '—'}
+                </TableCell>
+              </TableRow>
+            );
+          })}
+        </TableBody>
+      </Table>
+    </div>
+  );
+}
diff --git a/web/src/app/home/monitoring/components/executions/ExecutionValue.tsx b/web/src/app/home/monitoring/components/executions/ExecutionValue.tsx
new file mode 100644
index 000000000..72913462d
--- /dev/null
+++ b/web/src/app/home/monitoring/components/executions/ExecutionValue.tsx
@@ -0,0 +1,178 @@
+import { MessageContentRenderer } from '../MessageContentRenderer';
+
+// Keep arbitrary event properties, including false and zero. Empty transport
+// fields are not messages. This also applies to historical serialized payloads.
+export function hasExecutionValue(value: unknown): boolean {
+  if (value == null) return false;
+  if (typeof value === 'string') return value.trim().length > 0;
+  if (Array.isArray(value)) return value.some(hasExecutionValue);
+  if (typeof value === 'object')
+    return Object.values(value).some(hasExecutionValue);
+  return true;
+}
+
+function textProjection(value: unknown): string {
+  if (typeof value === 'string') return value;
+  if (Array.isArray(value)) return value.map(textProjection).join('');
+  if (value && typeof value === 'object') {
+    const item = value as Record<string, unknown>;
+    if (item.type === 'text' || item.type === 'Plain')
+      return typeof item.text === 'string' ? item.text : '';
+  }
+  return '';
+}
+
+const messageTypes = new Set([
+  'Plain',
+  'At',
+  'AtAll',
+  'Image',
+  'File',
+  'Voice',
+  'Quote',
+]);
+const componentFields: Record<string, string[]> = {
+  Plain: ['text'],
+  At: ['target', 'display'],
+  AtAll: [],
+  Image: ['url', 'base64', 'imageId'],
+  File: ['name', 'id', 'url', 'size'],
+  Voice: ['url', 'base64', 'voiceId', 'length'],
+  Quote: ['origin'],
+};
+
+export function ExecutionValue({ value }: { value: unknown }) {
+  if (!hasExecutionValue(value)) return null;
+  if (typeof value === 'string') {
+    // Do not reinterpret ordinary strings such as "false", "0", or quoted
+    // user text. Only historical structured JSON needs decoding.
+    if (/^\s*[\[{]/.test(value)) {
+      try {
+        return <ExecutionValue value={JSON.parse(value)} />;
+      } catch {
+        /* Text. */
+      }
+    }
+    return (
+      <pre className="whitespace-pre-wrap break-words font-sans text-sm [overflow-wrap:anywhere]">
+        {value}
+      </pre>
+    );
+  }
+  if (Array.isArray(value))
+    return (
+      <div className="space-y-2">
+        {value.map((item, index) => (
+          <ExecutionValue key={index} value={item} />
+        ))}
+      </div>
+    );
+  if (typeof value === 'object') {
+    const record = value as Record<string, unknown>;
+    if (typeof record.type === 'string' && messageTypes.has(record.type)) {
+      const fields = ['type', ...componentFields[record.type]];
+      return (
+        <>
+          <MessageContentRenderer
+            content={JSON.stringify([record])}
+            maxLines={12}
+          />
+          <ExecutionValue
+            value={Object.fromEntries(
+              Object.entries(record).filter(([key]) => !fields.includes(key)),
+            )}
+          />
+        </>
+      );
+    }
+    if (record.type === 'text') {
+      const rest = Object.fromEntries(
+        Object.entries(record).filter(
+          ([key, item]) =>
+            key !== 'type' && key !== 'text' && hasExecutionValue(item),
+        ),
+      );
+      return (
+        <>
+          <ExecutionValue value={record.text} />
+          <ExecutionValue value={rest} />
+        </>
+      );
+    }
+    if (['image', 'image_url', 'image_base64'].includes(String(record.type))) {
+      const image = record.image_url;
+      const url =
+        typeof image === 'string'
+          ? image
+          : image && typeof image === 'object'
+            ? (image as { url?: string }).url
+            : typeof record.image_base64 === 'string'
+              ? record.image_base64.startsWith('data:')
+                ? record.image_base64
+                : `data:image/png;base64,${record.image_base64}`
+              : null;
+      if (url)
+        return (
+          <>
+            <MessageContentRenderer
+              content={JSON.stringify([{ type: 'Image', url }])}
+              maxLines={12}
+            />
+            <ExecutionValue
+              value={Object.fromEntries(
+                Object.entries(record).filter(
+                  ([key]) =>
+                    !['type', 'image_url', 'image_base64'].includes(key),
+                ),
+              )}
+            />
+          </>
+        );
+    }
+    if (['file', 'file_url', 'file_base64'].includes(String(record.type))) {
+      const name = record.file_name ?? record.name;
+      return (
+        <>
+          <MessageContentRenderer
+            content={JSON.stringify([{ type: 'File', name }])}
+            maxLines={12}
+          />
+          <ExecutionValue
+            value={Object.fromEntries(
+              Object.entries(record).filter(
+                ([key]) =>
+                  !['type', 'file_name', 'name', 'file_base64'].includes(key),
+              ),
+            )}
+          />
+        </>
+      );
+    }
+    // Only equal text projections are aliases. Never deduplicate distinct
+    // records, or hide different text and structured content from one event.
+    const bodyKeys = ['chain', 'message_chain', 'contents', 'content'];
+    const bodyKey = bodyKeys.find((key) => hasExecutionValue(record[key]));
+    const body = bodyKey ? record[bodyKey] : undefined;
+    const duplicateText =
+      bodyKey &&
+      typeof record.text === 'string' &&
+      record.text === textProjection(body);
+    const entries = Object.entries(record).filter(
+      ([key, item]) =>
+        hasExecutionValue(item) && !(key === 'text' && duplicateText),
+    );
+    return (
+      <div className="space-y-2">
+        {entries.map(([key, item]) => (
+          <div key={key}>
+            {!['text', ...bodyKeys, 'attachments'].includes(key) && (
+              <div className="text-xs text-muted-foreground">{key}</div>
+            )}
+            <ExecutionValue value={item} />
+          </div>
+        ))}
+      </div>
+    );
+  }
+  return <span className="text-sm">{String(value)}</span>;
+}
diff --git a/web/src/app/home/monitoring/components/filters/MonitoringFilters.tsx b/web/src/app/home/monitoring/components/filters/MonitoringFilters.tsx
index 6cf204e14..ec3fb963b 100644
--- a/web/src/app/home/monitoring/components/filters/MonitoringFilters.tsx
+++ b/web/src/app/home/monitoring/components/filters/MonitoringFilters.tsx
@@ -1,4 +1,4 @@
-import React, { useEffect, useState } from 'react';
+import React, { useEffect, useRef, useState } from 'react';
 import { useTranslation } from 'react-i18next';
 import {
   Select,
@@ -7,44 +7,99 @@ import {
   SelectTrigger,
   SelectValue,
 } from '@/components/ui/select';
-import { backendClient } from '@/app/infra/http';
+import {
+  Popover,
+  PopoverContent,
+  PopoverTrigger,
+} from '@/components/ui/popover';
+import { Checkbox } from '@/components/ui/checkbox';
+import { Button } from '@/components/ui/button';
+import {
+  Dialog,
+  DialogContent,
+  DialogHeader,
+  DialogTitle,
+  DialogDescription,
+} from '@/components/ui/dialog';
+import { Input } from '@/components/ui/input';
+import { ChevronDown } from 'lucide-react';
+import { DateRange } from '../../types/monitoring';
+import { Label } from '@/components/ui/label';
+import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
 import { TimeRangeOption } from '../../types/monitoring';
+import type { ExecutionModeFilter } from '@/app/infra/entities/api/monitoring-executions';
 
 interface MonitoringFiltersProps {
   selectedBots: string[];
   selectedPipelines: string[];
   timeRange: TimeRangeOption;
+  customDateRange?: DateRange | null;
+  onCustomDateRangeChange?: (range: DateRange) => void;
   onBotsChange: (bots: string[]) => void;
   onPipelinesChange: (pipelines: string[]) => void;
   onTimeRangeChange: (timeRange: TimeRangeOption) => void;
+  mode?: ExecutionModeFilter;
+  onModeChange?: (value: ExecutionModeFilter) => void;
+  statusGroup?: string;
+  onStatusGroupChange?: (value: string) => void;
 }
 
+const STATUS_GROUP_OPTIONS = [
+  'completed',
+  'failed',
+  'running',
+  'queued',
+  'cancelled',
+  'ignored',
+];
+
+const STATUS_COLORS: Record<string, string> = {
+  completed: 'text-emerald-700 dark:text-emerald-400',
+  failed: 'text-red-600 dark:text-red-400',
+  running: 'text-blue-600 dark:text-blue-400',
+  queued: 'text-amber-700 dark:text-amber-400',
+  cancelled: 'text-slate-600 dark:text-slate-400',
+  ignored: 'text-muted-foreground',
+};
+
 interface Bot {
   uuid: string;
   name: string;
+  iconURL: string;
 }
 
-interface Pipeline {
+interface Processor {
   uuid: string;
   name: string;
+  emoji?: string;
+  kind: string;
 }
 
 export default function MonitoringFilters({
   selectedBots,
   selectedPipelines,
   timeRange,
+  customDateRange,
+  onCustomDateRangeChange,
   onBotsChange,
   onPipelinesChange,
   onTimeRangeChange,
+  mode,
+  onModeChange,
+  statusGroup,
+  onStatusGroupChange,
 }: MonitoringFiltersProps) {
   const { t } = useTranslation();
+  const workspace = useCurrentWorkspace()?.workspace.uuid;
   const [bots, setBots] = useState<Bot[]>([]);
-  const [pipelines, setPipelines] = useState<Pipeline[]>([]);
+  const [processors, setProcessors] = useState<Processor[]>([]);
   const [loadingBots, setLoadingBots] = useState(false);
-  const [loadingPipelines, setLoadingPipelines] = useState(false);
+  const [loadingProcessors, setLoadingProcessors] = useState(false);
 
   // Fetch bots list
   useEffect(() => {
+    let active = true;
+    setBots([]);
     const fetchBots = async () => {
       setLoadingBots(true);
       try {
@@ -52,41 +107,58 @@ export default function MonitoringFilters({
         // Filter out bots without uuid and map to local Bot interface
         const validBots = (response.bots || [])
           .filter((bot): bot is typeof bot & { uuid: string } => !!bot.uuid)
-          .map((bot) => ({ uuid: bot.uuid, name: bot.name }));
-        setBots(validBots);
+          .map((bot) => ({
+            uuid: bot.uuid,
+            name: bot.name,
+            iconURL: backendClient.getAdapterIconURL(bot.adapter),
+          }));
+        if (active) setBots(validBots);
       } catch (error) {
         console.error('Failed to fetch bots:', error);
       } finally {
-        setLoadingBots(false);
+        if (active) setLoadingBots(false);
       }
     };
 
     fetchBots();
-  }, []);
+    return () => {
+      active = false;
+    };
+  }, [workspace]);
 
-  // Fetch pipelines list
+  // Fetch all processor kinds from the same catalog as the sidebar
   useEffect(() => {
-    const fetchPipelines = async () => {
-      setLoadingPipelines(true);
+    let active = true;
+    setProcessors([]);
+    const fetchProcessors = async () => {
+      setLoadingProcessors(true);
       try {
-        const response = await backendClient.getPipelines();
-        // Filter out pipelines without uuid and map to local Pipeline interface
-        const validPipelines = (response.pipelines || [])
+        const response = await backendClient.getAgents();
+        // Preserve the custom icon displayed to the left of the sidebar name.
+        const validProcessors = (response.agents || [])
           .filter(
             (pipeline): pipeline is typeof pipeline & { uuid: string } =>
               !!pipeline.uuid,
           )
-          .map((pipeline) => ({ uuid: pipeline.uuid, name: pipeline.name }));
-        setPipelines(validPipelines);
+          .map((pipeline) => ({
+            uuid: pipeline.uuid,
+            name: pipeline.name,
+            emoji: pipeline.emoji,
+            kind: pipeline.kind,
+          }));
+        if (active) setProcessors(validProcessors);
       } catch (error) {
-        console.error('Failed to fetch pipelines:', error);
+        console.error('Failed to fetch processors:', error);
       } finally {
-        setLoadingPipelines(false);
+        if (active) setLoadingProcessors(false);
       }
     };
 
-    fetchPipelines();
-  }, []);
+    fetchProcessors();
+    return () => {
+      active = false;
+    };
+  }, [workspace]);
 
   const handleBotChange = (value: string) => {
     if (value === 'all') {
@@ -96,31 +168,57 @@ export default function MonitoringFilters({
     }
   };
 
-  const handlePipelineChange = (value: string) => {
-    if (value === 'all') {
-      onPipelinesChange([]);
-    } else {
-      onPipelinesChange([value]);
-    }
+  const customPending = useRef(false);
+  const [dateOpen, setDateOpen] = useState(false);
+  const [fromDate, setFromDate] = useState('');
+  const [toDate, setToDate] = useState('');
+  const localDate = (date: Date) =>
+    new Date(date.getTime() - date.getTimezoneOffset() * 60000)
+      .toISOString()
+      .slice(0, 16);
+  const openCustom = () => {
+    setFromDate(
+      localDate(customDateRange?.from ?? new Date(Date.now() - 86400000)),
+    );
+    setToDate(localDate(customDateRange?.to ?? new Date()));
+    setDateOpen(true);
   };
-
+  const from = new Date(fromDate);
+  const to = new Date(toDate);
+  const validDates =
+    Number.isFinite(from.getTime()) &&
+    Number.isFinite(to.getTime()) &&
+    from < to &&
+    to.getTime() - from.getTime() <= 365 * 86400000 &&
+    to.getTime() <= Date.now();
   const handleTimeRangeChange = (value: string) => {
-    onTimeRangeChange(value as TimeRangeOption);
+    if (value === 'custom') customPending.current = true;
+    else onTimeRangeChange(value as TimeRangeOption);
+  };
+  const toggleProcessors = (ids: string[], checked: boolean) => {
+    const next = new Set(selectedPipelines);
+    ids.forEach((id) => (checked ? next.add(id) : next.delete(id)));
+    onPipelinesChange([...next]);
   };
-
   return (
-    <div className="flex w-full flex-col gap-3 sm:w-auto sm:flex-row sm:flex-wrap sm:items-center sm:gap-6">
+    <div className="grid min-w-0 flex-1 grid-cols-2 gap-3 lg:grid-cols-5">
       {/* Bot Filter */}
-      <div className="flex items-center gap-2">
-        <label className="w-20 shrink-0 text-sm font-medium text-foreground sm:w-auto sm:whitespace-nowrap">
+      <div className="min-w-0 space-y-1.5">
+        <Label
+          htmlFor="monitoring-filter-bot"
+          className="text-xs font-medium text-muted-foreground"
+        >
           {t('monitoring.filters.bot')}
-        </label>
+        </Label>
         <Select
           value={selectedBots.length === 0 ? 'all' : selectedBots[0]}
           onValueChange={handleBotChange}
           disabled={loadingBots}
         >
-          <SelectTrigger className="h-9 w-full sm:w-[140px]">
+          <SelectTrigger
+            id="monitoring-filter-bot"
+            className="h-9 w-full min-w-0"
+          >
             <SelectValue
               placeholder={
                 loadingBots
@@ -135,55 +233,234 @@ export default function MonitoringFilters({
             </SelectItem>
             {bots.map((bot) => (
               <SelectItem key={bot.uuid} value={bot.uuid}>
-                {bot.name}
+                <span className="flex min-w-0 items-center gap-2">
+                  <img
+                    src={bot.iconURL}
+                    alt=""
+                    className="size-4 shrink-0 rounded"
+                  />
+                  <span className="truncate">{bot.name}</span>
+                </span>
               </SelectItem>
             ))}
           </SelectContent>
         </Select>
       </div>
 
-      {/* Pipeline Filter */}
-      <div className="flex items-center gap-2">
-        <label className="w-20 shrink-0 text-sm font-medium text-foreground sm:w-auto sm:whitespace-nowrap">
-          {t('monitoring.filters.pipeline')}
-        </label>
-        <Select
-          value={selectedPipelines.length === 0 ? 'all' : selectedPipelines[0]}
-          onValueChange={handlePipelineChange}
-          disabled={loadingPipelines}
+      {/* Processor Filter */}
+      <div className="min-w-0 space-y-1.5">
+        <Label
+          htmlFor="monitoring-filter-processor"
+          className="text-xs font-medium text-muted-foreground"
         >
-          <SelectTrigger className="h-9 w-full sm:w-[140px]">
-            <SelectValue
-              placeholder={
-                loadingPipelines
-                  ? t('monitoring.filters.loading')
-                  : t('monitoring.filters.selectPipeline')
-              }
-            />
-          </SelectTrigger>
-          <SelectContent>
-            <SelectItem value="all">
-              {t('monitoring.filters.allPipelines')}
-            </SelectItem>
-            {pipelines.map((pipeline) => (
-              <SelectItem key={pipeline.uuid} value={pipeline.uuid}>
-                {pipeline.name}
-              </SelectItem>
-            ))}
-          </SelectContent>
-        </Select>
+          {t('monitoring.filters.processor')}
+        </Label>
+        <Popover>
+          <PopoverTrigger asChild>
+            <Button
+              id="monitoring-filter-processor"
+              variant="outline"
+              disabled={loadingProcessors}
+              className="h-9 w-full justify-between font-normal"
+            >
+              <span className="truncate">
+                {selectedPipelines.length
+                  ? t('agents.apiToolsSelected', {
+                      count: selectedPipelines.length,
+                    })
+                  : t('monitoring.filters.allProcessors')}
+              </span>
+              <ChevronDown className="ml-2 size-4 shrink-0 text-muted-foreground" />
+            </Button>
+          </PopoverTrigger>
+          <PopoverContent
+            align="start"
+            className="max-h-96 w-72 overflow-y-auto p-2"
+          >
+            <Button
+              variant="ghost"
+              size="sm"
+              className="mb-1 w-full justify-start"
+              onClick={() => onPipelinesChange([])}
+            >
+              {t('monitoring.filters.allProcessors')}
+            </Button>
+            {['pipeline', 'agent', 'event_processor'].map((kind) => {
+              const group = processors.filter((item) => item.kind === kind);
+              if (!group.length) return null;
+              const selected = group.filter((item) =>
+                selectedPipelines.includes(item.uuid),
+              ).length;
+              return (
+                <div key={kind} className="border-t py-2">
+                  <label className="mb-1 flex cursor-pointer items-center gap-2 rounded px-2 py-1.5 text-xs font-medium text-muted-foreground hover:bg-muted">
+                    <Checkbox
+                      checked={
+                        selected === group.length
+                          ? true
+                          : selected
+                            ? 'indeterminate'
+                            : false
+                      }
+                      onCheckedChange={(checked) =>
+                        toggleProcessors(
+                          group.map((item) => item.uuid),
+                          checked === true,
+                        )
+                      }
+                    />
+                    <span className="flex-1">
+                      {t(
+                        `monitoring.execution.kind.${kind === 'event_processor' ? 'processor' : kind}`,
+                      )}
+                    </span>
+                    <span>{t('agents.apiToolsSelectAll')}</span>
+                  </label>
+                  {group.map((item) => (
+                    <label
+                      key={item.uuid}
+                      className="flex cursor-pointer items-center gap-2 rounded px-2 py-2 text-sm hover:bg-muted"
+                    >
+                      <Checkbox
+                        checked={selectedPipelines.includes(item.uuid)}
+                        onCheckedChange={(checked) =>
+                          toggleProcessors([item.uuid], checked === true)
+                        }
+                      />
+                      <span aria-hidden="true">{item.emoji}</span>
+                      <span className="truncate">{item.name}</span>
+                    </label>
+                  ))}
+                </div>
+              );
+            })}
+          </PopoverContent>
+        </Popover>
       </div>
 
+      {/* Mode Filter */}
+      {onModeChange && (
+        <div className="min-w-0 space-y-1.5">
+          <Label
+            htmlFor="monitoring-filter-mode"
+            className="text-xs font-medium text-muted-foreground"
+          >
+            {t('monitoring.execution.filters.mode')}
+          </Label>
+          <Select
+            value={mode ?? 'all'}
+            onValueChange={(value) =>
+              onModeChange(value as ExecutionModeFilter)
+            }
+          >
+            <SelectTrigger
+              id="monitoring-filter-mode"
+              className="h-9 w-full min-w-0"
+            >
+              <SelectValue />
+            </SelectTrigger>
+            <SelectContent>
+              <SelectItem value="all">
+                {t('monitoring.execution.filters.modeAll')}
+              </SelectItem>
+              <SelectItem value="real">
+                {t('monitoring.execution.filters.modeLive')}
+              </SelectItem>
+              <SelectItem value="debug">
+                {t('monitoring.execution.filters.modeDebug')}
+              </SelectItem>
+            </SelectContent>
+          </Select>
+        </div>
+      )}
+
+      {/* Status Filter */}
+      {onStatusGroupChange && (
+        <div className="min-w-0 space-y-1.5">
+          <Label
+            htmlFor="monitoring-filter-status"
+            className="text-xs font-medium text-muted-foreground"
+          >
+            {t('monitoring.execution.filters.status')}
+          </Label>
+          <Select
+            value={statusGroup ?? 'all'}
+            onValueChange={(value) => onStatusGroupChange(value)}
+          >
+            <SelectTrigger
+              id="monitoring-filter-status"
+              className="h-9 w-full min-w-0"
+            >
+              <SelectValue />
+            </SelectTrigger>
+            <SelectContent>
+              <SelectItem value="all">
+                {t('monitoring.execution.filters.statusAll')}
+              </SelectItem>
+              {STATUS_GROUP_OPTIONS.map((group) => (
+                <SelectItem key={group} value={group}>
+                  <span
+                    className={`flex items-center gap-2 ${STATUS_COLORS[group]}`}
+                  >
+                    <span
+                      aria-hidden="true"
+                      className="size-2 shrink-0 rounded-full bg-current"
+                    />
+                    {t(`monitoring.execution.status.${group}`)}
+                  </span>
+                </SelectItem>
+              ))}
+            </SelectContent>
+          </Select>
+        </div>
+      )}
+
       {/* Time Range Filter */}
-      <div className="flex items-center gap-2">
-        <label className="w-20 shrink-0 text-sm font-medium text-foreground sm:w-auto sm:whitespace-nowrap">
+      <div className="min-w-0 space-y-1.5">
+        <Label
+          htmlFor="monitoring-filter-time"
+          className="text-xs font-medium text-muted-foreground"
+        >
           {t('monitoring.filters.timeRange')}
-        </label>
-        <Select value={timeRange} onValueChange={handleTimeRangeChange}>
-          <SelectTrigger className="h-9 w-full sm:w-[150px]">
-            <SelectValue />
+        </Label>
+        <Select
+          value={timeRange === 'custom' ? '' : timeRange}
+          onValueChange={handleTimeRangeChange}
+        >
+          <SelectTrigger
+            id="monitoring-filter-time"
+            className="h-9 w-full min-w-0 gap-2 [&>svg]:shrink-0"
+            title={
+              timeRange === 'custom' && customDateRange
+                ? `${customDateRange.from.toLocaleString()} – ${customDateRange.to.toLocaleString()}`
+                : undefined
+            }
+          >
+            {timeRange === 'custom' && customDateRange ? (
+              <span className="min-w-0 truncate text-xs tabular-nums">
+                {customDateRange.from.toLocaleDateString(undefined, {
+                  month: '2-digit',
+                  day: '2-digit',
+                })}
+                {' – '}
+                {customDateRange.to.toLocaleDateString(undefined, {
+                  month: '2-digit',
+                  day: '2-digit',
+                })}
+              </span>
+            ) : (
+              <SelectValue />
+            )}
           </SelectTrigger>
-          <SelectContent>
+          <SelectContent
+            onCloseAutoFocus={(event) => {
+              if (customPending.current) {
+                event.preventDefault();
+                customPending.current = false;
+                openCustom();
+              }
+            }}
+          >
             <SelectItem value="lastHour">
               {t('monitoring.filters.lastHour')}
             </SelectItem>
@@ -196,11 +473,64 @@ export default function MonitoringFilters({
             <SelectItem value="last7Days">
               {t('monitoring.filters.last7Days')}
             </SelectItem>
-            <SelectItem value="last30Days">
-              {t('monitoring.filters.last30Days')}
-            </SelectItem>
+            {[30, 90, 180, 365].map((days) => (
+              <SelectItem key={days} value={`last${days}Days`}>
+                {t('monitoring.rangeDays', { days })}
+              </SelectItem>
+            ))}
+            {onCustomDateRangeChange && (
+              <SelectItem value="custom">
+                {t('monitoring.customRange')}
+              </SelectItem>
+            )}{' '}
           </SelectContent>
-        </Select>
+        </Select>{' '}
+        {onCustomDateRangeChange && (
+          <>
+            <Dialog open={dateOpen} onOpenChange={setDateOpen}>
+              <DialogContent className="sm:max-w-md space-y-3">
+                <DialogHeader>
+                  <DialogTitle>{t('monitoring.customRange')}</DialogTitle>
+                  <DialogDescription>
+                    {t('monitoring.rangeHint')}
+                  </DialogDescription>
+                </DialogHeader>
+                <label className="block space-y-1 text-xs">
+                  {t('monitoring.rangeStart')}
+                  <Input
+                    type="datetime-local"
+                    value={fromDate}
+                    max={toDate}
+                    onChange={(event) => setFromDate(event.target.value)}
+                  />
+                </label>
+                <label className="block space-y-1 text-xs">
+                  {t('monitoring.rangeEnd')}
+                  <Input
+                    type="datetime-local"
+                    value={toDate}
+                    min={fromDate}
+                    max={localDate(new Date())}
+                    onChange={(event) => setToDate(event.target.value)}
+                  />
+                </label>
+
+                <Button
+                  size="sm"
+                  className="w-full"
+                  disabled={!validDates}
+                  onClick={() => {
+                    onCustomDateRangeChange({ from, to });
+                    onTimeRangeChange('custom');
+                    setDateOpen(false);
+                  }}
+                >
+                  {t('common.confirm')}
+                </Button>
+              </DialogContent>
+            </Dialog>
+          </>
+        )}
       </div>
     </div>
   );
diff --git a/web/src/app/home/monitoring/components/overview-cards/ExecutionOverviewCards.tsx b/web/src/app/home/monitoring/components/overview-cards/ExecutionOverviewCards.tsx
new file mode 100644
index 000000000..ae3406e35
--- /dev/null
+++ b/web/src/app/home/monitoring/components/overview-cards/ExecutionOverviewCards.tsx
@@ -0,0 +1,175 @@
+import { useTranslation } from 'react-i18next';
+import {
+  Activity,
+  CheckCircle2,
+  Timer,
+  TriangleAlert,
+  Users,
+  Zap,
+} from 'lucide-react';
+import { Badge } from '@/components/ui/badge';
+import type { ExecutionSummary } from '@/app/infra/entities/api/monitoring-executions';
+import { formatRunDuration } from '@/app/home/agents/components/processor-run-timing';
+import { MetricCard } from './MetricCard';
+
+interface ExecutionOverviewCardsProps {
+  summary: ExecutionSummary | null;
+  loading?: boolean;
+  /** Active bot conversation sessions from the message rollup endpoint. */
+  activeSessions?: number | null;
+}
+
+function Stat({ label, value }: { label: string; value: string }) {
+  return (
+    <div className="space-y-1">
+      <div className="text-2xl font-bold tabular-nums text-foreground">
+        {value}
+      </div>
+      <div className="text-xs text-muted-foreground">{label}</div>
+    </div>
+  );
+}
+
+export default function ExecutionOverviewCards({
+  summary,
+  loading,
+  activeSessions,
+}: ExecutionOverviewCardsProps) {
+  const { t } = useTranslation();
+  const overview = summary?.executions ?? null;
+  const tokens = summary?.tokens ?? null;
+
+  // Bare cards, not a grid: the page owns the grid so every metric and the
+  // runtime status card land in the same rows.
+  return (
+    <>
+      <MetricCard
+        label={t('monitoring.execution.cards.totalExecutions')}
+        icon={<Activity className="h-4 w-4" />}
+        accent="#8b5cf6"
+        loading={loading}
+        value={(overview?.total ?? 0).toLocaleString()}
+        hint={t('monitoring.execution.cards.bySource', {
+          agent: (overview?.by_source.agent ?? 0).toLocaleString(),
+          pipeline: (overview?.by_source.pipeline ?? 0).toLocaleString(),
+        })}
+      >
+        {!!overview?.debug && (
+          <div className="text-xs text-muted-foreground">
+            {t('monitoring.execution.cards.debugRuns', {
+              count: overview.debug,
+            })}
+          </div>
+        )}
+      </MetricCard>
+
+      <MetricCard
+        label={t('monitoring.execution.cards.successRate')}
+        description={t('monitoring.execution.cards.successRateHint')}
+        icon={<CheckCircle2 className="h-4 w-4" />}
+        accent="#3b82f6"
+        loading={loading}
+        value={
+          overview?.success_rate != null ? `${overview.success_rate}%` : '—'
+        }
+        hint={t('monitoring.execution.cards.denominator', {
+          completed: (overview?.completed ?? 0).toLocaleString(),
+          denominator: (overview?.denominator ?? 0).toLocaleString(),
+        })}
+      />
+
+      <MetricCard
+        label={t('monitoring.execution.cards.running')}
+        icon={<Activity className="h-4 w-4" />}
+        accent="#06b6d4"
+        loading={loading}
+      >
+        <div className="grid grid-cols-3 gap-3">
+          <Stat
+            label={t('monitoring.execution.cards.running')}
+            value={(overview?.running ?? 0).toLocaleString()}
+          />
+          <Stat
+            label={t('monitoring.execution.cards.queued')}
+            value={(overview?.queued ?? 0).toLocaleString()}
+          />
+          <Stat
+            label={t('monitoring.execution.cards.waiting')}
+            value={(overview?.waiting ?? 0).toLocaleString()}
+          />
+        </div>
+      </MetricCard>
+
+      <MetricCard
+        label={t('monitoring.execution.cards.failed')}
+        icon={<TriangleAlert className="h-4 w-4" />}
+        accent="#ef4444"
+        loading={loading}
+        value={
+          <span className="flex items-center gap-3">
+            {(overview?.failed ?? 0).toLocaleString()}
+            <Badge variant="secondary">
+              {t('monitoring.execution.cards.ignored')}{' '}
+              {(overview?.ignored ?? 0).toLocaleString()}
+            </Badge>
+          </span>
+        }
+      />
+
+      <MetricCard
+        label={t('monitoring.execution.cards.latency')}
+        description={t('monitoring.execution.cards.latencyHint')}
+        icon={<Timer className="h-4 w-4" />}
+        accent="#f59e0b"
+        loading={loading}
+      >
+        <div className="grid grid-cols-2 gap-3">
+          <Stat
+            label={t('monitoring.execution.cards.p50')}
+            value={
+              overview?.p50_duration_ms != null
+                ? formatRunDuration(overview.p50_duration_ms)
+                : '—'
+            }
+          />
+          <Stat
+            label={t('monitoring.execution.cards.p95')}
+            value={
+              overview?.p95_duration_ms != null
+                ? formatRunDuration(overview.p95_duration_ms)
+                : '—'
+            }
+          />
+        </div>
+        {/* The sample size explains the percentiles above, so it stays under
+            them rather than in the card's hint slot. */}
+        <div className="text-xs text-muted-foreground">
+          {t('monitoring.execution.cards.sample', {
+            count: overview?.duration_sample ?? 0,
+          })}
+        </div>
+      </MetricCard>
+
+      <MetricCard
+        label={t('monitoring.execution.cards.tokens')}
+        description={t('monitoring.execution.cards.tokensHint')}
+        icon={<Zap className="h-4 w-4" />}
+        accent="#10b981"
+        loading={loading}
+        value={(tokens?.total_tokens ?? 0).toLocaleString()}
+        hint={t('monitoring.execution.cards.coverage', {
+          withUsage: (tokens?.calls_with_usage ?? 0).toLocaleString(),
+          calls: (tokens?.calls ?? 0).toLocaleString(),
+        })}
+      />
+
+      <MetricCard
+        label={t('monitoring.activeSessions')}
+        icon={<Users className="h-4 w-4" />}
+        accent="#6366f1"
+        loading={loading}
+        value={(activeSessions ?? 0).toLocaleString()}
+      />
+    </>
+  );
+}
diff --git a/web/src/app/home/monitoring/components/overview-cards/MetricCard.tsx b/web/src/app/home/monitoring/components/overview-cards/MetricCard.tsx
index 2a3e573de..d70d6b171 100644
--- a/web/src/app/home/monitoring/components/overview-cards/MetricCard.tsx
+++ b/web/src/app/home/monitoring/components/overview-cards/MetricCard.tsx
@@ -1,80 +1,116 @@
-import React from 'react';
-import { TrendingUp, TrendingDown } from 'lucide-react';
-import { Card, CardHeader, CardTitle, CardContent } from '@/components/ui/card';
+'use client';
 
-interface MetricCardProps {
-  title: string;
-  value: string | number;
-  icon: React.ReactNode;
-  trend?: {
-    value: number;
-    direction: 'up' | 'down';
-  };
+import React from 'react';
+import { Card } from '@/components/ui/card';
+import { Skeleton } from '@/components/ui/skeleton';
+import { cn } from '@/lib/utils';
+import { Info } from 'lucide-react';
+import { Button } from '@/components/ui/button';
+import {
+  Tooltip,
+  TooltipContent,
+  TooltipTrigger,
+} from '@/components/ui/tooltip';
+
+export interface MetricCardProps {
+  /** Small label, sitting next to the icon. */
+  label: string;
+  description?: string;
+  /** The card's headline value; a node so it can carry a badge. */
+  value?: React.ReactNode;
+  /** Icon of the tinted tile on the leading edge. */
+  icon?: React.ReactNode;
+  /** Muted line under the value. */
+  hint?: React.ReactNode;
+  /** Tile colour as a hex value; defaults to the muted token. */
+  accent?: string;
+  /** Rendered at the card's top-right (tooltip trigger, help dialog, badge). */
+  action?: React.ReactNode;
+  /** Extra body under the value, for composite cards (stat grids, status rows). */
+  children?: React.ReactNode;
   loading?: boolean;
+  className?: string;
 }
 
-export default function MetricCard({
-  title,
+/**
+ * The dashboard's one metric card: tinted icon tile and label on the leading
+ * edge, value, then a muted hint. Token, execution, feedback and runtime
+ * counters all render through it, so their rows and internal spacing match
+ * instead of each surface inventing its own card.
+ */
+export function MetricCard({
+  label,
+  description,
   value,
   icon,
-  trend,
+  hint,
+  accent,
+  action,
+  children,
   loading,
+  className,
 }: MetricCardProps) {
-  if (loading) {
-    return (
-      <Card className="transition-all duration-300">
-        <CardHeader className="flex flex-row items-center justify-between space-y-0 pb-3">
-          <CardTitle className="text-sm font-medium text-muted-foreground">
-            {title}
-          </CardTitle>
-          <div className="h-10 w-10 rounded-lg bg-gradient-to-br from-blue-50 to-blue-100 dark:from-blue-900/30 dark:to-blue-800/30 flex items-center justify-center">
-            <div className="h-5 w-5 text-blue-600 dark:text-blue-400">
-              {icon}
-            </div>
-          </div>
-        </CardHeader>
-        <CardContent>
-          <div className="h-9 w-28 bg-muted animate-pulse rounded"></div>
-          <div className="h-4 w-20 bg-muted animate-pulse rounded mt-2"></div>
-        </CardContent>
-      </Card>
-    );
-  }
-
   return (
-    <Card className="transition-all duration-300 group">
-      <CardHeader className="flex flex-row items-center justify-between space-y-0 pb-3">
-        <CardTitle className="text-sm font-medium text-muted-foreground">
-          {title}
-        </CardTitle>
-        <div className="h-10 w-10 rounded-lg bg-gradient-to-br from-blue-50 to-blue-100 dark:from-blue-900/30 dark:to-blue-800/30 flex items-center justify-center group-hover:scale-110 transition-transform duration-300">
-          <div className="h-5 w-5 text-blue-600 dark:text-blue-400">{icon}</div>
-        </div>
-      </CardHeader>
-      <CardContent>
-        <div className="text-3xl font-bold text-foreground mb-2">{value}</div>
-        {trend && (
-          <div className="flex items-center gap-1.5">
+    <Card className={cn('gap-3 p-4', className)}>
+      <div className="flex items-start justify-between gap-2">
+        <div className="flex min-w-0 items-center gap-2">
+          {icon ? (
             <span
-              className={`inline-flex items-center gap-1 text-xs font-medium px-2 py-0.5 rounded-full ${
-                trend.direction === 'up'
-                  ? 'bg-green-50 text-green-700 dark:bg-green-900/30 dark:text-green-400'
-                  : 'bg-red-50 text-red-700 dark:bg-red-900/30 dark:text-red-400'
-              }`}
+              className="flex h-7 w-7 shrink-0 items-center justify-center rounded-lg"
+              style={{
+                backgroundColor: accent ? `${accent}1a` : 'var(--muted)',
+                color: accent || 'var(--muted-foreground)',
+              }}
             >
-              {trend.direction === 'up' ? (
-                <TrendingUp className="w-3 h-3" />
-              ) : (
-                <TrendingDown className="w-3 h-3" />
-              )}
-              {Math.abs(trend.value)}%
+              {icon}
             </span>
-            <span className="text-xs text-muted-foreground">
-              vs previous period
-            </span>
-          </div>
-        )}
-      </CardContent>
+          ) : null}
+          <span className="text-sm text-muted-foreground">{label}</span>
+          {description && (
+            <Tooltip>
+              <TooltipTrigger asChild>
+                <Button
+                  type="button"
+                  variant="ghost"
+                  size="icon"
+                  className="h-6 w-6 shrink-0 text-muted-foreground hover:text-foreground"
+                  aria-label={label}
+                >
+                  <Info className="h-3.5 w-3.5" aria-hidden="true" />
+                </Button>
+              </TooltipTrigger>
+              <TooltipContent
+                sideOffset={6}
+                className="max-w-[min(20rem,calc(100vw-2rem))] text-sm leading-relaxed"
+              >
+                {description}
+              </TooltipContent>
+            </Tooltip>
+          )}
+        </div>
+        {action}
+      </div>
+
+      {loading ? (
+        <div className="space-y-2">
+          <Skeleton className="h-8 w-24" />
+          <Skeleton className="h-3 w-20" />
+        </div>
+      ) : (
+        <>
+          {value != null && value !== '' && (
+            <div className="text-2xl font-semibold tabular-nums text-foreground">
+              {value}
+            </div>
+          )}
+          {hint != null && hint !== '' && (
+            <div className="text-xs text-muted-foreground">{hint}</div>
+          )}
+          {children}
+        </>
+      )}
     </Card>
   );
 }
+
+export default MetricCard;
diff --git a/web/src/app/home/monitoring/components/overview-cards/OverviewCards.tsx b/web/src/app/home/monitoring/components/overview-cards/OverviewCards.tsx
deleted file mode 100644
index cb1e9c1d8..000000000
--- a/web/src/app/home/monitoring/components/overview-cards/OverviewCards.tsx
+++ /dev/null
@@ -1,100 +0,0 @@
-import React from 'react';
-import { useTranslation } from 'react-i18next';
-import { MessageSquare, Sparkles, Check, Users } from 'lucide-react';
-import MetricCard from './MetricCard';
-import SystemStatusCard from './SystemStatusCards';
-import TrafficChart from './TrafficChart';
-import { OverviewMetrics, MonitoringData } from '../../types/monitoring';
-
-interface OverviewCardsProps {
-  metrics: OverviewMetrics | null;
-  traffic?: MonitoringData['traffic'];
-  loading?: boolean;
-  refreshKey?: number;
-}
-
-export default function OverviewCards({
-  metrics,
-  traffic,
-  loading,
-  refreshKey,
-}: OverviewCardsProps) {
-  const { t } = useTranslation();
-
-  const cards = [
-    {
-      title: t('monitoring.totalMessages'),
-      value: metrics?.totalMessages || 0,
-      icon: <MessageSquare />,
-      trend: metrics?.trends
-        ? {
-            value: metrics.trends.messages,
-            direction: (metrics.trends.messages >= 0 ? 'up' : 'down') as
-              | 'up'
-              | 'down',
-          }
-        : undefined,
-    },
-    {
-      title: t('monitoring.modelCallsCount'),
-      value: metrics?.modelCalls || 0,
-      icon: <Sparkles />,
-      trend: metrics?.trends
-        ? {
-            value: metrics.trends.llmCalls,
-            direction: (metrics.trends.llmCalls >= 0 ? 'up' : 'down') as
-              | 'up'
-              | 'down',
-          }
-        : undefined,
-    },
-    {
-      title: t('monitoring.successRate'),
-      value: metrics ? `${metrics.successRate}%` : '0%',
-      icon: <Check />,
-      trend: metrics?.trends
-        ? {
-            value: metrics.trends.successRate,
-            direction: (metrics.trends.successRate >= 0 ? 'up' : 'down') as
-              | 'up'
-              | 'down',
-          }
-        : undefined,
-    },
-    {
-      title: t('monitoring.activeSessions'),
-      value: metrics?.activeSessions || 0,
-      icon: <Users />,
-      trend: metrics?.trends
-        ? {
-            value: metrics.trends.sessions,
-            direction: (metrics.trends.sessions >= 0 ? 'up' : 'down') as
-              | 'up'
-              | 'down',
-          }
-        : undefined,
-    },
-  ];
-
-  return (
-    <div className="space-y-6">
-      {/* Metric Cards + System Status */}
-      <div className="grid grid-cols-1 md:grid-cols-2 xl:grid-cols-5 gap-6">
-        {cards.map((card, index) => (
-          <MetricCard
-            key={index}
-            title={card.title}
-            value={card.value}
-            icon={card.icon}
-            trend={card.trend}
-            loading={loading}
-          />
-        ))}
-        <SystemStatusCard refreshKey={refreshKey} />
-      </div>
-
-      {/* Traffic Chart */}
-      <TrafficChart traffic={traffic} loading={loading} />
-    </div>
-  );
-}
diff --git a/web/src/app/home/monitoring/components/overview-cards/SystemStatusCards.tsx b/web/src/app/home/monitoring/components/overview-cards/SystemStatusCards.tsx
index ca49c6d05..f8bc9f138 100644
--- a/web/src/app/home/monitoring/components/overview-cards/SystemStatusCards.tsx
+++ b/web/src/app/home/monitoring/components/overview-cards/SystemStatusCards.tsx
@@ -5,7 +5,7 @@ import {
   Box,
   CircleCheck,
   CircleX,
-  Loader2,
+  Server,
   Info,
   Container,
   Clock,
@@ -15,7 +15,7 @@ import {
   Image,
   FolderOpen,
 } from 'lucide-react';
-import { Card, CardHeader, CardTitle, CardContent } from '@/components/ui/card';
+import { MetricCard } from './MetricCard';
 import { Button } from '@/components/ui/button';
 import {
   Dialog,
@@ -38,6 +38,10 @@ import { httpClient } from '@/app/infra/http/HttpClient';
 
 type StatusState = 'ok' | 'disabled' | 'failed' | null;
 
+// Neutral tile: the status dots already carry the health colour, so the card's
+// own tile must not encode a second, conflicting verdict.
+const SYSTEM_STATUS_ACCENT = '#64748b';
+
 function StatusDot({ state }: { state: StatusState }) {
   if (state === null)
     return <span className="w-2 h-2 rounded-full bg-muted-foreground/40" />;
@@ -118,28 +122,22 @@ export default function SystemStatusCard({
 
   if (loading) {
     return (
-      <Card className="transition-all duration-300">
-        <CardHeader className="flex flex-row items-center justify-between space-y-0 pb-3">
-          <CardTitle className="text-sm font-medium text-muted-foreground">
-            {t('monitoring.systemStatus')}
-          </CardTitle>
-        </CardHeader>
-        <CardContent>
-          <div className="flex items-center gap-2">
-            <Loader2 className="w-4 h-4 animate-spin text-muted-foreground" />
-          </div>
-        </CardContent>
-      </Card>
+      <MetricCard
+        label={t('monitoring.systemStatus')}
+        icon={<Server className="h-4 w-4" />}
+        accent={SYSTEM_STATUS_ACCENT}
+        loading
+      />
     );
   }
 
   return (
     <>
-      <Card className="transition-all duration-300 group">
-        <CardHeader className="flex flex-row items-center justify-between space-y-0 pb-3">
-          <CardTitle className="text-sm font-medium text-muted-foreground">
-            {t('monitoring.systemStatus')}
-          </CardTitle>
+      <MetricCard
+        label={t('monitoring.systemStatus')}
+        icon={<Server className="h-4 w-4" />}
+        accent={SYSTEM_STATUS_ACCENT}
+        action={
           <Button
             variant="ghost"
             size="icon"
@@ -148,8 +146,9 @@ export default function SystemStatusCard({
           >
             <Info className="w-4 h-4" />
           </Button>
-        </CardHeader>
-        <CardContent className="space-y-2">
+        }
+      >
+        <div className="space-y-2">
           <div className="flex items-center gap-2">
             <StatusDot state={pluginState} />
             <Plug className="w-3.5 h-3.5 text-muted-foreground" />
@@ -162,8 +161,8 @@ export default function SystemStatusCard({
               <span className="text-sm">{t('monitoring.boxRuntime')}</span>
             </div>
           )}
-        </CardContent>
-      </Card>
+        </div>
+      </MetricCard>
 
       <Dialog open={dialogOpen} onOpenChange={setDialogOpen}>
         <DialogContent className="max-w-2xl max-h-[80vh] overflow-hidden flex flex-col">
diff --git a/web/src/app/home/monitoring/components/overview-cards/TrafficChart.tsx b/web/src/app/home/monitoring/components/overview-cards/TrafficChart.tsx
index fd825299a..73be8c5f4 100644
--- a/web/src/app/home/monitoring/components/overview-cards/TrafficChart.tsx
+++ b/web/src/app/home/monitoring/components/overview-cards/TrafficChart.tsx
@@ -1,6 +1,6 @@
-import React, { useMemo } from 'react';
+import React, { useMemo, useState } from 'react';
 import { useTranslation } from 'react-i18next';
-import { BarChart3 } from 'lucide-react';
+import { BarChart3, ChevronDown } from 'lucide-react';
 import {
   AreaChart,
   Area,
@@ -9,8 +9,14 @@ import {
   CartesianGrid,
   Tooltip,
   ResponsiveContainer,
-  Legend,
 } from 'recharts';
+import { Card } from '@/components/ui/card';
+import {
+  Collapsible,
+  CollapsibleContent,
+  CollapsibleTrigger,
+} from '@/components/ui/collapsible';
+import { Skeleton } from '@/components/ui/skeleton';
 import { MonitoringData } from '../../types/monitoring';
 
 interface TrafficChartProps {
@@ -18,7 +24,66 @@ interface TrafficChartProps {
   loading?: boolean;
 }
 
-export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
+export default function TrafficChart(props: TrafficChartProps) {
+  const { t } = useTranslation();
+  const [open, setOpen] = useState(() => {
+    try {
+      return localStorage.getItem('langbot-traffic-chart-open') !== 'false';
+    } catch {
+      return true;
+    }
+  });
+  return (
+    <Collapsible
+      open={open}
+      onOpenChange={(value) => {
+        setOpen(value);
+        try {
+          localStorage.setItem('langbot-traffic-chart-open', String(value));
+        } catch {
+          /* Storage may be unavailable. */
+        }
+      }}
+      asChild
+    >
+      <Card className="gap-0 overflow-hidden py-0">
+        <CollapsibleTrigger asChild>
+          <button
+            type="button"
+            className="group flex w-full items-center gap-3 px-5 py-4 text-left transition-colors hover:bg-muted/40 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-inset focus-visible:ring-ring"
+          >
+            <span className="flex size-9 shrink-0 items-center justify-center rounded-lg bg-blue-500/10 text-blue-600 dark:text-blue-400">
+              <BarChart3 className="size-4" />
+            </span>
+            <span className="flex-1 text-sm font-semibold">
+              {t('monitoring.trafficChart.title')}
+            </span>
+            <span className="hidden items-center gap-4 text-xs text-muted-foreground sm:flex">
+              <span className="flex items-center gap-1.5">
+                <span className="size-1.5 rounded-full bg-blue-500" />
+                {t('monitoring.trafficChart.messages')}
+              </span>
+              <span className="flex items-center gap-1.5">
+                <span className="size-1.5 rounded-full bg-violet-500" />
+                {t('monitoring.trafficChart.llmCalls')}
+              </span>
+            </span>
+            <ChevronDown
+              className={`ml-2 size-4 text-muted-foreground transition-transform motion-reduce:transition-none ${open ? 'rotate-180' : ''}`}
+            />
+          </button>
+        </CollapsibleTrigger>
+        <CollapsibleContent>
+          <div className="border-t px-3 pb-4 pt-4 sm:px-5">
+            <TrafficChartContent {...props} />
+          </div>
+        </CollapsibleContent>
+      </Card>
+    </Collapsible>
+  );
+}
+
+function TrafficChartContent({ traffic, loading }: TrafficChartProps) {
   const { t } = useTranslation();
   const chartData = useMemo(
     () =>
@@ -39,18 +104,11 @@ export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
     [traffic],
   );
 
-  if (loading) {
+  if (loading && !traffic) {
     return (
-      <div className="bg-card rounded-xl border p-6">
-        <div className="flex items-center justify-between mb-4">
-          <div className="h-5 w-32 bg-muted animate-pulse rounded"></div>
-          <div className="flex gap-4">
-            <div className="h-4 w-24 bg-muted animate-pulse rounded"></div>
-            <div className="h-4 w-24 bg-muted animate-pulse rounded"></div>
-          </div>
-        </div>
+      <div>
         <div className="h-[300px] flex items-center justify-center">
-          <div className="animate-pulse w-full h-full bg-muted rounded"></div>
+          <Skeleton className="w-full h-full rounded" />
         </div>
       </div>
     );
@@ -58,10 +116,7 @@ export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
 
   if (chartData.length === 0) {
     return (
-      <div className="bg-card rounded-xl border p-6">
-        <h3 className="text-base font-semibold text-foreground mb-4">
-          {t('monitoring.trafficChart.title')}
-        </h3>
+      <div>
         <div className="h-[300px] flex flex-col items-center justify-center text-muted-foreground gap-2">
           <BarChart3 className="h-[3rem] w-[3rem]" />
           <div className="text-sm">
@@ -77,10 +132,7 @@ export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
   }
 
   return (
-    <div className="bg-card rounded-xl border p-6 transition-shadow duration-300">
-      <h3 className="text-base font-semibold text-foreground mb-6">
-        {t('monitoring.trafficChart.title')}
-      </h3>
+    <div>
       {traffic?.truncated && (
         <p role="status" className="text-sm text-muted-foreground mb-3">
           {t('monitoring.trafficChart.truncated')}
@@ -139,16 +191,8 @@ export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
               }}
               itemStyle={{ padding: '4px 0' }}
             />
-            <Legend
-              wrapperStyle={{
-                fontSize: '13px',
-                paddingTop: '16px',
-                fontWeight: 500,
-              }}
-              iconType="circle"
-              iconSize={10}
-            />
             <Area
+              isAnimationActive={false}
               type="monotone"
               dataKey="messages"
               name={t('monitoring.trafficChart.messages')}
@@ -160,6 +204,7 @@ export default function TrafficChart({ traffic, loading }: TrafficChartProps) {
               activeDot={{ r: 6, strokeWidth: 2 }}
             />
             <Area
+              isAnimationActive={false}
               type="monotone"
               dataKey="llmCalls"
               name={t('monitoring.trafficChart.llmCalls')}
diff --git a/web/src/app/home/monitoring/hooks/useExecutions.ts b/web/src/app/home/monitoring/hooks/useExecutions.ts
new file mode 100644
index 000000000..79d0eb65f
--- /dev/null
+++ b/web/src/app/home/monitoring/hooks/useExecutions.ts
@@ -0,0 +1,152 @@
+import { useState, useEffect, useCallback, useMemo, useRef } from 'react';
+import { TimeRangeOption, DateRange } from '../types/monitoring';
+import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
+import { getCurrentWorkspaceSnapshot } from '@/app/infra/http/currentWorkspaceStore';
+import { resolveMonitoringWindow } from '../utils/dateUtils';
+import type {
+  ExecutionListResult,
+  ExecutionModeFilter,
+  ExecutionSourceFilter,
+} from '@/app/infra/entities/api/monitoring-executions';
+
+/**
+ * Custom hook for fetching the unified execution list
+ */
+export function useExecutions(params: {
+  selectedBots: string[];
+  selectedPipelines: string[];
+  selectedAgents: string[];
+  source: ExecutionSourceFilter;
+  mode: ExecutionModeFilter;
+  statusGroup: string;
+  timeRange: TimeRangeOption;
+  customDateRange: DateRange | null;
+  limit: number;
+  offset: number;
+  refreshKey?: number;
+}) {
+  const [result, setResult] = useState<ExecutionListResult | null>(null);
+  const [loading, setLoading] = useState(false);
+  const [error, setError] = useState<Error | null>(null);
+  const [nonce, setNonce] = useState(0);
+  const workspaceUuid = useCurrentWorkspace()?.workspace.uuid;
+  const requestIdRef = useRef(0);
+  const scope = JSON.stringify([
+    workspaceUuid,
+    { ...params, refreshKey: undefined },
+  ]);
+  const [requestScope, setRequestScope] = useState<string | null>(null);
+
+  // Memoize filter parameters to prevent unnecessary re-renders
+  const selectedBotsStr = useMemo(
+    () => JSON.stringify(params.selectedBots),
+    [params.selectedBots],
+  );
+  const selectedPipelinesStr = useMemo(
+    () => JSON.stringify(params.selectedPipelines),
+    [params.selectedPipelines],
+  );
+  const selectedAgentsStr = useMemo(
+    () => JSON.stringify(params.selectedAgents),
+    [params.selectedAgents],
+  );
+  const customDateRangeStr = useMemo(
+    () => JSON.stringify(params.customDateRange),
+    [params.customDateRange],
+  );
+
+  // Fetch data based on filters
+  const fetchData = useCallback(async () => {
+    const requestId = ++requestIdRef.current;
+    const isCurrent = () =>
+      requestId === requestIdRef.current &&
+      getCurrentWorkspaceSnapshot()?.workspace.uuid === workspaceUuid;
+    setLoading(true);
+    setError(null);
+
+    try {
+      const { startTime, endTime } = resolveMonitoringWindow(
+        params.timeRange,
+        params.customDateRange,
+      );
+
+      const response = await backendClient.getExecutions({
+        botId: params.selectedBots.length > 0 ? params.selectedBots : undefined,
+        pipelineId:
+          params.selectedPipelines.length > 0
+            ? params.selectedPipelines
+            : undefined,
+        agentId:
+          params.selectedAgents.length > 0 ? params.selectedAgents : undefined,
+        status: params.statusGroup === 'all' ? undefined : [params.statusGroup],
+        source: params.source,
+        mode: params.mode,
+        startTime,
+        endTime,
+        limit: params.limit,
+        offset: params.offset,
+      });
+      if (!isCurrent()) return;
+
+      setResult(response);
+      setRequestScope(scope);
+    } catch (err) {
+      if (!isCurrent()) return;
+      setRequestScope(scope);
+      setResult((current) => (requestScope === scope ? current : null));
+      setError(err as Error);
+      console.error('Failed to fetch executions:', err);
+    } finally {
+      if (isCurrent()) setLoading(false);
+    }
+  }, [
+    params.selectedBots,
+    params.selectedPipelines,
+    params.selectedAgents,
+    params.source,
+    params.mode,
+    params.statusGroup,
+    params.timeRange,
+    params.customDateRange,
+    params.limit,
+    params.offset,
+    scope,
+    requestScope,
+    workspaceUuid,
+  ]);
+
+  // Fetch data when filter state changes
+  useEffect(() => {
+    fetchData();
+    return () => {
+      requestIdRef.current += 1;
+    };
+    // eslint-disable-next-line react-hooks/exhaustive-deps
+  }, [
+    selectedBotsStr,
+    selectedPipelinesStr,
+    selectedAgentsStr,
+    params.source,
+    params.mode,
+    params.statusGroup,
+    params.timeRange,
+    customDateRangeStr,
+    params.limit,
+    params.offset,
+    params.refreshKey,
+    nonce,
+    workspaceUuid,
+  ]);
+
+  // Manual refetch function
+  const refetch = () => {
+    setNonce((value) => value + 1);
+  };
+
+  return {
+    result: requestScope === scope ? result : null,
+    loading: requestScope !== scope || loading,
+    error: requestScope === scope ? error : null,
+    refetch,
+  };
+}
diff --git a/web/src/app/home/monitoring/hooks/useFeedbackData.ts b/web/src/app/home/monitoring/hooks/useFeedbackData.ts
index 9948bee89..cce559375 100644
--- a/web/src/app/home/monitoring/hooks/useFeedbackData.ts
+++ b/web/src/app/home/monitoring/hooks/useFeedbackData.ts
@@ -4,6 +4,8 @@ import { FeedbackRecord, FeedbackStats } from '../types/monitoring';
 import { parseUTCTimestamp } from '../utils/dateUtils';
 
 interface UseFeedbackDataParams {
+  mode?: 'all' | 'real' | 'debug';
+  statusGroup?: string;
   botIds?: string[];
   pipelineIds?: string[];
   startTime?: string;
@@ -60,6 +62,9 @@ export function useFeedbackData(params: UseFeedbackDataParams = {}) {
   const fetchStats = useCallback(async () => {
     try {
       const queryParams = new URLSearchParams();
+      if (params.mode) queryParams.append('mode', params.mode);
+      if (params.statusGroup && params.statusGroup !== 'all')
+        queryParams.append('status', params.statusGroup);
       if (params.botIds) {
         params.botIds.forEach((id) => queryParams.append('botId', id));
       }
@@ -99,7 +104,14 @@ export function useFeedbackData(params: UseFeedbackDataParams = {}) {
     } catch (err) {
       console.error('Failed to fetch feedback stats:', err);
     }
-  }, [params.botIds, params.pipelineIds, params.startTime, params.endTime]);
+  }, [
+    params.mode,
+    params.statusGroup,
+    params.botIds,
+    params.pipelineIds,
+    params.startTime,
+    params.endTime,
+  ]);
 
   const fetchFeedback = useCallback(async () => {
     setLoading(true);
@@ -107,6 +119,9 @@ export function useFeedbackData(params: UseFeedbackDataParams = {}) {
 
     try {
       const queryParams = new URLSearchParams();
+      if (params.mode) queryParams.append('mode', params.mode);
+      if (params.statusGroup && params.statusGroup !== 'all')
+        queryParams.append('status', params.statusGroup);
       if (params.botIds) {
         params.botIds.forEach((id) => queryParams.append('botId', id));
       }
diff --git a/web/src/app/home/monitoring/hooks/useMonitoringData.ts b/web/src/app/home/monitoring/hooks/useMonitoringData.ts
index a817c9b19..ee8610bfd 100644
--- a/web/src/app/home/monitoring/hooks/useMonitoringData.ts
+++ b/web/src/app/home/monitoring/hooks/useMonitoringData.ts
@@ -8,7 +8,7 @@ import {
 } from '../types/monitoring';
 import { backendClient, useCurrentWorkspace } from '@/app/infra/http';
 import { getCurrentWorkspaceSnapshot } from '@/app/infra/http/currentWorkspaceStore';
-import { parseUTCTimestamp } from '../utils/dateUtils';
+import { parseUTCTimestamp, resolveMonitoringWindow } from '../utils/dateUtils';
 
 /**
  * Custom hook for fetching and managing monitoring data
@@ -36,60 +36,28 @@ export function useMonitoringData(filterState: FilterState) {
     [filterState.customDateRange],
   );
 
-  // Convert time range to datetime strings
-  const getTimeRange = useCallback(() => {
-    const now = new Date();
-    let startTime: Date | null = null;
-
-    switch (filterState.timeRange) {
-      case 'lastHour':
-        startTime = new Date(now.getTime() - 60 * 60 * 1000);
-        break;
-      case 'last6Hours':
-        startTime = new Date(now.getTime() - 6 * 60 * 60 * 1000);
-        break;
-      case 'last24Hours':
-        startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
-        break;
-      case 'last7Days':
-        startTime = new Date(now.getTime() - 7 * 24 * 60 * 60 * 1000);
-        break;
-      case 'last30Days':
-        startTime = new Date(now.getTime() - 30 * 24 * 60 * 60 * 1000);
-        break;
-      case 'custom':
-        if (filterState.customDateRange) {
-          startTime = filterState.customDateRange.from;
-        }
-        break;
-    }
-
-    const endTime =
-      filterState.timeRange === 'custom' && filterState.customDateRange
-        ? filterState.customDateRange.to
-        : now;
-
-    return {
-      startTime: startTime?.toISOString(),
-      endTime: endTime.toISOString(),
-    };
-  }, [filterState.timeRange, filterState.customDateRange]);
-
   // Fetch data based on filters
   const fetchData = useCallback(async () => {
     const requestId = ++requestIdRef.current;
     const isCurrent = () =>
       requestId === requestIdRef.current &&
       getCurrentWorkspaceSnapshot()?.workspace.uuid === workspaceUuid;
-    setRequestScope(scope);
-    setData(null);
     setLoading(true);
-    setError(null);
+    // Keep the failure view during retries; restore the dashboard only after
+    // a successful response, without briefly restarting its polling timer.
 
     try {
-      const { startTime, endTime } = getTimeRange();
+      const { startTime, endTime } = resolveMonitoringWindow(
+        filterState.timeRange,
+        filterState.customDateRange,
+      );
 
       const response = await backendClient.getMonitoringData({
+        mode: filterState.mode,
+        status:
+          filterState.statusGroup && filterState.statusGroup !== 'all'
+            ? [filterState.statusGroup]
+            : undefined,
         botId:
           filterState.selectedBots.length > 0
             ? filterState.selectedBots
@@ -418,18 +386,26 @@ export function useMonitoringData(filterState: FilterState) {
       ].sort((a, b) => b.timestamp.getTime() - a.timestamp.getTime());
 
       setData(transformedData);
+      setError(null);
+      setRequestScope(scope);
     } catch (err) {
       if (!isCurrent()) return;
+      setRequestScope(scope);
+      setData((current) => (requestScope === scope ? current : null));
       setError(err as Error);
       console.error('Failed to fetch monitoring data:', err);
     } finally {
       if (isCurrent()) setLoading(false);
     }
   }, [
-    getTimeRange,
+    filterState.mode,
+    filterState.statusGroup,
+    filterState.timeRange,
+    filterState.customDateRange,
     filterState.selectedBots,
     filterState.selectedPipelines,
     scope,
+    requestScope,
     workspaceUuid,
   ]);
 
@@ -443,6 +419,8 @@ export function useMonitoringData(filterState: FilterState) {
   }, [
     selectedBotsStr,
     selectedPipelinesStr,
+    filterState.mode,
+    filterState.statusGroup,
     filterState.timeRange,
     customDateRangeStr,
     workspaceUuid,
diff --git a/web/src/app/home/monitoring/hooks/useMonitoringFilters.ts b/web/src/app/home/monitoring/hooks/useMonitoringFilters.ts
index 53545bab0..f8a39651d 100644
--- a/web/src/app/home/monitoring/hooks/useMonitoringFilters.ts
+++ b/web/src/app/home/monitoring/hooks/useMonitoringFilters.ts
@@ -1,65 +1,132 @@
-import { useState } from 'react';
+import { useEffect, useState } from 'react';
 import { useSearchParams } from 'react-router-dom';
+import { useCurrentWorkspace } from '@/app/infra/http';
 import { FilterState, TimeRangeOption, DateRange } from '../types/monitoring';
 import { getPresetDateRange } from '../utils/dateUtils';
+import type { ExecutionModeFilter } from '@/app/infra/entities/api/monitoring-executions';
 
-/**
- * Custom hook for managing monitoring filters
- */
-export function useMonitoringFilters() {
-  const [searchParams] = useSearchParams();
+const defaults = (): FilterState => ({
+  selectedBots: [],
+  selectedPipelines: [],
+  timeRange: 'last24Hours',
+  customDateRange: null,
+  mode: 'all',
+  statusGroup: 'all',
+});
+const timeRanges = [
+  'lastHour',
+  'last6Hours',
+  'last24Hours',
+  'last7Days',
+  'last30Days',
+  'last90Days',
+  'last180Days',
+  'last365Days',
+  'custom',
+];
+const statuses = [
+  'all',
+  'completed',
+  'failed',
+  'running',
+  'queued',
+  'cancelled',
+  'ignored',
+];
 
-  // Initialize filters from URL params
-  const [selectedBots, setSelectedBots] = useState<string[]>(() => {
-    const botId = searchParams.get('botId');
-    return botId ? [botId] : [];
-  });
-
-  const [selectedPipelines, setSelectedPipelines] = useState<string[]>(() => {
-    const pipelineId = searchParams.get('pipelineId');
-    return pipelineId ? [pipelineId] : [];
-  });
-
-  const [timeRange, setTimeRange] = useState<TimeRangeOption>('last24Hours');
-  const [customDateRange, setCustomDateRange] = useState<DateRange | null>(
-    null,
-  );
-
-  // Get the active date range (either preset or custom)
-  const getActiveDateRange = (): DateRange | null => {
-    if (timeRange === 'custom' && customDateRange) {
-      return customDateRange;
+function restore(key: string): FilterState {
+  const state = defaults();
+  try {
+    const saved = JSON.parse(localStorage.getItem(key) || 'null');
+    if (!saved || typeof saved !== 'object') return state;
+    for (const name of ['selectedBots', 'selectedPipelines'] as const) {
+      if (Array.isArray(saved[name]))
+        state[name] = saved[name].filter(
+          (id: unknown) => typeof id === 'string' && id.length > 0,
+        );
     }
-    return getPresetDateRange(timeRange);
-  };
+    if (timeRanges.includes(saved.timeRange)) state.timeRange = saved.timeRange;
+    if (['all', 'real', 'debug'].includes(saved.mode)) state.mode = saved.mode;
+    if (statuses.includes(saved.statusGroup))
+      state.statusGroup = saved.statusGroup;
+    if (saved.customDateRange) {
+      const from = new Date(saved.customDateRange.from);
+      const to = new Date(saved.customDateRange.to);
+      if (
+        Number.isFinite(from.getTime()) &&
+        Number.isFinite(to.getTime()) &&
+        from < to &&
+        to.getTime() - from.getTime() <= 365 * 86400000
+      )
+        state.customDateRange = { from, to };
+    }
+    if (state.timeRange === 'custom' && !state.customDateRange)
+      state.timeRange = 'last24Hours';
+  } catch {
+    // Invalid or unavailable browser storage must not prevent monitoring.
+  }
+  return state;
+}
 
-  // Reset all filters
+/** The page is keyed by Workspace so saved resource IDs never cross scopes. */
+export function useMonitoringFilters() {
+  const workspace = useCurrentWorkspace()?.workspace.uuid;
+  const key = `monitoring.filters.v1:${workspace ?? 'pending'}`;
+  const [searchParams, setSearchParams] = useSearchParams();
+  const [filterState, setFilters] = useState<FilterState>(() => {
+    const state = workspace ? restore(key) : defaults();
+    // Explicit links take precedence over saved resource filters.
+    if (searchParams.has('botId'))
+      state.selectedBots = searchParams.getAll('botId').filter(Boolean);
+    if (searchParams.has('pipelineId'))
+      state.selectedPipelines = searchParams
+        .getAll('pipelineId')
+        .filter(Boolean);
+    return state;
+  });
+  useEffect(() => {
+    if (!workspace) return;
+    try {
+      localStorage.setItem(key, JSON.stringify(filterState));
+    } catch {
+      /* Storage may be disabled. */
+    }
+  }, [key, workspace, filterState]);
+  const update = (value: Partial<FilterState>) =>
+    setFilters((current) => ({ ...current, ...value }));
+  const clearLinkedFilter = (name: string) => {
+    if (!searchParams.has(name)) return;
+    const next = new URLSearchParams(searchParams);
+    next.delete(name);
+    setSearchParams(next, { replace: true });
+  };
   const resetFilters = () => {
-    setSelectedBots([]);
-    setSelectedPipelines([]);
-    setTimeRange('last24Hours');
-    setCustomDateRange(null);
+    setFilters(defaults());
+    const next = new URLSearchParams(searchParams);
+    next.delete('botId');
+    next.delete('pipelineId');
+    setSearchParams(next, { replace: true });
   };
-
-  // Get the current filter state
-  const filterState: FilterState = {
-    selectedBots,
-    selectedPipelines,
-    timeRange,
-    customDateRange,
-  };
-
   return {
-    selectedBots,
-    setSelectedBots,
-    selectedPipelines,
-    setSelectedPipelines,
-    timeRange,
-    setTimeRange,
-    customDateRange,
-    setCustomDateRange,
-    getActiveDateRange,
-    resetFilters,
+    ...filterState,
     filterState,
+    setSelectedBots: (selectedBots: string[]) => {
+      update({ selectedBots });
+      clearLinkedFilter('botId');
+    },
+    setSelectedPipelines: (selectedPipelines: string[]) => {
+      update({ selectedPipelines });
+      clearLinkedFilter('pipelineId');
+    },
+    setTimeRange: (timeRange: TimeRangeOption) => update({ timeRange }),
+    setCustomDateRange: (customDateRange: DateRange | null) =>
+      update({ customDateRange }),
+    setExecutionMode: (mode: ExecutionModeFilter) => update({ mode }),
+    setExecutionStatus: (statusGroup: string) => update({ statusGroup }),
+    getActiveDateRange: () =>
+      filterState.timeRange === 'custom'
+        ? filterState.customDateRange
+        : getPresetDateRange(filterState.timeRange),
+    resetFilters,
   };
 }
diff --git a/web/src/app/home/monitoring/page.tsx b/web/src/app/home/monitoring/page.tsx
index a28d8978d..2547883a8 100644
--- a/web/src/app/home/monitoring/page.tsx
+++ b/web/src/app/home/monitoring/page.tsx
@@ -1,28 +1,53 @@
-import React, { Suspense, useState, useMemo, useCallback } from 'react';
+import React, {
+  Suspense,
+  useState,
+  useEffect,
+  useMemo,
+  useCallback,
+  useRef,
+} from 'react';
 import { useTranslation } from 'react-i18next';
 import { Tabs, TabsContent, TabsList, TabsTrigger } from '@/components/ui/tabs';
 import { Button } from '@/components/ui/button';
+import {
+  Select,
+  SelectContent,
+  SelectItem,
+  SelectTrigger,
+} from '@/components/ui/select';
+import LoadErrorState from '@/components/LoadErrorState';
+import { Card } from '@/components/ui/card';
+import { Badge } from '@/components/ui/badge';
 import {
   ChevronRight,
   ChevronDown,
   ExternalLink,
   RefreshCw,
-  MessageSquare,
+  RotateCcw,
   Sparkles,
   CheckCircle2,
+  Activity,
 } from 'lucide-react';
-import OverviewCards from './components/overview-cards/OverviewCards';
+import { TabState } from './components/TabState';
+import SystemStatusCard from './components/overview-cards/SystemStatusCards';
+import TrafficChart from './components/overview-cards/TrafficChart';
 import MonitoringFilters from './components/filters/MonitoringFilters';
 import TokenMonitoring from './components/TokenMonitoring';
 import { ExportDropdown } from './components/ExportDropdown';
 import { useMonitoringFilters } from './hooks/useMonitoringFilters';
 import { useMonitoringData } from './hooks/useMonitoringData';
 import { useFeedbackData } from './hooks/useFeedbackData';
-import { ConversationTurnList } from './components/ConversationTurnList';
 import { FeedbackStatsCards } from './components/FeedbackCard';
 import { FeedbackList } from './components/FeedbackList';
-import { buildConversationTurns } from './utils/conversationTurns';
-import { LoadingSpinner, LoadingPage } from '@/components/ui/loading-spinner';
+import ExecutionOverviewCards from './components/overview-cards/ExecutionOverviewCards';
+import ExecutionTable from './components/executions/ExecutionTable';
+import ExecutionDetailSheet, {
+  type ExecutionSelection,
+} from './components/executions/ExecutionDetailSheet';
+import { useExecutions } from './hooks/useExecutions';
+import { cn } from '@/lib/utils';
+import { resolveMonitoringWindow } from './utils/dateUtils';
+import { LoadingPage } from '@/components/ui/loading-spinner';
 import { useCurrentWorkspace } from '@/app/infra/http';
 
 function MonitoringPageContent() {
@@ -30,59 +55,85 @@ function MonitoringPageContent() {
   const currentWorkspace = useCurrentWorkspace();
   const canExport =
     currentWorkspace?.permissions.includes('data.export') ?? false;
-  const { filterState, setSelectedBots, setSelectedPipelines, setTimeRange } =
-    useMonitoringFilters();
-  const { data, loading, error, refetch } = useMonitoringData(filterState);
+  const {
+    filterState,
+    setSelectedBots,
+    setSelectedPipelines,
+    setTimeRange,
+    setCustomDateRange,
+    setExecutionMode,
+    setExecutionStatus,
+    resetFilters,
+  } = useMonitoringFilters();
+  const executionMode = filterState.mode ?? 'all';
+  const executionStatus = filterState.statusGroup ?? 'all';
+  const {
+    data,
+    loading: monitoringRefreshing,
+    error,
+    refetch,
+  } = useMonitoringData({
+    ...filterState,
+    mode: executionMode,
+    statusGroup: executionStatus,
+  });
+  const [executionOffset, setExecutionOffset] = useState(0);
+  const [selectedExecution, setSelectedExecution] =
+    useState<ExecutionSelection | null>(null);
+  const EXECUTIONS_PAGE_SIZE = 50;
+
+  const {
+    result: executionResult,
+    loading: executionRefreshing,
+    error: executionError,
+    refetch: refetchExecutions,
+  } = useExecutions({
+    selectedBots: filterState.selectedBots,
+    selectedPipelines: filterState.selectedPipelines,
+    selectedAgents: [],
+    source: 'all',
+    mode: executionMode,
+    statusGroup: executionStatus,
+    timeRange: filterState.timeRange,
+    customDateRange: filterState.customDateRange,
+    limit: EXECUTIONS_PAGE_SIZE,
+    offset: executionOffset,
+  });
+
+  // Reset paging whenever the execution filters narrow the result set.
+  useEffect(() => {
+    setExecutionOffset(0);
+  }, [
+    executionMode,
+    executionStatus,
+    filterState.selectedBots,
+    filterState.selectedPipelines,
+    filterState.timeRange,
+    filterState.customDateRange,
+  ]);
 
   // Counter to force feedbackTimeRange recomputation on manual refresh
   const [feedbackRefreshKey, setFeedbackRefreshKey] = useState(0);
 
   // Get time range for feedback data
-  const feedbackTimeRange = useMemo(() => {
-    const now = new Date();
-    let startTime: Date | null = null;
-
-    switch (filterState.timeRange) {
-      case 'lastHour':
-        startTime = new Date(now.getTime() - 60 * 60 * 1000);
-        break;
-      case 'last6Hours':
-        startTime = new Date(now.getTime() - 6 * 60 * 60 * 1000);
-        break;
-      case 'last24Hours':
-        startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
-        break;
-      case 'last7Days':
-        startTime = new Date(now.getTime() - 7 * 24 * 60 * 60 * 1000);
-        break;
-      case 'last30Days':
-        startTime = new Date(now.getTime() - 30 * 24 * 60 * 60 * 1000);
-        break;
-      case 'custom':
-        if (filterState.customDateRange) {
-          startTime = filterState.customDateRange.from;
-        }
-        break;
-    }
-
-    const endTime =
-      filterState.timeRange === 'custom' && filterState.customDateRange
-        ? filterState.customDateRange.to
-        : now;
-
-    return {
-      startTime: startTime?.toISOString(),
-      endTime: endTime.toISOString(),
-    };
+  const feedbackTimeRange = useMemo(
+    () =>
+      resolveMonitoringWindow(
+        filterState.timeRange,
+        filterState.customDateRange,
+      ),
     // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [filterState.timeRange, filterState.customDateRange, feedbackRefreshKey]);
+    [filterState.timeRange, filterState.customDateRange, feedbackRefreshKey],
+  );
 
   // Feedback data hook
   const {
     feedback: feedbackList,
     stats: feedbackStats,
-    loading: feedbackLoading,
+    loading: feedbackRefreshing,
   } = useFeedbackData({
+    mode: executionMode,
+    statusGroup: executionStatus,
     botIds:
       filterState.selectedBots.length > 0
         ? filterState.selectedBots
@@ -96,43 +147,68 @@ function MonitoringPageContent() {
     limit: 50,
   });
 
+  const loading = monitoringRefreshing && !data;
+  const executionLoading = executionRefreshing && !executionResult;
+  const feedbackLoading = feedbackRefreshing && !feedbackStats;
+  const refreshing =
+    monitoringRefreshing || executionRefreshing || feedbackRefreshing;
+  const countdown = useRef(0);
+  const intervalRef = useRef(0);
+  const [remainingSeconds, setRemainingSeconds] = useState(0);
   // Combined refresh handler for both monitoring and feedback data
   const handleRefresh = useCallback(() => {
+    countdown.current = intervalRef.current;
+    setRemainingSeconds(countdown.current);
     refetch();
+    refetchExecutions();
     setFeedbackRefreshKey((k) => k + 1);
-  }, [refetch]);
+  }, [refetch, refetchExecutions]);
 
-  const conversationTurns = useMemo(
-    () =>
-      buildConversationTurns(
-        data?.messages || [],
-        data?.llmCalls || [],
-        data?.errors || [],
-        data?.toolCalls || [],
-      ),
-    [data?.messages, data?.llmCalls, data?.errors, data?.toolCalls],
-  );
+  const [refreshInterval, setRefreshInterval] = useState(() => {
+    try {
+      const saved = Number(
+        localStorage.getItem('langbot-dashboard-refresh-interval'),
+      );
+      return [5, 15, 30, 60, 300].includes(saved) ? saved : 0;
+    } catch {
+      return 0;
+    }
+  });
+  const refreshState = useRef({ handleRefresh, busy: false });
+  useEffect(() => {
+    refreshState.current = { handleRefresh, busy: refreshing };
+  }, [handleRefresh, refreshing]);
+  useEffect(() => {
+    intervalRef.current = refreshInterval;
+    countdown.current = refreshInterval;
+    setRemainingSeconds(refreshInterval);
+    if (!refreshInterval || error) return;
+    const timer = window.setInterval(() => {
+      if (
+        document.visibilityState === 'visible' &&
+        !refreshState.current.busy
+      ) {
+        countdown.current = Math.max(0, countdown.current - 1);
+        setRemainingSeconds(countdown.current);
+        if (countdown.current === 0) {
+          refreshState.current.busy = true;
+          refreshState.current.handleRefresh();
+        }
+      }
+    }, 1000);
+    return () => window.clearInterval(timer);
+  }, [refreshInterval, error]);
 
   // State for expanded errors
   const [expandedErrorId, setExpandedErrorId] = useState<string | null>(null);
-  const [expandedTurnId, setExpandedTurnId] = useState<string | null>(null);
 
   // State for controlled tabs
-  const [activeTab, setActiveTab] = useState<string>('messages');
+  const [activeTab, setActiveTab] = useState<string>('executions');
 
-  // Function to jump to a message record
-  const jumpToMessage = (messageId: string) => {
-    setActiveTab('messages');
-    setTimeout(() => {
-      const turn = conversationTurns.find((item) =>
-        item.messages.some((message) => message.id === messageId),
-      );
-      setExpandedTurnId(turn?.id ?? messageId);
-    }, 100);
-  };
-
-  const toggleTurnExpand = (turnId: string) => {
-    setExpandedTurnId((current) => (current === turnId ? null : turnId));
+  // Resolve the owning record server-side, including records outside this page.
+  const jumpToMessage = (identifier: string) => {
+    setActiveTab('executions');
+    setSelectedExecution({ source: 'auto', id: identifier });
   };
 
   const toggleErrorExpand = (errorId: string) => {
@@ -146,52 +222,131 @@ function MonitoringPageContent() {
   return (
     <div className="w-full h-full overflow-y-auto overflow-x-hidden">
       {/* Filters and Refresh Button - Sticky */}
-      <div className="sticky top-0 z-10 -mt-1 pb-5 pt-1 bg-background">
-        <div>
-          <div className="flex flex-col gap-3 p-3 bg-card rounded-xl border sm:flex-row sm:flex-wrap sm:items-center sm:justify-between sm:gap-4 sm:p-4">
-            <MonitoringFilters
-              selectedBots={filterState.selectedBots}
-              selectedPipelines={filterState.selectedPipelines}
-              timeRange={filterState.timeRange}
-              onBotsChange={setSelectedBots}
-              onPipelinesChange={setSelectedPipelines}
-              onTimeRangeChange={setTimeRange}
-            />
-            <div className="flex items-center gap-2">
-              {canExport && <ExportDropdown filterState={filterState} />}
-              <Button
-                variant="outline"
-                size="sm"
-                onClick={handleRefresh}
-                className="flex-1 shadow-sm sm:flex-shrink-0 sm:flex-none"
-              >
-                <RefreshCw className="w-4 h-4 mr-2" />
-                {t('monitoring.refreshData')}
-              </Button>
-            </div>
+      {!error && (
+        <div className="relative z-10 -mt-1 pb-3 pt-1 bg-background lg:sticky lg:top-0 lg:pb-5">
+          <div>
+            <Card className="flex flex-col gap-4 p-4 xl:flex-row xl:items-end">
+              <MonitoringFilters
+                selectedBots={filterState.selectedBots}
+                selectedPipelines={filterState.selectedPipelines}
+                timeRange={filterState.timeRange}
+                onBotsChange={setSelectedBots}
+                onPipelinesChange={setSelectedPipelines}
+                onTimeRangeChange={setTimeRange}
+                customDateRange={filterState.customDateRange}
+                onCustomDateRangeChange={setCustomDateRange}
+                mode={executionMode}
+                onModeChange={setExecutionMode}
+                statusGroup={executionStatus}
+                onStatusGroupChange={setExecutionStatus}
+              />
+              <div className="flex shrink-0 flex-wrap items-center justify-end gap-2 border-t pt-3 xl:border-l xl:border-t-0 xl:pl-4 xl:pt-0 [&_button]:h-9">
+                <Button variant="outline" size="sm" onClick={resetFilters}>
+                  <RotateCcw className="mr-2 h-4 w-4" />
+                  {t('monitoring.filters.reset')}
+                </Button>
+                {canExport && <ExportDropdown filterState={filterState} />}
+                <div className="relative flex items-center overflow-hidden rounded-lg border bg-background shadow-xs">
+                  <Button
+                    variant="ghost"
+                    size="sm"
+                    onClick={handleRefresh}
+                    disabled={refreshing}
+                    className="gap-2 rounded-none px-3"
+                  >
+                    <RefreshCw
+                      className={cn('size-3.5', refreshing && 'animate-spin')}
+                    />
+                    {t('monitoring.refreshData')}
+                    {refreshInterval > 0 && (
+                      <span className="min-w-8 text-right text-xs tabular-nums text-muted-foreground">
+                        {remainingSeconds}s
+                      </span>
+                    )}
+                  </Button>
+                  <Select
+                    value={String(refreshInterval)}
+                    onValueChange={(value) => {
+                      setRefreshInterval(Number(value));
+                      try {
+                        localStorage.setItem(
+                          'langbot-dashboard-refresh-interval',
+                          value,
+                        );
+                      } catch {
+                        /* Storage may be unavailable. */
+                      }
+                    }}
+                  >
+                    <SelectTrigger
+                      className="w-11 justify-center rounded-l-none border-0 border-l px-2 shadow-none [&>svg:last-child]:hidden"
+                      aria-label={t('monitoring.autoRefreshLabel')}
+                      title={
+                        refreshInterval
+                          ? t('monitoring.autoRefreshEvery', {
+                              seconds: refreshInterval,
+                            })
+                          : t('monitoring.autoRefreshOff')
+                      }
+                    >
+                      <span aria-hidden="true">
+                        <ChevronDown className="size-3.5" />
+                      </span>
+                    </SelectTrigger>
+                    <SelectContent>
+                      <SelectItem value="0">
+                        {t('monitoring.autoRefreshOff')}
+                      </SelectItem>
+                      {[5, 15, 30, 60, 300].map((seconds) => (
+                        <SelectItem key={seconds} value={String(seconds)}>
+                          {t('monitoring.autoRefreshEvery', { seconds })}
+                        </SelectItem>
+                      ))}
+                    </SelectContent>
+                  </Select>
+                  {refreshInterval > 0 && (
+                    <div
+                      className="pointer-events-none absolute inset-x-0 bottom-0 h-0.5 bg-blue-500/10"
+                      aria-hidden="true"
+                    >
+                      <div
+                        className="h-full origin-left bg-blue-500/70 transition-transform duration-700 ease-linear motion-reduce:transition-none"
+                        style={{
+                          transform: `scaleX(${remainingSeconds / refreshInterval})`,
+                        }}
+                      />
+                    </div>
+                  )}
+                </div>
+              </div>
+            </Card>
           </div>
         </div>
-      </div>
+      )}
 
       {/* Content Area */}
       {error ? (
-        <div
-          role="alert"
-          className="rounded-xl border border-destructive p-6 space-y-3"
-        >
-          <p>{t('monitoring.loadError')}</p>
-          <Button variant="outline" onClick={handleRefresh}>
-            {t('common.retry')}
-          </Button>
-        </div>
+        <LoadErrorState
+          title={t('monitoring.loadError')}
+          onRetry={handleRefresh}
+          busy={refreshing}
+        />
       ) : (
         <div className="relative z-0 flex flex-col gap-6 pb-4 pt-3">
-          {/* Overview Section */}
-          <OverviewCards
-            metrics={data?.overview || null}
-            traffic={data?.traffic}
-            loading={loading}
-          />
+          {/* Overview Section: the execution metrics and the runtime status share
+              one grid, so a card always lands in the same row regardless of how
+              many metrics reported. */}
+          <div className="grid grid-cols-1 gap-6 md:grid-cols-2 xl:grid-cols-4">
+            <ExecutionOverviewCards
+              summary={executionResult?.summary ?? null}
+              loading={executionLoading}
+              activeSessions={data?.overview?.activeSessions ?? null}
+            />
+            <SystemStatusCard refreshKey={feedbackRefreshKey} />
+          </div>
+
+          {/* Traffic Chart */}
+          <TrafficChart traffic={data?.traffic} loading={loading} />
 
           {/* Tabs Section */}
           {!loading && data && (
@@ -199,14 +354,6 @@ function MonitoringPageContent() {
               className="text-sm text-muted-foreground space-y-1"
               role="status"
             >
-              {data.totalCount.messages > data.messages.length && (
-                <p>
-                  {t('monitoring.partialMessages', {
-                    shown: data.messages.length,
-                    total: data.totalCount.messages,
-                  })}
-                </p>
-              )}
               {data.totalCount.llmCalls + data.totalCount.embeddingCalls >
                 data.modelCalls.length && (
                 <p>
@@ -235,17 +382,21 @@ function MonitoringPageContent() {
               )}
             </div>
           )}
-          <div className="bg-card rounded-xl border overflow-hidden">
+          {/* Detail records region: one fixed-height surface for every tab. Tab
+              switches and expanded records scroll inside it, so the page never
+              resizes and nothing below the fold is pushed out of view. */}
+          <div className="bg-card flex h-[min(72vh,48rem)] min-h-[24rem] flex-col overflow-hidden rounded-xl border">
             <Tabs
               value={activeTab}
               onValueChange={setActiveTab}
-              className="w-full"
+              className="flex h-full min-h-0 w-full flex-col gap-0"
             >
-              <div className="px-3 pt-4 sm:px-6">
+              <div className="border-b px-3 pt-4 pb-3 sm:px-6">
                 <TabsList className="h-12 w-full justify-start gap-1 overflow-x-auto p-1 sm:w-auto">
-                  <TabsTrigger value="messages" className="px-3 py-2 sm:px-6">
-                    {t('monitoring.tabs.messages')}
+                  <TabsTrigger value="executions" className="px-3 py-2 sm:px-6">
+                    {t('monitoring.execution.tabs.executions')}
                   </TabsTrigger>
+
                   <TabsTrigger value="modelCalls" className="px-3 py-2 sm:px-6">
                     {t('monitoring.tabs.modelCalls')}
                   </TabsTrigger>
@@ -261,233 +412,331 @@ function MonitoringPageContent() {
                 </TabsList>
               </div>
 
-              <TabsContent value="messages" className="p-3 m-0 sm:p-6">
-                <div>
-                  {loading && (
-                    <div className="py-12 flex justify-center">
-                      <LoadingSpinner
-                        text={t('monitoring.messageList.loading')}
-                      />
-                    </div>
-                  )}
+              <TabsContent
+                value="executions"
+                className="m-0 flex min-h-0 flex-1 flex-col"
+              >
+                <p className="px-3 pt-4 text-sm text-muted-foreground sm:px-6">
+                  {t('monitoring.execution.subtitle')}
+                </p>
 
-                  {!loading && data && conversationTurns.length > 0 && (
-                    <ConversationTurnList
-                      turns={conversationTurns}
-                      expandedTurnId={expandedTurnId}
-                      onToggleTurn={toggleTurnExpand}
+                {executionError ? (
+                  <div className="px-3 pt-4 sm:px-6">
+                    <LoadErrorState
+                      title={t('monitoring.execution.detail.loadError')}
+                      onRetry={handleRefresh}
+                      busy={refreshing}
                     />
-                  )}
-
-                  {!loading && (!data || conversationTurns.length === 0) && (
-                    <div className="flex flex-col items-center justify-center text-muted-foreground py-16 gap-2">
-                      <MessageSquare className="h-[3rem] w-[3rem]" />
-                      <div className="text-sm">
-                        {t('monitoring.messageList.noMessages')}
-                      </div>
+                  </div>
+                ) : (
+                  <>
+                    <div className="min-h-0 flex-1 overflow-hidden px-3 pt-4 pb-4 sm:px-6">
+                      {executionLoading ? (
+                        <TabState loading rows={6} />
+                      ) : (executionResult?.items.length ?? 0) === 0 ? (
+                        <TabState
+                          icon={<Activity className="h-12 w-12" />}
+                          title={t('monitoring.execution.empty')}
+                        />
+                      ) : (
+                        <ExecutionTable
+                          rows={executionResult?.items ?? []}
+                          selectedId={selectedExecution?.id}
+                          onSelect={setSelectedExecution}
+                        />
+                      )}
                     </div>
-                  )}
-                </div>
+                    {executionResult &&
+                      executionResult.total > EXECUTIONS_PAGE_SIZE && (
+                        <div className="flex flex-wrap items-center justify-between gap-2 border-t px-3 py-3 sm:px-6">
+                          <Button
+                            variant="outline"
+                            size="sm"
+                            disabled={executionOffset === 0}
+                            onClick={() =>
+                              setExecutionOffset(
+                                Math.max(
+                                  0,
+                                  executionOffset - EXECUTIONS_PAGE_SIZE,
+                                ),
+                              )
+                            }
+                          >
+                            {t('operationTrace.previousPage')}
+                          </Button>
+                          <span className="text-xs text-muted-foreground">
+                            {executionOffset + 1}–
+                            {Math.min(
+                              executionOffset + EXECUTIONS_PAGE_SIZE,
+                              executionResult.total,
+                            )}{' '}
+                            / {executionResult.total}
+                          </span>
+                          <Button
+                            variant="outline"
+                            size="sm"
+                            disabled={!executionResult.has_more}
+                            onClick={() =>
+                              setExecutionOffset(
+                                executionOffset + EXECUTIONS_PAGE_SIZE,
+                              )
+                            }
+                          >
+                            {t('operationTrace.nextPage')}
+                          </Button>
+                        </div>
+                      )}
+                  </>
+                )}
               </TabsContent>
 
-              <TabsContent value="modelCalls" className="p-3 m-0 sm:p-6">
+              <TabsContent
+                value="modelCalls"
+                className="m-0 min-h-0 flex-1 overflow-y-auto p-3 sm:p-6"
+              >
                 <div>
-                  {loading && (
-                    <div className="py-12 flex justify-center">
-                      <LoadingSpinner text={t('common.loading')} />
-                    </div>
-                  )}
+                  {loading && <TabState loading rows={6} />}
 
                   {!loading &&
                     data &&
                     data.modelCalls &&
                     data.modelCalls.length > 0 && (
                       <div className="space-y-4">
-                        {data.modelCalls.map((call) => (
-                          <div
-                            key={call.id}
-                            className="border rounded-xl p-3 transition-all duration-200 sm:p-5"
-                          >
-                            <div className="flex justify-between items-start mb-3">
-                              <div className="flex-1">
-                                {/* Query ID - only show if messageId exists */}
-                                {call.messageId && (
-                                  <div className="flex items-center gap-2 mb-1">
-                                    <span className="text-xs text-muted-foreground font-mono">
-                                      Query ID: {call.messageId}
-                                    </span>
-                                    <Button
-                                      variant="ghost"
-                                      size="sm"
-                                      className="h-5 px-1.5 text-xs"
-                                      onClick={() =>
-                                        jumpToMessage(call.messageId!)
+                        {data.modelCalls.map((call) => {
+                          // A recorded call names the record it belongs to, so
+                          // the whole card opens it; without a match the card
+                          // stays static instead of pretending to be a link.
+                          const target = call.messageId;
+                          const openTarget = () => {
+                            if (target) jumpToMessage(target);
+                          };
+                          return (
+                            <Card
+                              key={call.id}
+                              role={target ? 'button' : undefined}
+                              tabIndex={target ? 0 : undefined}
+                              aria-label={
+                                target
+                                  ? t('monitoring.execution.detail.openObject')
+                                  : undefined
+                              }
+                              onClick={target ? openTarget : undefined}
+                              onKeyDown={
+                                target
+                                  ? (
+                                      event: React.KeyboardEvent<HTMLDivElement>,
+                                    ) => {
+                                      if (
+                                        event.key === 'Enter' ||
+                                        event.key === ' '
+                                      ) {
+                                        event.preventDefault();
+                                        openTarget();
                                       }
-                                    >
-                                      <ExternalLink className="w-3 h-3 mr-1" />
-                                      {t(
-                                        'monitoring.messageList.viewConversation',
-                                      )}
-                                    </Button>
-                                  </div>
-                                )}
-                                <div className="flex items-center gap-2 mb-2">
-                                  {/* Model Type Badge */}
-                                  <span
-                                    className={`text-xs px-2 py-1 rounded ${
-                                      call.modelType === 'llm'
-                                        ? 'bg-blue-100 text-blue-800 dark:bg-blue-900 dark:text-blue-200'
-                                        : 'bg-purple-100 text-purple-800 dark:bg-purple-900 dark:text-purple-200'
-                                    }`}
-                                  >
-                                    {call.modelType === 'llm'
-                                      ? t('monitoring.modelCalls.llmModel')
-                                      : t(
-                                          'monitoring.modelCalls.embeddingModel',
-                                        )}
-                                  </span>
-                                  {/* Call Type Badge for Embedding */}
-                                  {call.modelType === 'embedding' &&
-                                    call.callType && (
-                                      <span
-                                        className={`text-xs px-2 py-1 rounded ${
-                                          call.callType === 'retrieve'
-                                            ? 'bg-cyan-100 text-cyan-800 dark:bg-cyan-900 dark:text-cyan-200'
-                                            : 'bg-amber-100 text-amber-800 dark:bg-amber-900 dark:text-amber-200'
-                                        }`}
-                                      >
-                                        {call.callType === 'retrieve'
-                                          ? t(
-                                              'monitoring.modelCalls.retrieveCall',
-                                            )
-                                          : t(
-                                              'monitoring.modelCalls.embeddingCall',
-                                            )}
+                                    }
+                                  : undefined
+                              }
+                              className={cn(
+                                'gap-0 px-4 py-4 transition-all duration-200 sm:px-5 sm:py-5',
+                                target &&
+                                  'cursor-pointer hover:bg-muted/40 focus-visible:ring-ring/50 focus-visible:ring-[3px] focus-visible:outline-none',
+                              )}
+                            >
+                              <div className="flex justify-between items-start mb-3">
+                                <div className="flex-1">
+                                  {/* The stored id is the owning record: a
+                                      Pipeline's message, or an Agent run's id
+                                      that the card itself opens. */}
+                                  {call.messageId && (
+                                    <div className="flex items-center gap-2 mb-1">
+                                      <span className="text-xs text-muted-foreground font-mono">
+                                        Query ID: {call.messageId}
                                       </span>
-                                    )}
-                                  {/* Status Badge */}
-                                  <span
-                                    className={`text-xs px-2 py-1 rounded ${
-                                      call.status === 'success'
-                                        ? 'bg-green-100 text-green-800 dark:bg-green-900 dark:text-green-200'
-                                        : 'bg-red-100 text-red-800 dark:bg-red-900 dark:text-red-200'
-                                    }`}
-                                  >
-                                    {call.status}
-                                  </span>
-                                </div>
-                                {/* Model Name */}
-                                <div className="font-medium text-sm text-foreground mb-2">
-                                  {call.modelName}
-                                </div>
-                                {/* Context Info - only for LLM calls */}
-                                {call.modelType === 'llm' &&
-                                  call.botName &&
-                                  call.pipelineName && (
-                                    <div className="text-xs text-muted-foreground mb-1">
-                                      {call.botName} → {call.pipelineName}
+                                      <Button
+                                        variant="ghost"
+                                        size="sm"
+                                        className="h-5 px-1.5 text-xs"
+                                        onClick={(event) => {
+                                          // The card itself opens the call's
+                                          // record; this button is the more
+                                          // specific message target.
+                                          event.stopPropagation();
+                                          jumpToMessage(call.messageId!);
+                                        }}
+                                      >
+                                        <ExternalLink className="w-3 h-3 mr-1" />
+                                        {t(
+                                          'monitoring.messageList.viewConversation',
+                                        )}
+                                      </Button>
                                     </div>
                                   )}
-                                {/* Token Info */}
-                                <div className="text-xs text-muted-foreground space-y-1">
-                                  <div className="flex flex-wrap gap-4">
-                                    {call.modelType === 'llm' &&
-                                      call.tokens && (
+                                  <div className="flex items-center gap-2 mb-2">
+                                    {/* Model Type Badge */}
+                                    <Badge
+                                      variant="outline"
+                                      className={
+                                        call.modelType === 'llm'
+                                          ? 'bg-blue-100 text-blue-800 dark:bg-blue-900 dark:text-blue-200'
+                                          : 'bg-purple-100 text-purple-800 dark:bg-purple-900 dark:text-purple-200'
+                                      }
+                                    >
+                                      {call.modelType === 'llm'
+                                        ? t('monitoring.modelCalls.llmModel')
+                                        : t(
+                                            'monitoring.modelCalls.embeddingModel',
+                                          )}
+                                    </Badge>
+                                    {/* Call Type Badge for Embedding */}
+                                    {call.modelType === 'embedding' &&
+                                      call.callType && (
+                                        <Badge
+                                          variant="outline"
+                                          className={
+                                            call.callType === 'retrieve'
+                                              ? 'bg-cyan-100 text-cyan-800 dark:bg-cyan-900 dark:text-cyan-200'
+                                              : 'bg-amber-100 text-amber-800 dark:bg-amber-900 dark:text-amber-200'
+                                          }
+                                        >
+                                          {call.callType === 'retrieve'
+                                            ? t(
+                                                'monitoring.modelCalls.retrieveCall',
+                                              )
+                                            : t(
+                                                'monitoring.modelCalls.embeddingCall',
+                                              )}
+                                        </Badge>
+                                      )}
+                                    {/* Status Badge */}
+                                    <Badge
+                                      variant="outline"
+                                      className={
+                                        call.status === 'success'
+                                          ? 'bg-green-100 text-green-800 dark:bg-green-900 dark:text-green-200'
+                                          : 'bg-red-100 text-red-800 dark:bg-red-900 dark:text-red-200'
+                                      }
+                                    >
+                                      {call.status}
+                                    </Badge>
+                                  </div>
+                                  {/* Model Name */}
+                                  <div className="font-medium text-sm text-foreground mb-2">
+                                    {call.modelName}
+                                  </div>
+                                  {/* Context Info - only for LLM calls */}
+                                  {call.modelType === 'llm' &&
+                                    call.botName &&
+                                    call.pipelineName && (
+                                      <div className="text-xs text-muted-foreground mb-1">
+                                        {call.botName} → {call.pipelineName}
+                                      </div>
+                                    )}
+                                  {/* Token Info */}
+                                  <div className="text-xs text-muted-foreground space-y-1">
+                                    <div className="flex flex-wrap gap-4">
+                                      {call.modelType === 'llm' &&
+                                        call.tokens && (
+                                          <>
+                                            <span>
+                                              {t(
+                                                'monitoring.llmCalls.inputTokens',
+                                              )}
+                                              : {call.tokens.input}
+                                            </span>
+                                            <span>
+                                              {t(
+                                                'monitoring.llmCalls.outputTokens',
+                                              )}
+                                              : {call.tokens.output}
+                                            </span>
+                                            <span>
+                                              {t(
+                                                'monitoring.llmCalls.totalTokens',
+                                              )}
+                                              : {call.tokens.total}
+                                            </span>
+                                          </>
+                                        )}
+                                      {call.modelType === 'embedding' && (
                                         <>
                                           <span>
                                             {t(
-                                              'monitoring.llmCalls.inputTokens',
+                                              'monitoring.embeddingCalls.promptTokens',
                                             )}
-                                            : {call.tokens.input}
+                                            : {call.promptTokens}
                                           </span>
                                           <span>
                                             {t(
-                                              'monitoring.llmCalls.outputTokens',
+                                              'monitoring.embeddingCalls.totalTokens',
                                             )}
-                                            : {call.tokens.output}
+                                            : {call.totalTokens}
                                           </span>
                                           <span>
                                             {t(
-                                              'monitoring.llmCalls.totalTokens',
+                                              'monitoring.embeddingCalls.inputCount',
                                             )}
-                                            : {call.tokens.total}
+                                            : {call.inputCount}
                                           </span>
                                         </>
                                       )}
-                                    {call.modelType === 'embedding' && (
-                                      <>
-                                        <span>
-                                          {t(
-                                            'monitoring.embeddingCalls.promptTokens',
-                                          )}
-                                          : {call.promptTokens}
-                                        </span>
-                                        <span>
-                                          {t(
-                                            'monitoring.embeddingCalls.totalTokens',
-                                          )}
-                                          : {call.totalTokens}
-                                        </span>
-                                        <span>
-                                          {t(
-                                            'monitoring.embeddingCalls.inputCount',
-                                          )}
-                                          : {call.inputCount}
-                                        </span>
-                                      </>
-                                    )}
-                                    <span>
-                                      {t('monitoring.llmCalls.duration')}:{' '}
-                                      {call.duration}ms
-                                    </span>
-                                    {call.cost && (
                                       <span>
-                                        {t('monitoring.llmCalls.cost')}: $
-                                        {call.cost.toFixed(4)}
+                                        {t('monitoring.llmCalls.duration')}:{' '}
+                                        {call.duration}ms
                                       </span>
-                                    )}
-                                  </div>
-                                  {/* Knowledge Base Info for Embedding */}
-                                  {call.modelType === 'embedding' &&
-                                    call.knowledgeBaseId && (
-                                      <div>
-                                        {t(
-                                          'monitoring.embeddingCalls.knowledgeBase',
-                                        )}
-                                        : {call.knowledgeBaseId}
-                                      </div>
-                                    )}
-                                  {/* Query Text for Embedding Retrieve */}
-                                  {call.modelType === 'embedding' &&
-                                    call.queryText && (
-                                      <div className="mt-2 p-2 bg-muted rounded text-sm">
-                                        <span className="text-muted-foreground">
+                                      {call.cost && (
+                                        <span>
+                                          {t('monitoring.llmCalls.cost')}: $
+                                          {call.cost.toFixed(4)}
+                                        </span>
+                                      )}
+                                    </div>
+                                    {/* Knowledge Base Info for Embedding */}
+                                    {call.modelType === 'embedding' &&
+                                      call.knowledgeBaseId && (
+                                        <div>
                                           {t(
-                                            'monitoring.embeddingCalls.queryText',
+                                            'monitoring.embeddingCalls.knowledgeBase',
                                           )}
-                                          :{' '}
-                                        </span>
-                                        <span className="text-foreground">
-                                          {call.queryText.length > 100
-                                            ? call.queryText.substring(0, 100) +
-                                              '...'
-                                            : call.queryText}
-                                        </span>
-                                      </div>
-                                    )}
-                                </div>
-                                {call.errorMessage && (
-                                  <div className="mt-2 text-xs text-red-600 dark:text-red-400">
-                                    Error: {call.errorMessage}
+                                          : {call.knowledgeBaseId}
+                                        </div>
+                                      )}
+                                    {/* Query Text for Embedding Retrieve */}
+                                    {call.modelType === 'embedding' &&
+                                      call.queryText && (
+                                        <div className="mt-2 p-2 bg-muted rounded text-sm">
+                                          <span className="text-muted-foreground">
+                                            {t(
+                                              'monitoring.embeddingCalls.queryText',
+                                            )}
+                                            :{' '}
+                                          </span>
+                                          <span className="text-foreground">
+                                            {call.queryText.length > 100
+                                              ? call.queryText.substring(
+                                                  0,
+                                                  100,
+                                                ) + '...'
+                                              : call.queryText}
+                                          </span>
+                                        </div>
+                                      )}
                                   </div>
-                                )}
+                                  {call.errorMessage && (
+                                    <div className="mt-2 text-xs text-red-600 dark:text-red-400">
+                                      Error: {call.errorMessage}
+                                    </div>
+                                  )}
+                                </div>
+                                <span className="ml-4 flex items-center gap-1 text-xs text-muted-foreground whitespace-nowrap">
+                                  {call.timestamp.toLocaleString()}
+                                  {target && (
+                                    <ChevronRight className="h-4 w-4" />
+                                  )}
+                                </span>
                               </div>
-                              <span className="text-xs text-muted-foreground whitespace-nowrap ml-4">
-                                {call.timestamp.toLocaleString()}
-                              </span>
-                            </div>
-                          </div>
-                        ))}
+                            </Card>
+                          );
+                        })}
                       </div>
                     )}
 
@@ -495,18 +744,21 @@ function MonitoringPageContent() {
                     (!data ||
                       !data.modelCalls ||
                       data.modelCalls.length === 0) && (
-                      <div className="flex flex-col items-center justify-center text-muted-foreground py-16 gap-2">
-                        <Sparkles className="h-[3rem] w-[3rem]" />
-                        <div className="text-sm">
-                          {t('monitoring.modelCalls.noData')}
-                        </div>
-                      </div>
+                      <TabState
+                        icon={<Sparkles className="h-12 w-12" />}
+                        title={t('monitoring.modelCalls.noData')}
+                      />
                     )}
                 </div>
               </TabsContent>
 
-              <TabsContent value="tokens" className="p-3 m-0 sm:p-6">
+              <TabsContent
+                value="tokens"
+                className="m-0 min-h-0 flex-1 overflow-y-auto p-3 sm:p-6"
+              >
                 <TokenMonitoring
+                  mode={executionMode}
+                  statusGroup={executionStatus}
                   botIds={
                     filterState.selectedBots.length > 0
                       ? filterState.selectedBots
@@ -523,13 +775,12 @@ function MonitoringPageContent() {
                 />
               </TabsContent>
 
-              <TabsContent value="feedback" className="p-3 m-0 sm:p-6">
+              <TabsContent
+                value="feedback"
+                className="m-0 min-h-0 flex-1 overflow-y-auto p-3 sm:p-6"
+              >
                 <div>
-                  {loading && (
-                    <div className="py-12 flex justify-center">
-                      <LoadingSpinner text={t('common.loading')} />
-                    </div>
-                  )}
+                  {loading && <TabState loading rows={6} />}
 
                   {!loading && (
                     <>
@@ -542,7 +793,7 @@ function MonitoringPageContent() {
                       </div>
 
                       {/* Feedback List */}
-                      <h3 className="text-lg font-semibold text-gray-900 dark:text-white mb-4">
+                      <h3 className="text-lg font-semibold text-foreground mb-4">
                         {t('monitoring.feedback.feedbackList')}
                       </h3>
                       <FeedbackList
@@ -555,13 +806,12 @@ function MonitoringPageContent() {
                 </div>
               </TabsContent>
 
-              <TabsContent value="errors" className="p-3 m-0 sm:p-6">
+              <TabsContent
+                value="errors"
+                className="m-0 min-h-0 flex-1 overflow-y-auto p-3 sm:p-6"
+              >
                 <div>
-                  {loading && (
-                    <div className="py-12 flex justify-center">
-                      <LoadingSpinner text={t('common.loading')} />
-                    </div>
-                  )}
+                  {loading && <TabState loading rows={6} />}
 
                   {!loading &&
                     data &&
@@ -569,9 +819,9 @@ function MonitoringPageContent() {
                     data.errors.length > 0 && (
                       <div className="space-y-4">
                         {data.errors.map((error) => (
-                          <div
+                          <Card
                             key={error.id}
-                            className="border border-red-200 dark:border-red-900 rounded-xl overflow-hidden transition-all duration-200"
+                            className="gap-0 overflow-hidden border-red-200 py-0 transition-all duration-200 dark:border-red-900"
                           >
                             {/* Error Header - Always Visible */}
                             <div
@@ -704,19 +954,19 @@ function MonitoringPageContent() {
                                 </div>
                               </div>
                             )}
-                          </div>
+                          </Card>
                         ))}
                       </div>
                     )}
 
                   {!loading &&
                     (!data || !data.errors || data.errors.length === 0) && (
-                      <div className="flex flex-col items-center justify-center text-muted-foreground py-16 gap-2">
-                        <CheckCircle2 className="h-[3rem] w-[3rem] text-green-500 dark:text-green-600" />
-                        <div className="text-sm text-green-600 dark:text-green-400">
-                          {t('monitoring.errors.noErrors')}
-                        </div>
-                      </div>
+                      <TabState
+                        icon={
+                          <CheckCircle2 className="h-12 w-12 text-green-500 dark:text-green-600" />
+                        }
+                        title={t('monitoring.errors.noErrors')}
+                      />
                     )}
                 </div>
               </TabsContent>
@@ -724,14 +974,22 @@ function MonitoringPageContent() {
           </div>
         </div>
       )}
+
+      {/* Execution trace drawer */}
+      <ExecutionDetailSheet
+        row={selectedExecution}
+        onSelect={setSelectedExecution}
+        onClose={() => setSelectedExecution(null)}
+      />
     </div>
   );
 }
 
 export default function MonitoringPage() {
+  const workspace = useCurrentWorkspace()?.workspace.uuid;
   return (
     <Suspense fallback={<LoadingPage />}>
-      <MonitoringPageContent />
+      <MonitoringPageContent key={workspace ?? 'pending'} />
     </Suspense>
   );
 }
diff --git a/web/src/app/home/monitoring/types/monitoring.ts b/web/src/app/home/monitoring/types/monitoring.ts
index 8a1f4dc53..21fe9fcee 100644
--- a/web/src/app/home/monitoring/types/monitoring.ts
+++ b/web/src/app/home/monitoring/types/monitoring.ts
@@ -164,6 +164,8 @@ export interface OverviewMetrics {
 }
 
 export interface FilterState {
+  mode?: 'all' | 'real' | 'debug';
+  statusGroup?: string;
   selectedBots: string[];
   selectedPipelines: string[];
   timeRange: TimeRangeOption;
@@ -176,6 +178,9 @@ export type TimeRangeOption =
   | 'last24Hours'
   | 'last7Days'
   | 'last30Days'
+  | 'last90Days'
+  | 'last180Days'
+  | 'last365Days'
   | 'custom';
 
 export interface DateRange {
diff --git a/web/src/app/home/monitoring/utils/callLinks.ts b/web/src/app/home/monitoring/utils/callLinks.ts
new file mode 100644
index 000000000..91d29e6ce
--- /dev/null
+++ b/web/src/app/home/monitoring/utils/callLinks.ts
@@ -0,0 +1,99 @@
+import type { ExecutionRow } from '@/app/infra/entities/api/monitoring-executions';
+
+/**
+ * The identity a recorded call carries: every call kind (model, tool) has a time,
+ * a session and - when a Pipeline recorded it - the message it belongs to.
+ */
+export interface CallLinkSource {
+  timestamp: Date;
+  sessionId?: string | null;
+  messageId?: string | null;
+}
+
+/** A record the dashboard can open for a recorded model or tool call. */
+export type CallTarget =
+  { kind: 'execution'; row: ExecutionRow } | { kind: 'message'; id: string };
+
+/** Calls are matched against a run's window; clock skew must not lose the link. */
+const WINDOW_SLACK_MS = 2000;
+
+function containsCall(row: ExecutionRow, callTime: number): boolean {
+  if (row.started_at_ms == null) {
+    return false;
+  }
+  const started = row.started_at_ms;
+  if (callTime < started - WINDOW_SLACK_MS) {
+    return false;
+  }
+  if (row.finished_at_ms == null) {
+    // A run with no end time is still open: anything after its start belongs to it.
+    return true;
+  }
+  return callTime <= row.finished_at_ms + WINDOW_SLACK_MS;
+}
+
+/**
+ * The execution a recorded call ran inside.
+ *
+ * A Pipeline call names its message, which is the same id the execution list
+ * uses for that query. An Agent call names nothing, so it is matched to the run
+ * sharing its session whose window contains the call; the latest such start
+ * wins, so a session with consecutive runs links to the run that was live.
+ */
+export function findCallExecution(
+  call: CallLinkSource,
+  executions: ExecutionRow[],
+): ExecutionRow | null {
+  if (executions.length === 0) {
+    return null;
+  }
+
+  const messageId = call.messageId;
+  if (messageId) {
+    const direct = executions.find((row) => row.id === messageId);
+    if (direct) {
+      return direct;
+    }
+  }
+
+  const sessionId = call.sessionId;
+  const callTime = call.timestamp.getTime();
+  let matched: ExecutionRow | null = null;
+  for (const row of executions) {
+    if (
+      !sessionId ||
+      row.session_id !== sessionId ||
+      !containsCall(row, callTime)
+    ) {
+      continue;
+    }
+    if (
+      matched == null ||
+      (row.started_at_ms ?? 0) > (matched.started_at_ms ?? 0)
+    ) {
+      matched = row;
+    }
+  }
+  return matched;
+}
+
+/**
+ * Where a call's card jumps to: its execution record when the loaded list knows
+ * it, else the pipeline message it names. An Agent call carries its run id, so
+ * the message fallback applies only to ids the message list actually holds.
+ */
+export function resolveCallTarget(
+  call: CallLinkSource,
+  executions: ExecutionRow[],
+  messageIds?: ReadonlySet<string>,
+): CallTarget | null {
+  const row = findCallExecution(call, executions);
+  if (row) {
+    return { kind: 'execution', row };
+  }
+  const messageId = call.messageId;
+  if (messageId && messageIds?.has(messageId)) {
+    return { kind: 'message', id: messageId };
+  }
+  return null;
+}
diff --git a/web/src/app/home/monitoring/utils/dateUtils.ts b/web/src/app/home/monitoring/utils/dateUtils.ts
index b246d3cf9..cce636e22 100644
--- a/web/src/app/home/monitoring/utils/dateUtils.ts
+++ b/web/src/app/home/monitoring/utils/dateUtils.ts
@@ -22,8 +22,13 @@ export function getPresetDateRange(option: TimeRangeOption): DateRange | null {
     case 'last7Days':
       from.setDate(now.getDate() - 7);
       break;
+    case 'last90Days':
+    case 'last180Days':
+    case 'last365Days':
     case 'last30Days':
-      from.setDate(now.getDate() - 30);
+      from.setTime(
+        now.getTime() - Number(option.match(/\d+/)?.[0] ?? 30) * 86400000,
+      );
       break;
     default:
       return null;
@@ -116,3 +121,53 @@ export function parseUTCTimestamp(timestamp: string): Date {
   }
   return new Date(timestamp + 'Z');
 }
+
+/**
+ * Resolve a monitoring time window to ISO start/end strings.
+ *
+ * Single source of the window semantics for the dashboard, the unified
+ * execution list, and the feedback queries.
+ */
+export function resolveMonitoringWindow(
+  timeRange: TimeRangeOption,
+  customDateRange: DateRange | null,
+): { startTime: string; endTime: string } {
+  const now = new Date();
+  let startTime: Date | null = null;
+
+  switch (timeRange) {
+    case 'lastHour':
+      startTime = new Date(now.getTime() - 60 * 60 * 1000);
+      break;
+    case 'last6Hours':
+      startTime = new Date(now.getTime() - 6 * 60 * 60 * 1000);
+      break;
+    case 'last24Hours':
+      startTime = new Date(now.getTime() - 24 * 60 * 60 * 1000);
+      break;
+    case 'last7Days':
+      startTime = new Date(now.getTime() - 7 * 24 * 60 * 60 * 1000);
+      break;
+    case 'last90Days':
+    case 'last180Days':
+    case 'last365Days':
+    case 'last30Days':
+      startTime = new Date(
+        now.getTime() - Number(timeRange.match(/\d+/)?.[0] ?? 30) * 86400000,
+      );
+      break;
+    case 'custom':
+      if (customDateRange) {
+        startTime = customDateRange.from;
+      }
+      break;
+  }
+
+  const endTime =
+    timeRange === 'custom' && customDateRange ? customDateRange.to : now;
+
+  return {
+    startTime: (startTime ?? now).toISOString(),
+    endTime: endTime.toISOString(),
+  };
+}
diff --git a/web/src/app/home/pipelines/PipelineDetailContent.tsx b/web/src/app/home/pipelines/PipelineDetailContent.tsx
index 1066342df..3183328b7 100644
--- a/web/src/app/home/pipelines/PipelineDetailContent.tsx
+++ b/web/src/app/home/pipelines/PipelineDetailContent.tsx
@@ -1,7 +1,7 @@
 import EntityLoadState from '@/components/EntityLoadState';
 import { isCurrentPipelineConfig } from './pipeline-config-safety';
 import { useEffect, useRef, useState } from 'react';
-import { useNavigate } from 'react-router-dom';
+import { useNavigate, useSearchParams } from 'react-router-dom';
 import { toast } from 'sonner';
 import { Button } from '@/components/ui/button';
 import { Alert, AlertDescription } from '@/components/ui/alert';
@@ -51,10 +51,9 @@ export default function PipelineDetailContent({
     return () => setDetailEntityName(null);
   }, [id, isCreateMode, pipelines, setDetailEntityName, t]);
 
-  const [activeView, setActiveView] = useState<'workbench' | 'monitoring'>(
-    'workbench',
-  );
-  useEffect(() => setActiveView('workbench'), [id]);
+  const [searchParams] = useSearchParams();
+  const activeView =
+    searchParams.get('tab') === 'logs' ? 'monitoring' : 'workbench';
   const [isWebSocketConnected, setIsWebSocketConnected] = useState(false);
   const [formDirty, setFormDirty] = useState(false);
   const [formSaving, setFormSaving] = useState(false);
@@ -187,7 +186,6 @@ export default function PipelineDetailContent({
     <>
       <ProcessorDetailWorkbench
         key={id}
-        onViewChange={setActiveView}
         title={`${pipelineEmoji} ${pipelineName}`}
         titleAction={
           canManage ? (
diff --git a/web/src/app/home/pipelines/PipelineMigration.tsx b/web/src/app/home/pipelines/PipelineMigration.tsx
index 892022f37..a578ff9d5 100644
--- a/web/src/app/home/pipelines/PipelineMigration.tsx
+++ b/web/src/app/home/pipelines/PipelineMigration.tsx
@@ -147,7 +147,15 @@ export default function PipelineMigration({
       ['migrated', 'already_current'].includes(item.state),
     );
   const rows = preview?.items ?? [];
+  // Custom 4.10 runners cannot be migrated automatically; they are listed
+  // for attention but never submitted.
   const count = rows.filter(
+    (item) =>
+      !['already_current', 'not_legacy', 'unsupported_runner'].includes(
+        item.state,
+      ),
+  ).length;
+  const attention = rows.filter(
     (item) => !['already_current', 'not_legacy'].includes(item.state),
   ).length;
 
@@ -189,7 +197,10 @@ export default function PipelineMigration({
     )
       return;
     let items = rows.filter(
-      (item) => !['already_current', 'not_legacy'].includes(item.state),
+      (item) =>
+        !['already_current', 'not_legacy', 'unsupported_runner'].includes(
+          item.state,
+        ),
     );
     setInstallations([]);
     setDataOnly(!installPlugins);
@@ -340,16 +351,12 @@ export default function PipelineMigration({
 
   return (
     <>
-      {(count > 0 || previewError || results.length > 0) && (
+      {(attention > 0 || results.length > 0) && (
         <Alert className="mb-4 shrink-0 border-amber-200 bg-amber-50 text-amber-900 dark:border-amber-900/50 dark:bg-amber-950/30 dark:text-amber-200">
           <AlertTriangle aria-hidden="true" />
           <AlertTitle>{t('pipelineMigration.title')}</AlertTitle>
           <AlertDescription className="flex items-center justify-between gap-3 text-amber-800 dark:text-amber-200">
-            <span>
-              {previewError
-                ? t('pipelineMigration.previewError')
-                : t('pipelineMigration.detected', { count })}
-            </span>
+            <span>{t('pipelineMigration.detected', { count: attention })}</span>
             <Button
               variant="outline"
               size="sm"
@@ -408,7 +415,7 @@ export default function PipelineMigration({
               </div>
             ) : (
               <p className="text-sm text-muted-foreground">
-                {t('pipelineMigration.detected', { count })}
+                {t('pipelineMigration.detected', { count: attention })}
               </p>
             )}
             <MigrationInstallProgress installations={installations} />
@@ -421,7 +428,7 @@ export default function PipelineMigration({
                 })}
               </p>
             )}
-            {(count > 0 || results.length > 0) && (
+            {(attention > 0 || results.length > 0) && (
               <Collapsible>
                 <CollapsibleTrigger asChild>
                   <Button
@@ -474,6 +481,11 @@ export default function PipelineMigration({
                                   {item.target_plugin.name}
                                 </p>
                               )}
+                              {item.state === 'unsupported_runner' && (
+                                <p className="text-xs text-destructive break-words">
+                                  {t('pipelineMigration.notices.customRunner')}
+                                </p>
+                              )}
                               {item.warnings.some(
                                 (issue) =>
                                   issue.code === 'local.box_state_reset',
diff --git a/web/src/app/home/pipelines/components/debug-dialog/DebugDialog.tsx b/web/src/app/home/pipelines/components/debug-dialog/DebugDialog.tsx
index 9548a16ca..4b8c7be09 100644
--- a/web/src/app/home/pipelines/components/debug-dialog/DebugDialog.tsx
+++ b/web/src/app/home/pipelines/components/debug-dialog/DebugDialog.tsx
@@ -231,6 +231,7 @@ export default function DebugDialog({
         wsClient
           .onConnected(() => {
             if (wsClientRef.current !== wsClient) return;
+            errorReported = false;
             setIsConnected(true);
             isInitializingRef.current = false;
           })
@@ -261,11 +262,12 @@ export default function DebugDialog({
           })
           .onError((error) => {
             if (wsClientRef.current !== wsClient) return;
+            const notify = !errorReported;
             errorReported = true;
             console.error('WebSocket error:', error);
             setIsConnected(false);
             isInitializingRef.current = false;
-            toast.error(t('pipelines.debugDialog.connectionError'));
+            if (notify) toast.error(t('pipelines.debugDialog.connectionError'));
           })
           .onClose(() => {
             if (wsClientRef.current !== wsClient) return;
@@ -332,7 +334,22 @@ export default function DebugDialog({
       loadMessages(selectedPipelineId);
       initWebSocket(selectedPipelineId);
     }
-  }, [sessionType, selectedPipelineId, open, loadMessages, initWebSocket]);
+    return () => {
+      invalidateHistoryRequests();
+      const client = wsClientRef.current;
+      wsClientRef.current = null;
+      client?.disconnect();
+      isInitializingRef.current = false;
+      setIsConnected(false);
+    };
+  }, [
+    invalidateHistoryRequests,
+    sessionType,
+    selectedPipelineId,
+    open,
+    loadMessages,
+    initWebSocket,
+  ]);
 
   // Notify parent of connection status changes
   useEffect(() => {
diff --git a/web/src/app/home/pipelines/components/monitoring-tab/PipelineMonitoringTab.tsx b/web/src/app/home/pipelines/components/monitoring-tab/PipelineMonitoringTab.tsx
index a57c5cdd8..a0d2ae2af 100644
--- a/web/src/app/home/pipelines/components/monitoring-tab/PipelineMonitoringTab.tsx
+++ b/web/src/app/home/pipelines/components/monitoring-tab/PipelineMonitoringTab.tsx
@@ -57,7 +57,7 @@ export default function PipelineMonitoringTab({
   const [activeTab, setActiveTab] = useState<string>('messages');
 
   const toggleTurnExpand = (turnId: string) => {
-    setExpandedTurnId((current) => (current === turnId ? null : turnId));
+    setExpandedTurnId(turnId);
   };
 
   const toggleErrorExpand = (errorId: string) => {
@@ -83,7 +83,7 @@ export default function PipelineMonitoringTab({
   return (
     <div className="w-full h-full flex flex-col">
       {/* Header with refresh button */}
-      <div className="flex items-center justify-between mb-4 pb-4 border-b border-gray-200 dark:border-gray-700">
+      <div className="mb-4 flex flex-wrap items-center justify-between gap-3 border-b pb-4">
         <p className="text-sm text-gray-500 dark:text-gray-400">
           {t('pipelines.monitoring.description')}
         </p>
@@ -113,8 +113,8 @@ export default function PipelineMonitoringTab({
 
       {/* Overview Stats */}
       {data && (
-        <div className="grid grid-cols-3 gap-4 mb-6">
-          <div className="bg-white dark:bg-[#2a2a2e] rounded-lg border border-gray-200 dark:border-gray-700 p-4">
+        <div className="mb-4 grid grid-cols-3 divide-x rounded-xl border bg-card">
+          <div className="p-3">
             <div className="text-sm text-gray-500 dark:text-gray-400">
               {t('monitoring.totalMessages')}
             </div>
@@ -122,7 +122,7 @@ export default function PipelineMonitoringTab({
               {data.overview.totalMessages}
             </div>
           </div>
-          <div className="bg-white dark:bg-[#2a2a2e] rounded-lg border border-gray-200 dark:border-gray-700 p-4">
+          <div className="p-3">
             <div className="text-sm text-gray-500 dark:text-gray-400">
               {t('monitoring.successRate')}
             </div>
@@ -130,7 +130,7 @@ export default function PipelineMonitoringTab({
               {data.overview.successRate.toFixed(1)}%
             </div>
           </div>
-          <div className="bg-white dark:bg-[#2a2a2e] rounded-lg border border-gray-200 dark:border-gray-700 p-4">
+          <div className="p-3">
             <div className="text-sm text-gray-500 dark:text-gray-400">
               {t('monitoring.tabs.errors')}
             </div>
diff --git a/web/src/app/home/pipelines/components/pipeline-extensions/PipelineExtension.tsx b/web/src/app/home/pipelines/components/pipeline-extensions/PipelineExtension.tsx
index 52dd26082..636bd5ced 100644
--- a/web/src/app/home/pipelines/components/pipeline-extensions/PipelineExtension.tsx
+++ b/web/src/app/home/pipelines/components/pipeline-extensions/PipelineExtension.tsx
@@ -12,7 +12,15 @@ import {
   DialogFooter,
 } from '@/components/ui/dialog';
 import { Checkbox } from '@/components/ui/checkbox';
-import { CircleHelp, Plus, X, Server, Wrench, Sparkles } from 'lucide-react';
+import {
+  CircleHelp,
+  Plus,
+  X,
+  Server,
+  Wrench,
+  Sparkles,
+  Puzzle,
+} from 'lucide-react';
 import { Badge } from '@/components/ui/badge';
 import { Switch } from '@/components/ui/switch';
 import { Label } from '@/components/ui/label';
@@ -24,6 +32,27 @@ import {
 import { Plugin } from '@/app/infra/entities/plugin';
 import { MCPServer, Skill } from '@/app/infra/entities/api';
 import PluginComponentList from '@/app/home/plugins/components/plugin-installed/PluginComponentList';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
+
+function PluginThumbnail({ author, name }: { author: string; name: string }) {
+  const iconURL = useInstalledPluginIcon(author, name);
+
+  if (!iconURL) {
+    return (
+      <div className="w-10 h-10 rounded-lg border bg-muted flex items-center justify-center flex-shrink-0">
+        <Puzzle className="size-5 text-muted-foreground" />
+      </div>
+    );
+  }
+
+  return (
+    <img
+      src={iconURL}
+      alt={name}
+      className="w-10 h-10 rounded-lg border bg-muted object-cover flex-shrink-0"
+    />
+  );
+}
 
 function InfoTooltip({ label }: { label: string }) {
   return (
@@ -366,13 +395,9 @@ export default function PipelineExtension({
                     className="flex items-center justify-between rounded-lg border p-3 hover:bg-accent"
                   >
                     <div className="flex-1 flex items-center gap-3">
-                      <img
-                        src={backendClient.getPluginIconURL(
-                          metadata.author || '',
-                          metadata.name,
-                        )}
-                        alt={metadata.name}
-                        className="w-10 h-10 rounded-lg border bg-muted object-cover flex-shrink-0"
+                      <PluginThumbnail
+                        author={metadata.author || ''}
+                        name={metadata.name}
                       />
                       <div className="flex-1">
                         <div className="font-medium">{metadata.name}</div>
@@ -653,13 +678,9 @@ export default function PipelineExtension({
                     onClick={() => handleTogglePlugin(pluginId)}
                   >
                     <Checkbox checked={isSelected} />
-                    <img
-                      src={backendClient.getPluginIconURL(
-                        metadata.author || '',
-                        metadata.name,
-                      )}
-                      alt={metadata.name}
-                      className="w-10 h-10 rounded-lg border bg-muted object-cover flex-shrink-0"
+                    <PluginThumbnail
+                      author={metadata.author || ''}
+                      name={metadata.name}
                     />
                     <div className="flex-1">
                       <div className="font-medium">{metadata.name}</div>
diff --git a/web/src/app/home/pipelines/components/pipeline-form/PipelineFormComponent.tsx b/web/src/app/home/pipelines/components/pipeline-form/PipelineFormComponent.tsx
index d27e9d9dc..309f33bb1 100644
--- a/web/src/app/home/pipelines/components/pipeline-form/PipelineFormComponent.tsx
+++ b/web/src/app/home/pipelines/components/pipeline-form/PipelineFormComponent.tsx
@@ -1,3 +1,4 @@
+import { useSnapshotDirty } from '@/app/infra/hooks/useSnapshotDirty';
 import GuidedTour, {
   type GuidedTourStep,
 } from '@/app/home/components/guided-tour/GuidedTour';
@@ -25,7 +26,7 @@ import DynamicFormComponent from '@/app/home/components/dynamic-form/DynamicForm
 import { getDefaultValues } from '@/app/home/components/dynamic-form/DynamicFormItemConfig';
 import { Button } from '@/components/ui/button';
 import { Tabs, TabsList, TabsTrigger } from '@/components/ui/tabs';
-import { useForm } from 'react-hook-form';
+import { useForm, useWatch } from 'react-hook-form';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { z } from 'zod';
 import { Input } from '@/components/ui/input';
@@ -324,6 +325,13 @@ const PipelineFormComponent = forwardRef<
       output: {},
     },
   });
+  useWatch({
+    control: form.control,
+    compute: (values) => [
+      values.ai?.runner?.id,
+      values.output?.misc?.['remove-think'],
+    ],
+  });
   const runnerInstallScope = `pipeline:${pipelineId || 'new'}`;
   const applyInstalledRunner = useCallback((installed: InstalledRunner) => {
     setAIConfigTabSchema(installed.configTab);
@@ -337,15 +345,11 @@ const PipelineFormComponent = forwardRef<
   const savedSnapshotRef = useRef<string>('');
   // Track which dynamic form stages have completed their initial mount emission.
   const initializedStagesRef = useRef<Set<string>>(new Set());
-  const watchedValues = form.watch();
-  const hasUnsavedChanges = (() => {
-    if (!isEditMode || !savedSnapshotRef.current) return false;
-    return JSON.stringify(watchedValues) !== savedSnapshotRef.current;
-  })();
-  // Keep a ref so that non-reactive callbacks (handleDynamicFormEmit) can
-  // read the latest dirty state without stale closures.
-  const hasUnsavedChangesRef = useRef(hasUnsavedChanges);
-  hasUnsavedChangesRef.current = hasUnsavedChanges;
+  const {
+    isDirty: hasUnsavedChanges,
+    dirtyRef: hasUnsavedChangesRef,
+    refreshDirty,
+  } = useSnapshotDirty(form, savedSnapshotRef, isEditMode);
 
   // Notify parent when dirty state changes
   useEffect(() => {
@@ -405,6 +409,7 @@ const PipelineFormComponent = forwardRef<
           };
           form.reset(loadedValues);
           savedSnapshotRef.current = JSON.stringify(loadedValues);
+          refreshDirty();
           initializedStagesRef.current.clear();
           setPipelineLoaded(true);
         })
@@ -415,7 +420,14 @@ const PipelineFormComponent = forwardRef<
     return () => {
       cancelled = true;
     };
-  }, [form, isEditMode, pipelineId, loadAttempt, onLegacyPipeline]);
+  }, [
+    form,
+    isEditMode,
+    pipelineId,
+    loadAttempt,
+    onLegacyPipeline,
+    refreshDirty,
+  ]);
 
   useEffect(() => {
     if (
@@ -538,6 +550,7 @@ const PipelineFormComponent = forwardRef<
     try {
       await httpClient.updatePipeline(pipelineId || '', pipeline);
       savedSnapshotRef.current = submittedSnapshot;
+      refreshDirty();
       onFinish();
       toast.success(t('pipelines.saveSuccess'));
       return true;
@@ -572,6 +585,7 @@ const PipelineFormComponent = forwardRef<
           emoji: values.emoji || '⚙️',
         };
         savedSnapshotRef.current = JSON.stringify(snapshot);
+        refreshDirty();
       }
     },
     async save() {
@@ -595,6 +609,7 @@ const PipelineFormComponent = forwardRef<
     stageName: string,
     values: object,
   ) {
+    const wasDirty = hasUnsavedChangesRef.current;
     const stageKey = `${String(formName)}.${stageName}`;
     const isFirstEmission =
       !initializedStagesRef.current.has(stageKey) &&
@@ -665,8 +680,9 @@ const PipelineFormComponent = forwardRef<
       // runner), the snapshot must remain at the last-saved state so that
       // hasUnsavedChanges stays true.
       const currentSnapshot = JSON.stringify(form.getValues());
-      if (savedSnapshotRef.current === '' || !hasUnsavedChangesRef.current) {
+      if (savedSnapshotRef.current === '' || !wasDirty) {
         savedSnapshotRef.current = currentSnapshot;
+        refreshDirty();
       }
     }
   }
@@ -698,7 +714,7 @@ const PipelineFormComponent = forwardRef<
   ) {
     // Special handling for AI config section
     if (formName === 'ai') {
-      const runnerConfig = (form.watch('ai.runner') as any) || {};
+      const runnerConfig = (form.getValues('ai.runner') as any) || {};
       const currentRunner = runnerConfig.id;
 
       // If this is the runner selector stage, render it directly
@@ -717,7 +733,7 @@ const PipelineFormComponent = forwardRef<
               <DynamicFormComponent
                 itemConfigList={stage.config}
                 initialValues={
-                  (form.watch(formName) as Record<string, unknown>)?.[
+                  (form.getValues(formName) as Record<string, unknown>)?.[
                     stage.name
                   ] || {}
                 }
@@ -752,7 +768,7 @@ const PipelineFormComponent = forwardRef<
       const isPluginRunner =
         currentRunner && currentRunner.startsWith('plugin:');
       if (isPluginRunner) {
-        const runnerConfigs = (form.watch('ai.runner_config') as any) || {};
+        const runnerConfigs = (form.getValues('ai.runner_config') as any) || {};
         const stageInitialValues = runnerConfigs[stage.name] || {};
         return (
           <Card key={stage.name}>
@@ -784,7 +800,7 @@ const PipelineFormComponent = forwardRef<
         ? PersistedRunnerForm
         : DynamicFormComponent;
     const stageInitialValues: Record<string, any> =
-      (form.watch(formName) as Record<string, any>)?.[stage.name] || {};
+      (form.getValues(formName) as Record<string, any>)?.[stage.name] || {};
 
     return (
       <Card key={stage.name}>
diff --git a/web/src/app/home/pipelines/pipeline-migration-issues.ts b/web/src/app/home/pipelines/pipeline-migration-issues.ts
index 3cc925d12..aebeedabe 100644
--- a/web/src/app/home/pipelines/pipeline-migration-issues.ts
+++ b/web/src/app/home/pipelines/pipeline-migration-issues.ts
@@ -31,6 +31,21 @@ const issueKeys: Record<string, string> = {
   runner_schema_incompatible: 'schemaChanged',
   plugin_version_incompatible: 'pluginVersion',
   'runtime.pending_interaction': 'pendingInteraction',
+  'legacy.custom_runner_unsupported': 'customRunner',
+  'migration.unknown_field_dropped': 'fieldDropped',
+  'migration.unknown_section_archived': 'fieldDropped',
+  'migration.value_coerced': 'valueRepaired',
+  'migration.invalid_value_defaulted': 'valueRepaired',
+  'migration.missing_default_filled': 'valueRepaired',
+  'migration.alias_conflict_resolved': 'aliasRepaired',
+  'local.prompt_normalized': 'valueRepaired',
+  'local.reasoning_dropped': 'valueRepaired',
+  'local.box_template_defaulted': 'valueRepaired',
+  'langflow.tweaks_invalid_defaulted': 'valueRepaired',
+  'migration.mixed_runner_selection': 'mixedRunner',
+  'migration.runner_config_replaced': 'runnerConfigReplaced',
+  'extensions.runner_added': 'runnerAdded',
+  invalid_extension_preferences: 'extensionPrefs',
 };
 
 /** Only locally reviewed codes may select localized copy. */
diff --git a/web/src/app/home/plugin-pages/page.tsx b/web/src/app/home/plugin-pages/page.tsx
index 520240791..3e4029973 100644
--- a/web/src/app/home/plugin-pages/page.tsx
+++ b/web/src/app/home/plugin-pages/page.tsx
@@ -1,4 +1,5 @@
 import EntityLoadState from '@/components/EntityLoadState';
+import LoadErrorState from '@/components/LoadErrorState';
 import { useSearchParams } from 'react-router-dom';
 import { httpClient } from '@/app/infra/http/HttpClient';
 import { useEffect, useRef, useState, useCallback } from 'react';
@@ -202,9 +203,11 @@ function PluginPageIframe({
   return (
     <div className="flex flex-col h-full w-full">
       {assetError ? (
-        <div className="flex items-center justify-center h-full text-muted-foreground">
-          {t('plugins.loadFailed')}
-        </div>
+        <LoadErrorState
+          title={t('plugins.loadFailed')}
+          className="flex-1"
+          onRetry={() => window.location.reload()}
+        />
       ) : loading || !assetUrl ? (
         <EntityLoadState />
       ) : null}
diff --git a/web/src/app/home/plugins/components/plugin-installed/ExtensionCardComponent.tsx b/web/src/app/home/plugins/components/plugin-installed/ExtensionCardComponent.tsx
index e3c030091..2add99f3e 100644
--- a/web/src/app/home/plugins/components/plugin-installed/ExtensionCardComponent.tsx
+++ b/web/src/app/home/plugins/components/plugin-installed/ExtensionCardComponent.tsx
@@ -13,7 +13,7 @@ import {
   Puzzle,
 } from 'lucide-react';
 import { getCloudServiceClientSync, systemInfo } from '@/app/infra/http';
-import { useAuthenticatedPluginIcon } from '@/hooks/useAuthenticatedPluginResource';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 import { Button } from '@/components/ui/button';
 import { Card } from '@/components/ui/card';
 import {
@@ -39,10 +39,9 @@ export default function ExtensionCardComponent({
   const { t } = useTranslation();
   const [dropdownOpen, setDropdownOpen] = useState(false);
   const [iconFailed, setIconFailed] = useState(false);
-  const authenticatedIcon = useAuthenticatedPluginIcon(
-    cardVO.author,
-    cardVO.name,
-    cardVO.type === 'plugin',
+  const pluginIconURL = useInstalledPluginIcon(
+    cardVO.type === 'plugin' ? cardVO.author : null,
+    cardVO.type === 'plugin' ? cardVO.name : null,
   );
 
   const FallbackIcon =
@@ -51,9 +50,8 @@ export default function ExtensionCardComponent({
       : cardVO.type === 'skill'
         ? Sparkles
         : Puzzle;
-  const iconSrc =
-    cardVO.type === 'plugin' ? authenticatedIcon.url : cardVO.iconURL;
-  const showFallback = iconFailed || authenticatedIcon.error || !iconSrc;
+  const iconSrc = cardVO.type === 'plugin' ? pluginIconURL : cardVO.iconURL;
+  const showFallback = iconFailed || !iconSrc;
 
   const getTypeLabel = (type: ExtensionType) => {
     switch (type) {
diff --git a/web/src/app/home/plugins/components/plugin-installed/PluginInstalledComponent.tsx b/web/src/app/home/plugins/components/plugin-installed/PluginInstalledComponent.tsx
index 76e7ad956..be575d9b2 100644
--- a/web/src/app/home/plugins/components/plugin-installed/PluginInstalledComponent.tsx
+++ b/web/src/app/home/plugins/components/plugin-installed/PluginInstalledComponent.tsx
@@ -157,7 +157,9 @@ const PluginInstalledComponent = forwardRef<
               version: meta.version ?? '',
               enabled: plugin.enabled,
               type: marketplacePlugin?.type || 'plugin',
-              iconURL: httpClient.getPluginIconURL(author, name),
+              // Icon is resolved by ExtensionCardComponent through the shared
+              // installed-plugin hook (authenticated fetch in multi-Workspace
+              // deployments); the public /icon route 404s there.
               install_source: plugin.install_source,
               install_info: plugin.install_info,
               status: plugin.status,
diff --git a/web/src/app/home/plugins/components/plugin-installed/plugin-card/PluginCardComponent.tsx b/web/src/app/home/plugins/components/plugin-installed/plugin-card/PluginCardComponent.tsx
index e5e8bf14c..f53b73db7 100644
--- a/web/src/app/home/plugins/components/plugin-installed/plugin-card/PluginCardComponent.tsx
+++ b/web/src/app/home/plugins/components/plugin-installed/plugin-card/PluginCardComponent.tsx
@@ -2,9 +2,16 @@ import { PluginCardVO } from '@/app/home/plugins/components/plugin-installed/Plu
 import { useState } from 'react';
 import { Badge } from '@/components/ui/badge';
 import { useTranslation } from 'react-i18next';
-import { BugIcon, ExternalLink, Ellipsis, Trash, ArrowUp } from 'lucide-react';
+import {
+  BugIcon,
+  ExternalLink,
+  Ellipsis,
+  Trash,
+  ArrowUp,
+  Puzzle,
+} from 'lucide-react';
 import { getCloudServiceClientSync, systemInfo } from '@/app/infra/http';
-import { httpClient } from '@/app/infra/http/HttpClient';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 import { Button } from '@/components/ui/button';
 import {
   DropdownMenu,
@@ -27,6 +34,7 @@ export default function PluginCardComponent({
 }) {
   const { t } = useTranslation();
   const [dropdownOpen, setDropdownOpen] = useState(false);
+  const iconURL = useInstalledPluginIcon(cardVO.author, cardVO.name);
 
   return (
     <>
@@ -36,11 +44,17 @@ export default function PluginCardComponent({
       >
         <div className="w-full h-full flex flex-row items-start justify-start gap-[1.2rem]">
           {/* Icon - fixed width */}
-          <img
-            src={httpClient.getPluginIconURL(cardVO.author, cardVO.name)}
-            alt="plugin icon"
-            className="w-16 h-16 rounded-[8%] flex-shrink-0"
-          />
+          {iconURL ? (
+            <img
+              src={iconURL}
+              alt="plugin icon"
+              className="w-16 h-16 rounded-[8%] flex-shrink-0"
+            />
+          ) : (
+            <div className="w-16 h-16 flex-shrink-0 flex items-center justify-center">
+              <Puzzle className="w-12 h-12 text-blue-500" />
+            </div>
+          )}
 
           {/* Content area - flexible width with min-width to prevent overflow */}
           <div className="flex-1 min-w-0 h-full flex flex-col items-start justify-between gap-[0.6rem]">
diff --git a/web/src/app/home/plugins/components/plugin-market/PluginMarketComponent.tsx b/web/src/app/home/plugins/components/plugin-market/PluginMarketComponent.tsx
index 1dd640e0e..92520d8b2 100644
--- a/web/src/app/home/plugins/components/plugin-market/PluginMarketComponent.tsx
+++ b/web/src/app/home/plugins/components/plugin-market/PluginMarketComponent.tsx
@@ -298,6 +298,7 @@ function MarketPageContent({
         iconURL,
         githubURL: plugin.repository,
         version: plugin.latest_version,
+        certificationStatus: plugin.certification_status,
         components: plugin.components || {},
         tags: plugin.tags || [],
         type: plugin.type,
diff --git a/web/src/app/home/plugins/components/plugin-market/RecommendationLists.tsx b/web/src/app/home/plugins/components/plugin-market/RecommendationLists.tsx
index 6f8dbfe11..a1861312d 100644
--- a/web/src/app/home/plugins/components/plugin-market/RecommendationLists.tsx
+++ b/web/src/app/home/plugins/components/plugin-market/RecommendationLists.tsx
@@ -57,6 +57,7 @@ function pluginToVO(
     iconURL,
     githubURL: plugin.repository,
     version: plugin.latest_version,
+    certificationStatus: plugin.certification_status,
     components: plugin.components,
     tags: plugin.tags || [],
     type: plugin.type,
diff --git a/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardComponent.tsx b/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardComponent.tsx
index c8393ac1b..cddd29c64 100644
--- a/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardComponent.tsx
+++ b/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardComponent.tsx
@@ -1,3 +1,4 @@
+import { CertifiedPluginBadge } from '@/components/market/CertifiedPluginBadge';
 import { PluginMarketCardVO } from './PluginMarketCardVO';
 import { useRef, useState, useEffect } from 'react';
 import { useTranslation } from 'react-i18next';
@@ -276,6 +277,9 @@ export default function PluginMarketCardComponent({
                     <span className="truncate">{typeLabel}</span>
                   </Badge>
                 )}
+                {cardVO.certificationStatus === 'issued' && (
+                  <CertifiedPluginBadge compact />
+                )}
               </div>
             </div>
 
diff --git a/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardVO.ts b/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardVO.ts
index 4816d5d3a..c71e2d482 100644
--- a/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardVO.ts
+++ b/web/src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardVO.ts
@@ -9,6 +9,7 @@ export interface IPluginMarketCardVO {
   iconURL: string;
   githubURL: string;
   version: string;
+  certificationStatus?: string;
   components?: Record<string, number>;
   tags?: string[];
   type?: 'plugin' | 'mcp' | 'skill';
@@ -29,6 +30,7 @@ export class PluginMarketCardVO implements IPluginMarketCardVO {
   installCount: number;
   likeCount: number;
   version: string;
+  certificationStatus?: string;
   components?: Record<string, number>;
   tags?: string[];
   type?: 'plugin' | 'mcp' | 'skill';
@@ -46,6 +48,7 @@ export class PluginMarketCardVO implements IPluginMarketCardVO {
     this.likeCount = prop.likeCount ?? 0;
     this.pluginId = prop.pluginId;
     this.version = prop.version;
+    this.certificationStatus = prop.certificationStatus;
     this.components = prop.components;
     this.tags = prop.tags;
     this.type = prop.type;
diff --git a/web/src/app/home/plugins/page.tsx b/web/src/app/home/plugins/page.tsx
index 0cf603042..c76b8eac6 100644
--- a/web/src/app/home/plugins/page.tsx
+++ b/web/src/app/home/plugins/page.tsx
@@ -17,6 +17,7 @@ import {
   PopoverTrigger,
 } from '@/components/ui/popover';
 import { Alert, AlertTitle, AlertDescription } from '@/components/ui/alert';
+import LoadErrorState from '@/components/LoadErrorState';
 import { Skeleton } from '@/components/ui/skeleton';
 import { Input } from '@/components/ui/input';
 import React, { useState, useRef, useEffect } from 'react';
@@ -54,6 +55,8 @@ function PluginListView() {
   const [pluginSystemStatus, setPluginSystemStatus] =
     useState<ApiRespPluginSystemStatus | null>(null);
   const [statusLoading, setStatusLoading] = useState(true);
+  const [statusError, setStatusError] = useState(false);
+  const [statusRevision, setStatusRevision] = useState(0);
   const [debugInfo, setDebugInfo] = useState<{
     debug_url: string;
     plugin_debug_key: string;
@@ -70,10 +73,12 @@ function PluginListView() {
     const fetchPluginSystemStatus = async () => {
       try {
         setStatusLoading(true);
+        setStatusError(false);
         const status = await httpClient.getPluginSystemStatus();
         setPluginSystemStatus(status);
       } catch (error) {
         console.error('Failed to fetch plugin system status:', error);
+        setStatusError(true);
         toast.error(t('plugins.failedToGetStatus'));
       } finally {
         setStatusLoading(false);
@@ -81,7 +86,7 @@ function PluginListView() {
     };
 
     void fetchPluginSystemStatus();
-  }, [t]);
+  }, [t, statusRevision]);
 
   useEffect(() => {
     const onComplete = (
@@ -141,13 +146,13 @@ function PluginListView() {
   );
 
   const renderPluginConnectionErrorState = () => (
-    <div className="flex justify-center pt-[10vh] px-4">
-      <Alert variant="destructive" className="max-w-md">
-        <Unlink />
-        <AlertTitle>{t('plugins.connectionError')}</AlertTitle>
-        <AlertDescription>{t('plugins.connectionErrorDesc')}</AlertDescription>
-      </Alert>
-    </div>
+    <LoadErrorState
+      className="pt-[10vh]"
+      icon={<Unlink className="size-5" aria-hidden="true" />}
+      title={t('plugins.connectionError')}
+      description={t('plugins.connectionErrorDesc')}
+      onRetry={() => setStatusRevision((value) => value + 1)}
+    />
   );
 
   const renderLoadingState = () => (
@@ -162,6 +167,15 @@ function PluginListView() {
     return renderLoadingState();
   }
 
+  if (statusError) {
+    return (
+      <LoadErrorState
+        title={t('plugins.failedToGetStatus')}
+        onRetry={() => setStatusRevision((value) => value + 1)}
+      />
+    );
+  }
+
   if (!pluginSystemStatus?.is_enable) {
     return renderPluginDisabledState();
   }
diff --git a/web/src/app/home/skills/components/skill-form/SkillForm.tsx b/web/src/app/home/skills/components/skill-form/SkillForm.tsx
index 8df14b163..c4455d44d 100644
--- a/web/src/app/home/skills/components/skill-form/SkillForm.tsx
+++ b/web/src/app/home/skills/components/skill-form/SkillForm.tsx
@@ -829,6 +829,7 @@ export default function SkillForm({
           ...baseSkillData,
           base_revision: skill.revision || '',
         });
+        setSkill(resp.skill);
         toast.success(t('skills.saveSuccess'));
         onSkillUpdated(resp.skill.name);
       } else {
diff --git a/web/src/app/infra/entities/api/index.ts b/web/src/app/infra/entities/api/index.ts
index 664bc21f2..a91ff09b0 100644
--- a/web/src/app/infra/entities/api/index.ts
+++ b/web/src/app/infra/entities/api/index.ts
@@ -566,6 +566,8 @@ export interface ApiRespSystemInfo {
   debug: boolean;
   version: string;
   edition: string;
+  beta: boolean;
+  deployment_mode: 'cloud' | 'oss';
   /** Independent instance-level gate for local stdio MCP transports. */
   mcp_stdio_enabled: boolean;
   cloud_service_url: string;
@@ -949,3 +951,20 @@ export interface ApiRespSkills {
 export interface ApiRespSkill {
   skill: Skill;
 }
+
+export type {
+  ExecutionSource,
+  ExecutionSourceFilter,
+  ExecutionModeFilter,
+  ExecutionUsage,
+  ExecutionRow,
+  ExecutionOverview,
+  TokenCoverage,
+  ExecutionSummary,
+  ExecutionListResult,
+  PipelineMessageRow,
+  ExecutionCallRow,
+  AgentExecutionDetail,
+  PipelineExecutionDetail,
+  ExecutionDetail,
+} from './monitoring-executions';
diff --git a/web/src/app/infra/entities/api/monitoring-executions.ts b/web/src/app/infra/entities/api/monitoring-executions.ts
new file mode 100644
index 000000000..04d042dca
--- /dev/null
+++ b/web/src/app/infra/entities/api/monitoring-executions.ts
@@ -0,0 +1,174 @@
+export type ExecutionSource = 'agent' | 'pipeline' | 'event';
+export type ExecutionSourceFilter = 'all' | ExecutionSource;
+export type ExecutionModeFilter = 'all' | 'real' | 'debug';
+
+export interface ExecutionUsage {
+  input_tokens?: number | null;
+  output_tokens?: number | null;
+  total_tokens?: number | null;
+}
+
+export interface ExecutionRow {
+  source: ExecutionSource;
+  id: string;
+  event_id: string | null;
+  status: string;
+  status_group:
+    | 'completed'
+    | 'failed'
+    | 'running'
+    | 'queued'
+    | 'cancelled'
+    | 'ignored'
+    | string;
+  title: string;
+  input_preview?: string;
+  event_type?: string;
+  status_reason?: string | null;
+  target_kind: 'agent' | 'pipeline' | 'processor' | string;
+  target_id: string | null;
+  target_name: string | null;
+  bot_id: string | null;
+  bot_name: string | null;
+  pipeline_id: string | null;
+  pipeline_name: string | null;
+  runner_id: string | null;
+  conversation_id: string | null;
+  session_id: string | null;
+  created_at_ms: number | null;
+  started_at_ms: number | null;
+  finished_at_ms: number | null;
+  duration_ms: number | null;
+  usage: ExecutionUsage | null;
+  cost: Record<string, unknown> | null;
+  queue_name: string | null;
+  debug: boolean;
+  has_error: boolean;
+  platform?: string | null;
+  user_id?: string | null;
+  user_name?: string | null;
+}
+
+export interface ExecutionOverview {
+  total: number;
+  by_source: { agent: number; pipeline: number };
+  by_status: Record<string, number>;
+  completed: number;
+  failed: number;
+  running: number;
+  queued: number;
+  cancelled: number;
+  ignored: number;
+  waiting: number;
+  success_rate: number | null;
+  denominator: number;
+  p50_duration_ms: number | null;
+  p95_duration_ms: number | null;
+  duration_sample: number;
+  debug: number;
+  real: number;
+}
+
+export interface TokenCoverage {
+  total_tokens: number;
+  calls: number;
+  calls_with_usage: number;
+}
+
+export interface ExecutionSummary {
+  executions: ExecutionOverview;
+  tokens: TokenCoverage;
+}
+
+export interface ExecutionListResult {
+  items: ExecutionRow[];
+  total: number;
+  limit: number;
+  offset: number;
+  has_more: boolean;
+  summary: ExecutionSummary;
+}
+
+export interface PipelineMessageRow {
+  id: string;
+  timestamp: string;
+  bot_id: string;
+  bot_name: string;
+  pipeline_id: string;
+  pipeline_name: string;
+  message_content: string;
+  session_id: string;
+  status: string;
+  level: string;
+  platform?: string | null;
+  user_id?: string | null;
+  user_name?: string | null;
+  runner_name?: string | null;
+  role?: string | null;
+}
+
+export interface ExecutionCallRow {
+  id: string;
+  timestamp: string;
+  /** LLM/tool calls report a status; error records carry `error_type` instead. */
+  status?: string | null;
+  duration?: number | null;
+  error_type?: string | null;
+  error_message?: string | null;
+  [key: string]: unknown;
+}
+
+export type ExecutionSection =
+  | 'inputs'
+  | 'outputs'
+  | 'deliveries'
+  | 'conversation'
+  | 'related'
+  | 'events'
+  | 'llm_calls'
+  | 'tool_calls'
+  | 'errors';
+
+export interface ExecutionDetailItem {
+  id: string | number;
+  timestamp_ms?: number | null;
+  role?: string;
+  origin?: string;
+  event_type?: string;
+  actor_name?: string;
+  actor_id?: string;
+  content?: unknown;
+  attachments?: unknown[];
+  status?: string;
+  type?: string;
+  sequence?: number;
+  data?: Record<string, unknown>;
+  metadata?: Record<string, unknown>;
+  delivery?: Record<string, unknown>;
+  [key: string]: unknown;
+}
+
+export interface ExecutionDetailPage {
+  items: ExecutionDetailItem[];
+  has_more: boolean;
+  next_offset: number;
+}
+
+export interface ExecutionDetail {
+  source: ExecutionSource;
+  row: ExecutionRow;
+  pages: Partial<Record<ExecutionSection, ExecutionDetailPage>>;
+  legacy_context: boolean;
+}
+
+export type AgentExecutionDetail = ExecutionDetail & { source: 'agent' };
+export type PipelineExecutionDetail = ExecutionDetail & { source: 'pipeline' };
+
+export interface InflightSnapshot {
+  items: (ExecutionRow & { progress_event?: string })[];
+  active_total: number;
+  truncated: boolean;
+}
+export type InflightFrame =
+  | { kind: 'snapshot'; data: InflightSnapshot }
+  | { kind: 'heartbeat' | 'unavailable' };
diff --git a/web/src/app/infra/entities/api/pipeline-migration.ts b/web/src/app/infra/entities/api/pipeline-migration.ts
index b2a103c25..620575359 100644
--- a/web/src/app/infra/entities/api/pipeline-migration.ts
+++ b/web/src/app/infra/entities/api/pipeline-migration.ts
@@ -4,6 +4,7 @@ export type PipelineMigrationState =
   | 'blocked'
   | 'already_current'
   | 'not_legacy'
+  | 'unsupported_runner'
   | 'activation_pending';
 
 export interface PipelineMigrationIssue {
diff --git a/web/src/app/infra/entities/operation-log.ts b/web/src/app/infra/entities/operation-log.ts
index 651b2f957..f9b0b618a 100644
--- a/web/src/app/infra/entities/operation-log.ts
+++ b/web/src/app/infra/entities/operation-log.ts
@@ -73,10 +73,7 @@ export interface OperationLogRecord {
 
 /** Verification filter accepted by the listing endpoint. */
 export type OperationIntegrityFilter =
-  | 'all'
-  | 'issues'
-  | 'hash_mismatch'
-  | 'chain_broken';
+  'all' | 'issues' | 'hash_mismatch' | 'chain_broken';
 
 export interface OperationLogPage {
   records: OperationLogRecord[];
diff --git a/web/src/app/infra/entities/plugin/index.ts b/web/src/app/infra/entities/plugin/index.ts
index 64f19396e..8ade9b80f 100644
--- a/web/src/app/infra/entities/plugin/index.ts
+++ b/web/src/app/infra/entities/plugin/index.ts
@@ -36,6 +36,7 @@ export enum PluginV4Status {
 }
 
 export interface PluginV4 {
+  certification_status?: string;
   id: number;
   plugin_id: string;
   mcp_id?: string;
diff --git a/web/src/app/infra/entities/workspace.ts b/web/src/app/infra/entities/workspace.ts
index c369961a2..f5b720558 100644
--- a/web/src/app/infra/entities/workspace.ts
+++ b/web/src/app/infra/entities/workspace.ts
@@ -1,9 +1,5 @@
 export type WorkspaceRole =
-  | 'owner'
-  | 'admin'
-  | 'developer'
-  | 'operator'
-  | 'viewer';
+  'owner' | 'admin' | 'developer' | 'operator' | 'viewer';
 
 export interface Workspace {
   uuid: string;
@@ -28,12 +24,21 @@ export interface WorkspaceMembership {
 }
 
 export interface CurrentWorkspace {
+  current_plan?: {
+    subscription_started_at?: number | null;
+    name: string;
+    revision: number;
+    limits: Record<string, number>;
+    entitlement_features: Record<string, boolean>;
+    monthly_bonus_credits: number;
+  } | null;
   workspace: Workspace;
   membership: WorkspaceMembership;
   permissions: string[];
   placement_generation: number;
   /** Signed Cloud display metadata; never used for client-side authorization. */
   plan_name?: string | null;
+  limitation?: Record<string, number> | null;
 }
 
 export interface WorkspaceSpaceBilling {
diff --git a/web/src/app/infra/hooks/useInstalledPluginIcon.ts b/web/src/app/infra/hooks/useInstalledPluginIcon.ts
new file mode 100644
index 000000000..5d8fad696
--- /dev/null
+++ b/web/src/app/infra/hooks/useInstalledPluginIcon.ts
@@ -0,0 +1,118 @@
+import { useEffect, useState } from 'react';
+
+import { httpClient } from '@/app/infra/http';
+
+/**
+ * Resolve the icon of an *installed* plugin in every deployment.
+ *
+ * The Core's public `/icon` route only exists in single-workspace (OSS)
+ * installs; in multi-Workspace deployments it 404s, so installed-plugin
+ * surfaces must go through the authenticated `/authenticated-icon` route and
+ * stream the bytes as a blob object URL.
+ *
+ * - Core backend client: blob-fetch `getAuthenticatedPluginIconURL` and cache
+ *   the resulting object URL per `author/name` (module-level, shared across
+ *   every mounted consumer of the same icon).
+ * - Cloud/Space marketplace surfaces: callers that already resolved a
+ *   marketplace URL pass it as `marketplaceURL`; it is returned verbatim and
+ *   no blob fetch happens.
+ * - Returns `null` while loading / when the plugin has no icon, so callers
+ *   keep rendering their existing fallback.
+ *
+ * Object URLs are revoked once the last subscriber unmounts (or the last
+ * subscriber's fetch settles with no remaining subscriber).
+ */
+
+type CachedIcon = { url: string };
+
+// Resolved object URLs, keyed by `author/name`.
+const iconCache = new Map<string, CachedIcon>();
+// In-flight authenticated fetches, de-duplicated per key.
+const inflightIcons = new Map<string, Promise<string | null>>();
+// Mounted subscriber count per key, so we only revoke when nobody needs it.
+const iconRefCounts = new Map<string, number>();
+
+function loadIcon(
+  key: string,
+  author: string,
+  name: string,
+): Promise<string | null> {
+  const existing = inflightIcons.get(key);
+  if (existing) return existing;
+
+  const request = httpClient
+    .getAuthenticatedPluginIconURL(author, name)
+    .catch(() => null)
+    .finally(() => {
+      inflightIcons.delete(key);
+    });
+  inflightIcons.set(key, request);
+  return request;
+}
+
+export function useInstalledPluginIcon(
+  author?: string | null,
+  name?: string | null,
+  marketplaceURL?: string | null,
+): string | null {
+  const key = author && name ? `${author}/${name}` : null;
+  const [url, setUrl] = useState<string | null>(
+    () => marketplaceURL ?? (key ? (iconCache.get(key)?.url ?? null) : null),
+  );
+
+  useEffect(() => {
+    if (marketplaceURL) {
+      // Cloud/Space marketplace surface: already resolved upstream.
+      setUrl(marketplaceURL);
+      return;
+    }
+    if (!key || !author || !name) {
+      setUrl(null);
+      return;
+    }
+
+    let active = true;
+    iconRefCounts.set(key, (iconRefCounts.get(key) ?? 0) + 1);
+
+    const cached = iconCache.get(key);
+    if (cached) {
+      setUrl(cached.url);
+    } else {
+      setUrl(null);
+      void loadIcon(key, author, name).then((objectURL) => {
+        if (!objectURL) return;
+        // The last subscriber may have left while the fetch was in flight.
+        if (iconRefCounts.get(key) === undefined) {
+          URL.revokeObjectURL(objectURL);
+          return;
+        }
+        const current = iconCache.get(key);
+        if (current) {
+          // Another subscriber resolved first; share its object URL.
+          if (current.url !== objectURL) URL.revokeObjectURL(objectURL);
+          if (active) setUrl(current.url);
+          return;
+        }
+        iconCache.set(key, { url: objectURL });
+        if (active) setUrl(objectURL);
+      });
+    }
+
+    return () => {
+      active = false;
+      const remaining = (iconRefCounts.get(key) ?? 1) - 1;
+      if (remaining > 0) {
+        iconRefCounts.set(key, remaining);
+        return;
+      }
+      iconRefCounts.delete(key);
+      const entry = iconCache.get(key);
+      if (entry) {
+        iconCache.delete(key);
+        URL.revokeObjectURL(entry.url);
+      }
+    };
+  }, [key, author, name, marketplaceURL]);
+
+  return marketplaceURL ?? url;
+}
diff --git a/web/src/app/infra/hooks/useSnapshotDirty.ts b/web/src/app/infra/hooks/useSnapshotDirty.ts
new file mode 100644
index 000000000..e9065e523
--- /dev/null
+++ b/web/src/app/infra/hooks/useSnapshotDirty.ts
@@ -0,0 +1,32 @@
+import { useCallback, useEffect, useRef, useState } from 'react';
+import type { RefObject } from 'react';
+import type { FieldValues, UseFormReturn } from 'react-hook-form';
+
+/** Subscribe without rendering the form for every keystroke. Values remain synchronous. */
+export function useSnapshotDirty<T extends FieldValues>(
+  form: UseFormReturn<T>,
+  savedSnapshot: RefObject<string>,
+  enabled = true,
+) {
+  const [isDirty, setIsDirty] = useState(false);
+  const dirtyRef = useRef(false);
+  const refreshDirty = useCallback(() => {
+    const next =
+      enabled &&
+      !!savedSnapshot.current &&
+      JSON.stringify(form.getValues()) !== savedSnapshot.current;
+    if (next !== dirtyRef.current) {
+      dirtyRef.current = next;
+      setIsDirty(next);
+    }
+    return next;
+  }, [enabled, form, savedSnapshot]);
+
+  useEffect(() => {
+    refreshDirty();
+    const subscription = form.watch(refreshDirty);
+    return () => subscription.unsubscribe();
+  }, [form, refreshDirty]);
+
+  return { isDirty, dirtyRef, refreshDirty };
+}
diff --git a/web/src/app/infra/http/BackendClient.ts b/web/src/app/infra/http/BackendClient.ts
index d17327cf8..517b294d3 100644
--- a/web/src/app/infra/http/BackendClient.ts
+++ b/web/src/app/infra/http/BackendClient.ts
@@ -1,9 +1,18 @@
 import { BaseHttpClient, type RequestConfig } from './BaseHttpClient';
+import { notifyAfterModelConfigurationChange } from './modelConfigEvents';
 import type {
   PipelineMigrationPreview,
   PipelineMigrationRequest,
 } from '@/app/infra/entities/api/pipeline-migration';
 import type { DebugExecutionEvent } from '@/app/infra/entities/api/agent-debug';
+import type {
+  InflightFrame,
+  ExecutionDetail,
+  ExecutionListResult,
+  ExecutionModeFilter,
+  ExecutionSource,
+  ExecutionSourceFilter,
+} from '../entities/api/monitoring-executions';
 import type {
   CodexAuthStatus,
   CodexDeviceAuthorization,
@@ -102,6 +111,17 @@ import type {
  * 负责与后端 API 的所有交互
  */
 export class BackendClient extends BaseHttpClient {
+  public getWorkspaceLanguage(): Promise<{
+    language: string;
+    supported_languages: string[];
+  }> {
+    return this.get('/api/v1/workspaces/current/language');
+  }
+
+  public setWorkspaceLanguage(language: string): Promise<{ language: string }> {
+    return this.put('/api/v1/workspaces/current/language', { language });
+  }
+
   constructor(baseURL: string) {
     super(baseURL, false);
   }
@@ -141,19 +161,25 @@ export class BackendClient extends BaseHttpClient {
   public createModelProvider(
     provider: Omit<ModelProvider, 'uuid'>,
   ): Promise<{ uuid: string }> {
-    return this.post('/api/v1/provider/providers', provider);
+    return notifyAfterModelConfigurationChange(
+      this.post<{ uuid: string }>('/api/v1/provider/providers', provider),
+    );
   }
 
   public updateModelProvider(
     uuid: string,
     provider: Partial<ModelProvider>,
   ): Promise<object> {
-    return this.put(`/api/v1/provider/providers/${uuid}`, provider);
+    return notifyAfterModelConfigurationChange(
+      this.put<object>(`/api/v1/provider/providers/${uuid}`, provider),
+    );
   }
 
   public deleteModelProvider(uuid: string, cascade = false): Promise<object> {
-    return this.delete(
-      `/api/v1/provider/providers/${uuid}${cascade ? '?cascade=true' : ''}`,
+    return notifyAfterModelConfigurationChange(
+      this.delete<object>(
+        `/api/v1/provider/providers/${uuid}${cascade ? '?cascade=true' : ''}`,
+      ),
     );
   }
 
@@ -222,18 +248,24 @@ export class BackendClient extends BaseHttpClient {
   public createProviderLLMModel(
     model: Omit<LLMModel, 'uuid'>,
   ): Promise<{ uuid: string }> {
-    return this.post('/api/v1/provider/models/llm', model);
+    return notifyAfterModelConfigurationChange(
+      this.post<{ uuid: string }>('/api/v1/provider/models/llm', model),
+    );
   }
 
   public deleteProviderLLMModel(uuid: string): Promise<object> {
-    return this.delete(`/api/v1/provider/models/llm/${uuid}`);
+    return notifyAfterModelConfigurationChange(
+      this.delete<object>(`/api/v1/provider/models/llm/${uuid}`),
+    );
   }
 
   public updateProviderLLMModel(
     uuid: string,
     model: LLMModel,
   ): Promise<object> {
-    return this.put(`/api/v1/provider/models/llm/${uuid}`, model);
+    return notifyAfterModelConfigurationChange(
+      this.put<object>(`/api/v1/provider/models/llm/${uuid}`, model),
+    );
   }
 
   public testLLMModel(uuid: string, model: LLMModel): Promise<object> {
@@ -352,6 +384,18 @@ export class BackendClient extends BaseHttpClient {
     return this.post('/api/v1/agents', agent);
   }
 
+  public getStarredModel(): Promise<{ uuid: string | null }> {
+    return this.get('/api/v1/provider/models/llm/_/starred');
+  }
+
+  public setStarredModel(uuid: string | null): Promise<object> {
+    return this.put('/api/v1/provider/models/llm/_/starred', { uuid });
+  }
+
+  public getDefaultModel(): Promise<{ uuid: string | null }> {
+    return this.get('/api/v1/provider/models/llm/_/default');
+  }
+
   public updateAgent(uuid: string, agent: Partial<Agent>): Promise<object> {
     return this.put(`/api/v1/agents/${uuid}`, agent);
   }
@@ -432,6 +476,58 @@ export class BackendClient extends BaseHttpClient {
     return result;
   }
 
+  public async streamAssistant<T>(
+    uuid: string,
+    payload: {
+      revision: number;
+      text?: string;
+      approved?: boolean;
+      model_uuid?: string;
+    },
+    onEvent: (event: { kind: string; data: unknown }) => void,
+    signal: AbortSignal,
+  ): Promise<T> {
+    let offset = 0;
+    let result: T | undefined;
+    let failure: { code: string; msg: string } | undefined;
+    const consume = (text: string) => {
+      let end: number;
+      while ((end = text.indexOf('\n', offset)) !== -1) {
+        const line = text.slice(offset, end).trim();
+        offset = end + 1;
+        if (!line) continue;
+        const frame = JSON.parse(line);
+        if (frame.kind === 'completed') result = frame.data;
+        else if (frame.kind === 'error') failure = frame;
+        else onEvent(frame);
+      }
+    };
+    const response = await this.instance.post<string>(
+      `/api/v1/assistant/conversations/${encodeURIComponent(uuid)}/turn/stream`,
+      payload,
+      {
+        adapter: 'xhr',
+        responseType: 'text',
+        timeout: 0,
+        signal,
+        headers: { Accept: 'application/x-ndjson' },
+        transformResponse: [(data) => data],
+        onDownloadProgress: (progress) => {
+          const xhr = progress.event?.target as XMLHttpRequest | undefined;
+          if (xhr?.status === 200) consume(xhr.responseText);
+        },
+      },
+    );
+    consume(response.data);
+    if (failure) throw failure;
+    if (!result)
+      throw {
+        code: 'assistant_protocol_error',
+        msg: 'Assistant stream ended before completion',
+      };
+    return result;
+  }
+
   public getGeneralPipelineMetadata(): Promise<GetPipelineMetadataResponseData> {
     // as designed, this method will be deprecated, and only for developer to check the prefered config schema
     return this.get('/api/v1/pipelines/_/metadata');
@@ -737,6 +833,16 @@ export class BackendClient extends BaseHttpClient {
     return this.get(`/api/v1/monitoring/sessions?${queryParams.toString()}`);
   }
 
+  public resetSessionContext(
+    botId: string,
+    sessionId: string,
+  ): Promise<{ reset: boolean }> {
+    return this.post('/api/v1/monitoring/sessions/reset-context', {
+      bot_id: botId,
+      session_id: sessionId,
+    });
+  }
+
   public getSessionAnalysis<T>(
     sessionId: string,
     botId: string,
@@ -1973,6 +2079,8 @@ export class BackendClient extends BaseHttpClient {
 
   // ============ Monitoring API ============
   public getMonitoringData(params: {
+    mode?: ExecutionModeFilter;
+    status?: string[];
     botId?: string[];
     pipelineId?: string[];
     startTime?: string;
@@ -2096,6 +2204,8 @@ export class BackendClient extends BaseHttpClient {
     };
   }> {
     const queryParams = new URLSearchParams();
+    if (params.mode) queryParams.append('mode', params.mode);
+    params.status?.forEach((status) => queryParams.append('status', status));
     if (params.botId) {
       params.botId.forEach((id) => queryParams.append('botId', id));
     }
@@ -2115,6 +2225,104 @@ export class BackendClient extends BaseHttpClient {
     return this.get(`/api/v1/monitoring/data?${queryParams.toString()}`);
   }
 
+  public async streamInflight(
+    onFrame: (frame: InflightFrame) => void,
+    signal: AbortSignal,
+  ): Promise<void> {
+    const response = await this.instance.get<ReadableStream<Uint8Array>>(
+      '/api/v1/monitoring/in-flight/stream',
+      {
+        adapter: 'fetch',
+        responseType: 'stream',
+        timeout: 0,
+        signal,
+        headers: { Accept: 'text/event-stream' },
+      },
+    );
+    const reader = response.data.getReader();
+    const decoder = new TextDecoder();
+    let buffer = '';
+    try {
+      while (!signal.aborted) {
+        const { value, done } = await reader.read();
+        if (done) break;
+        buffer += decoder.decode(value, { stream: true });
+        let end: number;
+        while ((end = buffer.indexOf('\n\n')) !== -1) {
+          const frame = buffer.slice(0, end);
+          buffer = buffer.slice(end + 2);
+          if (frame.startsWith('data: '))
+            onFrame(JSON.parse(frame.slice(6)) as InflightFrame);
+        }
+        if (buffer.length > 2_000_000)
+          throw new Error('Monitoring frame exceeds limit');
+      }
+    } finally {
+      await reader.cancel().catch(() => undefined);
+      reader.releaseLock();
+    }
+  }
+
+  public getExecutions(params: {
+    botId?: string[];
+    pipelineId?: string[];
+    agentId?: string[];
+    status?: string[];
+    source?: ExecutionSourceFilter;
+    mode?: ExecutionModeFilter;
+    startTime?: string;
+    endTime?: string;
+    limit?: number;
+    offset?: number;
+  }): Promise<ExecutionListResult> {
+    const queryParams = new URLSearchParams();
+    if (params.botId) {
+      params.botId.forEach((id) => queryParams.append('botId', id));
+    }
+    if (params.pipelineId) {
+      params.pipelineId.forEach((id) => queryParams.append('pipelineId', id));
+    }
+    if (params.agentId) {
+      params.agentId.forEach((id) => queryParams.append('agentId', id));
+    }
+    if (params.status) {
+      params.status.forEach((status) => queryParams.append('status', status));
+    }
+    if (params.source) {
+      queryParams.append('source', params.source);
+    }
+    if (params.mode) {
+      queryParams.append('mode', params.mode);
+    }
+    if (params.startTime) {
+      queryParams.append('startTime', params.startTime);
+    }
+    if (params.endTime) {
+      queryParams.append('endTime', params.endTime);
+    }
+    if (params.limit) {
+      queryParams.append('limit', params.limit.toString());
+    }
+    if (params.offset) {
+      queryParams.append('offset', params.offset.toString());
+    }
+
+    return this.get(`/api/v1/monitoring/executions?${queryParams.toString()}`);
+  }
+
+  public getExecutionDetail(
+    source: ExecutionSource | 'auto',
+    executionId: string,
+    options: { section?: string; offset?: number; limit?: number } = {},
+  ): Promise<ExecutionDetail> {
+    const params = new URLSearchParams(
+      Object.entries(options).map(([key, value]) => [key, String(value)]),
+    );
+    return this.get(
+      `/api/v1/monitoring/executions/${encodeURIComponent(source)}/${encodeURIComponent(executionId)}?${params}`,
+    );
+  }
+
   public getMonitoringOverview(params: {
     botId?: string[];
     pipelineId?: string[];
@@ -2144,6 +2352,8 @@ export class BackendClient extends BaseHttpClient {
   }
 
   public getTokenStatistics(params: {
+    mode?: ExecutionModeFilter;
+    status?: string[];
     botId?: string[];
     pipelineId?: string[];
     startTime?: string;
@@ -2184,6 +2394,8 @@ export class BackendClient extends BaseHttpClient {
     bucket: string;
   }> {
     const queryParams = new URLSearchParams();
+    if (params.mode) queryParams.append('mode', params.mode);
+    params.status?.forEach((status) => queryParams.append('status', status));
     if (params.botId) {
       params.botId.forEach((id) => queryParams.append('botId', id));
     }
diff --git a/web/src/app/infra/http/BaseHttpClient.ts b/web/src/app/infra/http/BaseHttpClient.ts
index 0243c9a3c..d1c86022a 100644
--- a/web/src/app/infra/http/BaseHttpClient.ts
+++ b/web/src/app/infra/http/BaseHttpClient.ts
@@ -4,6 +4,7 @@ import axios, {
   AxiosResponse,
   AxiosError,
 } from 'axios';
+import i18n from 'i18next';
 import {
   clearActiveWorkspaceUuid,
   getActiveWorkspaceUuid,
@@ -79,6 +80,12 @@ export abstract class BaseHttpClient {
     // 请求拦截
     this.instance.interceptors.request.use(
       async (config) => {
+        // Account-only endpoints have no selected Workspace. Workspace routes
+        // resolve their persisted language and override this presentation hint.
+        // This is a LangBot backend hint, not part of the public Space API.
+        if (!this.disableToken) {
+          config.headers['X-LangBot-Language'] = i18n.language || 'en-US';
+        }
         // 客户端添加认证头
         if (typeof window !== 'undefined' && !this.disableToken) {
           const session = this.getSessionSync();
diff --git a/web/src/app/infra/http/index.ts b/web/src/app/infra/http/index.ts
index 0014bc272..e7fcf4d03 100644
--- a/web/src/app/infra/http/index.ts
+++ b/web/src/app/infra/http/index.ts
@@ -25,6 +25,8 @@ export const systemInfo: ApiRespSystemInfo = {
   debug: false,
   version: '',
   edition: 'community',
+  beta: false,
+  deployment_mode: 'oss',
   mcp_stdio_enabled: false,
   enable_marketplace: true,
   cloud_service_url: '',
diff --git a/web/src/app/infra/http/modelConfigEvents.ts b/web/src/app/infra/http/modelConfigEvents.ts
new file mode 100644
index 000000000..a8fbfb1c2
--- /dev/null
+++ b/web/src/app/infra/http/modelConfigEvents.ts
@@ -0,0 +1,12 @@
+export const MODEL_CONFIGURATION_CHANGED =
+  'langbot:model-configuration-changed';
+
+export async function notifyAfterModelConfigurationChange<T>(
+  request: Promise<T>,
+): Promise<T> {
+  const result = await request;
+  if (typeof window !== 'undefined') {
+    window.dispatchEvent(new Event(MODEL_CONFIGURATION_CHANGED));
+  }
+  return result;
+}
diff --git a/web/src/app/infra/websocket/WebSocketClient.ts b/web/src/app/infra/websocket/WebSocketClient.ts
index 9e930b723..2d79041af 100644
--- a/web/src/app/infra/websocket/WebSocketClient.ts
+++ b/web/src/app/infra/websocket/WebSocketClient.ts
@@ -3,6 +3,7 @@
  * 用于管理WebSocket连接和消息处理
  */
 import { getActiveWorkspaceUuid } from '@/app/infra/http/workspaceContext';
+import { backendWebSocketURL } from './url';
 import type { MessageChainComponent } from '@/app/infra/entities/message';
 
 export interface WebSocketMessage {
@@ -17,12 +18,7 @@ export interface WebSocketMessage {
 
 export interface WebSocketResponse {
   type:
-    | 'connected'
-    | 'response'
-    | 'user_message'
-    | 'pong'
-    | 'broadcast'
-    | 'error';
+    'connected' | 'response' | 'user_message' | 'pong' | 'broadcast' | 'error';
   connection_id?: string;
   pipeline_uuid?: string;
   session_type?: string;
@@ -43,6 +39,25 @@ export class WebSocketClient {
   private isConnecting = false; // 防止重复连接
   private shouldReconnect = true;
   private disconnectedByUser = false;
+  private workspaceUuid = getActiveWorkspaceUuid();
+  private handshakeTimeout: ReturnType<typeof setTimeout> | null = null;
+  private recoveryListenersAttached = false;
+  private recover = () => {
+    if (
+      document.visibilityState === 'hidden' ||
+      this.disconnectedByUser ||
+      !this.shouldReconnect ||
+      this.isConnecting ||
+      this.isConnected()
+    )
+      return;
+    if (getActiveWorkspaceUuid() !== this.workspaceUuid) return;
+    void this.connect().catch(() => {});
+  };
+  private clearHandshakeTimeout() {
+    if (this.handshakeTimeout) clearTimeout(this.handshakeTimeout);
+    this.handshakeTimeout = null;
+  }
 
   // 事件回调
   private onConnectedCallback?: (data: WebSocketResponse) => void;
@@ -89,17 +104,32 @@ export class WebSocketClient {
 
         this.isConnecting = true;
 
-        // 构建WebSocket URL
-        const protocol = window.location.protocol === 'https:' ? 'wss:' : 'ws:';
-        // extract host from import.meta.env.VITE_API_BASE_URL
-        // If env var is undefined, use current page host (for production)
-        const host =
-          import.meta.env.VITE_API_BASE_URL?.split('://')[1] ||
-          window.location.host;
-        const url = `${protocol}//${host}/api/v1/pipelines/${this.pipelineId}/ws/connect?session_type=${this.sessionType}`;
-
-        const socket = new WebSocket(url);
+        if (getActiveWorkspaceUuid() !== this.workspaceUuid) {
+          this.isConnecting = false;
+          this.shouldReconnect = false;
+          reject(new Error('Workspace changed'));
+          return;
+        }
+        const url = backendWebSocketURL(
+          import.meta.env.VITE_API_BASE_URL,
+          window.location.href,
+          `api/v1/pipelines/${encodeURIComponent(this.pipelineId)}/ws/connect`,
+        );
+        url.searchParams.set('session_type', this.sessionType);
+        const socket = new WebSocket(url.toString());
         this.ws = socket;
+        if (!this.recoveryListenersAttached) {
+          window.addEventListener('online', this.recover);
+          window.addEventListener('focus', this.recover);
+          document.addEventListener('visibilitychange', this.recover);
+          this.recoveryListenersAttached = true;
+        }
+        this.clearHandshakeTimeout();
+        this.handshakeTimeout = setTimeout(() => {
+          if (this.ws !== socket || this.connectionId) return;
+          reject(new Error('WebSocket authentication timed out'));
+          socket.close();
+        }, 15000);
 
         // 连接打开
         socket.onopen = () => {
@@ -107,9 +137,8 @@ export class WebSocketClient {
             socket.close();
             return;
           }
-          this.isConnecting = false;
           const token = this.token || localStorage.getItem('token');
-          const workspaceUuid = getActiveWorkspaceUuid();
+          const workspaceUuid = this.workspaceUuid;
           if (!token || !workspaceUuid) {
             const error = new Error('WebSocket认证信息缺失');
             this.shouldReconnect = false;
@@ -144,6 +173,8 @@ export class WebSocketClient {
               // Only a fully authenticated runtime connection should reset
               // the retry budget. Resetting on TCP open makes server-side
               // errors (for example a pipeline still loading) retry forever.
+              this.clearHandshakeTimeout();
+              this.isConnecting = false;
               this.reconnectAttempts = 0;
               this.connectionId = data.connection_id;
               this.startHeartbeat();
@@ -157,6 +188,9 @@ export class WebSocketClient {
 
         // 连接关闭
         socket.onclose = () => {
+          reject(new Error('WebSocket connection closed'));
+          if (this.ws !== socket) return;
+          this.clearHandshakeTimeout();
           if (this.ws === socket) {
             this.ws = null;
             this.connectionId = null;
@@ -169,9 +203,12 @@ export class WebSocketClient {
           // 自动重连
           if (
             this.shouldReconnect &&
-            this.reconnectAttempts < this.maxReconnectAttempts
+            getActiveWorkspaceUuid() === this.workspaceUuid
           ) {
-            this.reconnectAttempts++;
+            this.reconnectAttempts = Math.min(
+              this.reconnectAttempts + 1,
+              this.maxReconnectAttempts,
+            );
             this.reconnectTimeout = setTimeout(() => {
               this.reconnectTimeout = null;
               if (!this.shouldReconnect || this.disconnectedByUser) return;
@@ -320,6 +357,11 @@ export class WebSocketClient {
    */
   public disconnect() {
     this.disconnectedByUser = true;
+    this.clearHandshakeTimeout();
+    window.removeEventListener('online', this.recover);
+    window.removeEventListener('focus', this.recover);
+    document.removeEventListener('visibilitychange', this.recover);
+    this.recoveryListenersAttached = false;
     this.shouldReconnect = false;
     this.reconnectAttempts = this.maxReconnectAttempts;
     if (this.reconnectTimeout) {
@@ -354,7 +396,11 @@ export class WebSocketClient {
    * 获取连接状态
    */
   public isConnected(): boolean {
-    return this.ws !== null && this.ws.readyState === WebSocket.OPEN;
+    return (
+      this.ws !== null &&
+      this.ws.readyState === WebSocket.OPEN &&
+      this.connectionId !== null
+    );
   }
 
   // ===== 事件回调设置 =====
diff --git a/web/src/app/infra/websocket/url.ts b/web/src/app/infra/websocket/url.ts
new file mode 100644
index 000000000..60b92f2ff
--- /dev/null
+++ b/web/src/app/infra/websocket/url.ts
@@ -0,0 +1,15 @@
+/** Match the HTTP backend base, including reverse-proxy path prefixes. */
+export function backendWebSocketURL(
+  base: string | undefined,
+  pageURL: string,
+  path: string,
+): URL {
+  const url = new URL(base || '/', pageURL);
+  url.protocol =
+    url.protocol === 'https:' || url.protocol === 'wss:' ? 'wss:' : 'ws:';
+  url.pathname =
+    url.pathname.replace(/\/$/, '') + '/' + path.replace(/^\//, '');
+  url.search = '';
+  url.hash = '';
+  return url;
+}
diff --git a/web/src/app/login/page.tsx b/web/src/app/login/page.tsx
index d0134bb4c..d342909ed 100644
--- a/web/src/app/login/page.tsx
+++ b/web/src/app/login/page.tsx
@@ -212,6 +212,12 @@ export default function Login() {
         );
       const authResp = await startAuthentication({ optionsJSON: options });
       const res = await httpClient.verifyPasskeyAuth(challenge_token, authResp);
+      // A rejected assertion answers with an error payload and no session, so a
+      // missing token means the login failed rather than that the user cancelled.
+      if (!res?.token) {
+        toast.error(t('common.passkeyLoginFailed'));
+        return;
+      }
       if (await finishLogin(res.token, res.user)) {
         toast.success(t('common.passkeyLoginSuccess'));
       }
diff --git a/web/src/app/wizard/page.tsx b/web/src/app/wizard/page.tsx
index 8e9f51dee..d9bb49d10 100644
--- a/web/src/app/wizard/page.tsx
+++ b/web/src/app/wizard/page.tsx
@@ -29,6 +29,7 @@ import {
 } from 'lucide-react';
 
 import { httpClient } from '@/app/infra/http/HttpClient';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 import {
   systemInfo,
   bootstrapWorkspaceSession,
@@ -390,8 +391,7 @@ export default function WizardPage() {
 
             // Restore webhook URLs
             const runtimeValues = botData.bot.adapter_runtime_values as
-              | Record<string, unknown>
-              | undefined;
+              Record<string, unknown> | undefined;
             setWebhookUrl((runtimeValues?.webhook_full_url as string) || '');
             setExtraWebhookUrl(
               (runtimeValues?.extra_webhook_full_url as string) || '',
@@ -734,8 +734,7 @@ export default function WizardPage() {
       try {
         const botData = await httpClient.getBot(resp.uuid);
         const runtimeValues = botData.bot.adapter_runtime_values as
-          | Record<string, unknown>
-          | undefined;
+          Record<string, unknown> | undefined;
         setWebhookUrl((runtimeValues?.webhook_full_url as string) || '');
         setExtraWebhookUrl(
           (runtimeValues?.extra_webhook_full_url as string) || '',
@@ -823,7 +822,7 @@ export default function WizardPage() {
 
       if (!previewPipelineUuid) {
         const pipelineResp = await httpClient.createPipeline({
-          name: `${botName} Pipeline`,
+          name: `${botName} ${t('agents.pipelineType')}`,
           description: botDescription || '',
           config: {},
         });
@@ -901,8 +900,7 @@ export default function WizardPage() {
       try {
         const botData = await httpClient.getBot(createdBotUuid);
         const runtimeValues = botData.bot.adapter_runtime_values as
-          | Record<string, unknown>
-          | undefined;
+          Record<string, unknown> | undefined;
         setWebhookUrl((runtimeValues?.webhook_full_url as string) || '');
         setExtraWebhookUrl(
           (runtimeValues?.extra_webhook_full_url as string) || '',
@@ -1816,6 +1814,10 @@ function StepBotConfig({
                   systemContext={{
                     is_wizard: true,
                     webhook_url: webhookUrl,
+                    onebot_ws_url: webhookUrl
+                      ? webhookUrl.replace(/^http/, 'ws').replace(/\/$/, '') +
+                        '/onebot/v11/ws'
+                      : '',
                     extra_webhook_url: extraWebhookUrl,
                     outbound_ips: systemInfo.outbound_ips,
                   }}
@@ -1898,6 +1900,30 @@ function runnerPluginId(runner: string) {
     .join('/');
 }
 
+function InstalledRunnerIcon({
+  author,
+  name,
+}: {
+  author: string;
+  name: string;
+}) {
+  const iconURL = useInstalledPluginIcon(author, name);
+
+  if (!iconURL) {
+    return (
+      <Blocks className="size-10 shrink-0 rounded-md border bg-muted p-2 text-muted-foreground" />
+    );
+  }
+
+  return (
+    <img
+      src={iconURL}
+      alt=""
+      className="size-10 shrink-0 rounded-md border bg-muted object-cover"
+    />
+  );
+}
+
 function StepAIEngine({
   choice,
   onChoiceChange,
@@ -2292,11 +2318,7 @@ function StepAIEngine({
             >
               <CardHeader className="flex flex-row items-start gap-3 px-4">
                 {author && name ? (
-                  <img
-                    src={httpClient.getPluginIconURL(author, name)}
-                    alt=""
-                    className="size-10 shrink-0 rounded-md border bg-muted object-cover"
-                  />
+                  <InstalledRunnerIcon author={author} name={name} />
                 ) : (
                   <Blocks className="size-10 shrink-0 rounded-md border bg-muted p-2 text-muted-foreground" />
                 )}
diff --git a/web/src/components/EntityLoadState.tsx b/web/src/components/EntityLoadState.tsx
index 9b57a2440..9ee258d85 100644
--- a/web/src/components/EntityLoadState.tsx
+++ b/web/src/components/EntityLoadState.tsx
@@ -1,6 +1,6 @@
 import { Loader2 } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
-import { Button } from '@/components/ui/button';
+import LoadErrorState from '@/components/LoadErrorState';
 
 export default function EntityLoadState({
   error = false,
@@ -10,27 +10,15 @@ export default function EntityLoadState({
   onRetry?: () => void;
 }) {
   const { t } = useTranslation();
+  if (error) return <LoadErrorState onRetry={onRetry} className="flex-1" />;
   return (
     <div
-      role={error ? 'alert' : 'status'}
-      aria-busy={!error}
+      role="status"
+      aria-busy="true"
       className="flex min-h-40 flex-1 flex-col items-center justify-center gap-3 p-6 text-sm text-muted-foreground"
     >
-      {error ? (
-        <>
-          <p>{t('common.loadFailed')}</p>
-          {onRetry && (
-            <Button type="button" variant="outline" onClick={onRetry}>
-              {t('common.retry')}
-            </Button>
-          )}
-        </>
-      ) : (
-        <>
-          <Loader2 aria-hidden="true" className="size-5 animate-spin" />
-          <p>{t('common.loading')}</p>
-        </>
-      )}
+      <Loader2 aria-hidden="true" className="size-5 animate-spin" />
+      <p>{t('common.loading')}</p>
     </div>
   );
 }
diff --git a/web/src/components/LoadErrorState.tsx b/web/src/components/LoadErrorState.tsx
new file mode 100644
index 000000000..0e65adfb7
--- /dev/null
+++ b/web/src/components/LoadErrorState.tsx
@@ -0,0 +1,85 @@
+import { useState, type ReactNode } from 'react';
+import { FileWarning, RefreshCw } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+import { Button } from '@/components/ui/button';
+import { cn } from '@/lib/utils';
+
+export default function LoadErrorState({
+  title,
+  description,
+  icon,
+  onRetry,
+  busy = false,
+  compact = false,
+  className,
+}: {
+  title?: string;
+  description?: string;
+  icon?: ReactNode;
+  onRetry?: () => unknown;
+  busy?: boolean;
+  compact?: boolean;
+  className?: string;
+}) {
+  const { t } = useTranslation();
+  const [retrying, setRetrying] = useState(false);
+
+  async function retry() {
+    if (!onRetry || retrying || busy) return;
+    setRetrying(true);
+    try {
+      await onRetry();
+    } catch {
+      // The owning view retains its error state if the request fails again.
+    } finally {
+      setRetrying(false);
+    }
+  }
+
+  return (
+    <div
+      role="alert"
+      aria-busy={retrying || busy}
+      className={cn(
+        'flex w-full items-center justify-center px-4',
+        compact ? 'py-4' : 'min-h-64 py-10',
+        className,
+      )}
+    >
+      <div className="flex w-full max-w-md items-start gap-4 rounded-xl border bg-card p-5 text-card-foreground shadow-sm sm:p-6">
+        <div className="flex size-10 shrink-0 items-center justify-center rounded-xl bg-amber-500/10 text-amber-600 dark:text-amber-400">
+          {icon ?? <FileWarning className="size-5" aria-hidden="true" />}
+        </div>
+        <div className="min-w-0 flex-1 space-y-2">
+          <p className="break-words text-sm font-medium leading-6">
+            {title ?? t('common.loadFailedTitle')}
+          </p>
+          <p className="break-words text-sm leading-6 text-muted-foreground">
+            {description ?? t('common.loadFailedHint')}
+          </p>
+          {onRetry && (
+            <div className="pt-2">
+              <Button
+                type="button"
+                variant="outline"
+                size="sm"
+                disabled={retrying || busy}
+                onClick={() => void retry()}
+              >
+                <RefreshCw
+                  aria-hidden="true"
+                  className={cn(
+                    'size-3.5',
+                    (retrying || busy) &&
+                      'animate-spin motion-reduce:animate-none',
+                  )}
+                />
+                {retrying || busy ? t('common.loading') : t('common.retry')}
+              </Button>
+            </div>
+          )}
+        </div>
+      </div>
+    </div>
+  );
+}
diff --git a/web/src/components/market/CertifiedPluginBadge.tsx b/web/src/components/market/CertifiedPluginBadge.tsx
new file mode 100644
index 000000000..eb9041fe2
--- /dev/null
+++ b/web/src/components/market/CertifiedPluginBadge.tsx
@@ -0,0 +1,58 @@
+'use client';
+
+import { BadgeCheck } from 'lucide-react';
+import { useTranslation } from 'react-i18next';
+import {
+  Tooltip,
+  TooltipContent,
+  TooltipProvider,
+  TooltipTrigger,
+} from '@/components/ui/tooltip';
+
+export function CertifiedPluginBadge({
+  compact = false,
+}: {
+  compact?: boolean;
+}) {
+  const { t, i18n } = useTranslation();
+  const language = i18n.language.startsWith('zh')
+    ? 'zh'
+    : i18n.language.startsWith('ja')
+      ? 'ja'
+      : 'en';
+  const docsURL = `https://langbot.app/docs/${language}/plugin/certified-plugins`;
+  return (
+    <TooltipProvider delayDuration={200}>
+      <Tooltip>
+        <TooltipTrigger asChild>
+          <a
+            href={docsURL}
+            target="_blank"
+            rel="noopener noreferrer"
+            aria-label={`${t('market.certifiedPlugin')}: ${t('market.certificationLearnMore')}`}
+            onClick={(event) => event.stopPropagation()}
+            className={`inline-flex shrink-0 items-center gap-1 rounded border border-emerald-600/25 bg-emerald-50 text-emerald-700 dark:bg-emerald-950/40 dark:text-emerald-400 focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-emerald-500 ${compact ? 'h-4 px-0.5' : 'px-2 py-0.5 text-xs font-medium'}`}
+          >
+            <BadgeCheck
+              aria-hidden="true"
+              className={compact ? 'h-3 w-3' : 'h-3.5 w-3.5'}
+            />
+            {!compact && t('market.certifiedPlugin')}
+          </a>
+        </TooltipTrigger>
+        <TooltipContent className="max-w-72 text-xs">
+          <p>{t('market.certificationTooltip')}</p>
+          <a
+            href={docsURL}
+            target="_blank"
+            rel="noopener noreferrer"
+            className="mt-1 inline-block underline underline-offset-2"
+            onClick={(event) => event.stopPropagation()}
+          >
+            {t('market.certificationLearnMore')}
+          </a>
+        </TooltipContent>
+      </Tooltip>
+    </TooltipProvider>
+  );
+}
diff --git a/web/src/components/ui/badge.tsx b/web/src/components/ui/badge.tsx
index b0c1fc933..81eceb835 100644
--- a/web/src/components/ui/badge.tsx
+++ b/web/src/components/ui/badge.tsx
@@ -28,19 +28,52 @@ const badgeVariants = cva(
 function Badge({
   className,
   variant,
+  status,
   asChild = false,
   ...props
 }: React.ComponentProps<'span'> &
-  VariantProps<typeof badgeVariants> & { asChild?: boolean }) {
+  VariantProps<typeof badgeVariants> & { asChild?: boolean; status?: string }) {
   const Comp = asChild ? Slot : 'span';
 
   return (
     <Comp
       data-slot="badge"
-      className={cn(badgeVariants({ variant }), className)}
+      className={cn(
+        badgeVariants({ variant: status ? 'outline' : variant }),
+        status && statusBadgeClass(status),
+        className,
+      )}
       {...props}
     />
   );
 }
 
+function statusBadgeClass(status: string): string {
+  if (
+    [
+      'completed',
+      'success',
+      'succeeded',
+      'delivered',
+      'ready',
+      'enabled',
+      'connected',
+    ].includes(status)
+  )
+    return 'border-emerald-500/25 bg-emerald-500/10 text-emerald-700 dark:text-emerald-300';
+  if (['failed', 'error', 'timeout', 'disconnected'].includes(status))
+    return 'border-red-500/25 bg-red-500/10 text-red-700 dark:text-red-300';
+  if (['running', 'active', 'connecting'].includes(status))
+    return 'border-blue-500/25 bg-blue-500/10 text-blue-700 dark:text-blue-300';
+  if (
+    ['queued', 'pending', 'created', 'claimed', 'waiting', 'warning'].includes(
+      status,
+    )
+  )
+    return 'border-amber-500/25 bg-amber-500/10 text-amber-800 dark:text-amber-300';
+  if (['mock', 'simulated', 'debug'].includes(status))
+    return 'border-violet-500/25 bg-violet-500/10 text-violet-700 dark:text-violet-300';
+  return 'border-border bg-muted text-muted-foreground';
+}
+
 export { Badge, badgeVariants };
diff --git a/web/src/components/ui/sidebar.tsx b/web/src/components/ui/sidebar.tsx
index fd0daa9dc..d142aaf03 100644
--- a/web/src/components/ui/sidebar.tsx
+++ b/web/src/components/ui/sidebar.tsx
@@ -203,6 +203,7 @@ function SidebarProvider({
             {
               '--sidebar-width': `${sidebarWidth}px`,
               '--sidebar-width-icon': SIDEBAR_WIDTH_ICON,
+              height: '100svh',
               ...style,
             } as React.CSSProperties
           }
@@ -298,6 +299,10 @@ function Sidebar({
       />
       <div
         data-slot="sidebar-container"
+        style={{
+          top: 0,
+          height: '100svh',
+        }}
         className={cn(
           'fixed inset-y-0 z-10 hidden h-svh w-(--sidebar-width) transition-[left,right,width] duration-200 ease-linear md:flex',
           isSidebarResizing && 'transition-none',
diff --git a/web/src/components/ui/table.tsx b/web/src/components/ui/table.tsx
index 64efa928b..91961a1d0 100644
--- a/web/src/components/ui/table.tsx
+++ b/web/src/components/ui/table.tsx
@@ -2,11 +2,18 @@ import * as React from 'react';
 
 import { cn } from '@/lib/utils';
 
-function Table({ className, ...props }: React.ComponentProps<'table'>) {
+// `containerClassName` styles the scroll container shadcn wraps every table in.
+// A caller that owns the scrolling region (a fixed-height panel) uses it to make
+// that container the scroller, which is what keeps a sticky header working.
+function Table({
+  className,
+  containerClassName,
+  ...props
+}: React.ComponentProps<'table'> & { containerClassName?: string }) {
   return (
     <div
       data-slot="table-container"
-      className="relative w-full overflow-x-auto"
+      className={cn('relative w-full overflow-x-auto', containerClassName)}
     >
       <table
         data-slot="table"
diff --git a/web/src/components/ui/tooltip.tsx b/web/src/components/ui/tooltip.tsx
index fb7ee36a5..39ab05ebe 100644
--- a/web/src/components/ui/tooltip.tsx
+++ b/web/src/components/ui/tooltip.tsx
@@ -109,10 +109,13 @@ function TooltipTrigger({
 
 function TooltipContent({
   className,
+  arrowClassName,
   sideOffset = 0,
   children,
   ...props
-}: React.ComponentProps<typeof TooltipPrimitive.Content>) {
+}: React.ComponentProps<typeof TooltipPrimitive.Content> & {
+  arrowClassName?: string;
+}) {
   return (
     <TooltipPrimitive.Portal>
       <TooltipPrimitive.Content
@@ -125,7 +128,12 @@ function TooltipContent({
         {...props}
       >
         {children}
-        <TooltipPrimitive.Arrow className="bg-primary fill-primary z-50 size-2.5 translate-y-[calc(-50%_-_2px)] rotate-45 rounded-[2px]" />
+        <TooltipPrimitive.Arrow
+          className={cn(
+            'bg-primary fill-primary z-50 size-2.5 translate-y-[calc(-50%_-_2px)] rotate-45 rounded-[2px]',
+            arrowClassName,
+          )}
+        />
       </TooltipPrimitive.Content>
     </TooltipPrimitive.Portal>
   );
diff --git a/web/src/hooks/useAuthenticatedPluginResource.ts b/web/src/hooks/useAuthenticatedPluginResource.ts
index c271d99ff..3b534df2e 100644
--- a/web/src/hooks/useAuthenticatedPluginResource.ts
+++ b/web/src/hooks/useAuthenticatedPluginResource.ts
@@ -1,6 +1,7 @@
 import { useEffect, useState } from 'react';
 import { httpClient } from '@/app/infra/http/HttpClient';
 import { useCurrentWorkspace } from '@/app/infra/http';
+import { useInstalledPluginIcon } from '@/app/infra/hooks/useInstalledPluginIcon';
 
 type AuthenticatedResourceState = {
   key: string;
@@ -14,50 +15,21 @@ const EMPTY_RESOURCE: AuthenticatedResourceState = {
   error: false,
 };
 
+/**
+ * Backwards-compatible wrapper around the shared installed-plugin icon hook.
+ * Kept for callers that want a `{ url, error }` shape; the url is '' while
+ * loading or when the icon is unavailable.
+ */
 export function useAuthenticatedPluginIcon(
   author: string,
   name: string,
   enabled = true,
 ): { url: string; error: boolean } {
-  const [resource, setResource] =
-    useState<AuthenticatedResourceState>(EMPTY_RESOURCE);
-  const currentWorkspace = useCurrentWorkspace();
-  const workspaceUuid = currentWorkspace?.workspace.uuid;
-  const resourceKey = `${workspaceUuid ?? ''}:${author}/${name}`;
-
-  useEffect(() => {
-    if (!enabled) {
-      setResource({ key: resourceKey, url: '', error: false });
-      return;
-    }
-    let active = true;
-    let objectURL = '';
-    setResource({ key: resourceKey, url: '', error: false });
-    httpClient
-      .getAuthenticatedPluginIconURL(author, name)
-      .then((nextURL) => {
-        objectURL = nextURL;
-        if (active) {
-          setResource({ key: resourceKey, url: nextURL, error: false });
-        } else {
-          URL.revokeObjectURL(nextURL);
-        }
-      })
-      .catch(() => {
-        if (active) {
-          setResource({ key: resourceKey, url: '', error: true });
-        }
-      });
-    return () => {
-      active = false;
-      if (objectURL) URL.revokeObjectURL(objectURL);
-    };
-  }, [author, enabled, name, resourceKey]);
-
-  return {
-    url: resource.key === resourceKey ? resource.url : '',
-    error: resource.key === resourceKey && resource.error,
-  };
+  const url = useInstalledPluginIcon(
+    enabled ? author : null,
+    enabled ? name : null,
+  );
+  return { url: url ?? '', error: false };
 }
 
 export function useAuthenticatedPluginAsset(
diff --git a/web/src/i18n/index.ts b/web/src/i18n/index.ts
index 5e7826821..39c950b9a 100644
--- a/web/src/i18n/index.ts
+++ b/web/src/i18n/index.ts
@@ -3,6 +3,7 @@ import { initReactI18next } from 'react-i18next';
 import LanguageDetector from 'i18next-browser-languagedetector';
 
 import enUS from './locales/en-US';
+import botSetup from './locales/bot-setup';
 import zhHans from './locales/zh-Hans';
 import zhHant from './locales/zh-Hant';
 import jaJP from './locales/ja-JP';
@@ -17,28 +18,28 @@ i18n
   .init({
     resources: {
       'en-US': {
-        translation: enUS,
+        translation: { ...enUS, botSetup: botSetup['en-US'] },
       },
       'zh-Hans': {
-        translation: zhHans,
+        translation: { ...zhHans, botSetup: botSetup['zh-Hans'] },
       },
       'zh-Hant': {
-        translation: zhHant,
+        translation: { ...zhHant, botSetup: botSetup['zh-Hant'] },
       },
       'ja-JP': {
-        translation: jaJP,
+        translation: { ...jaJP, botSetup: botSetup['ja-JP'] },
       },
       'th-TH': {
-        translation: thTH,
+        translation: { ...thTH, botSetup: botSetup['th-TH'] },
       },
       'vi-VN': {
-        translation: viVN,
+        translation: { ...viVN, botSetup: botSetup['vi-VN'] },
       },
       'es-ES': {
-        translation: esES,
+        translation: { ...esES, botSetup: botSetup['es-ES'] },
       },
       'ru-RU': {
-        translation: ruRU,
+        translation: { ...ruRU, botSetup: botSetup['ru-RU'] },
       },
     },
     fallbackLng: 'zh-Hans',
diff --git a/web/src/i18n/locales/bot-setup.ts b/web/src/i18n/locales/bot-setup.ts
new file mode 100644
index 000000000..36c430ea7
--- /dev/null
+++ b/web/src/i18n/locales/bot-setup.ts
@@ -0,0 +1,67 @@
+export default {
+  'en-US': {
+    warning:
+      'No active routing. Add a rule and choose a processor to handle events.',
+    create: 'Create new processor',
+    createDescription:
+      'Create a processor and select it for this rule. Save the bot to apply the change.',
+    incompatible:
+      'The processor was created, but it does not support this event. Choose a compatible processor.',
+  },
+  'zh-Hans': {
+    warning: '尚未配置事件处理，请添加规则并选择处理器。',
+    create: '创建新处理器',
+    createDescription: '创建成功后会选入当前规则，保存机器人后生效。',
+    incompatible: '处理器已创建，但不支持当前事件，请选择兼容的处理器。',
+  },
+  'zh-Hant': {
+    warning: '尚未設定事件處理，請新增規則並選擇處理器。',
+    create: '建立新處理器',
+    createDescription: '建立成功後會選入目前規則，儲存機器人後生效。',
+    incompatible: '處理器已建立，但不支援目前事件，請選擇相容的處理器。',
+  },
+  'ja-JP': {
+    warning:
+      'イベント処理が未設定です。ルールを追加し、処理を選択してください。',
+    create: '新しい処理を作成',
+    createDescription:
+      '作成した処理を現在のルールに選択します。ボットを保存すると反映されます。',
+    incompatible:
+      '処理は作成されましたが、このイベントに対応していません。対応する処理を選択してください。',
+  },
+  'es-ES': {
+    warning:
+      'Sin procesamiento de eventos. Añade una regla y elige un procesador.',
+    create: 'Crear procesador',
+    createDescription:
+      'El procesador se seleccionará para esta regla. Guarda el bot para aplicar el cambio.',
+    incompatible:
+      'El procesador se creó, pero no admite este evento. Elige uno compatible.',
+  },
+  'ru-RU': {
+    warning:
+      'Обработка событий не настроена. Добавьте правило и выберите обработчик.',
+    create: 'Создать обработчик',
+    createDescription:
+      'Созданный обработчик будет выбран для правила. Сохраните бота для применения.',
+    incompatible:
+      'Обработчик создан, но не поддерживает это событие. Выберите совместимый обработчик.',
+  },
+  'vi-VN': {
+    warning: 'Chưa thiết lập xử lý sự kiện. Hãy thêm quy tắc và chọn bộ xử lý.',
+    create: 'Tạo bộ xử lý mới',
+    createDescription:
+      'Bộ xử lý sẽ được chọn cho quy tắc này. Lưu bot để áp dụng.',
+    incompatible:
+      'Đã tạo bộ xử lý nhưng không hỗ trợ sự kiện này. Hãy chọn bộ xử lý tương thích.',
+  },
+  'th-TH': {
+    warning:
+      'ยังไม่ได้ตั้งค่าการจัดการเหตุการณ์ โปรดเพิ่มกฎและเลือกตัวประมวลผล',
+    create: 'สร้างตัวประมวลผลใหม่',
+    createDescription:
+      'ระบบจะเลือกตัวประมวลผลให้กฎนี้ บันทึกบอตเพื่อใช้การเปลี่ยนแปลง',
+    incompatible:
+      'สร้างตัวประมวลผลแล้ว แต่ไม่รองรับเหตุการณ์นี้ โปรดเลือกตัวประมวลผลที่รองรับ',
+  },
+};
diff --git a/web/src/i18n/locales/en-US.ts b/web/src/i18n/locales/en-US.ts
index 62213e9b2..05f9ee3dc 100644
--- a/web/src/i18n/locales/en-US.ts
+++ b/web/src/i18n/locales/en-US.ts
@@ -1,6 +1,179 @@
+import slackSetup from './slack-setup/en-US';
 import pipelineMigration from './pipeline-migration/en-US';
 const enUS = {
+  planRevision: {
+    subscribedOn: 'Subscribed on {{date}}',
+    latestQuota: 'Latest quotas for this plan',
+    showLatest: 'Show latest quotas for this plan',
+    showCurrent: 'Show your current subscription',
+    latestTab: 'Latest',
+    currentTab: 'Current',
+    cancelToSubscribe:
+      'Cancel your subscription to subscribe to the latest quota next period',
+
+    current: 'Current subscription: {{plan}} · revision {{revision}}',
+    owned: 'Purchased revision {{revision}}',
+    latest: 'Latest revision {{revision}}',
+    changed:
+      'Your purchased benefits and the latest version differ. Highlighted rows show the changes.',
+    same: 'Benefits match latest revision {{revision}}. Your purchased revision remains unchanged.',
+    unavailable:
+      'The latest plan is unavailable. Your purchased benefits remain in effect.',
+    benefit: 'Benefit',
+    change: 'Changed',
+    enabled: 'Enabled',
+    disabled: 'Disabled',
+    monthlyCredits: 'Monthly bonus credits',
+    retention: 'Chat history retention (days)',
+    unknown:
+      'Current subscription details are unavailable. Refresh or manage your subscription in Space; latest catalog benefits are not your purchased benefits.',
+    handoff:
+      'Manage your subscription in Space. Cancel automatic renewal before purchasing a new version. Prepaid subscriptions end automatically unless renewed before expiry.',
+    cancelFirst:
+      'Cancel automatic renewal in subscription management first. Cancellation preserves paid benefits until the end of the period, without a refund. After confirmation, you can purchase the latest version, effective when payment succeeds.',
+    canceled:
+      'Renewal is canceled. Paid benefits remain until the end of the period, without a refund. A new purchase applies its benefits when payment succeeds.',
+    prepaid:
+      'This prepaid subscription ends automatically. You can renew your current plan before expiry; after expiry, subscriptions use the latest quotas for each plan.',
+    renew: 'Renew current subscription',
+    buying: 'Purchasing revision {{revision}}.',
+    fields: {
+      bots_max: 'Bots',
+      members_max: 'Members',
+      pipelines_max: 'Processors',
+      plugins_max: 'Extensions',
+      knowledge_bases_max: 'Knowledge bases',
+      managed_sandbox_sessions_max: 'Sandbox sessions',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: 'Not available to subscribe',
+    subscriptionUnavailableHint:
+      'This plan is not accepting new subscriptions. Please choose another plan. Existing subscriptions and scheduled automatic renewals are unaffected.',
+    platformNames: {
+      qqofficial: 'QQ Official API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComCustomerService',
+      officialaccount: 'Official Account',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'Manage subscription',
+    downgradeUnavailable: 'Cannot downgrade to Free',
+    features: {
+      multiUserWorkspace: 'Multi-user workspace',
+      remoteE2BSandbox: 'Remote E2B sandbox configuration',
+      communitySupport: 'Community support',
+      everythingInFree: 'Everything in Free',
+      alwaysOnSandbox: 'One always-on global sandbox',
+      higherWorkspaceLimits: 'Higher workspace limits',
+    },
+    title: 'Compare plans',
+    description: 'Choose a plan for {{workspace}}',
+    billing: 'Billing cycle',
+    monthly: 'Monthly',
+    annual: 'Yearly',
+    currency: 'Currency',
+    loading: 'Loading plans…',
+    error: 'Could not load plans',
+    retry: 'Retry',
+    empty: 'No plans available',
+    current: 'Current plan',
+    recommended: 'Recommended',
+    free: 'Free',
+    annualNote: 'Equivalent to {{price}} / month',
+    ownerOnly: 'Owner or admin only',
+    choose: 'Continue to payment',
+    handoff:
+      'Continue in Space to sign in and pay. Your selection will be preserved.',
+    quotas: 'Workspace quotas',
+    unlimited: 'Unlimited',
+    retention: 'Message history · {{count}} days',
+    retentionLabel: 'Message history retention',
+    retentionHelp:
+      'Message history is automatically removed after this period.',
+    credits: '{{count}} credits every month',
+    creditsLabel: 'Monthly credits',
+    creditsHelp:
+      'Credits are granted each subscription month, including on yearly plans.',
+    platforms: 'Mainstream platforms supported',
+    excludedPlatforms: 'Excluded: {{platforms}}',
+    allPlatforms: 'All currently available platforms',
+    limits: {
+      bots: 'Bots',
+      pipelines: 'Processors',
+      knowledge_bases: 'Knowledge bases',
+      plugins: 'Extensions',
+      members: 'Members',
+      managed_sandbox_sessions: 'Managed sandbox sessions',
+    },
+  },
+  workspaceLanguage: {
+    label: 'Workspace language',
+    description:
+      'Language for backend messages and default bot replies. Your interface language stays independent.',
+    saved: 'Language saved',
+    saveFailed: 'Failed to save language',
+  },
+  slackSetup,
+  inflight: {
+    autoCollapse: 'Auto-collapse',
+    empty: 'No events are being processed',
+    move: 'Drag vertically; use arrow keys to move',
+    botLogs: 'Robot logs',
+    processorLogs: 'Processor logs',
+
+    title: 'In flight',
+    expand: 'Expand',
+    collapse: 'Collapse',
+    reconnecting: 'Reconnecting; showing the last received state.',
+    details: 'View details',
+    debug: 'Debug event',
+    limited: 'Showing a limited set. Open the dashboard for all executions.',
+    queued: 'Queued',
+    running: 'Processing',
+    waiting: 'Waiting for input',
+    tool: 'Calling a tool',
+    thinking: 'Thinking',
+    replying: 'Preparing reply',
+    generating: 'Generating reply',
+    completed: 'Completed',
+    failed: 'Failed',
+    cancelled: 'Cancelled',
+    ignored: 'Ignored',
+  },
   assistant: {
+    movePosition: 'Move assistant horizontally',
+    resizeHeight: 'Resize assistant height',
+    sessions: 'Conversations',
+    currentChat: 'Current conversation',
+    stop: 'Stop task',
+    stopped:
+      'Stopped. An in-progress tool may already have taken effect. Check its resource before retrying.',
+    setupRequired:
+      'Sign in to LangBot Account or configure a model to start using the workspace assistant.',
+    loginAccount: 'Sign in to LangBot Account',
+    configureModels: 'Configure models',
+    providerCheckFailed: 'Could not check model providers. Please try again.',
+    retrySetup: 'Retry',
+    preparing: 'Preparing assistant…',
+    thinking: 'Model responding · Round {{round}}',
+    executing: 'Executing tool…',
+    toolRunning: 'Running',
+    recommendationFailed:
+      'Could not load a recommended model. Please select a model.',
+    expand: 'Expand workspace assistant',
+    collapse: 'Collapse workspace assistant',
     modelHint: 'Switch assistant model for your next message',
     details: 'View raw data',
     found: 'Found {{count}} items',
@@ -21,7 +194,7 @@ const enUS = {
       create_knowledge_base: 'Create knowledge base',
       get_pipeline: 'Read Pipeline',
       get_knowledge_schema: 'Read knowledge engine schema',
-      list_operation_logs: 'Read operation trace',
+      list_operation_logs: 'Read operation logs',
     },
     resources: {
       models: 'Find chat models',
@@ -35,7 +208,7 @@ const enUS = {
     newChat: 'New conversation',
     close: 'Close',
     welcome:
-      'Tell me what you want to build. I can inspect this Workspace and propose knowledge-base and Pipeline changes for you to confirm.',
+      'Tell me what you want to build. I can inspect this Workspace and propose configurations for bots, Agents, knowledge bases, Pipelines and more, then apply them after your confirmation.',
     discover: 'What models, knowledge bases and Pipelines are available?',
     build: 'Help me create a knowledge-base question-answering application.',
     toolResult: 'Operation result',
@@ -55,6 +228,14 @@ const enUS = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: 'Last {{days}} days',
+        customRange: 'Custom time range',
+        rangeStart: 'Start time',
+        rangeEnd: 'End time',
+        rangeHint: 'Start must precede end. Maximum range: 365 days.',
+        autoRefreshLabel: 'Auto refresh',
+        autoRefreshOff: 'Auto refresh: off',
+        autoRefreshEvery: 'Refresh every {{seconds}} s',
         title: 'Review activity',
         description:
           'Review bot activity, model usage, message volume, and system performance at a glance.',
@@ -136,6 +317,14 @@ const enUS = {
           'Send test messages to check triggers, AI replies, and output processing. Any configuration changes are saved before testing.',
       },
       monitoring: {
+        rangeDays: 'Last {{days}} days',
+        customRange: 'Custom time range',
+        rangeStart: 'Start time',
+        rangeEnd: 'End time',
+        rangeHint: 'Start must precede end. Maximum range: 365 days.',
+        autoRefreshLabel: 'Auto refresh',
+        autoRefreshOff: 'Auto refresh: off',
+        autoRefreshEvery: 'Refresh every {{seconds}} s',
         title: 'Review run history',
         description:
           'Use the tab beside the title to review past conversations, processing steps, and errors when replies are missing or unexpected.',
@@ -303,6 +492,11 @@ const enUS = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'The failure reason was not recorded. Check the bot logs or retry to capture more details.',
+    loadFailedTitle: 'Unable to load this content',
+    loadFailedHint:
+      'Please try again in a moment. If this continues, check your connection and service status.',
     customValue: 'Custom',
     loadFailed: 'Failed to load. Please try again.',
     search: 'Search',
@@ -501,6 +695,9 @@ const enUS = {
     help: 'Get Help',
   },
   models: {
+    starModel: 'Star model (default for new processors)',
+    unstarModel: 'Unstar model',
+    starModelFailed: 'Could not update starred model',
     codex: {
       account: 'ChatGPT subscription',
       description:
@@ -760,10 +957,13 @@ const enUS = {
     deleteConfirmation: 'Are you sure you want to delete this bot?',
     platformAdapter: 'Platform/Adapter Selection',
     selectAdapter: 'Select Adapter',
+    saveAndEnable: 'Save and enable',
+    searchAdapters: 'Search adapters',
+    noMatchingAdapters: 'No matching adapters found',
     legacyAdapters: 'Legacy adapters',
     legacyAdapterBadge: 'Legacy',
     legacyAdaptersHint:
-      'These adapters have newer event-driven counterparts.\nThey are kept only for existing configurations and are not recommended for new bots.',
+      'Legacy adapters mainly send and receive messages and remain available for existing bot configurations.\nNew adapters connect to the same platform using events, including messages and supported friend requests or group changes, routed to processors.\nPrefer the new adapter for new bots. Existing bots can keep using legacy adapters. When switching, review connection settings and event routes; this does not upgrade your platform account.',
     adapterConfig: 'Adapter Configuration',
     viewAdapterDocs: 'View Docs',
     bindPipeline: 'Bind Pipeline',
@@ -776,6 +976,10 @@ const enUS = {
     earlier: 'Earlier',
     dateFormat: '{{month}}/{{day}}',
     setBotEnableError: 'Failed to set bot enable status',
+    enabling: 'Enabling…',
+    disabling: 'Disabling…',
+    enableConfirmed: 'Enabled',
+    disableConfirmed: 'Disabled',
     log: 'Log',
     configuration: 'Configuration',
     logs: 'Logs',
@@ -784,6 +988,28 @@ const enUS = {
     routingConnection: 'Routing & Connection',
     routingConnectionDescription:
       'Bind the pipeline that processes messages for this bot',
+    routingHelp: {
+      title: 'What is event routing?',
+      intro:
+        'Bots receive more than messages: friend requests, group invitations and other events too.',
+      events: 'Incoming events',
+      agents: 'Different agents',
+      message: 'Message received',
+      friend: 'Friend request',
+      group: 'Joined a group',
+      messageAction: 'Reply agent → Answer',
+      friendAction: 'Review agent → Review',
+      groupAction: 'Welcome agent → Greet',
+      rules: 'Your routing rules · Example',
+      instructions:
+        'Agents follow your instructions and use tools to reply, review requests and more.',
+      plugin:
+        'For advanced workflows, plugin processors register events in bulk and handle them with code. Add a plugin processor configuration below.',
+      docs: 'Develop your own plugin processor',
+      runnerDocs: 'Develop your own Agent runner',
+      pause: 'Pause',
+      play: 'Play',
+    },
     eventRouting: 'Event Routing',
     eventRoutingDescription: 'Choose which processor handles each event.',
     eventBindings: 'Event Routes',
@@ -822,7 +1048,8 @@ const enUS = {
     enabled: 'Enabled',
     eventBindingDescriptionPlaceholder: 'Rule description',
     noEventBindings: 'No event routes',
-    unsupportedPipelineEvent: 'Pipelines can only be used for message.* events',
+    unsupportedPipelineEvent:
+      'Pipelines can only be used for message.received events',
     disable: 'Disable',
     enable: 'Enable',
     disabledBindings: 'Disabled',
@@ -832,6 +1059,7 @@ const enUS = {
     advancedEventValues: 'View all',
     eventGroup: 'Group',
     eventGroupNames: {
+      wecomcs: 'WeCom Customer Service',
       bot: 'Bot status',
       feedback: 'Feedback',
       friend: 'Friends',
@@ -846,6 +1074,9 @@ const enUS = {
     routeFallbackCatchAll: '{{route}} is the catch-all route.',
     routeFallbackIgnored:
       'Events that match no route are ignored. Add a catch-all route only when every event needs an explicit outcome.',
+    matchRecords: 'Match records',
+    matchRecordsDescription:
+      'Latest result for each rule and recent unmatched events, from bot memory logs. Cleared on restart.',
     testRoute: 'Check route',
     adapterEventDebugAction: 'Test listener',
     adapterEventDebugTitle: 'Platform event debugging',
@@ -931,6 +1162,15 @@ const enUS = {
     eventWildcard: 'All events',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'User entered bot chat',
+      wecombot_template_card_event: 'Template card interaction',
+      dingtalk_card_action: 'DingTalk card interaction',
+      wecomcs_enter_session: 'Customer entered conversation',
+      wecomcs_msg_send_fail: 'Customer service message failed',
+      wecomcs_servicer_status_change: 'Representative availability changed',
+      wecomcs_session_status_change: 'Customer service conversation changed',
+      wecomcs_reject_customer_msg_switch_change:
+        'Customer message rejection changed',
       message_received: 'Message received',
       message_edited: 'Message edited',
       message_deleted: 'Message deleted',
@@ -940,6 +1180,7 @@ const enUS = {
       friend_added: 'Friend added',
       group_member_joined: 'Member joined group',
       group_member_left: 'Member left group',
+      group_info_updated: 'Group information updated',
       group_member_banned: 'Member banned',
       bot_invited_to_group: 'Bot invited to group',
       bot_removed_from_group: 'Bot removed from group',
@@ -948,6 +1189,22 @@ const enUS = {
       platform_specific: 'Platform-specific event',
     },
     eventDescriptions: {
+      wecombot_enter_chat:
+        'A user opened the bot chat. Use this event for welcome workflows.',
+      wecombot_template_card_event:
+        'A user clicked or submitted a template card. Managed form callbacks are handled separately.',
+      dingtalk_card_action:
+        'Card clicks and submissions; feedback and managed forms are handled separately.',
+      wecomcs_enter_session:
+        'A customer opened a service link; includes entry source and welcome code.',
+      wecomcs_msg_send_fail:
+        'WeChat reported an outgoing message failure, with message ID and failure type.',
+      wecomcs_servicer_status_change:
+        'A representative started, stopped or suspended service.',
+      wecomcs_session_status_change:
+        'A representative accepted, transferred, closed or reopened a conversation.',
+      wecomcs_reject_customer_msg_switch_change:
+        'A representative enabled or disabled rejection of customer messages.',
       all: 'Matches every event received by this adapter.',
       namespace: 'Matches all {{group}} events.',
       namespace_bot:
@@ -955,7 +1212,7 @@ const enUS = {
       namespace_feedback: 'Matches feedback events from a platform or user.',
       namespace_friend: 'Matches friend requests and friendship changes.',
       namespace_group:
-        'Matches member joins, leaves, removals, and other group events.',
+        'Matches group events such as membership changes and group information updates.',
       namespace_message:
         'Matches received, edited, deleted, and reaction message events.',
       namespace_platform:
@@ -971,6 +1228,8 @@ const enUS = {
       friend_added: 'A friend relationship was created.',
       group_member_joined: 'A member joins a group where the bot is present.',
       group_member_left: 'A member leaves a group where the bot is present.',
+      group_info_updated:
+        'Group details such as its name or owner have changed.',
       group_member_banned: 'A group member is banned or removed.',
       bot_invited_to_group: 'The bot is invited to a group.',
       bot_removed_from_group: 'The bot is removed from a group.',
@@ -1057,6 +1316,14 @@ const enUS = {
     imagesAttached: 'image(s) attached',
     noLogs: 'No logs yet',
     sessionMonitor: {
+      resetContext: 'Reset context',
+      resetDescription:
+        'New messages will start with an empty context. Monitoring history is kept; files and long-term memory are unchanged. Wait for active tasks to finish before resetting.',
+      resetSuccess:
+        'Context reset. New messages will start a fresh conversation.',
+      resetError:
+        'Could not reset context. Wait for active tasks to finish and try again.',
+
       title: 'Sessions',
       sessions: 'Sessions',
       noSessions: 'No sessions found',
@@ -1104,7 +1371,26 @@ const enUS = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'Select all',
     monitoring: {
+      rangeDays: 'Last {{days}} days',
+      customRange: 'Custom time range',
+      rangeStart: 'Start time',
+      rangeEnd: 'End time',
+      rangeHint: 'Start must precede end. Maximum range: 365 days.',
+      autoRefreshLabel: 'Auto refresh',
+      autoRefreshOff: 'Auto refresh: off',
+      autoRefreshEvery: 'Refresh every {{seconds}} s',
+      runStatus: 'Run status',
+      toolCalls: 'Tool calls',
+      delivery: 'Message delivery',
+      sendFailed: 'Send failed',
+      sendCount: '{{count}} send calls succeeded',
+      noSend: 'No send recorded',
+      noSteps: 'No execution steps recorded yet.',
+      recipient: 'Recipient',
+      actionCompleted: 'Tool completed successfully.',
+      simulatedAction: 'Simulated action; no real message was sent.',
       description:
         'Follow each task from its triggering event through model output and tool execution.',
       empty:
@@ -1209,6 +1495,9 @@ const enUS = {
       'Handle message events only, with AI generating replies directly and practical features such as knowledge bases and plugins.',
     allEvents: 'Supports all events',
     messageEventsOnly: 'Message events only',
+    selectAgentRunner: 'Select Agent runner',
+    selectProcessorPlugin: 'Select processor plugin',
+    createBasicInfoDescription: 'Set a name, icon and plugin',
     chooseType: 'Choose how it works',
     chooseTypeDescription:
       'Choose the processor model first. You can configure it further after creation.',
@@ -1624,6 +1913,11 @@ const enUS = {
     },
   },
   market: {
+    certifiedPlugin: 'Certified plugin',
+    certificationTooltip:
+      'This version has passed a review of data isolation and compatibility with shared runtimes, and its package is signed.',
+    certificationLearnMore: 'Learn about certification',
+
     runnerUsage: 'Runner usage',
     runnerUsageAll: 'All',
     runnerUsageAgent: 'Agent / Pipeline',
@@ -2022,6 +2316,14 @@ const enUS = {
       uploading: 'Uploading...',
     },
     monitoring: {
+      rangeDays: 'Last {{days}} days',
+      customRange: 'Custom time range',
+      rangeStart: 'Start time',
+      rangeEnd: 'End time',
+      rangeHint: 'Start must precede end. Maximum range: 365 days.',
+      autoRefreshLabel: 'Auto refresh',
+      autoRefreshOff: 'Auto refresh: off',
+      autoRefreshEvery: 'Refresh every {{seconds}} s',
       title: 'Run logs',
       workbench: 'Configure & debug',
       description:
@@ -2316,7 +2618,7 @@ const enUS = {
     totpStatusDisabled: 'Not enabled',
     totpCodesRemaining: '{{count}} recovery codes left',
     totpManagerSectionDesc:
-      'Owners and admins can review and revoke the second factor of any Account.',
+      'Owners and admins can review and revoke the second factor of Accounts in this Workspace.',
     revokeTotp: 'Re-bind',
     totpAdminResetTitle: 'Re-bind two-factor authentication for {{user}}',
     totpAdminResetDesc:
@@ -2381,8 +2683,7 @@ const enUS = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2451,6 +2752,124 @@ const enUS = {
     },
   },
   monitoring: {
+    rangeDays: 'Last {{days}} days',
+    customRange: 'Custom time range',
+    rangeStart: 'Start time',
+    rangeEnd: 'End time',
+    rangeHint: 'Start must precede end. Maximum range: 365 days.',
+    autoRefreshLabel: 'Auto refresh',
+    autoRefreshOff: 'Auto refresh: off',
+    autoRefreshEvery: 'Refresh every {{seconds}} s',
+    unified: {
+      inputs: 'Event / input',
+      outputs: 'Generated results',
+      deliveries: 'Delivery records',
+      conversation: 'Conversation context',
+      related: 'Related executions',
+      events: 'Execution events',
+      llm_calls: 'Model calls',
+      tool_calls: 'Tool calls',
+      errors: 'Errors',
+      payload: 'Details',
+      metadata: 'Technical details',
+      loadMore: 'Load more',
+      noRecords: 'No records',
+      contextHint:
+        'Conversation context is not necessarily part of this execution.',
+      unhandled: 'No processor run',
+      roles: {
+        user: 'User',
+        assistant: 'Assistant',
+        system: 'System',
+        tool: 'Tool',
+      },
+    },
+    execution: {
+      title: 'Executions',
+      subtitle:
+        'Every inbound event and the execution it triggered, across agents and pipelines.',
+      tabs: { executions: 'Executions' },
+      filters: {
+        source: 'Type',
+        sourceAll: 'All types',
+        sourceAgent: 'Agents',
+        sourcePipeline: 'Pipelines',
+        mode: 'Traffic',
+        modeAll: 'Live + debug',
+        modeLive: 'Live only',
+        modeDebug: 'Debug only',
+        status: 'Status',
+        statusAll: 'All statuses',
+      },
+      status: {
+        completed: 'Completed',
+        failed: 'Failed',
+        running: 'Running',
+        queued: 'Queued',
+        cancelled: 'Cancelled',
+        ignored: 'Ignored',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Agent',
+        pipeline: 'Pipeline',
+        processor: 'Processor',
+      },
+      columns: {
+        time: 'Time',
+        target: 'Processor',
+        status: 'Status',
+        duration: 'Duration',
+        tokens: 'Tokens',
+      },
+      debug: 'Debug',
+      empty: 'No executions in this range',
+      loading: 'Loading executions...',
+      triggeredBy: 'Triggered by {{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95: about 50%/95% of runs finish within this time. Based on recent completed Agent runs, excluding queue time.',
+        tokensHint:
+          'Only recorded usage is included. Totals may be incomplete when models do not return usage.',
+        totalExecutions: 'Executions',
+        bySource: '{{agent}} agent · {{pipeline}} pipeline',
+        successRate: 'Execution Success Rate',
+        successRateHint:
+          'Successes ÷ (successes + failures). In-progress, ignored and cancelled runs are excluded.',
+        denominator: '{{completed}} succeeded out of {{denominator}} finished',
+        running: 'Running',
+        queued: 'Queued',
+        waiting: 'Waiting input',
+        failed: 'Failed',
+        ignored: 'Ignored',
+        latency: 'Execution time',
+        p50: 'P50',
+        p95: 'P95',
+        sample: 'Based on {{count}} runs',
+        tokens: 'Tokens',
+        coverage: 'usage on {{withUsage}}/{{calls}} calls',
+        debugRuns: 'Includes {{count}} debug runs',
+      },
+      detail: {
+        title: 'Execution',
+        timeline: 'Timeline',
+        executionId: 'Execution ID',
+        eventId: 'Event',
+        queue: 'Queue',
+        runner: 'Runner',
+        created: 'Created',
+        duration: 'Duration',
+        usage: 'Usage',
+        noEvents: 'No trace events recorded',
+        query: 'Inbound message',
+        modelCalls: 'Model calls',
+        toolCalls: 'Tool calls',
+        errors: 'Errors',
+        loadError: 'Failed to load execution trace',
+        openObject: 'Open object',
+        a11yDescription: 'Full trace of this execution',
+      },
+    },
     title: 'Dashboard',
     description: 'Monitor bot activities, LLM calls, and system performance',
     overview: 'Overview',
@@ -2466,6 +2885,9 @@ const enUS = {
       pipeline: 'Pipeline',
       allBots: 'All Bots',
       selectBot: 'Select Bot',
+      processor: 'Processor',
+      allProcessors: 'All Processors',
+      selectProcessor: 'Select Processor',
       allPipelines: 'All Pipelines',
       selectPipeline: 'Select Pipeline',
       loading: 'Loading...',
@@ -2491,6 +2913,15 @@ const enUS = {
       sessions: 'Session Analysis',
       errors: 'Error Logs',
     },
+    pipelineTrace: {
+      received: 'Received message',
+      processing: 'Processing',
+      replies: 'Replies',
+      noSteps: 'No model or tool calls recorded',
+      success: 'Success',
+      error: 'Failed',
+      pending: 'Pending',
+    },
     messageList: {
       timestamp: 'Timestamp',
       bot: 'Bot',
@@ -2713,7 +3144,7 @@ const enUS = {
     title: 'Settings',
     nav: {
       workspace: 'Workspace',
-      operationTrace: 'Traceability',
+      operationTrace: 'Operation Log',
       models: 'Models',
       api: 'API',
       storage: 'Storage',
@@ -2721,21 +3152,21 @@ const enUS = {
     },
   },
   operationTrace: {
-    title: 'Operation traceability',
+    title: 'Operation Log',
     description: 'Trace admin and owner edits and views, and what changed.',
+    reset: 'Reset',
+    settingsUpdated: 'Settings saved',
+    settingsUpdateFailed: 'Failed to save settings',
+    levelBadge: 'Level: {{level}}',
+    collectionSettings: 'Collection settings',
     captureLevel: 'Capture level',
     retention: 'Retention',
-    retentionSummary: 'Keep {{days}} days · up to {{rows}} records',
     retentionDays: 'Retention days',
     maxRows: 'Max records',
     dedupeWindow: 'Dedupe window (s)',
     records: 'Operation records',
     changesCount: '{{count}} change(s)',
     hideDetails: 'Hide details',
-    levelUpdated: 'Capture level updated',
-    levelUpdateFailed: 'Failed to update capture level',
-    retentionUpdated: 'Retention updated',
-    retentionUpdateFailed: 'Failed to update retention',
     export: 'Download logs',
     exportFailed: 'Failed to download operation logs',
     empty: 'No operation records yet',
@@ -2749,7 +3180,7 @@ const enUS = {
     filterAllResources: 'All resources',
     filterActor: 'Actor',
     filterAllActors: 'All actors',
-    loadFailed: 'Failed to load traceability data',
+    loadFailed: 'Failed to load operation logs',
     refresh: 'Refresh',
     save: 'Save',
     redacted: 'Redacted',
@@ -2768,7 +3199,7 @@ const enUS = {
     elidedValue: '{{count}} chars',
     levels: {
       off: {
-        label: 'Tracing off',
+        label: 'Logging off',
       },
       mutation: {
         label: 'Mutations only',
diff --git a/web/src/i18n/locales/es-ES.ts b/web/src/i18n/locales/es-ES.ts
index 43073f37b..7f10d6055 100644
--- a/web/src/i18n/locales/es-ES.ts
+++ b/web/src/i18n/locales/es-ES.ts
@@ -1,6 +1,178 @@
 import pipelineMigration from './pipeline-migration/es-ES';
 const esES = {
+  planRevision: {
+    subscribedOn: 'Fecha de suscripción: {{date}}',
+    latestQuota: 'Cuotas más recientes de este plan',
+    showLatest: 'Ver las cuotas más recientes de este plan',
+    showCurrent: 'Ver tu suscripción actual',
+    latestTab: 'Más reciente',
+    currentTab: 'Actual',
+    cancelToSubscribe:
+      'Cancela tu suscripción para acceder a las últimas cuotas el próximo período',
+
+    current: 'Current subscription: {{plan}} · revision {{revision}}',
+    owned: 'Purchased revision {{revision}}',
+    latest: 'Latest revision {{revision}}',
+    changed:
+      'Your purchased benefits and the latest version differ. Highlighted rows show the changes.',
+    same: 'Benefits match latest revision {{revision}}. Your purchased revision remains unchanged.',
+    unavailable:
+      'The latest plan is unavailable. Your purchased benefits remain in effect.',
+    benefit: 'Benefit',
+    change: 'Changed',
+    enabled: 'Enabled',
+    disabled: 'Disabled',
+    monthlyCredits: 'Monthly bonus credits',
+    retention: 'Chat history retention (days)',
+    unknown:
+      'Current subscription details are unavailable. Refresh or manage your subscription in Space; latest catalog benefits are not your purchased benefits.',
+    handoff:
+      'Manage your subscription in Space. Cancel automatic renewal before purchasing a new version. Prepaid subscriptions end automatically unless renewed before expiry.',
+    cancelFirst:
+      'Cancel automatic renewal in subscription management first. Cancellation preserves paid benefits until the end of the period, without a refund. After confirmation, you can purchase the latest version, effective when payment succeeds.',
+    canceled:
+      'Renewal is canceled. Paid benefits remain until the end of the period, without a refund. A new purchase applies its benefits when payment succeeds.',
+    prepaid:
+      'This prepaid subscription ends automatically. You can renew your current plan before expiry; after expiry, subscriptions use the latest quotas for each plan.',
+    renew: 'Renovar la suscripción actual',
+    buying: 'Purchasing revision {{revision}}.',
+    fields: {
+      bots_max: 'Bots',
+      members_max: 'Members',
+      pipelines_max: 'Processors',
+      plugins_max: 'Extensions',
+      knowledge_bases_max: 'Knowledge bases',
+      managed_sandbox_sessions_max: 'Sandbox sessions',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: 'Suscripción no disponible',
+    subscriptionUnavailableHint:
+      'Este plan no acepta nuevas suscripciones. Elige otro plan. Las suscripciones existentes y sus renovaciones automáticas no se ven afectadas.',
+    platformNames: {
+      qqofficial: 'QQ Official API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComCustomerService',
+      officialaccount: 'Cuenta Oficial',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'Gestionar suscripción',
+    downgradeUnavailable: 'No se puede cambiar al plan gratuito',
+    features: {
+      multiUserWorkspace: 'Espacio de trabajo multiusuario',
+      remoteE2BSandbox: 'Configuración de sandbox E2B remoto',
+      communitySupport: 'Soporte de la comunidad',
+      everythingInFree: 'Todas las funciones del plan gratuito',
+      alwaysOnSandbox: 'Un sandbox global siempre activo',
+      higherWorkspaceLimits: 'Límites superiores del espacio de trabajo',
+    },
+    title: 'Comparar planes',
+    description: 'Elige un plan para {{workspace}}',
+    billing: 'Ciclo de facturación',
+    monthly: 'Mensual',
+    annual: 'Anual',
+    currency: 'Moneda',
+    loading: 'Cargando planes…',
+    error: 'No se pudieron cargar los planes',
+    retry: 'Reintentar',
+    empty: 'No hay planes disponibles',
+    current: 'Plan actual',
+    recommended: 'Recomendado',
+    free: 'Gratis',
+    annualNote: 'Equivale a {{price}} / mes',
+    ownerOnly: 'Solo propietario o administrador',
+    choose: 'Continuar al pago',
+    handoff: 'Inicia sesión y paga en Space. Se conservará tu selección.',
+    quotas: 'Cuotas del espacio',
+    unlimited: 'Sin límite',
+    retention: 'Historial · {{count}} días',
+    retentionLabel: 'Retención del historial',
+    retentionHelp:
+      'El historial se elimina automáticamente al superar este periodo.',
+    credits: '{{count}} créditos al mes',
+    creditsLabel: 'Créditos mensuales',
+    creditsHelp:
+      'Los créditos se conceden cada mes de suscripción, también en los planes anuales.',
+    platforms: 'Plataformas principales compatibles',
+    excludedPlatforms: 'No compatibles: {{platforms}}',
+    allPlatforms: 'Todas las plataformas disponibles',
+    limits: {
+      bots: 'Bots',
+      pipelines: 'Procesadores',
+      knowledge_bases: 'Bases de conocimiento',
+      plugins: 'Extensiones',
+      members: 'Miembros',
+      managed_sandbox_sessions: 'Sesiones de sandbox gestionado',
+    },
+  },
+  workspaceLanguage: {
+    label: 'Idioma del espacio de trabajo',
+    description:
+      'Idioma de los mensajes del servidor y respuestas predeterminadas del bot. No cambia tu interfaz.',
+    saved: 'Idioma guardado',
+    saveFailed: 'No se pudo guardar el idioma',
+  },
+  inflight: {
+    autoCollapse: 'Plegar auto.',
+    empty: 'No hay eventos en proceso',
+    move: 'Arrastra verticalmente o usa las flechas',
+    botLogs: 'Registros del bot',
+    processorLogs: 'Registros del procesador',
+
+    title: 'En curso',
+    expand: 'Expandir',
+    collapse: 'Contraer',
+    reconnecting: 'Reconectando; se muestra el último estado recibido.',
+    details: 'Ver detalles',
+    debug: 'Evento de depuración',
+    limited:
+      'Se muestra un subconjunto. Consulte todas las ejecuciones en el panel.',
+    queued: 'En cola',
+    running: 'Procesando',
+    waiting: 'Esperando entrada',
+    tool: 'Ejecutando herramienta',
+    thinking: 'Pensando',
+    replying: 'Preparando respuesta',
+    generating: 'Generando respuesta',
+    completed: 'Completado',
+    failed: 'Fallido',
+    cancelled: 'Cancelado',
+    ignored: 'Ignorado',
+  },
   assistant: {
+    movePosition: 'Mover el asistente horizontalmente',
+    resizeHeight: 'Ajustar la altura del asistente',
+    sessions: 'Conversaciones',
+    currentChat: 'Conversación actual',
+    stop: 'Detener tarea',
+    stopped:
+      'Detenido. Una herramienta en curso puede haber surtido efecto. Comprueba el recurso antes de reintentar.',
+    setupRequired:
+      'Inicia sesión con tu cuenta de LangBot o configura un modelo para empezar a usar el asistente del espacio de trabajo.',
+    loginAccount: 'Iniciar sesión en LangBot Account',
+    configureModels: 'Configurar modelos',
+    providerCheckFailed:
+      'No se pudieron comprobar los proveedores. Inténtalo de nuevo.',
+    retrySetup: 'Reintentar',
+    preparing: 'Preparando asistente…',
+    thinking: 'Modelo respondiendo · Ronda {{round}}',
+    executing: 'Ejecutando herramienta…',
+    toolRunning: 'En ejecución',
+    recommendationFailed:
+      'No se pudo cargar el modelo recomendado. Selecciona un modelo.',
+    expand: 'Expandir asistente del espacio de trabajo',
+    collapse: 'Contraer asistente del espacio de trabajo',
     modelHint: 'Cambia el modelo del asistente para el próximo mensaje',
     details: 'Ver datos sin procesar',
     found: 'Se encontraron {{count}} elementos',
@@ -21,7 +193,7 @@ const esES = {
       create_knowledge_base: 'Crear base de conocimiento',
       get_pipeline: 'Consultar Pipeline',
       get_knowledge_schema: 'Consultar el esquema del motor de conocimiento',
-      list_operation_logs: 'Consultar la trazabilidad de operaciones',
+      list_operation_logs: 'Consultar el registro de operaciones',
     },
     resources: {
       models: 'Buscar modelos de chat',
@@ -35,7 +207,7 @@ const esES = {
     newChat: 'Nueva conversación',
     close: 'Cerrar',
     welcome:
-      'Dime qué quieres crear. Puedo inspeccionar este espacio de trabajo y proponer cambios en la base de conocimiento y el Pipeline para que los confirmes.',
+      'Dime qué quieres crear. Puedo revisar este espacio de trabajo y proponer configuraciones para bots, Agents, bases de conocimiento, Pipelines y más, y aplicarlas tras tu confirmación.',
     discover:
       '¿Qué modelos, bases de conocimiento y Pipelines hay disponibles?',
     build:
@@ -57,6 +229,14 @@ const esES = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: 'Últimos {{days}} días',
+        customRange: 'Periodo personalizado',
+        rangeStart: 'Inicio',
+        rangeEnd: 'Fin',
+        rangeHint: 'El inicio debe preceder al fin. Máximo: 365 días.',
+        autoRefreshLabel: 'Actualización automática',
+        autoRefreshOff: 'Actualización: desactivada',
+        autoRefreshEvery: 'Actualizar cada {{seconds}} s',
         title: 'Revisar la actividad',
         description:
           'Consulta la actividad de los bots, el uso de modelos, los mensajes y el rendimiento del sistema.',
@@ -138,6 +318,14 @@ const esES = {
           'Envía mensajes de prueba para comprobar los activadores, las respuestas de IA y el procesamiento de salida. Los cambios se guardan antes de la prueba.',
       },
       monitoring: {
+        rangeDays: 'Últimos {{days}} días',
+        customRange: 'Periodo personalizado',
+        rangeStart: 'Inicio',
+        rangeEnd: 'Fin',
+        rangeHint: 'El inicio debe preceder al fin. Máximo: 365 días.',
+        autoRefreshLabel: 'Actualización automática',
+        autoRefreshOff: 'Actualización: desactivada',
+        autoRefreshEvery: 'Actualizar cada {{seconds}} s',
         title: 'Revisar el historial',
         description:
           'Usa la pestaña junto al título para revisar conversaciones, pasos de procesamiento y errores cuando falten respuestas o sean inesperadas.',
@@ -305,6 +493,11 @@ const esES = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'No se registró el motivo del fallo. Consulta los registros del bot o vuelve a enviar el mensaje para obtener más detalles.',
+    loadFailedTitle: 'No se pudo cargar el contenido',
+    loadFailedHint:
+      'Inténtalo de nuevo. Si el problema persiste, comprueba la conexión y el estado del servicio.',
     customValue: 'Personalizado',
     loadFailed: 'No se pudo cargar. Inténtalo de nuevo.',
     search: 'Buscar',
@@ -507,6 +700,9 @@ const esES = {
     help: 'Obtener ayuda',
   },
   models: {
+    starModel: 'Marcar como favorito para nuevos procesadores',
+    unstarModel: 'Quitar de favoritos',
+    starModelFailed: 'No se pudo actualizar el modelo favorito',
     codex: {
       account: 'Suscripción de ChatGPT',
       description:
@@ -775,10 +971,13 @@ const esES = {
     deleteConfirmation: '¿Estás seguro de que deseas eliminar este Bot?',
     platformAdapter: 'Selección de plataforma/adaptador',
     selectAdapter: 'Seleccionar adaptador',
+    saveAndEnable: 'Guardar y activar',
+    searchAdapters: 'Buscar adaptadores',
+    noMatchingAdapters: 'No se encontraron adaptadores coincidentes',
     legacyAdapters: 'Adaptadores heredados',
     legacyAdapterBadge: 'Heredado',
     legacyAdaptersHint:
-      'Estos adaptadores tienen versiones nuevas orientadas a eventos.\nSe mantienen solo para configuraciones existentes y no se recomiendan para Bots nuevos.',
+      'Los adaptadores antiguos se centran en mensajes y conservan la compatibilidad con bots existentes.\nLos nuevos conectan la misma plataforma mediante eventos, incluidos mensajes y solicitudes de amistad o cambios de grupo compatibles.\nUsa los nuevos para bots nuevos. Los existentes pueden seguir con los antiguos. Al cambiar, revisa la conexión y las rutas; no se actualiza la cuenta de la plataforma.',
     adapterConfig: 'Configuración del adaptador',
     viewAdapterDocs: 'Ver documentación',
     bindPipeline: 'Vincular Pipeline',
@@ -791,6 +990,10 @@ const esES = {
     earlier: 'Anterior',
     dateFormat: '{{day}}/{{month}}',
     setBotEnableError: 'Error al establecer el estado de activación del Bot',
+    enabling: 'Activando…',
+    disabling: 'Desactivando…',
+    enableConfirmed: 'Activado',
+    disableConfirmed: 'Desactivado',
     log: 'Registro',
     configuration: 'Configuración',
     logs: 'Registros',
@@ -799,6 +1002,28 @@ const esES = {
     routingConnection: 'Enrutamiento y conexión',
     routingConnectionDescription:
       'Vincula el Pipeline que procesa los mensajes de este Bot',
+    routingHelp: {
+      title: '¿Qué es el enrutamiento de eventos?',
+      intro:
+        'Los bots reciben mensajes, solicitudes de amistad y eventos como unirse a un grupo.',
+      events: 'Eventos recibidos',
+      agents: 'Distintos agentes',
+      message: 'Mensaje recibido',
+      friend: 'Solicitud de amistad',
+      group: 'Entrada a un grupo',
+      messageAction: 'Agente de respuestas',
+      friendAction: 'Agente de revisión',
+      groupAction: 'Agente de bienvenida',
+      rules: 'Según tus reglas · Ejemplo',
+      instructions:
+        'Los agentes siguen tus instrucciones y usan herramientas para responder o revisar solicitudes.',
+      plugin:
+        'Para casos avanzados, los procesadores de plugins registran eventos en lote y los gestionan con código. Añade su configuración abajo.',
+      docs: 'Desarrolla tu propio procesador de plugins',
+      runnerDocs: 'Desarrolla tu propio ejecutor de Agent',
+      pause: 'Pausar',
+      play: 'Reproducir',
+    },
     eventRouting: 'Enrutamiento de eventos',
     eventRoutingDescription:
       'Elige qué procesador maneja cada evento recibido por este Bot. Edita la lógica en la configuración del Agent o Pipeline correspondiente. Los Pipelines solo admiten eventos de mensaje.',
@@ -839,7 +1064,8 @@ const esES = {
     enabled: 'Activado',
     eventBindingDescriptionPlaceholder: 'Descripción de la regla',
     noEventBindings: 'No hay rutas de eventos',
-    unsupportedPipelineEvent: 'Los flujos solo admiten eventos message.*',
+    unsupportedPipelineEvent:
+      'Los flujos solo admiten eventos message.received',
     disable: 'Desactivar',
     enable: 'Activar',
     disabledBindings: 'Desactivado',
@@ -849,6 +1075,7 @@ const esES = {
     advancedEventValues: 'Ver todos',
     eventGroup: 'Grupo',
     eventGroupNames: {
+      wecomcs: 'Atención al cliente de WeCom',
       bot: 'Estado del bot',
       feedback: 'Comentarios',
       friend: 'Amigos',
@@ -863,6 +1090,9 @@ const esES = {
     routeFallbackCatchAll: '{{route}} es la ruta general.',
     routeFallbackIgnored:
       'Se ignoran los eventos sin ruta coincidente. Añade una ruta general solo si todos los eventos necesitan un resultado explícito.',
+    matchRecords: 'Registros de coincidencias',
+    matchRecordsDescription:
+      'Último resultado de cada regla y eventos recientes sin coincidencia. Los registros en memoria se borran al reiniciar.',
     testRoute: 'Comprobar ruta',
     adapterEventDebugAction: 'Probar escucha',
     adapterEventDebugTitle: 'Depuración de eventos de plataforma',
@@ -949,6 +1179,15 @@ const esES = {
     eventWildcard: 'Todos los eventos',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'Usuario entró al chat del bot',
+      wecombot_template_card_event: 'Interacción con tarjeta',
+      dingtalk_card_action: 'Interacción con tarjeta de DingTalk',
+      wecomcs_enter_session: 'Cliente entró en la conversación',
+      wecomcs_msg_send_fail: 'Fallo de envío de atención al cliente',
+      wecomcs_servicer_status_change: 'Disponibilidad del agente modificada',
+      wecomcs_session_status_change: 'Estado de conversación modificado',
+      wecomcs_reject_customer_msg_switch_change:
+        'Rechazo de mensajes modificado',
       message_received: 'Mensaje recibido',
       message_edited: 'Mensaje editado',
       message_deleted: 'Mensaje eliminado',
@@ -958,6 +1197,7 @@ const esES = {
       friend_added: 'Amigo añadido',
       group_member_joined: 'Miembro unido al grupo',
       group_member_left: 'Miembro salió del grupo',
+      group_info_updated: 'Información del grupo actualizada',
       group_member_banned: 'Miembro bloqueado',
       bot_invited_to_group: 'Bot invitado al grupo',
       bot_removed_from_group: 'Bot eliminado del grupo',
@@ -966,6 +1206,22 @@ const esES = {
       platform_specific: 'Evento específico de la plataforma',
     },
     eventDescriptions: {
+      wecombot_enter_chat:
+        'Un usuario abrió el chat del bot. Úsalo para flujos de bienvenida.',
+      wecombot_template_card_event:
+        'Un usuario pulsó o envió una tarjeta. Las respuestas de formularios gestionados se procesan por separado.',
+      dingtalk_card_action:
+        'Clics y envíos de tarjetas; las valoraciones y los formularios gestionados se procesan por separado.',
+      wecomcs_enter_session:
+        'Un cliente abrió el enlace de atención; incluye origen y código de bienvenida.',
+      wecomcs_msg_send_fail:
+        'WeChat notificó un fallo de envío con el ID y el tipo de error.',
+      wecomcs_servicer_status_change:
+        'Un agente inició, detuvo o suspendió la atención.',
+      wecomcs_session_status_change:
+        'Un agente aceptó, transfirió, cerró o reabrió la conversación.',
+      wecomcs_reject_customer_msg_switch_change:
+        'Un agente activó o desactivó el rechazo de mensajes del cliente.',
       all: 'Coincide con todos los eventos recibidos por este adaptador.',
       namespace: 'Coincide con todos los eventos de {{group}}.',
       namespace_bot:
@@ -974,7 +1230,7 @@ const esES = {
         'Coincide con comentarios recibidos de la plataforma o de usuarios.',
       namespace_friend: 'Coincide con solicitudes y cambios de amistad.',
       namespace_group:
-        'Coincide con entradas, salidas, expulsiones y otros eventos de grupo.',
+        'Coincide con cambios de miembros y actualizaciones de información del grupo.',
       namespace_message:
         'Coincide con mensajes recibidos, editados, eliminados y reacciones.',
       namespace_platform:
@@ -991,6 +1247,8 @@ const esES = {
       friend_added: 'Se ha creado una relación de amistad.',
       group_member_joined: 'Un miembro se une a un grupo donde está el bot.',
       group_member_left: 'Un miembro abandona un grupo donde está el bot.',
+      group_info_updated:
+        'Han cambiado datos del grupo, como su nombre o propietario.',
       group_member_banned: 'Un miembro del grupo es bloqueado o expulsado.',
       bot_invited_to_group: 'El bot es invitado a un grupo.',
       bot_removed_from_group: 'El bot es expulsado de un grupo.',
@@ -1078,6 +1336,13 @@ const esES = {
     imagesAttached: 'imagen(es) adjunta(s)',
     noLogs: 'Aún no hay registros',
     sessionMonitor: {
+      resetContext: 'Restablecer contexto',
+      resetDescription:
+        'Los mensajes nuevos comenzarán sin contexto previo. Se conservan el historial de monitoreo, los archivos y la memoria a largo plazo. Espere a que terminen las tareas activas.',
+      resetSuccess: 'Contexto restablecido.',
+      resetError:
+        'No se pudo restablecer. Espere a que terminen las tareas activas y vuelva a intentarlo.',
+
       title: 'Sesiones',
       sessions: 'Sesiones',
       noSessions: 'No se encontraron sesiones',
@@ -1125,7 +1390,26 @@ const esES = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'Seleccionar todo',
     monitoring: {
+      rangeDays: 'Últimos {{days}} días',
+      customRange: 'Periodo personalizado',
+      rangeStart: 'Inicio',
+      rangeEnd: 'Fin',
+      rangeHint: 'El inicio debe preceder al fin. Máximo: 365 días.',
+      autoRefreshLabel: 'Actualización automática',
+      autoRefreshOff: 'Actualización: desactivada',
+      autoRefreshEvery: 'Actualizar cada {{seconds}} s',
+      runStatus: 'Estado de ejecución',
+      toolCalls: 'Llamadas a herramientas',
+      delivery: 'Envío de mensajes',
+      sendFailed: 'Error de envío',
+      sendCount: '{{count}} llamadas de envío correctas',
+      noSend: 'Sin envíos registrados',
+      noSteps: 'Aún no hay pasos registrados.',
+      recipient: 'Destinatario',
+      actionCompleted: 'Herramienta ejecutada correctamente.',
+      simulatedAction: 'Acción simulada; no se envió ningún mensaje real.',
       description:
         'Consulta el evento, la salida del modelo y las herramientas de cada tarea.',
       empty:
@@ -1232,6 +1516,9 @@ const esES = {
       'Sigue un flujo fijo: recibir un mensaje, consultar a la IA y responder al usuario, con bases de conocimiento y plugins configurables. Solo procesa eventos de mensaje, para tareas con pasos claros y control del proceso.',
     allEvents: 'Compatible con todos los eventos',
     messageEventsOnly: 'Solo eventos de mensaje',
+    selectAgentRunner: 'Seleccionar ejecutor de Agent',
+    selectProcessorPlugin: 'Seleccionar plugin de procesamiento',
+    createBasicInfoDescription: 'Configura el nombre, icono y plugin',
     chooseType: 'Elegir cómo funciona',
     chooseTypeDescription:
       'Elige primero el tipo de procesador. Podrás configurarlo después de crearlo.',
@@ -1655,6 +1942,11 @@ const esES = {
     },
   },
   market: {
+    certifiedPlugin: 'Plugin certificado',
+    certificationTooltip:
+      'Esta versión ha superado la revisión de aislamiento de datos y compatibilidad con entornos compartidos. El paquete está firmado.',
+    certificationLearnMore: 'Acerca de la certificación',
+
     runnerUsage: 'Uso del ejecutor',
     runnerUsageAll: 'Todos',
     runnerUsageAgent: 'Agent / Flujo',
@@ -2058,6 +2350,14 @@ const esES = {
       uploading: 'Subiendo...',
     },
     monitoring: {
+      rangeDays: 'Últimos {{days}} días',
+      customRange: 'Periodo personalizado',
+      rangeStart: 'Inicio',
+      rangeEnd: 'Fin',
+      rangeHint: 'El inicio debe preceder al fin. Máximo: 365 días.',
+      autoRefreshLabel: 'Actualización automática',
+      autoRefreshOff: 'Actualización: desactivada',
+      autoRefreshEvery: 'Actualizar cada {{seconds}} s',
       title: 'Registros de ejecución',
       workbench: 'Configurar y depurar',
       description:
@@ -2367,7 +2667,7 @@ const esES = {
     totpStatusDisabled: 'No activada',
     totpCodesRemaining: 'Quedan {{count}} códigos de recuperación',
     totpManagerSectionDesc:
-      'Los propietarios y administradores pueden revisar y restablecer el segundo factor de cualquier cuenta.',
+      'Los propietarios y administradores pueden revisar y restablecer el segundo factor de las cuentas de este espacio de trabajo.',
     revokeTotp: 'Reasignar',
     totpAdminResetTitle: 'Reasignar la verificación en dos pasos de {{user}}',
     totpAdminResetDesc:
@@ -2433,8 +2733,7 @@ const esES = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2503,6 +2802,124 @@ const esES = {
     },
   },
   monitoring: {
+    rangeDays: 'Últimos {{days}} días',
+    customRange: 'Periodo personalizado',
+    rangeStart: 'Inicio',
+    rangeEnd: 'Fin',
+    rangeHint: 'El inicio debe preceder al fin. Máximo: 365 días.',
+    autoRefreshLabel: 'Actualización automática',
+    autoRefreshOff: 'Actualización: desactivada',
+    autoRefreshEvery: 'Actualizar cada {{seconds}} s',
+    unified: {
+      inputs: 'Evento y entrada',
+      outputs: 'Resultados generados',
+      deliveries: 'Registros de envío',
+      conversation: 'Contexto de conversación',
+      related: 'Ejecuciones relacionadas',
+      events: 'Eventos de ejecución',
+      llm_calls: 'Llamadas al modelo',
+      tool_calls: 'Llamadas a herramientas',
+      errors: 'Errores',
+      payload: 'Detalles',
+      metadata: 'Información técnica',
+      loadMore: 'Cargar más',
+      noRecords: 'Sin registros',
+      contextHint:
+        'Otros mensajes no pertenecen necesariamente a esta ejecución.',
+      unhandled: 'Sin ejecución',
+      roles: {
+        user: 'Usuario',
+        assistant: 'Asistente',
+        system: 'Sistema',
+        tool: 'Herramienta',
+      },
+    },
+    execution: {
+      title: 'Ejecuciones',
+      subtitle:
+        'Cada evento entrante y la ejecución que activó, en agentes y pipelines.',
+      tabs: { executions: 'Ejecuciones' },
+      filters: {
+        source: 'Tipo',
+        sourceAll: 'Todos los tipos',
+        sourceAgent: 'Agentes',
+        sourcePipeline: 'Pipelines',
+        mode: 'Tráfico',
+        modeAll: 'Real + depuración',
+        modeLive: 'Solo real',
+        modeDebug: 'Solo depuración',
+        status: 'Estado',
+        statusAll: 'Todos los estados',
+      },
+      status: {
+        completed: 'Completada',
+        failed: 'Fallida',
+        running: 'En curso',
+        queued: 'En cola',
+        cancelled: 'Cancelada',
+        ignored: 'Ignorada',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Agente',
+        pipeline: 'Pipeline',
+        processor: 'Procesador',
+      },
+      columns: {
+        time: 'Hora',
+        target: 'Procesador',
+        status: 'Estado',
+        duration: 'Duración',
+        tokens: 'Tokens',
+      },
+      debug: 'Depuración',
+      empty: 'No hay ejecuciones en este intervalo',
+      loading: 'Cargando ejecuciones...',
+      triggeredBy: 'Activada por {{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95: cerca del 50%/95% termina dentro de este tiempo. Solo ejecuciones recientes de Agent completadas, sin espera en cola.',
+        tokensHint:
+          'Solo se suma el uso registrado. El total puede estar incompleto si el modelo no devuelve el uso.',
+        totalExecutions: 'Ejecuciones',
+        bySource: '{{agent}} de agentes · {{pipeline}} de pipelines',
+        successRate: 'Tasa de éxito de ejecución',
+        successRateHint:
+          'Éxitos ÷ (éxitos + fallos). Se excluyen ejecuciones en curso, ignoradas y canceladas.',
+        denominator: '{{completed}} de {{denominator}} finalizadas',
+        running: 'En curso',
+        queued: 'En cola',
+        waiting: 'Esperando entrada',
+        failed: 'Fallida',
+        ignored: 'Ignorada',
+        latency: 'Tiempo de ejecución',
+        p50: 'P50',
+        p95: 'P95',
+        sample: 'muestra de {{count}} ejecuciones',
+        tokens: 'Tokens',
+        coverage: 'uso registrado en {{withUsage}}/{{calls}} llamadas',
+        debugRuns: 'Incluye {{count}} ejecuciones de depuración',
+      },
+      detail: {
+        title: 'Ejecución',
+        timeline: 'Cronología',
+        executionId: 'ID de ejecución',
+        eventId: 'Evento',
+        queue: 'Cola',
+        runner: 'Runner',
+        created: 'Creada',
+        duration: 'Duración',
+        usage: 'Uso',
+        noEvents: 'No se registraron eventos de traza',
+        query: 'Mensaje entrante',
+        modelCalls: 'Llamadas al modelo',
+        toolCalls: 'Llamadas a herramientas',
+        errors: 'Errores',
+        loadError: 'No se pudo cargar la traza de ejecución',
+        openObject: 'Abrir objeto',
+        a11yDescription: 'Traza completa de esta ejecución',
+      },
+    },
     title: 'Panel de control',
     description:
       'Monitorea las actividades de los Bots, llamadas LLM y rendimiento del sistema',
@@ -2519,6 +2936,9 @@ const esES = {
       pipeline: 'Pipeline',
       allBots: 'Todos los Bots',
       selectBot: 'Seleccionar Bot',
+      processor: 'Procesador',
+      allProcessors: 'Todos los procesadores',
+      selectProcessor: 'Seleccionar procesador',
       allPipelines: 'Todos los Pipelines',
       selectPipeline: 'Seleccionar Pipeline',
       loading: 'Cargando...',
@@ -2544,6 +2964,15 @@ const esES = {
       sessions: 'Análisis de sesiones',
       errors: 'Registros de errores',
     },
+    pipelineTrace: {
+      received: 'Mensaje recibido',
+      processing: 'Procesamiento',
+      replies: 'Respuestas',
+      noSteps: 'No se registraron llamadas a modelos o herramientas',
+      success: 'Correcto',
+      error: 'Error',
+      pending: 'Pendiente',
+    },
     messageList: {
       timestamp: 'Marca de tiempo',
       bot: 'Bot',
@@ -2768,7 +3197,7 @@ const esES = {
     title: 'Configuración',
     nav: {
       workspace: 'Workspace',
-      operationTrace: 'Trazabilidad',
+      operationTrace: 'Registro',
       models: 'Modelos',
       api: 'API',
       storage: 'Almacenamiento',
@@ -2776,21 +3205,21 @@ const esES = {
     },
   },
   operationTrace: {
-    title: 'Trazabilidad de operaciones',
+    title: 'Registro de operaciones',
     description: 'Rastrea cambios y consultas de los administradores',
+    reset: 'Restablecer',
+    settingsUpdated: 'Configuración guardada',
+    settingsUpdateFailed: 'No se pudo guardar la configuración',
+    levelBadge: 'Nivel: {{level}}',
+    collectionSettings: 'Configuración de recopilación',
     captureLevel: 'Nivel de captura',
     retention: 'Retención',
-    retentionSummary: 'Conservar {{days}} días · hasta {{rows}} registros',
     retentionDays: 'Días de retención',
     maxRows: 'Registros máximos',
     dedupeWindow: 'Ventana de deduplicación (s)',
     records: 'Registros de operación',
     changesCount: '{{count}} cambio(s)',
     hideDetails: 'Ocultar detalles',
-    levelUpdated: 'Nivel de captura actualizado',
-    levelUpdateFailed: 'No se pudo actualizar el nivel de captura',
-    retentionUpdated: 'Retención actualizada',
-    retentionUpdateFailed: 'No se pudo actualizar la retención',
     export: 'Descargar registros',
     exportFailed: 'No se pudieron descargar los registros',
     empty: 'Aún no hay registros de operación',
@@ -2804,7 +3233,7 @@ const esES = {
     filterAllResources: 'Todos los recursos',
     filterActor: 'Actor',
     filterAllActors: 'Todos los actores',
-    loadFailed: 'No se pudieron cargar los datos de trazabilidad',
+    loadFailed: 'No se pudieron cargar los registros de operaciones',
     refresh: 'Actualizar',
     save: 'Guardar',
     redacted: 'Oculto',
@@ -2823,7 +3252,7 @@ const esES = {
     elidedValue: '{{count}} caracteres',
     levels: {
       off: {
-        label: 'Sin trazabilidad',
+        label: 'Sin registro',
       },
       mutation: {
         label: 'Solo cambios',
diff --git a/web/src/i18n/locales/ja-JP.ts b/web/src/i18n/locales/ja-JP.ts
index 0a79af4a0..ec7e6fb77 100644
--- a/web/src/i18n/locales/ja-JP.ts
+++ b/web/src/i18n/locales/ja-JP.ts
@@ -1,6 +1,178 @@
+import slackSetup from './slack-setup/ja-JP';
 import pipelineMigration from './pipeline-migration/ja-JP';
 const jaJP = {
+  planRevision: {
+    subscribedOn: '契約日：{{date}}',
+    latestQuota: 'このプランの最新の利用枠',
+    showLatest: 'このプランの最新クォータを表示',
+    showCurrent: '現在のサブスクリプションを表示',
+    latestTab: '最新',
+    currentTab: '現在',
+    cancelToSubscribe:
+      '現在のサブスクリプションをキャンセルすると、次期から最新のクォータに登録できます',
+
+    current: '現在の契約：{{plan}} · revision {{revision}}',
+    owned: '購入済み revision {{revision}}',
+    latest: '最新 revision {{revision}}',
+    changed:
+      '購入済みの特典と最新バージョンに違いがあります。変更箇所を強調表示しています。',
+    same: '特典は最新 revision {{revision}} と同じです。購入済みバージョンは変更されません。',
+    unavailable:
+      '最新プランは現在利用できません。購入済みの特典は引き続き有効です。',
+    benefit: '特典',
+    change: '変更あり',
+    enabled: '有効',
+    disabled: '無効',
+    monthlyCredits: '毎月のボーナスクレジット',
+    retention: 'チャット履歴の保存日数',
+    unknown:
+      '現在の契約情報を取得できません。再読み込みするか Space で契約を管理してください。最新カタログの特典は購入済み特典ではありません。',
+    handoff:
+      'Space で契約を管理してください。新バージョンの購入前に自動更新を解除してください。前払い契約は期限前に更新しなければ自動終了します。',
+    cancelFirst:
+      'まず契約管理で自動更新を解除してください。支払済みの特典は期間終了まで保持され、返金はありません。解除の確認後、最新バージョンを購入でき、決済成功時に新しい特典が有効になります。',
+    canceled:
+      '更新は解除済みです。支払済みの特典は期間終了まで保持され、返金はありません。新しい購入の特典は決済成功時に有効になります。',
+    prepaid:
+      'この前払い契約は期間終了時に自動終了します。期限前は現在の内容で更新でき、期限後は各プランの最新の利用枠でのみ契約できます。',
+    renew: '現在の契約を更新',
+    buying: 'revision {{revision}} を購入します。',
+    fields: {
+      bots_max: 'ボット',
+      members_max: 'メンバー',
+      pipelines_max: 'プロセッサー',
+      plugins_max: '拡張機能',
+      knowledge_bases_max: 'ナレッジベース',
+      managed_sandbox_sessions_max: 'サンドボックスセッション',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: '現在お申し込みできません',
+    subscriptionUnavailableHint:
+      'このプランは新規申込を停止しています。他のプランをお選びください。既存の契約と自動更新には影響しません。',
+    platformNames: {
+      qqofficial: 'QQ公式API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComカスタマーサービス',
+      officialaccount: 'WeChat公式アカウント',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'サブスクリプションを管理',
+    downgradeUnavailable: '無料プランへの変更はできません',
+    features: {
+      multiUserWorkspace: '複数ユーザーのワークスペース',
+      remoteE2BSandbox: 'リモート E2B サンドボックスの設定',
+      communitySupport: 'コミュニティサポート',
+      everythingInFree: '無料プランのすべての機能',
+      alwaysOnSandbox: '常時稼働のグローバルサンドボックス 1 個',
+      higherWorkspaceLimits: 'ワークスペースの上限を拡大',
+    },
+    title: 'プランを比較',
+    description: '{{workspace}} のプランを選択',
+    billing: '請求周期',
+    monthly: '月払い',
+    annual: '年払い',
+    currency: '通貨',
+    loading: 'プランを読み込み中…',
+    error: 'プランを読み込めませんでした',
+    retry: '再試行',
+    empty: '利用可能なプランはありません',
+    current: '現在のプラン',
+    recommended: 'おすすめ',
+    free: '無料',
+    annualNote: '月額 {{price}} 相当',
+    ownerOnly: '所有者または管理者のみ',
+    choose: '支払いに進む',
+    handoff:
+      'LangBot アカウントでログインしてお支払いください。選択内容は保持されます。',
+    quotas: 'ワークスペースの上限',
+    unlimited: '無制限',
+    retention: 'メッセージ履歴 · {{count}} 日間',
+    retentionLabel: 'メッセージ履歴の保存期間',
+    retentionHelp: '保存期間を過ぎた履歴は自動的に削除されます。',
+    credits: '毎月 {{count}} クレジット',
+    creditsLabel: '毎月のクレジット',
+    creditsHelp: '年払いでも、契約の各月にクレジットが付与されます。',
+    platforms: '主要プラットフォームに対応',
+    excludedPlatforms: '対象外：{{platforms}}',
+    allPlatforms: '現在利用可能なすべてのプラットフォーム',
+    limits: {
+      bots: 'ボット',
+      pipelines: 'プロセッサー',
+      knowledge_bases: 'ナレッジベース',
+      plugins: '拡張機能',
+      members: 'メンバー',
+      managed_sandbox_sessions: '管理サンドボックスセッション',
+    },
+  },
+  workspaceLanguage: {
+    label: 'ワークスペースの言語',
+    description:
+      'バックエンドの通知とボットの既定の返信に使用します。個人の表示言語には影響しません。',
+    saved: '言語設定を保存しました',
+    saveFailed: '言語設定の保存に失敗しました',
+  },
+  slackSetup,
+  inflight: {
+    autoCollapse: '自動折りたたみ',
+    empty: '処理中のイベントはありません',
+    move: '上下にドラッグ、または矢印キーで移動',
+    botLogs: 'ボットのログ',
+    processorLogs: 'プロセッサーのログ',
+
+    title: '実行中',
+    expand: '展開',
+    collapse: '折りたたむ',
+    reconnecting: '再接続中。最後に受信した状態を表示しています。',
+    details: '詳細を見る',
+    debug: 'デバッグイベント',
+    limited: '一部のみ表示しています。全実行はダッシュボードで確認できます。',
+    queued: '待機中',
+    running: '処理中',
+    waiting: '入力待ち',
+    tool: 'ツール呼び出し中',
+    thinking: '思考中',
+    replying: '返信準備中',
+    generating: '返信生成中',
+    completed: '完了',
+    failed: '失敗',
+    cancelled: 'キャンセル済み',
+    ignored: '無視',
+  },
   assistant: {
+    movePosition: 'アシスタントを左右に移動',
+    resizeHeight: 'アシスタントの高さを調整',
+    sessions: '会話を切り替え',
+    currentChat: '現在の会話',
+    stop: 'タスクを停止',
+    stopped:
+      '停止しました。実行中のツールが反映済みの場合があります。再試行前にリソースを確認してください。',
+    setupRequired:
+      'LangBot アカウントにログインするか、モデルを設定すると、ワークスペースアシスタントを利用できます。',
+    loginAccount: 'LangBot Account にログイン',
+    configureModels: 'モデルを設定',
+    providerCheckFailed:
+      'モデルプロバイダーを確認できません。再試行してください。',
+    retrySetup: '再試行',
+    preparing: '準備中…',
+    thinking: 'モデル応答中 · {{round}} 回目',
+    executing: 'ツール実行中…',
+    toolRunning: '実行中',
+    recommendationFailed:
+      '推奨モデルを取得できません。モデルを選択してください。',
+    expand: 'ワークスペースアシスタントを展開',
+    collapse: 'ワークスペースアシスタントを折りたたむ',
     modelHint: '次のメッセージからモデルを切り替え',
     details: '元のデータを見る',
     found: '{{count}} 件見つかりました',
@@ -21,7 +193,7 @@ const jaJP = {
       create_knowledge_base: 'ナレッジベース作成',
       get_pipeline: 'Pipeline 参照',
       get_knowledge_schema: 'エンジン設定の参照',
-      list_operation_logs: '操作トレーサビリティの参照',
+      list_operation_logs: '操作ログの参照',
     },
     resources: {
       models: 'チャットモデル検索',
@@ -35,7 +207,7 @@ const jaJP = {
     newChat: '新しい会話',
     close: '閉じる',
     welcome:
-      '作りたいアプリを教えてください。現在のリソースを確認し、承認後にナレッジベースや Pipeline を設定します。',
+      '作りたいアプリを教えてください。現在のワークスペースを確認し、ボット、Agent、ナレッジベース、Pipeline などの設定を提案し、承認後に適用します。',
     discover: '利用できるモデル、ナレッジベース、Pipeline は？',
     build: 'ナレッジベースを使う質問応答アプリを作りたい。',
     toolResult: '操作結果',
@@ -55,6 +227,14 @@ const jaJP = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: '過去 {{days}} 日間',
+        customRange: '期間を指定',
+        rangeStart: '開始日時',
+        rangeEnd: '終了日時',
+        rangeHint: '開始は終了より前、期間は最大365日です。',
+        autoRefreshLabel: '自動更新',
+        autoRefreshOff: '自動更新：オフ',
+        autoRefreshEvery: '{{seconds}} 秒ごとに更新',
         title: '稼働状況を確認する',
         description:
           'ボットの稼働状況、モデルの使用量、メッセージ数、システムの状態をまとめて確認できます。',
@@ -136,6 +316,14 @@ const jaJP = {
           'テストメッセージで起動条件、AI の返信、出力処理を確認します。設定の変更はテスト前に保存されます。',
       },
       monitoring: {
+        rangeDays: '過去 {{days}} 日間',
+        customRange: '期間を指定',
+        rangeStart: '開始日時',
+        rangeEnd: '終了日時',
+        rangeHint: '開始は終了より前、期間は最大365日です。',
+        autoRefreshLabel: '自動更新',
+        autoRefreshOff: '自動更新：オフ',
+        autoRefreshEvery: '{{seconds}} 秒ごとに更新',
         title: '実行履歴を確認する',
         description:
           'タイトル横のタブで過去の会話、処理過程、エラーを確認し、返信がない場合や想定外の回答を調べます。',
@@ -301,6 +489,11 @@ const jaJP = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'この記録には失敗の理由が保存されていません。ボットのログを確認するか、再送信して詳細を確認してください。',
+    loadFailedTitle: '読み込めませんでした',
+    loadFailedHint:
+      'しばらくしてから再試行してください。解決しない場合は、接続とサービスの状態を確認してください。',
     customValue: 'カスタム',
     loadFailed: '読み込みに失敗しました。再試行してください。',
     search: '検索',
@@ -501,6 +694,9 @@ const jaJP = {
     help: 'ヘルプドキュメントを見る',
   },
   models: {
+    starModel: '新規処理器の既定モデルに設定',
+    unstarModel: 'スターを解除',
+    starModelFailed: 'スター付きモデルを更新できませんでした',
     codex: {
       account: 'ChatGPT サブスクリプション',
       description:
@@ -762,10 +958,13 @@ const jaJP = {
     deleteConfirmation: '本当にこのボットを削除しますか？',
     platformAdapter: 'プラットフォーム/アダプター選択',
     selectAdapter: 'アダプターを選択',
+    saveAndEnable: '保存して有効化',
+    searchAdapters: 'アダプターを検索',
+    noMatchingAdapters: '一致するアダプターが見つかりません',
     legacyAdapters: '旧式アダプター',
     legacyAdapterBadge: '旧式',
     legacyAdaptersHint:
-      'これらのアダプターには新しいイベント駆動の接続方式があります。\n既存設定との互換性のために残しており、新規ボットには推奨されません。',
+      '旧版は主にメッセージ送受信用で、既存の設定との互換性のために残されています。\n新版は同じプラットフォームにイベント方式で接続し、対応する友達申請やグループの変化も処理器へ振り分けます。\n新規ボットには新版を推奨します。既存ボットは旧版を継続利用できます。切り替え時は接続設定とイベントルートを確認してください。アカウントの更新ではありません。',
     adapterConfig: 'アダプター設定',
     viewAdapterDocs: 'ドキュメントを見る',
     bindPipeline: 'パイプラインを紐付け',
@@ -778,6 +977,10 @@ const jaJP = {
     earlier: 'それ以前',
     dateFormat: '{{month}}月{{day}}日',
     setBotEnableError: 'ボットの有効状態の設定に失敗しました',
+    enabling: '有効にしています…',
+    disabling: '無効にしています…',
+    enableConfirmed: '有効',
+    disableConfirmed: '無効',
     log: 'ログ',
     configuration: '設定',
     logs: 'ログ',
@@ -786,6 +989,28 @@ const jaJP = {
     routingConnection: 'ルーティングと接続',
     routingConnectionDescription:
       'このボットのメッセージを処理するパイプラインを紐付け',
+    routingHelp: {
+      title: 'イベントルーティングとは？',
+      intro:
+        'ボットはメッセージのほか、友達リクエストやグループ参加などのイベントも受け取ります。',
+      events: '受信イベント',
+      agents: '各 Agent へ',
+      message: 'メッセージ受信',
+      friend: '友達リクエスト',
+      group: 'グループ参加',
+      messageAction: '応答 Agent → 返信',
+      friendAction: '審査 Agent → 確認',
+      groupAction: '歓迎 Agent → 挨拶',
+      rules: '設定したルールで振り分け · 例',
+      instructions:
+        'Agent は指示に従ってツールを使い、返信や申請の確認を行います。',
+      plugin:
+        '高度な用途にはプラグイン処理器を利用できます。イベントを一括登録してコードで処理し、下の欄で設定を追加します。',
+      docs: '独自のプラグイン処理器を開発',
+      runnerDocs: '独自の Agent ランナーを開発',
+      pause: '一時停止',
+      play: '再生',
+    },
     eventRouting: 'イベントルーティング',
     eventRoutingDescription: 'イベントごとの処理先を設定します。',
     eventBindings: 'イベントルート',
@@ -826,7 +1051,7 @@ const jaJP = {
     eventBindingDescriptionPlaceholder: 'ルール説明',
     noEventBindings: 'イベントルートはありません',
     unsupportedPipelineEvent:
-      'Pipeline は message.* イベントにのみ使用できます',
+      'Pipeline は message.received イベントにのみ使用できます',
     disable: '無効化',
     enable: '有効化',
     disabledBindings: '無効',
@@ -836,6 +1061,7 @@ const jaJP = {
     advancedEventValues: 'すべて表示',
     eventGroup: 'グループ',
     eventGroupNames: {
+      wecomcs: 'WeCom カスタマーサービス',
       bot: 'ボットの状態',
       feedback: 'フィードバック',
       friend: '友だち',
@@ -850,6 +1076,9 @@ const jaJP = {
     routeFallbackCatchAll: '{{route}} はフォールバックルートです。',
     routeFallbackIgnored:
       'どのルートにも一致しないイベントは無視されます。すべてのイベントに明示的な結果が必要な場合のみ、フォールバックを追加してください。',
+    matchRecords: 'マッチ記録',
+    matchRecordsDescription:
+      '各ルールの最新結果と最近の未一致イベントを表示します。ボットのメモリ内ログは再起動時に消去されます。',
     testRoute: 'ルートを確認',
     adapterEventDebugAction: 'イベント監視テスト',
     adapterEventDebugTitle: 'プラットフォームイベントのデバッグ',
@@ -935,6 +1164,15 @@ const jaJP = {
     eventWildcard: 'すべてのイベント',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'ボットとのチャットに参加',
+      wecombot_template_card_event: 'テンプレートカード操作',
+      dingtalk_card_action: 'DingTalk カード操作',
+      wecomcs_enter_session: '顧客が会話に参加',
+      wecomcs_msg_send_fail: 'カスタマーサービスの送信失敗',
+      wecomcs_servicer_status_change: '担当者の受付状態変更',
+      wecomcs_session_status_change: 'カスタマーサービスの会話状態変更',
+      wecomcs_reject_customer_msg_switch_change:
+        '顧客メッセージの受信拒否設定変更',
       message_received: 'メッセージ受信',
       message_edited: 'メッセージ編集',
       message_deleted: 'メッセージ削除',
@@ -944,6 +1182,7 @@ const jaJP = {
       friend_added: '友達追加',
       group_member_joined: 'メンバー参加',
       group_member_left: 'メンバー退出',
+      group_info_updated: 'グループ情報更新',
       group_member_banned: 'メンバーBAN',
       bot_invited_to_group: 'ボットがグループに招待された',
       bot_removed_from_group: 'ボットがグループから削除された',
@@ -952,6 +1191,22 @@ const jaJP = {
       platform_specific: 'プラットフォーム固有イベント',
     },
     eventDescriptions: {
+      wecombot_enter_chat:
+        'ユーザーがボットのチャットを開きました。歓迎や案内のフローに使用できます。',
+      wecombot_template_card_event:
+        'ユーザーがカードをクリックまたは送信しました。管理対象フォームの応答は別途処理します。',
+      dingtalk_card_action:
+        'カードのクリックや送信。フィードバックと管理対象フォームは別途処理します。',
+      wecomcs_enter_session:
+        '顧客がリンクから会話を開きました。流入元と歓迎メッセージ用コードを含みます。',
+      wecomcs_msg_send_fail:
+        'WeChat が送信失敗を通知しました。メッセージ ID と失敗種別を含みます。',
+      wecomcs_servicer_status_change:
+        '担当者が受付を開始、停止、または一時停止しました。',
+      wecomcs_session_status_change:
+        '担当者が会話を受付、転送、終了、または再開しました。',
+      wecomcs_reject_customer_msg_switch_change:
+        '担当者が顧客メッセージの受信拒否を有効または無効にしました。',
       all: 'このアダプターが受信するすべてのイベントに一致します。',
       namespace: 'すべての{{group}}イベントに一致します。',
       namespace_bot:
@@ -959,8 +1214,7 @@ const jaJP = {
       namespace_feedback:
         'プラットフォームまたはユーザーからのフィードバックイベントに一致します。',
       namespace_friend: '友達リクエストや友達関係の変更イベントに一致します。',
-      namespace_group:
-        'メンバーの参加、退出、削除などのグループイベントに一致します。',
+      namespace_group: 'メンバーの変更やグループ情報の更新などに一致します。',
       namespace_message:
         'メッセージの受信、編集、削除、リアクションイベントに一致します。',
       namespace_platform:
@@ -980,6 +1234,7 @@ const jaJP = {
       friend_added: '友達関係が作成されました。',
       group_member_joined: 'ボットがいるグループにメンバーが参加しました。',
       group_member_left: 'ボットがいるグループからメンバーが退出しました。',
+      group_info_updated: 'グループ名や所有者などの情報が変更されました。',
       group_member_banned: 'グループメンバーがBANまたは削除されました。',
       bot_invited_to_group: 'ボットがグループに招待されました。',
       bot_removed_from_group: 'ボットがグループから削除されました。',
@@ -1066,6 +1321,13 @@ const jaJP = {
     imagesAttached: '枚の画像が添付されています',
     noLogs: 'ログはまだありません',
     sessionMonitor: {
+      resetContext: 'コンテキストをリセット',
+      resetDescription:
+        '新しいメッセージは空のコンテキストから開始します。監視履歴、ファイル、長期記憶は保持されます。実行中のタスクが完了してからリセットしてください。',
+      resetSuccess: 'コンテキストをリセットしました。',
+      resetError:
+        'リセットできませんでした。実行中のタスクが完了してから再試行してください。',
+
       title: 'セッション監視',
       sessions: 'セッション一覧',
       noSessions: 'セッションが見つかりません',
@@ -1113,7 +1375,27 @@ const jaJP = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'すべて選択',
     monitoring: {
+      rangeDays: '過去 {{days}} 日間',
+      customRange: '期間を指定',
+      rangeStart: '開始日時',
+      rangeEnd: '終了日時',
+      rangeHint: '開始は終了より前、期間は最大365日です。',
+      autoRefreshLabel: '自動更新',
+      autoRefreshOff: '自動更新：オフ',
+      autoRefreshEvery: '{{seconds}} 秒ごとに更新',
+      runStatus: '実行状態',
+      toolCalls: 'ツール呼び出し',
+      delivery: 'メッセージ送信',
+      sendFailed: '送信失敗',
+      sendCount: '送信 API 呼び出し成功：{{count}} 回',
+      noSend: '送信記録なし',
+      noSteps: '実行ステップはまだありません。',
+      recipient: '送信先',
+      actionCompleted: 'ツールの実行が完了しました。',
+      simulatedAction:
+        'シミュレーションです。実際のメッセージは送信されていません。',
       description:
         '各タスクのトリガーイベント、モデル出力、ツール実行を確認します。',
       empty:
@@ -1220,6 +1502,9 @@ const jaJP = {
       'メッセージイベントのみを処理し、AI が直接返信を生成します。ナレッジベースやプラグインなどの便利な機能も利用できます。',
     allEvents: 'すべてのイベントに対応',
     messageEventsOnly: 'メッセージイベントのみ',
+    selectAgentRunner: 'Agent ランナーを選択',
+    selectProcessorPlugin: '処理プラグインを選択',
+    createBasicInfoDescription: '名前・アイコン・プラグインを設定',
     chooseType: '処理方法を選択',
     chooseTypeDescription:
       '最初にプロセッサーの動作方式を選びます。作成後に詳しく設定できます。',
@@ -1639,6 +1924,11 @@ const jaJP = {
     },
   },
   market: {
+    certifiedPlugin: '認証済みプラグイン',
+    certificationTooltip:
+      'このバージョンはデータ分離と共有実行環境との互換性の審査に合格し、パッケージに署名されています。',
+    certificationLearnMore: 'プラグイン認証について',
+
     runnerUsage: 'ランナーの用途',
     runnerUsageAll: 'すべて',
     runnerUsageAgent: 'Agent / パイプライン',
@@ -2036,6 +2326,14 @@ const jaJP = {
       uploading: 'アップロード中...',
     },
     monitoring: {
+      rangeDays: '過去 {{days}} 日間',
+      customRange: '期間を指定',
+      rangeStart: '開始日時',
+      rangeEnd: '終了日時',
+      rangeHint: '開始は終了より前、期間は最大365日です。',
+      autoRefreshLabel: '自動更新',
+      autoRefreshOff: '自動更新：オフ',
+      autoRefreshEvery: '{{seconds}} 秒ごとに更新',
       title: '実行ログ',
       workbench: '設定とデバッグ',
       description: 'このパイプラインの実行ログとエラー情報を表示（過去24時間）',
@@ -2331,7 +2629,7 @@ const jaJP = {
     totpStatusDisabled: '未設定',
     totpCodesRemaining: 'リカバリーコード残り {{count}} 個',
     totpManagerSectionDesc:
-      'オーナーと管理者はすべてのアカウントの二段階認証を確認・解除できます。',
+      'オーナーと管理者はこのワークスペースのアカウントの二段階認証を確認・解除できます。',
     revokeTotp: '再バインド',
     totpAdminResetTitle: '{{user}} の二段階認証を再バインド',
     totpAdminResetDesc:
@@ -2398,7 +2696,7 @@ const jaJP = {
     ossSingletonDescription:
       'このセルフホストインスタンスには1つのワークスペースがあり、複数のユーザーを追加できます。',
     cloudManagedDescription:
-      'このワークスペースは LangBot Cloud でホストされています。メンバーはここで管理し、請求は Cloud で開きます。',
+      'このワークスペースは LangBot Cloud でホストされています。',
     loadFailed: 'ワークスペース情報の読み込みに失敗しました',
     members: 'メンバー',
     you: 'あなた',
@@ -2467,6 +2765,123 @@ const jaJP = {
     },
   },
   monitoring: {
+    rangeDays: '過去 {{days}} 日間',
+    customRange: '期間を指定',
+    rangeStart: '開始日時',
+    rangeEnd: '終了日時',
+    rangeHint: '開始は終了より前、期間は最大365日です。',
+    autoRefreshLabel: '自動更新',
+    autoRefreshOff: '自動更新：オフ',
+    autoRefreshEvery: '{{seconds}} 秒ごとに更新',
+    unified: {
+      inputs: 'イベントと入力',
+      outputs: '生成結果',
+      deliveries: '送信記録',
+      conversation: '会話の前後関係',
+      related: '関連する実行',
+      events: '実行イベント',
+      llm_calls: 'モデル呼び出し',
+      tool_calls: 'ツール呼び出し',
+      errors: 'エラー',
+      payload: '詳細',
+      metadata: '技術情報',
+      loadMore: 'さらに読み込む',
+      noRecords: '記録なし',
+      contextHint: '会話内の他のメッセージがこの実行に属するとは限りません。',
+      unhandled: '処理なし',
+      roles: {
+        user: 'ユーザー',
+        assistant: 'アシスタント',
+        system: 'システム',
+        tool: 'ツール',
+      },
+    },
+    execution: {
+      title: '実行履歴',
+      subtitle:
+        'エージェントとパイプラインに届いた各イベントと、それによって開始された実行を確認します。',
+      tabs: { executions: '実行履歴' },
+      filters: {
+        source: '種類',
+        sourceAll: 'すべての種類',
+        sourceAgent: 'エージェント',
+        sourcePipeline: 'パイプライン',
+        mode: '実行モード',
+        modeAll: '通常＋デバッグ',
+        modeLive: '通常のみ',
+        modeDebug: 'デバッグのみ',
+        status: '状態',
+        statusAll: 'すべての状態',
+      },
+      status: {
+        completed: '完了',
+        failed: '失敗',
+        running: '実行中',
+        queued: '待機中',
+        cancelled: 'キャンセル済み',
+        ignored: '無視',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'エージェント',
+        pipeline: 'パイプライン',
+        processor: '処理対象',
+      },
+      columns: {
+        time: '時刻',
+        target: '処理対象',
+        status: '状態',
+        duration: '所要時間',
+        tokens: 'トークン',
+      },
+      debug: 'デバッグ',
+      empty: 'この期間の実行履歴はありません',
+      loading: '実行履歴を読み込み中...',
+      triggeredBy: '{{trigger}} により開始',
+      cards: {
+        latencyHint:
+          'P50/P95：実行の約 50%/95% がこの時間内に完了します。最近の完了済み Agent 実行が対象で、キュー待ちは除きます。',
+        tokensHint:
+          '記録済みの使用量のみ集計します。モデルが使用量を返さない場合、合計は不完全になります。',
+        totalExecutions: '実行履歴',
+        bySource: 'エージェント {{agent}} 件 · パイプライン {{pipeline}} 件',
+        successRate: '実行成功率',
+        successRateHint:
+          '成功数 ÷（成功数 + 失敗数）。実行中・無視・キャンセルは除きます。',
+        denominator: '終了済み {{denominator}} 件中 {{completed}} 件が完了',
+        running: '実行中',
+        queued: '待機中',
+        waiting: '入力待ち',
+        failed: '失敗',
+        ignored: '無視',
+        latency: '実行時間',
+        p50: 'P50',
+        p95: 'P95',
+        sample: '{{count}} 件の実行を集計',
+        tokens: 'トークン',
+        coverage: '{{calls}} 回中 {{withUsage}} 回で使用量を取得',
+        debugRuns: 'うちデバッグ実行 {{count}} 件',
+      },
+      detail: {
+        title: '実行',
+        timeline: 'タイムライン',
+        executionId: '実行 ID',
+        eventId: 'イベント',
+        queue: 'キュー',
+        runner: 'ランナー',
+        created: '作成日時',
+        duration: '所要時間',
+        usage: '使用量',
+        noEvents: '記録されたトレースイベントはありません',
+        query: '受信メッセージ',
+        modelCalls: 'モデル呼び出し',
+        toolCalls: 'ツール呼び出し',
+        errors: 'エラー',
+        loadError: '実行トレースを読み込めませんでした',
+        openObject: '対象を開く',
+        a11yDescription: 'この実行の完全なトレース',
+      },
+    },
     title: 'ダッシュボード',
     description:
       'ボットアクティビティ、LLM呼び出し、システムパフォーマンスを監視',
@@ -2483,6 +2898,9 @@ const jaJP = {
       pipeline: 'パイプライン',
       allBots: 'すべてのボット',
       selectBot: 'ボットを選択',
+      processor: 'プロセッサー',
+      allProcessors: 'すべてのプロセッサー',
+      selectProcessor: 'プロセッサーを選択',
       allPipelines: 'すべてのパイプライン',
       selectPipeline: 'パイプラインを選択',
       loading: '読み込み中...',
@@ -2508,6 +2926,15 @@ const jaJP = {
       sessions: 'セッション分析',
       errors: 'エラーログ',
     },
+    pipelineTrace: {
+      received: '受信メッセージ',
+      processing: '処理の流れ',
+      replies: '返信',
+      noSteps: 'モデル・ツールの呼び出し記録はありません',
+      success: '成功',
+      error: '失敗',
+      pending: '処理待ち',
+    },
     messageList: {
       timestamp: 'タイムスタンプ',
       bot: 'ボット',
@@ -2731,7 +3158,7 @@ const jaJP = {
     title: '設定',
     nav: {
       workspace: 'ワークスペース',
-      operationTrace: '操作トレーサビリティ',
+      operationTrace: '操作ログ',
       models: 'モデル',
       api: 'API',
       storage: 'ストレージ',
@@ -2739,21 +3166,21 @@ const jaJP = {
     },
   },
   operationTrace: {
-    title: '操作トレーサビリティ',
+    title: '操作ログ',
     description: '管理者とオーナーの操作を追跡します。',
+    reset: 'リセット',
+    settingsUpdated: '設定を保存しました',
+    settingsUpdateFailed: '設定の保存に失敗しました',
+    levelBadge: 'レベル：{{level}}',
+    collectionSettings: '収集設定',
     captureLevel: 'キャプチャレベル',
     retention: '保持ポリシー',
-    retentionSummary: '{{days}} 日間保持 · 最大 {{rows}} 件',
     retentionDays: '保持日数',
     maxRows: '最大レコード数',
     dedupeWindow: '重複排除ウィンドウ（秒）',
     records: '操作レコード',
     changesCount: '変更 {{count}} 件',
     hideDetails: '詳細を隠す',
-    levelUpdated: 'キャプチャレベルを更新しました',
-    levelUpdateFailed: 'キャプチャレベルの更新に失敗しました',
-    retentionUpdated: '保持ポリシーを更新しました',
-    retentionUpdateFailed: '保持ポリシーの更新に失敗しました',
     export: 'ログをダウンロード',
     exportFailed: '操作ログのダウンロードに失敗しました',
     empty: '操作レコードはまだありません',
@@ -2767,7 +3194,7 @@ const jaJP = {
     filterAllResources: 'すべてのリソース',
     filterActor: '実行者',
     filterAllActors: 'すべての実行者',
-    loadFailed: 'トレーサビリティデータの読み込みに失敗しました',
+    loadFailed: '操作ログデータの読み込みに失敗しました',
     refresh: '更新',
     save: '保存',
     redacted: 'マスク済み',
@@ -2786,7 +3213,7 @@ const jaJP = {
     elidedValue: '{{count}} 文字',
     levels: {
       off: {
-        label: 'トレース無効',
+        label: 'ログ無効',
       },
       mutation: {
         label: '変更のみ',
diff --git a/web/src/i18n/locales/pipeline-migration/en-US.ts b/web/src/i18n/locales/pipeline-migration/en-US.ts
index 60e675d0f..e3a24a27f 100644
--- a/web/src/i18n/locales/pipeline-migration/en-US.ts
+++ b/web/src/i18n/locales/pipeline-migration/en-US.ts
@@ -37,6 +37,19 @@ export default {
     'After checking the runtime, refresh, select this pipeline and confirm to retry activation only. Its saved configuration will not be migrated again.',
   details: 'Migration details',
   notices: {
+    customRunner:
+      'Legacy custom runners must be rewritten as plugin runners before they can be bound. See the plugin docs: https://langbot.app/docs/en/plugin/dev/tutor',
+    fieldDropped:
+      'Unrecognized legacy fields or sections were removed from the active config; the original values are kept in the migration backup.',
+    valueRepaired:
+      'Malformed or missing values were repaired with the 4.10 defaults. Review them after migration.',
+    mixedRunner:
+      'Both a new and a legacy runner selection exist; the usable one was kept.',
+    runnerConfigReplaced:
+      'The existing new-style runner config is replaced by the migration result; the original is kept in the backup.',
+    runnerAdded:
+      'The required runner plugin was added to the extension list of this pipeline.',
+    extensionPrefs: 'Invalid entries in the extension settings were ignored.',
     runtimeUnavailable:
       'The plugin runtime is disconnected. Restore the connection and retry, or migrate data only.',
     executionFailed: 'Migration failed. Check the server logs and retry.',
@@ -119,6 +132,7 @@ export default {
     blocked: 'Blocked',
     already_current: 'Already current',
     not_legacy: 'Not legacy',
+    unsupported_runner: 'Unsupported runner',
     activation_pending: 'Activation pending',
     pending: 'Pending',
     migrated: 'Migrated',
diff --git a/web/src/i18n/locales/pipeline-migration/es-ES.ts b/web/src/i18n/locales/pipeline-migration/es-ES.ts
index 12293bc8f..ac64e1fd0 100644
--- a/web/src/i18n/locales/pipeline-migration/es-ES.ts
+++ b/web/src/i18n/locales/pipeline-migration/es-ES.ts
@@ -36,6 +36,20 @@ export default {
     'Tras comprobar el entorno, actualiza, selecciona esta canalización y confirma para reintentar solo la activación. No se volverá a migrar la configuración guardada.',
   details: 'Detalles de migración',
   notices: {
+    customRunner:
+      'Los runners personalizados antiguos deben reescribirse como runners de plugin antes de vincularlos. Consulta la documentación de plugins: https://langbot.app/docs/en/plugin/dev/tutor',
+    fieldDropped:
+      'Los campos o secciones antiguos no reconocidos se eliminaron de la configuración activa; los valores originales se conservan en la copia de seguridad.',
+    valueRepaired:
+      'Los valores mal formados o ausentes se repararon con los valores predeterminados de 4.10. Revísalos tras la migración.',
+    mixedRunner:
+      'Existen una selección de runner nueva y otra antigua; se conservó la utilizable.',
+    runnerConfigReplaced:
+      'La configuración de runner nueva existente se reemplaza por el resultado de la migración; la original se conserva en la copia de seguridad.',
+    runnerAdded:
+      'Se añadió el plugin de runner necesario a la lista de extensiones de esta pipeline.',
+    extensionPrefs:
+      'Se ignoraron entradas no válidas de la configuración de extensiones.',
     runtimeUnavailable:
       'El entorno de plugins está desconectado. Restablece la conexión y reintenta, o migra solo los datos.',
     executionFailed:
@@ -124,6 +138,7 @@ export default {
     blocked: 'Bloqueado',
     already_current: 'Ya actualizado',
     not_legacy: 'No es antiguo',
+    unsupported_runner: 'Runner no compatible',
     activation_pending: 'Activación pendiente',
     pending: 'Pendiente',
     migrated: 'Migrado',
diff --git a/web/src/i18n/locales/pipeline-migration/ja-JP.ts b/web/src/i18n/locales/pipeline-migration/ja-JP.ts
index 729fd1df4..7b6b6c5e7 100644
--- a/web/src/i18n/locales/pipeline-migration/ja-JP.ts
+++ b/web/src/i18n/locales/pipeline-migration/ja-JP.ts
@@ -37,6 +37,19 @@ export default {
     '実行環境を確認して再読み込みし、このパイプラインを選択して確定すると、有効化のみを再試行します。保存済み設定は再移行しません。',
   details: '移行の詳細',
   notices: {
+    customRunner:
+      '旧形式のカスタム Runner は、プラグイン Runner として書き直してからバインドしてください。プラグイン開発ドキュメント：https://langbot.app/docs/ja/plugin/dev/tutor',
+    fieldDropped:
+      '認識できない旧フィールドやセクションは有効な設定から削除され、元の値は移行バックアップに保存されます。',
+    valueRepaired:
+      '不正または欠落した値は 4.10 の既定値で自動修復されました。移行後に確認してください。',
+    mixedRunner:
+      '新旧両方の Runner 指定があるため、利用可能な方を採用しました。',
+    runnerConfigReplaced:
+      '既存の新形式 Runner 設定は移行結果で置き換えられ、元の値はバックアップに保存されます。',
+    runnerAdded:
+      '必要な Runner プラグインをこのパイプラインの拡張リストに追加しました。',
+    extensionPrefs: '拡張設定の無効な項目は無視されました。',
     runtimeUnavailable:
       'プラグインランタイムが未接続です。接続を復旧して再試行するか、データのみ移行してください。',
     executionFailed:
@@ -119,6 +132,7 @@ export default {
     blocked: '移行不可',
     already_current: '移行済みの形式',
     not_legacy: '旧形式ではない',
+    unsupported_runner: '非対応の Runner',
     activation_pending: '有効化待ち',
     pending: '待機中',
     migrated: '移行済み',
diff --git a/web/src/i18n/locales/pipeline-migration/ru-RU.ts b/web/src/i18n/locales/pipeline-migration/ru-RU.ts
index 25fba1b04..82fd98414 100644
--- a/web/src/i18n/locales/pipeline-migration/ru-RU.ts
+++ b/web/src/i18n/locales/pipeline-migration/ru-RU.ts
@@ -36,6 +36,19 @@ export default {
     'Проверьте среду выполнения, обновите список, выберите конвейер и подтвердите повторную активацию. Сохранённая конфигурация не будет мигрировать повторно.',
   details: 'Подробности переноса',
   notices: {
+    customRunner:
+      'Устаревшие пользовательские runner нужно переписать как runner-плагины перед привязкой. См. документацию по плагинам: https://langbot.app/docs/en/plugin/dev/tutor',
+    fieldDropped:
+      'Нераспознанные устаревшие поля или разделы удалены из активной конфигурации; исходные значения сохранены в резервной копии миграции.',
+    valueRepaired:
+      'Некорректные или отсутствующие значения исправлены на значения по умолчанию 4.10. Проверьте их после миграции.',
+    mixedRunner: 'Указан и новый, и устаревший runner; оставлен пригодный.',
+    runnerConfigReplaced:
+      'Существующая конфигурация runner нового формата заменяется результатом миграции; исходная сохранена в резервной копии.',
+    runnerAdded:
+      'Нужный runner-плагин добавлен в список расширений этого конвейера.',
+    extensionPrefs:
+      'Недопустимые записи в настройках расширений проигнорированы.',
     runtimeUnavailable:
       'Среда плагинов отключена. Восстановите подключение и повторите попытку или перенесите только данные.',
     executionFailed:
@@ -121,6 +134,7 @@ export default {
     blocked: 'Заблокирован',
     already_current: 'Уже актуален',
     not_legacy: 'Не устаревший',
+    unsupported_runner: 'Неподдерживаемый runner',
     activation_pending: 'Ожидает активации',
     pending: 'Ожидает',
     migrated: 'Перенесён',
diff --git a/web/src/i18n/locales/pipeline-migration/th-TH.ts b/web/src/i18n/locales/pipeline-migration/th-TH.ts
index 40ca5c7f0..637e7609f 100644
--- a/web/src/i18n/locales/pipeline-migration/th-TH.ts
+++ b/web/src/i18n/locales/pipeline-migration/th-TH.ts
@@ -36,6 +36,19 @@ export default {
     'หลังตรวจสอบสภาพแวดล้อมการทำงาน ให้รีเฟรช เลือกไปป์ไลน์นี้และยืนยันเพื่อลองเปิดใช้งานอีกครั้งเท่านั้น การตั้งค่าที่บันทึกไว้จะไม่ถูกย้ายซ้ำ',
   details: 'รายละเอียดการย้าย',
   notices: {
+    customRunner:
+      'Runner แบบกำหนดเองรุ่นเก่าต้องเขียนใหม่เป็น Runner แบบปลั๊กอินก่อนจึงจะผูกได้ ดูเอกสารปลั๊กอิน: https://langbot.app/docs/en/plugin/dev/tutor',
+    fieldDropped:
+      'ฟิลด์หรือส่วนเก่าที่ไม่รู้จักถูกนำออกจากการตั้งค่าที่ใช้งาน ค่าเดิมถูกเก็บไว้ในข้อมูลสำรองการย้าย',
+    valueRepaired:
+      'ค่าที่ผิดรูปแบบหรือขาดหายถูกซ่อมด้วยค่าเริ่มต้นของ 4.10 โปรดตรวจสอบหลังการย้าย',
+    mixedRunner:
+      'มีการเลือก Runner ทั้งแบบใหม่และแบบเก่า ระบบเลือกอันที่ใช้งานได้',
+    runnerConfigReplaced:
+      'การตั้งค่า Runner แบบใหม่ที่มีอยู่จะถูกแทนที่ด้วยผลการย้าย ค่าเดิมถูกเก็บไว้ในข้อมูลสำรอง',
+    runnerAdded:
+      'เพิ่มปลั๊กอิน Runner ที่จำเป็นลงในรายการส่วนขยายของไปป์ไลน์นี้แล้ว',
+    extensionPrefs: 'รายการที่ไม่ถูกต้องในการตั้งค่าส่วนขยายถูกละเว้น',
     runtimeUnavailable:
       'ไม่ได้เชื่อมต่อรันไทม์ปลั๊กอิน โปรดเชื่อมต่อใหม่แล้วลองอีกครั้ง หรือย้ายเฉพาะข้อมูล',
     executionFailed:
@@ -115,6 +128,7 @@ export default {
     blocked: 'ถูกบล็อก',
     already_current: 'เป็นรุ่นใหม่แล้ว',
     not_legacy: 'ไม่ใช่รุ่นเก่า',
+    unsupported_runner: 'Runner ที่ไม่รองรับ',
     activation_pending: 'รอเปิดใช้งาน',
     pending: 'รอดำเนินการ',
     migrated: 'ย้ายแล้ว',
diff --git a/web/src/i18n/locales/pipeline-migration/vi-VN.ts b/web/src/i18n/locales/pipeline-migration/vi-VN.ts
index dbad54151..dc34e1d78 100644
--- a/web/src/i18n/locales/pipeline-migration/vi-VN.ts
+++ b/web/src/i18n/locales/pipeline-migration/vi-VN.ts
@@ -36,6 +36,18 @@ export default {
     'Sau khi kiểm tra môi trường chạy, làm mới, chọn pipeline này và xác nhận để chỉ thử kích hoạt lại. Cấu hình đã lưu sẽ không được di chuyển lần nữa.',
   details: 'Chi tiết chuyển đổi',
   notices: {
+    customRunner:
+      'Runner tùy chỉnh kiểu cũ cần được viết lại thành runner dạng plugin trước khi liên kết. Xem tài liệu plugin: https://langbot.app/docs/en/plugin/dev/tutor',
+    fieldDropped:
+      'Các trường hoặc mục cũ không nhận dạng được đã bị loại khỏi cấu hình đang dùng; giá trị gốc được lưu trong bản sao lưu di chuyển.',
+    valueRepaired:
+      'Các giá trị sai định dạng hoặc thiếu đã được sửa bằng giá trị mặc định của 4.10. Hãy kiểm tra sau khi di chuyển.',
+    mixedRunner: 'Có cả lựa chọn runner mới và cũ; đã giữ lựa chọn dùng được.',
+    runnerConfigReplaced:
+      'Cấu hình runner kiểu mới hiện có được thay bằng kết quả di chuyển; bản gốc được lưu trong bản sao lưu.',
+    runnerAdded:
+      'Đã thêm plugin runner cần thiết vào danh sách tiện ích của pipeline này.',
+    extensionPrefs: 'Các mục không hợp lệ trong cài đặt tiện ích đã bị bỏ qua.',
     runtimeUnavailable:
       'Môi trường chạy plugin bị ngắt kết nối. Khôi phục kết nối rồi thử lại, hoặc chỉ di chuyển dữ liệu.',
     executionFailed:
@@ -118,6 +130,7 @@ export default {
     blocked: 'Bị chặn',
     already_current: 'Đã cập nhật',
     not_legacy: 'Không phải bản cũ',
+    unsupported_runner: 'Runner không được hỗ trợ',
     activation_pending: 'Chờ kích hoạt',
     pending: 'Đang chờ',
     migrated: 'Đã di chuyển',
diff --git a/web/src/i18n/locales/pipeline-migration/zh-Hans.ts b/web/src/i18n/locales/pipeline-migration/zh-Hans.ts
index 09143fa3c..48e2e1084 100644
--- a/web/src/i18n/locales/pipeline-migration/zh-Hans.ts
+++ b/web/src/i18n/locales/pipeline-migration/zh-Hans.ts
@@ -31,6 +31,17 @@ export default {
     '检查运行环境后刷新，选中此流水线并确认，即可仅重试激活，不会再次迁移已保存的配置。',
   details: '迁移详情',
   notices: {
+    customRunner:
+      '旧版自定义 Runner 需要改写为插件 Runner 后再绑定。参见插件开发文档：https://langbot.app/docs/zh/plugin/dev/tutor',
+    fieldDropped:
+      '未识别的旧字段或配置段已从活动配置中移除，原值保存在迁移备份中。',
+    valueRepaired:
+      '格式不正确或缺失的值已按 4.10 默认值自动修复，请在迁移后检查。',
+    mixedRunner: '同时存在新旧 Runner 选择，已采用可用的那一项。',
+    runnerConfigReplaced:
+      '已有的新版 Runner 配置会被迁移结果替换，原值保存在迁移备份中。',
+    runnerAdded: '已将所需 Runner 插件加入此流水线的扩展列表。',
+    extensionPrefs: '扩展设置中有无效项，已忽略。',
     runtimeUnavailable:
       '插件运行时未连接，请恢复连接后重试，或选择仅迁移数据。',
     executionFailed: '迁移执行失败，请查看服务日志后重试。',
@@ -96,6 +107,7 @@ export default {
     blocked: '已阻止',
     already_current: '已是新版',
     not_legacy: '非旧版',
+    unsupported_runner: '不支持的 Runner',
     activation_pending: '等待激活',
     pending: '等待中',
     migrated: '已迁移',
diff --git a/web/src/i18n/locales/pipeline-migration/zh-Hant.ts b/web/src/i18n/locales/pipeline-migration/zh-Hant.ts
index 8eb48f92b..fd6216886 100644
--- a/web/src/i18n/locales/pipeline-migration/zh-Hant.ts
+++ b/web/src/i18n/locales/pipeline-migration/zh-Hant.ts
@@ -32,6 +32,17 @@ export default {
     '檢查執行環境後重新整理，選取此流水線並確認，即可只重試啟用，不會再次遷移已儲存的設定。',
   details: '遷移詳情',
   notices: {
+    customRunner:
+      '舊版自訂 Runner 需要改寫為外掛 Runner 後再綁定。參見外掛開發文件：https://langbot.app/docs/zh/plugin/dev/tutor',
+    fieldDropped:
+      '未識別的舊欄位或設定段已從使用中設定移除，原值保存在遷移備份中。',
+    valueRepaired:
+      '格式不正確或缺少的值已依 4.10 預設值自動修復，請在遷移後檢查。',
+    mixedRunner: '同時存在新舊 Runner 選擇，已採用可用的那一項。',
+    runnerConfigReplaced:
+      '既有的新版 Runner 設定會被遷移結果取代，原值保存在遷移備份中。',
+    runnerAdded: '已將所需 Runner 外掛加入此流水線的擴充列表。',
+    extensionPrefs: '擴充設定中有無效項目，已忽略。',
     runtimeUnavailable:
       '外掛執行環境未連線，請恢復連線後重試，或選擇僅遷移資料。',
     executionFailed: '遷移執行失敗，請查看服務日誌後重試。',
@@ -97,6 +108,7 @@ export default {
     blocked: '已阻止',
     already_current: '已是新版',
     not_legacy: '非舊版',
+    unsupported_runner: '不支援的 Runner',
     activation_pending: '等待啟用',
     pending: '等待中',
     migrated: '已遷移',
diff --git a/web/src/i18n/locales/ru-RU.ts b/web/src/i18n/locales/ru-RU.ts
index 04dbec827..f987f12b1 100644
--- a/web/src/i18n/locales/ru-RU.ts
+++ b/web/src/i18n/locales/ru-RU.ts
@@ -1,6 +1,178 @@
 import pipelineMigration from './pipeline-migration/ru-RU';
 const ruRU = {
+  planRevision: {
+    subscribedOn: 'Дата подписки: {{date}}',
+    latestQuota: 'Актуальные квоты этого плана',
+    showLatest: 'Показать актуальные квоты этого тарифа',
+    showCurrent: 'Показать текущую подписку',
+    latestTab: 'Новые',
+    currentTab: 'Текущая',
+    cancelToSubscribe:
+      'Отмените подписку, чтобы получить доступ к новым квотам со следующего периода',
+
+    current: 'Current subscription: {{plan}} · revision {{revision}}',
+    owned: 'Purchased revision {{revision}}',
+    latest: 'Latest revision {{revision}}',
+    changed:
+      'Your purchased benefits and the latest version differ. Highlighted rows show the changes.',
+    same: 'Benefits match latest revision {{revision}}. Your purchased revision remains unchanged.',
+    unavailable:
+      'The latest plan is unavailable. Your purchased benefits remain in effect.',
+    benefit: 'Benefit',
+    change: 'Changed',
+    enabled: 'Enabled',
+    disabled: 'Disabled',
+    monthlyCredits: 'Monthly bonus credits',
+    retention: 'Chat history retention (days)',
+    unknown:
+      'Current subscription details are unavailable. Refresh or manage your subscription in Space; latest catalog benefits are not your purchased benefits.',
+    handoff:
+      'Manage your subscription in Space. Cancel automatic renewal before purchasing a new version. Prepaid subscriptions end automatically unless renewed before expiry.',
+    cancelFirst:
+      'Cancel automatic renewal in subscription management first. Cancellation preserves paid benefits until the end of the period, without a refund. After confirmation, you can purchase the latest version, effective when payment succeeds.',
+    canceled:
+      'Renewal is canceled. Paid benefits remain until the end of the period, without a refund. A new purchase applies its benefits when payment succeeds.',
+    prepaid:
+      'This prepaid subscription ends automatically. You can renew your current plan before expiry; after expiry, subscriptions use the latest quotas for each plan.',
+    renew: 'Продлить текущую подписку',
+    buying: 'Purchasing revision {{revision}}.',
+    fields: {
+      bots_max: 'Bots',
+      members_max: 'Members',
+      pipelines_max: 'Processors',
+      plugins_max: 'Extensions',
+      knowledge_bases_max: 'Knowledge bases',
+      managed_sandbox_sessions_max: 'Sandbox sessions',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: 'Подписка недоступна',
+    subscriptionUnavailableHint:
+      'Этот план не принимает новые подписки. Выберите другой план. Текущие подписки и автоматические продления не затронуты.',
+    platformNames: {
+      qqofficial: 'QQ Official API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComCustomerService',
+      officialaccount: 'Official Account',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'Управление подпиской',
+    downgradeUnavailable: 'Переход на бесплатный план недоступен',
+    features: {
+      multiUserWorkspace: 'Многопользовательское рабочее пространство',
+      remoteE2BSandbox: 'Настройка удалённой песочницы E2B',
+      communitySupport: 'Поддержка сообщества',
+      everythingInFree: 'Все возможности бесплатного плана',
+      alwaysOnSandbox: 'Одна постоянно работающая глобальная песочница',
+      higherWorkspaceLimits: 'Повышенные лимиты рабочего пространства',
+    },
+    title: 'Сравнить планы',
+    description: 'Выберите план для {{workspace}}',
+    billing: 'Период оплаты',
+    monthly: 'За месяц',
+    annual: 'За год',
+    currency: 'Валюта',
+    loading: 'Загрузка планов…',
+    error: 'Не удалось загрузить планы',
+    retry: 'Повторить',
+    empty: 'Нет доступных планов',
+    current: 'Текущий план',
+    recommended: 'Рекомендуем',
+    free: 'Бесплатно',
+    annualNote: 'Эквивалент {{price}} / месяц',
+    ownerOnly: 'Только владелец или администратор',
+    choose: 'Перейти к оплате',
+    handoff: 'Войдите в Space и оплатите. Ваш выбор будет сохранён.',
+    quotas: 'Лимиты рабочего пространства',
+    unlimited: 'Без ограничений',
+    retention: 'История сообщений · {{count}} дней',
+    retentionLabel: 'Срок хранения сообщений',
+    retentionHelp:
+      'По истечении этого срока сообщения автоматически удаляются.',
+    credits: '{{count}} кредитов в месяц',
+    creditsLabel: 'Ежемесячные кредиты',
+    creditsHelp:
+      'Кредиты начисляются каждый месяц подписки, в том числе при годовой оплате.',
+    platforms: 'Поддержка популярных платформ',
+    excludedPlatforms: 'Недоступны: {{platforms}}',
+    allPlatforms: 'Все доступные платформы',
+    limits: {
+      bots: 'Боты',
+      pipelines: 'Обработчики',
+      knowledge_bases: 'Базы знаний',
+      plugins: 'Расширения',
+      members: 'Участники',
+      managed_sandbox_sessions: 'Управляемые сеансы песочницы',
+    },
+  },
+  workspaceLanguage: {
+    label: 'Язык рабочего пространства',
+    description:
+      'Язык сообщений сервера и стандартных ответов бота. Язык интерфейса не меняется.',
+    saved: 'Язык сохранён',
+    saveFailed: 'Не удалось сохранить язык',
+  },
+  inflight: {
+    autoCollapse: 'Автосворачивание',
+    empty: 'Нет событий в обработке',
+    move: 'Перетащите по вертикали или используйте стрелки',
+    botLogs: 'Журнал бота',
+    processorLogs: 'Журнал обработчика',
+
+    title: 'В работе',
+    expand: 'Развернуть',
+    collapse: 'Свернуть',
+    reconnecting: 'Переподключение; показано последнее полученное состояние.',
+    details: 'Подробнее',
+    debug: 'Отладочное событие',
+    limited:
+      'Показана часть запусков. Все запуски доступны на панели мониторинга.',
+    queued: 'В очереди',
+    running: 'Обработка',
+    waiting: 'Ожидание ввода',
+    tool: 'Вызов инструмента',
+    thinking: 'Размышление',
+    replying: 'Подготовка ответа',
+    generating: 'Генерация ответа',
+    completed: 'Завершено',
+    failed: 'Ошибка',
+    cancelled: 'Отменено',
+    ignored: 'Пропущено',
+  },
   assistant: {
+    movePosition: 'Переместить помощника по горизонтали',
+    resizeHeight: 'Изменить высоту помощника',
+    sessions: 'Диалоги',
+    currentChat: 'Текущий диалог',
+    stop: 'Остановить задачу',
+    stopped:
+      'Остановлено. Инструмент мог уже внести изменения. Проверьте ресурс перед повтором.',
+    setupRequired:
+      'Войдите в аккаунт LangBot или настройте модель, чтобы начать пользоваться помощником рабочего пространства.',
+    loginAccount: 'Войти в LangBot Account',
+    configureModels: 'Настроить модели',
+    providerCheckFailed:
+      'Не удалось проверить провайдеров. Попробуйте ещё раз.',
+    retrySetup: 'Повторить',
+    preparing: 'Подготовка помощника…',
+    thinking: 'Ответ модели · Шаг {{round}}',
+    executing: 'Выполнение инструмента…',
+    toolRunning: 'Выполняется',
+    recommendationFailed:
+      'Не удалось загрузить рекомендуемую модель. Выберите модель.',
+    expand: 'Развернуть помощника рабочего пространства',
+    collapse: 'Свернуть помощника рабочего пространства',
     modelHint: 'Сменить модель ассистента для следующего сообщения',
     details: 'Показать исходные данные',
     found: 'Найдено элементов: {{count}}',
@@ -21,7 +193,7 @@ const ruRU = {
       create_knowledge_base: 'Создать базу знаний',
       get_pipeline: 'Посмотреть Pipeline',
       get_knowledge_schema: 'Посмотреть схему движка знаний',
-      list_operation_logs: 'Просмотр трассировки операций',
+      list_operation_logs: 'Просмотр журнала операций',
     },
     resources: {
       models: 'Найти чат-модели',
@@ -35,7 +207,7 @@ const ruRU = {
     newChat: 'Новый диалог',
     close: 'Закрыть',
     welcome:
-      'Расскажите, что вы хотите создать. Я изучу это рабочее пространство и предложу изменения для базы знаний и Pipeline, которые вы подтвердите.',
+      'Расскажите, что вы хотите создать. Я могу изучить рабочее пространство и предложить настройки ботов, Agents, баз знаний, Pipelines и других ресурсов, а затем применить их после вашего подтверждения.',
     discover: 'Какие модели, базы знаний и Pipelines доступны?',
     build: 'Помогите создать приложение для ответов на вопросы по базе знаний.',
     toolResult: 'Результат операции',
@@ -55,6 +227,14 @@ const ruRU = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: 'Последние {{days}} дней',
+        customRange: 'Свой период',
+        rangeStart: 'Начало',
+        rangeEnd: 'Конец',
+        rangeHint: 'Начало должно быть раньше конца. Максимум 365 дней.',
+        autoRefreshLabel: 'Автообновление',
+        autoRefreshOff: 'Автообновление: выкл.',
+        autoRefreshEvery: 'Обновлять каждые {{seconds}} с',
         title: 'Проверьте активность',
         description:
           'Просматривайте активность ботов, использование моделей, объём сообщений и производительность системы.',
@@ -136,6 +316,14 @@ const ruRU = {
           'Отправьте тестовые сообщения, чтобы проверить условия запуска, ответы ИИ и обработку вывода. Изменения сохраняются перед тестом.',
       },
       monitoring: {
+        rangeDays: 'Последние {{days}} дней',
+        customRange: 'Свой период',
+        rangeStart: 'Начало',
+        rangeEnd: 'Конец',
+        rangeHint: 'Начало должно быть раньше конца. Максимум 365 дней.',
+        autoRefreshLabel: 'Автообновление',
+        autoRefreshOff: 'Автообновление: выкл.',
+        autoRefreshEvery: 'Обновлять каждые {{seconds}} с',
         title: 'Посмотрите историю запусков',
         description:
           'На вкладке рядом с заголовком доступны прошлые диалоги, этапы обработки и ошибки для диагностики отсутствующих или неожиданных ответов.',
@@ -304,6 +492,11 @@ const ruRU = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'Причина сбоя не сохранена. Проверьте журнал бота или отправьте сообщение повторно для получения подробностей.',
+    loadFailedTitle: 'Не удалось загрузить данные',
+    loadFailedHint:
+      'Повторите попытку позже. Если проблема остаётся, проверьте подключение и состояние сервиса.',
     customValue: 'Свой вариант',
     loadFailed: 'Не удалось загрузить. Попробуйте ещё раз.',
     search: 'Поиск',
@@ -503,6 +696,9 @@ const ruRU = {
     help: 'Помощь',
   },
   models: {
+    starModel: 'Отметить модель для новых обработчиков',
+    unstarModel: 'Снять отметку',
+    starModelFailed: 'Не удалось изменить избранную модель',
     codex: {
       account: 'Подписка ChatGPT',
       description:
@@ -770,10 +966,13 @@ const ruRU = {
     deleteConfirmation: 'Вы уверены, что хотите удалить этого бота?',
     platformAdapter: 'Выбор платформы/адаптера',
     selectAdapter: 'Выберите адаптер',
+    saveAndEnable: 'Сохранить и включить',
+    searchAdapters: 'Поиск адаптеров',
+    noMatchingAdapters: 'Подходящие адаптеры не найдены',
     legacyAdapters: 'Устаревшие адаптеры',
     legacyAdapterBadge: 'Устаревший',
     legacyAdaptersHint:
-      'У этих адаптеров есть новые событийные версии.\nОни сохранены только для существующих конфигураций и не рекомендуются для новых ботов.',
+      'Старые адаптеры в основном обмениваются сообщениями и сохраняются для совместимости.\nНовые подключают ту же платформу через события: сообщения, поддерживаемые запросы в друзья и изменения в группах, направляя их обработчикам.\nДля новых ботов выбирайте новые адаптеры. Старые боты могут работать как прежде. При переходе проверьте подключение и маршруты; аккаунт платформы не обновляется.',
     adapterConfig: 'Настройка адаптера',
     viewAdapterDocs: 'Документация',
     bindPipeline: 'Привязать конвейер',
@@ -786,6 +985,10 @@ const ruRU = {
     earlier: 'Ранее',
     dateFormat: '{{day}}.{{month}}',
     setBotEnableError: 'Не удалось изменить статус бота',
+    enabling: 'Включение…',
+    disabling: 'Отключение…',
+    enableConfirmed: 'Включён',
+    disableConfirmed: 'Отключён',
     log: 'Журнал',
     configuration: 'Настройки',
     logs: 'Журналы',
@@ -794,6 +997,28 @@ const ruRU = {
     routingConnection: 'Маршрутизация и подключение',
     routingConnectionDescription:
       'Привяжите конвейер, обрабатывающий сообщения для этого бота',
+    routingHelp: {
+      title: 'Что такое маршрутизация событий?',
+      intro:
+        'Бот получает сообщения, запросы в друзья и события вступления в группу.',
+      events: 'Входящие события',
+      agents: 'Разные агенты',
+      message: 'Сообщение',
+      friend: 'Запрос в друзья',
+      group: 'Вступление в группу',
+      messageAction: 'Агент ответов',
+      friendAction: 'Агент проверки',
+      groupAction: 'Агент приветствия',
+      rules: 'По вашим правилам · Пример',
+      instructions:
+        'Агенты следуют вашим инструкциям и вызывают инструменты для ответов и обработки запросов.',
+      plugin:
+        'Для сложных задач обработчики плагинов регистрируют события пакетно и обрабатывают их кодом. Добавьте конфигурацию обработчика ниже.',
+      docs: 'Разработать свой обработчик плагина',
+      runnerDocs: 'Разработать свой исполнитель Agent',
+      pause: 'Пауза',
+      play: 'Воспроизвести',
+    },
     eventRouting: 'Маршрутизация событий',
     eventRoutingDescription:
       'Выберите обработчик для каждого события бота. Изменяйте логику в настройках соответствующего Agent или Pipeline. Pipeline поддерживает только события сообщений.',
@@ -834,7 +1059,8 @@ const ruRU = {
     enabled: 'Включено',
     eventBindingDescriptionPlaceholder: 'Описание правила',
     noEventBindings: 'Нет маршрутов событий',
-    unsupportedPipelineEvent: 'Конвейеры поддерживают только события message.*',
+    unsupportedPipelineEvent:
+      'Конвейеры поддерживают только события message.received',
     disable: 'Отключить',
     enable: 'Включить',
     disabledBindings: 'Отключено',
@@ -844,6 +1070,7 @@ const ruRU = {
     advancedEventValues: 'Показать все',
     eventGroup: 'Группа',
     eventGroupNames: {
+      wecomcs: 'Поддержка WeCom',
       bot: 'Состояние бота',
       feedback: 'Обратная связь',
       friend: 'Друзья',
@@ -858,6 +1085,9 @@ const ruRU = {
     routeFallbackCatchAll: '{{route}} — маршрут для всех событий.',
     routeFallbackIgnored:
       'События без подходящего маршрута игнорируются. Добавляйте общий маршрут, только если каждое событие требует явного результата.',
+    matchRecords: 'История совпадений',
+    matchRecordsDescription:
+      'Последний результат каждого правила и недавние события без совпадений. Журнал в памяти очищается при перезапуске.',
     testRoute: 'Проверить маршрут',
     adapterEventDebugAction: 'Тест прослушивания',
     adapterEventDebugTitle: 'Отладка событий платформы',
@@ -945,6 +1175,15 @@ const ruRU = {
     eventWildcard: 'Все события',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'Пользователь открыл чат бота',
+      wecombot_template_card_event: 'Взаимодействие с карточкой',
+      dingtalk_card_action: 'Взаимодействие с карточкой DingTalk',
+      wecomcs_enter_session: 'Клиент вошёл в диалог',
+      wecomcs_msg_send_fail: 'Ошибка отправки сообщения поддержки',
+      wecomcs_servicer_status_change: 'Статус оператора изменён',
+      wecomcs_session_status_change: 'Статус диалога поддержки изменён',
+      wecomcs_reject_customer_msg_switch_change:
+        'Приём сообщений клиента изменён',
       message_received: 'Получено сообщение',
       message_edited: 'Сообщение изменено',
       message_deleted: 'Сообщение удалено',
@@ -954,6 +1193,7 @@ const ruRU = {
       friend_added: 'Друг добавлен',
       group_member_joined: 'Участник вступил в группу',
       group_member_left: 'Участник покинул группу',
+      group_info_updated: 'Информация о группе обновлена',
       group_member_banned: 'Участник заблокирован',
       bot_invited_to_group: 'Бот приглашён в группу',
       bot_removed_from_group: 'Бот удалён из группы',
@@ -962,6 +1202,22 @@ const ruRU = {
       platform_specific: 'Событие конкретной платформы',
     },
     eventDescriptions: {
+      wecombot_enter_chat:
+        'Пользователь открыл чат бота. Событие подходит для сценариев приветствия.',
+      wecombot_template_card_event:
+        'Пользователь нажал или отправил карточку. Ответы управляемых форм обрабатываются отдельно.',
+      dingtalk_card_action:
+        'Нажатия и отправка карточек; отзывы и управляемые формы обрабатываются отдельно.',
+      wecomcs_enter_session:
+        'Клиент открыл ссылку поддержки; событие содержит источник и код приветствия.',
+      wecomcs_msg_send_fail:
+        'WeChat сообщил об ошибке отправки с ID сообщения и типом ошибки.',
+      wecomcs_servicer_status_change:
+        'Оператор начал, остановил или приостановил приём обращений.',
+      wecomcs_session_status_change:
+        'Оператор принял, передал, завершил или возобновил диалог.',
+      wecomcs_reject_customer_msg_switch_change:
+        'Оператор включил или отключил отказ в приёме сообщений клиента.',
       all: 'Соответствует всем событиям, полученным адаптером.',
       namespace: 'Соответствует всем событиям {{group}}.',
       namespace_bot:
@@ -970,8 +1226,7 @@ const ruRU = {
         'Соответствует обратной связи от платформы или пользователя.',
       namespace_friend:
         'Соответствует заявкам в друзья и изменениям дружеских связей.',
-      namespace_group:
-        'Соответствует вступлениям, выходам, удалениям участников и другим событиям группы.',
+      namespace_group: 'Соответствует изменениям состава и информации группы.',
       namespace_message:
         'Соответствует получению, редактированию, удалению сообщений и реакциям.',
       namespace_platform:
@@ -990,6 +1245,8 @@ const ruRU = {
       friend_added: 'Установлена дружеская связь.',
       group_member_joined: 'Участник вступает в группу, где есть бот.',
       group_member_left: 'Участник покидает группу, где есть бот.',
+      group_info_updated:
+        'Изменились сведения о группе, например название или владелец.',
       group_member_banned: 'Участник группы заблокирован или удалён.',
       bot_invited_to_group: 'Бот приглашён в группу.',
       bot_removed_from_group: 'Бот удалён из группы.',
@@ -1076,6 +1333,13 @@ const ruRU = {
     imagesAttached: 'изображений прикреплено',
     noLogs: 'Журналов пока нет',
     sessionMonitor: {
+      resetContext: 'Сбросить контекст',
+      resetDescription:
+        'Новые сообщения начнутся с пустого контекста. История мониторинга, файлы и долговременная память сохранятся. Дождитесь завершения активных задач.',
+      resetSuccess: 'Контекст сброшен.',
+      resetError:
+        'Не удалось сбросить контекст. Дождитесь завершения активных задач и повторите попытку.',
+
       title: 'Сессии',
       sessions: 'Сессии',
       noSessions: 'Сессии не найдены',
@@ -1123,7 +1387,27 @@ const ruRU = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'Выбрать все',
     monitoring: {
+      rangeDays: 'Последние {{days}} дней',
+      customRange: 'Свой период',
+      rangeStart: 'Начало',
+      rangeEnd: 'Конец',
+      rangeHint: 'Начало должно быть раньше конца. Максимум 365 дней.',
+      autoRefreshLabel: 'Автообновление',
+      autoRefreshOff: 'Автообновление: выкл.',
+      autoRefreshEvery: 'Обновлять каждые {{seconds}} с',
+      runStatus: 'Статус выполнения',
+      toolCalls: 'Вызовы инструментов',
+      delivery: 'Отправка сообщений',
+      sendFailed: 'Ошибка отправки',
+      sendCount: 'Успешных вызовов отправки: {{count}}',
+      noSend: 'Отправки не зарегистрированы',
+      noSteps: 'Шаги выполнения ещё не зарегистрированы.',
+      recipient: 'Получатель',
+      actionCompleted: 'Инструмент выполнен успешно.',
+      simulatedAction:
+        'Действие смоделировано; реальное сообщение не отправлено.',
       description:
         'Просмотр события, ответа модели и вызовов инструментов для каждой задачи.',
       empty:
@@ -1228,6 +1512,9 @@ const ruRU = {
       'Работает по заданному процессу: получить сообщение, вызвать AI и ответить пользователю. Поддерживает настройку баз знаний и плагинов; обрабатывает только события сообщений и подходит для задач с четкими шагами и контролем обработки.',
     allEvents: 'Поддерживает все события',
     messageEventsOnly: 'Только события сообщений',
+    selectAgentRunner: 'Выберите исполнитель Agent',
+    selectProcessorPlugin: 'Выберите плагин обработки',
+    createBasicInfoDescription: 'Задайте имя, значок и плагин',
     chooseType: 'Выберите способ работы',
     chooseTypeDescription:
       'Сначала выберите тип обработчика. Остальные настройки доступны после создания.',
@@ -1651,6 +1938,11 @@ const ruRU = {
     },
   },
   market: {
+    certifiedPlugin: 'Сертифицированный плагин',
+    certificationTooltip:
+      'Эта версия прошла проверку изоляции данных и совместимости с общей средой выполнения. Пакет подписан.',
+    certificationLearnMore: 'О сертификации',
+
     runnerUsage: 'Назначение исполнителя',
     runnerUsageAll: 'Все',
     runnerUsageAgent: 'Agent / Конвейер',
@@ -2048,6 +2340,14 @@ const ruRU = {
       uploading: 'Загрузка...',
     },
     monitoring: {
+      rangeDays: 'Последние {{days}} дней',
+      customRange: 'Свой период',
+      rangeStart: 'Начало',
+      rangeEnd: 'Конец',
+      rangeHint: 'Начало должно быть раньше конца. Максимум 365 дней.',
+      autoRefreshLabel: 'Автообновление',
+      autoRefreshOff: 'Автообновление: выкл.',
+      autoRefreshEvery: 'Обновлять каждые {{seconds}} с',
       title: 'Журнал выполнения',
       workbench: 'Настройка и отладка',
       description:
@@ -2339,7 +2639,7 @@ const ruRU = {
     totpStatusDisabled: 'Не включено',
     totpCodesRemaining: 'Осталось кодов восстановления: {{count}}',
     totpManagerSectionDesc:
-      'Владельцы и администраторы могут просматривать и сбрасывать второй фактор любого аккаунта.',
+      'Владельцы и администраторы могут просматривать и сбрасывать второй фактор аккаунтов этого рабочего пространства.',
     revokeTotp: 'Перепривязать',
     totpAdminResetTitle:
       'Перепривязать двухфакторную аутентификацию для {{user}}',
@@ -2405,8 +2705,7 @@ const ruRU = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2475,6 +2774,123 @@ const ruRU = {
     },
   },
   monitoring: {
+    rangeDays: 'Последние {{days}} дней',
+    customRange: 'Свой период',
+    rangeStart: 'Начало',
+    rangeEnd: 'Конец',
+    rangeHint: 'Начало должно быть раньше конца. Максимум 365 дней.',
+    autoRefreshLabel: 'Автообновление',
+    autoRefreshOff: 'Автообновление: выкл.',
+    autoRefreshEvery: 'Обновлять каждые {{seconds}} с',
+    unified: {
+      inputs: 'Событие и ввод',
+      outputs: 'Сгенерированные результаты',
+      deliveries: 'Записи отправки',
+      conversation: 'Контекст диалога',
+      related: 'Связанные запуски',
+      events: 'События выполнения',
+      llm_calls: 'Вызовы модели',
+      tool_calls: 'Вызовы инструментов',
+      errors: 'Ошибки',
+      payload: 'Подробности',
+      metadata: 'Технические сведения',
+      loadMore: 'Загрузить ещё',
+      noRecords: 'Нет записей',
+      contextHint: 'Другие сообщения могут не относиться к этому запуску.',
+      unhandled: 'Обработчик не запущен',
+      roles: {
+        user: 'Пользователь',
+        assistant: 'Ассистент',
+        system: 'Система',
+        tool: 'Инструмент',
+      },
+    },
+    execution: {
+      title: 'Выполнения',
+      subtitle:
+        'Все входящие события и вызванные ими выполнения агентов и конвейеров.',
+      tabs: { executions: 'Выполнения' },
+      filters: {
+        source: 'Тип',
+        sourceAll: 'Все типы',
+        sourceAgent: 'Агенты',
+        sourcePipeline: 'Конвейеры',
+        mode: 'Режим выполнения',
+        modeAll: 'Обычные и отладочные',
+        modeLive: 'Только обычные',
+        modeDebug: 'Только отладочные',
+        status: 'Статус',
+        statusAll: 'Все статусы',
+      },
+      status: {
+        completed: 'Завершено',
+        failed: 'Ошибка',
+        running: 'Выполняется',
+        queued: 'В очереди',
+        cancelled: 'Отменено',
+        ignored: 'Пропущено',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Агент',
+        pipeline: 'Конвейер',
+        processor: 'Обработчик',
+      },
+      columns: {
+        time: 'Время',
+        target: 'Обработчик',
+        status: 'Статус',
+        duration: 'Длительность',
+        tokens: 'Токены',
+      },
+      debug: 'Отладка',
+      empty: 'За этот период выполнений нет',
+      loading: 'Загрузка выполнений...',
+      triggeredBy: 'Запущено: {{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95: около 50%/95% запусков завершаются за это время. Только недавние завершённые запуски Agent, без ожидания в очереди.',
+        tokensHint:
+          'Учитывается только записанный расход. Итог может быть неполным, если модель не возвращает расход.',
+        totalExecutions: 'Выполнения',
+        bySource: 'Агенты: {{agent}} · Конвейеры: {{pipeline}}',
+        successRate: 'Доля успешных выполнений',
+        successRateHint:
+          'Успешные ÷ (успешные + неудачные). Текущие, пропущенные и отменённые запуски исключены.',
+        denominator: 'Успешно: {{completed}} из {{denominator}} завершившихся',
+        running: 'Выполняется',
+        queued: 'В очереди',
+        waiting: 'Ожидание ввода',
+        failed: 'Ошибка',
+        ignored: 'Пропущено',
+        latency: 'Время выполнения',
+        p50: 'P50',
+        p95: 'P95',
+        sample: 'выборка: {{count}} выполнений',
+        tokens: 'Токены',
+        coverage: 'расход указан для {{withUsage}}/{{calls}} вызовов',
+        debugRuns: 'В том числе отладочных выполнений: {{count}}',
+      },
+      detail: {
+        title: 'Выполнение',
+        timeline: 'Хронология',
+        executionId: 'ID выполнения',
+        eventId: 'Событие',
+        queue: 'Очередь',
+        runner: 'Исполнитель',
+        created: 'Создано',
+        duration: 'Длительность',
+        usage: 'Расход',
+        noEvents: 'События трассировки не записаны',
+        query: 'Входящее сообщение',
+        modelCalls: 'Вызовы моделей',
+        toolCalls: 'Вызовы инструментов',
+        errors: 'Ошибки',
+        loadError: 'Не удалось загрузить трассировку выполнения',
+        openObject: 'Открыть объект',
+        a11yDescription: 'Полная трассировка этого выполнения',
+      },
+    },
     title: 'Мониторинг',
     description:
       'Мониторинг активности ботов, вызовов LLM и производительности системы',
@@ -2491,6 +2907,9 @@ const ruRU = {
       pipeline: 'Конвейер',
       allBots: 'Все боты',
       selectBot: 'Выберите бота',
+      processor: 'Обработчик',
+      allProcessors: 'Все обработчики',
+      selectProcessor: 'Выберите обработчик',
       allPipelines: 'Все конвейеры',
       selectPipeline: 'Выберите конвейер',
       loading: 'Загрузка...',
@@ -2516,6 +2935,15 @@ const ruRU = {
       sessions: 'Анализ сессий',
       errors: 'Журнал ошибок',
     },
+    pipelineTrace: {
+      received: 'Входящее сообщение',
+      processing: 'Обработка',
+      replies: 'Ответы',
+      noSteps: 'Вызовы моделей и инструментов не зарегистрированы',
+      success: 'Успешно',
+      error: 'Ошибка',
+      pending: 'Ожидание',
+    },
     messageList: {
       timestamp: 'Время',
       bot: 'Бот',
@@ -2738,7 +3166,7 @@ const ruRU = {
     title: 'Настройки',
     nav: {
       workspace: 'Workspace',
-      operationTrace: 'Трассировка',
+      operationTrace: 'Журнал',
       models: 'Модели',
       api: 'API',
       storage: 'Хранилище',
@@ -2746,21 +3174,21 @@ const ruRU = {
     },
   },
   operationTrace: {
-    title: 'Трассировка операций',
+    title: 'Журнал операций',
     description: 'Отслеживание изменений и просмотров администраторов',
+    reset: 'Сбросить',
+    settingsUpdated: 'Настройки сохранены',
+    settingsUpdateFailed: 'Не удалось сохранить настройки',
+    levelBadge: 'Уровень: {{level}}',
+    collectionSettings: 'Настройки сбора',
     captureLevel: 'Уровень записи',
     retention: 'Хранение',
-    retentionSummary: 'Хранить {{days}} дн. · до {{rows}} записей',
     retentionDays: 'Дней хранения',
     maxRows: 'Максимум записей',
     dedupeWindow: 'Окно дедупликации (с)',
     records: 'Записи операций',
     changesCount: 'Изменений: {{count}}',
     hideDetails: 'Скрыть детали',
-    levelUpdated: 'Уровень записи обновлён',
-    levelUpdateFailed: 'Не удалось обновить уровень записи',
-    retentionUpdated: 'Хранение обновлено',
-    retentionUpdateFailed: 'Не удалось обновить хранение',
     export: 'Скачать журнал',
     exportFailed: 'Не удалось скачать журнал операций',
     empty: 'Записей операций пока нет',
@@ -2774,7 +3202,7 @@ const ruRU = {
     filterAllResources: 'Все ресурсы',
     filterActor: 'Исполнитель',
     filterAllActors: 'Все исполнители',
-    loadFailed: 'Не удалось загрузить данные трассировки',
+    loadFailed: 'Не удалось загрузить журнал операций',
     refresh: 'Обновить',
     save: 'Сохранить',
     redacted: 'Скрыто',
@@ -2793,7 +3221,7 @@ const ruRU = {
     elidedValue: '{{count}} символов',
     levels: {
       off: {
-        label: 'Трассировка выключена',
+        label: 'Журнал выключен',
       },
       mutation: {
         label: 'Только изменения',
diff --git a/web/src/i18n/locales/slack-setup/en-US.ts b/web/src/i18n/locales/slack-setup/en-US.ts
new file mode 100644
index 000000000..d15c65503
--- /dev/null
+++ b/web/src/i18n/locales/slack-setup/en-US.ts
@@ -0,0 +1,31 @@
+export default {
+  permissionsHint:
+    'Requests message, channel, user, file and reaction permissions. Subscribes to edits/deletions, reactions, membership, channel updates, App Home and app lifecycle events; enables button/form callbacks. Channel replies still require an @mention. Existing apps need updated subscriptions and reinstallation after adding scopes.',
+  title: 'Quick Slack setup',
+  description:
+    'Create and configure a Slack app, authorize installation, then fill the adapter credentials.',
+  saveFirst: 'Create the bot draft first, then open this assistant.',
+  generate: 'Generate App Configuration Tokens',
+  name: 'App name',
+  redirect: 'OAuth return URL',
+  webhook: 'Webhook URL',
+  redirectHint:
+    'Only your browser needs to reach this URL. HTTPS or http://localhost is supported; no public event webhook is needed.',
+  webhookHint: 'Use your public HTTPS URL and keep the /bots/… path unchanged.',
+  create: 'Create and configure',
+  creating: 'Creating and configuring the app…',
+  authorizing: 'Completing authorization…',
+  waiting:
+    'The app is configured. Authorize its installation in Slack, then return here.',
+  authorize: 'Authorize in Slack',
+  appTokenHint:
+    'Socket Mode also needs an App-Level Token. In Basic Information → App-Level Tokens, generate a token with connections:write and paste it below.',
+  openApp: 'Open app settings',
+  failed: 'Setup failed:',
+  appCreated:
+    'An app has already been created. Retrying creates another app; you can manage the existing one in Slack:',
+  retry: 'Start a new setup',
+  success:
+    'Authorization completed. Apply the credentials below, then save the bot configuration.',
+  apply: 'Fill adapter configuration',
+};
diff --git a/web/src/i18n/locales/slack-setup/ja-JP.ts b/web/src/i18n/locales/slack-setup/ja-JP.ts
new file mode 100644
index 000000000..4609f8923
--- /dev/null
+++ b/web/src/i18n/locales/slack-setup/ja-JP.ts
@@ -0,0 +1,30 @@
+export default {
+  permissionsHint:
+    'メッセージ、チャンネル、ユーザー、ファイル、リアクションの権限を要求します。編集・削除、リアクション、メンバー、チャンネル更新、App Home、アプリのライフサイクルを購読し、ボタン・フォームのコールバックを有効にします。チャンネルの通常返信には @メンションが必要です。既存アプリは購読を追加し、権限追加後に再インストールしてください。',
+  title: 'Slack クイック設定',
+  description:
+    'アプリを自動作成・設定し、インストール承認後に認証情報を入力します。',
+  saveFirst: '先にボットの下書きを作成してください。',
+  generate: 'App Configuration Tokens を生成',
+  name: 'アプリ名',
+  redirect: 'OAuth 戻り先 URL',
+  webhook: 'Webhook URL',
+  redirectHint:
+    'ブラウザーからアクセス可能な HTTPS または http://localhost を指定します。公開イベント Webhook は不要です。',
+  webhookHint: '公開 HTTPS URL を指定し、末尾の /bots/… を維持してください。',
+  create: '作成して自動設定',
+  creating: 'アプリを作成・設定中…',
+  authorizing: '承認を完了中…',
+  waiting: '設定が完了しました。Slack でインストールを承認して戻ってください。',
+  authorize: 'Slack で承認',
+  appTokenHint:
+    'Socket Mode には App-Level Token が必要です。Basic Information → App-Level Tokens で connections:write 権限のトークンを生成し、入力してください。',
+  openApp: 'アプリ設定を開く',
+  failed: '設定に失敗しました：',
+  appCreated:
+    'アプリは作成済みです。再試行すると別のアプリが作成されます。既存のアプリは Slack で管理できます：',
+  retry: '新しい設定を開始',
+  success:
+    '承認が完了しました。認証情報を入力してボット設定を保存してください。',
+  apply: 'アダプター設定に入力',
+};
diff --git a/web/src/i18n/locales/slack-setup/zh-Hans.ts b/web/src/i18n/locales/slack-setup/zh-Hans.ts
new file mode 100644
index 000000000..06c684977
--- /dev/null
+++ b/web/src/i18n/locales/slack-setup/zh-Hans.ts
@@ -0,0 +1,28 @@
+export default {
+  permissionsHint:
+    '申请消息、频道、用户、文件和表情反应权限，订阅消息编辑/删除、表情反应、成员变化、频道更新、应用首页和生命周期事件，并启用按钮/表单回调。频道普通回复仍需 @ 机器人。已有应用需补齐订阅，新增权限后重新安装授权。',
+  title: 'Slack 快速配置',
+  description: '自动创建和配置应用，授权安装后填入适配器凭据。',
+  saveFirst: '请先创建机器人草稿，再打开快速配置助手。',
+  generate: '生成 App Configuration Tokens',
+  name: '应用名称',
+  redirect: 'OAuth 授权返回地址',
+  webhook: 'Webhook 地址',
+  redirectHint:
+    '此地址只需浏览器能够访问，支持 HTTPS 或 http://localhost；不需要公网事件回调。',
+  webhookHint: '填写公网 HTTPS 地址，保留末尾 /bots/… 路径。',
+  create: '创建并自动配置',
+  creating: '正在创建和配置应用…',
+  authorizing: '正在完成授权…',
+  waiting: '应用配置完成。请在 Slack 确认安装授权，然后返回此处。',
+  authorize: '前往 Slack 授权',
+  appTokenHint:
+    '长连接还需要 App-Level Token。在 Basic Information → App-Level Tokens 中生成带 connections:write 权限的 Token，填入下方。',
+  openApp: '打开应用设置',
+  failed: '配置失败：',
+  appCreated:
+    '应用已创建。重新开始会创建另一个应用，已有应用可在 Slack 后台管理：',
+  retry: '重新开始配置',
+  success: '授权完成。填入凭据后，请保存机器人配置。',
+  apply: '填入适配器配置',
+};
diff --git a/web/src/i18n/locales/slack-setup/zh-Hant.ts b/web/src/i18n/locales/slack-setup/zh-Hant.ts
new file mode 100644
index 000000000..975ec7ce8
--- /dev/null
+++ b/web/src/i18n/locales/slack-setup/zh-Hant.ts
@@ -0,0 +1,28 @@
+export default {
+  permissionsHint:
+    '申請訊息、頻道、使用者、檔案和表情反應權限，訂閱訊息編輯/刪除、表情反應、成員變化、頻道更新、應用首頁和生命週期事件，並啟用按鈕/表單回呼。頻道一般回覆仍需 @ 機器人。既有應用需補齊訂閱，新增權限後重新安裝授權。',
+  title: 'Slack 快速設定',
+  description: '自動建立及設定應用程式，授權安裝後填入介接器憑證。',
+  saveFirst: '請先建立機器人草稿，再開啟設定助手。',
+  generate: '產生 App Configuration Tokens',
+  name: '應用程式名稱',
+  redirect: 'OAuth 授權返回網址',
+  webhook: 'Webhook 網址',
+  redirectHint:
+    '此網址只需瀏覽器能存取，支援 HTTPS 或 http://localhost；無需公開事件回呼。',
+  webhookHint: '填寫公開 HTTPS 網址，保留結尾 /bots/… 路徑。',
+  create: '建立並自動設定',
+  creating: '正在建立及設定應用程式…',
+  authorizing: '正在完成授權…',
+  waiting: '設定完成。請在 Slack 確認安裝授權，再返回此處。',
+  authorize: '前往 Slack 授權',
+  appTokenHint:
+    '長連線還需要 App-Level Token。在 Basic Information → App-Level Tokens 產生具有 connections:write 權限的 Token，填入下方。',
+  openApp: '開啟應用程式設定',
+  failed: '設定失敗：',
+  appCreated:
+    '應用程式已建立。重新開始會建立另一個應用程式，可在 Slack 後台管理現有應用程式：',
+  retry: '重新開始設定',
+  success: '授權完成。填入憑證後，請儲存機器人設定。',
+  apply: '填入介接器設定',
+};
diff --git a/web/src/i18n/locales/th-TH.ts b/web/src/i18n/locales/th-TH.ts
index 78abae6e4..5f11856bc 100644
--- a/web/src/i18n/locales/th-TH.ts
+++ b/web/src/i18n/locales/th-TH.ts
@@ -1,6 +1,172 @@
 import pipelineMigration from './pipeline-migration/th-TH';
 const thTH = {
+  planRevision: {
+    subscribedOn: 'วันที่สมัคร: {{date}}',
+    latestQuota: 'โควตาล่าสุดของแผนนี้',
+    showLatest: 'ดูโควตาล่าสุดของแผนนี้',
+    showCurrent: 'ดูการสมัครสมาชิกปัจจุบัน',
+    latestTab: 'ล่าสุด',
+    currentTab: 'ปัจจุบัน',
+    cancelToSubscribe: 'ยกเลิกการสมัครสมาชิกเพื่อรับโควต้าล่าสุดในรอบถัดไป',
+
+    current: 'Current subscription: {{plan}} · revision {{revision}}',
+    owned: 'Purchased revision {{revision}}',
+    latest: 'Latest revision {{revision}}',
+    changed:
+      'Your purchased benefits and the latest version differ. Highlighted rows show the changes.',
+    same: 'Benefits match latest revision {{revision}}. Your purchased revision remains unchanged.',
+    unavailable:
+      'The latest plan is unavailable. Your purchased benefits remain in effect.',
+    benefit: 'Benefit',
+    change: 'Changed',
+    enabled: 'Enabled',
+    disabled: 'Disabled',
+    monthlyCredits: 'Monthly bonus credits',
+    retention: 'Chat history retention (days)',
+    unknown:
+      'Current subscription details are unavailable. Refresh or manage your subscription in Space; latest catalog benefits are not your purchased benefits.',
+    handoff:
+      'Manage your subscription in Space. Cancel automatic renewal before purchasing a new version. Prepaid subscriptions end automatically unless renewed before expiry.',
+    cancelFirst:
+      'Cancel automatic renewal in subscription management first. Cancellation preserves paid benefits until the end of the period, without a refund. After confirmation, you can purchase the latest version, effective when payment succeeds.',
+    canceled:
+      'Renewal is canceled. Paid benefits remain until the end of the period, without a refund. A new purchase applies its benefits when payment succeeds.',
+    prepaid:
+      'This prepaid subscription ends automatically. You can renew your current plan before expiry; after expiry, subscriptions use the latest quotas for each plan.',
+    renew: 'ต่ออายุการสมัครปัจจุบัน',
+    buying: 'Purchasing revision {{revision}}.',
+    fields: {
+      bots_max: 'Bots',
+      members_max: 'Members',
+      pipelines_max: 'Processors',
+      plugins_max: 'Extensions',
+      knowledge_bases_max: 'Knowledge bases',
+      managed_sandbox_sessions_max: 'Sandbox sessions',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: 'ยังไม่เปิดรับการสมัคร',
+    subscriptionUnavailableHint:
+      'แผนนี้ยังไม่รับการสมัครใหม่ โปรดเลือกแผนอื่น การสมัครเดิมและการต่ออายุอัตโนมัติจะไม่ได้รับผลกระทบ',
+    platformNames: {
+      qqofficial: 'QQ Official API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComCustomerService',
+      officialaccount: 'Official Account',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'จัดการการสมัครสมาชิก',
+    downgradeUnavailable: 'ไม่สามารถลดระดับเป็นแผนฟรีได้',
+    features: {
+      multiUserWorkspace: 'พื้นที่ทำงานสำหรับผู้ใช้หลายคน',
+      remoteE2BSandbox: 'การตั้งค่าแซนด์บ็อกซ์ E2B ระยะไกล',
+      communitySupport: 'การสนับสนุนจากชุมชน',
+      everythingInFree: 'ฟีเจอร์ทั้งหมดในแผนฟรี',
+      alwaysOnSandbox: 'แซนด์บ็อกซ์ส่วนกลางที่ทำงานตลอดเวลา 1 รายการ',
+      higherWorkspaceLimits: 'โควตาพื้นที่ทำงานที่สูงขึ้น',
+    },
+    title: 'เปรียบเทียบแผน',
+    description: 'เลือกแผนสำหรับ {{workspace}}',
+    billing: 'รอบการเรียกเก็บเงิน',
+    monthly: 'รายเดือน',
+    annual: 'รายปี',
+    currency: 'สกุลเงิน',
+    loading: 'กำลังโหลดแผน…',
+    error: 'ไม่สามารถโหลดแผนได้',
+    retry: 'ลองอีกครั้ง',
+    empty: 'ไม่มีแผนที่ใช้ได้',
+    current: 'แผนปัจจุบัน',
+    recommended: 'แนะนำ',
+    free: 'ฟรี',
+    annualNote: 'เทียบเท่า {{price}} / เดือน',
+    ownerOnly: 'เฉพาะเจ้าของหรือผู้ดูแลระบบ',
+    choose: 'ดำเนินการชำระเงิน',
+    handoff: 'เข้าสู่ระบบและชำระเงินใน Space โดยจะเก็บตัวเลือกของคุณไว้',
+    quotas: 'โควตาพื้นที่ทำงาน',
+    unlimited: 'ไม่จำกัด',
+    retention: 'ประวัติข้อความ · {{count}} วัน',
+    retentionLabel: 'ระยะเวลาเก็บประวัติข้อความ',
+    retentionHelp: 'ประวัติข้อความจะถูกลบโดยอัตโนมัติเมื่อเกินระยะเวลานี้',
+    credits: '{{count}} เครดิตต่อเดือน',
+    creditsLabel: 'เครดิตรายเดือน',
+    creditsHelp: 'ได้รับเครดิตทุกเดือนของการสมัคร รวมถึงแผนรายปี',
+    platforms: 'รองรับแพลตฟอร์มหลัก',
+    excludedPlatforms: 'ไม่รองรับ: {{platforms}}',
+    allPlatforms: 'ทุกแพลตฟอร์มที่พร้อมใช้งาน',
+    limits: {
+      bots: 'บอต',
+      pipelines: 'ตัวประมวลผล',
+      knowledge_bases: 'ฐานความรู้',
+      plugins: 'ส่วนขยาย',
+      members: 'สมาชิก',
+      managed_sandbox_sessions: 'เซสชันแซนด์บ็อกซ์ที่จัดการ',
+    },
+  },
+  workspaceLanguage: {
+    label: 'ภาษาของพื้นที่ทำงาน',
+    description:
+      'ใช้สำหรับข้อความจากระบบและคำตอบเริ่มต้นของบอต โดยไม่เปลี่ยนภาษาอินเทอร์เฟซของคุณ',
+    saved: 'บันทึกภาษาแล้ว',
+    saveFailed: 'บันทึกภาษาไม่สำเร็จ',
+  },
+  inflight: {
+    autoCollapse: 'ยุบอัตโนมัติ',
+    empty: 'ไม่มีเหตุการณ์ที่กำลังประมวลผล',
+    move: 'ลากขึ้นลงหรือใช้ปุ่มลูกศรเพื่อย้าย',
+    botLogs: 'บันทึกบอต',
+    processorLogs: 'บันทึกตัวประมวลผล',
+
+    title: 'กำลังดำเนินการ',
+    expand: 'ขยาย',
+    collapse: 'ย่อ',
+    reconnecting: 'กำลังเชื่อมต่อใหม่ แสดงสถานะล่าสุดที่ได้รับ',
+    details: 'ดูรายละเอียด',
+    debug: 'เหตุการณ์ทดสอบ',
+    limited: 'แสดงบางรายการ ดูทั้งหมดได้ที่แดชบอร์ด',
+    queued: 'อยู่ในคิว',
+    running: 'กำลังประมวลผล',
+    waiting: 'รอข้อมูล',
+    tool: 'กำลังเรียกเครื่องมือ',
+    thinking: 'กำลังคิด',
+    replying: 'กำลังเตรียมคำตอบ',
+    generating: 'กำลังสร้างคำตอบ',
+    completed: 'เสร็จสิ้น',
+    failed: 'ล้มเหลว',
+    cancelled: 'ยกเลิกแล้ว',
+    ignored: 'ละเว้นแล้ว',
+  },
   assistant: {
+    movePosition: 'เลื่อนผู้ช่วยในแนวนอน',
+    resizeHeight: 'ปรับความสูงของผู้ช่วย',
+    sessions: 'การสนทนา',
+    currentChat: 'การสนทนาปัจจุบัน',
+    stop: 'หยุดงาน',
+    stopped:
+      'หยุดแล้ว เครื่องมืออาจดำเนินการไปแล้ว โปรดตรวจสอบทรัพยากรก่อนลองอีกครั้ง',
+    setupRequired:
+      'เข้าสู่ระบบด้วยบัญชี LangBot หรือตั้งค่าโมเดลเพื่อเริ่มใช้ผู้ช่วยพื้นที่ทำงาน',
+    loginAccount: 'เข้าสู่ระบบ LangBot Account',
+    configureModels: 'ตั้งค่าโมเดล',
+    providerCheckFailed: 'ไม่สามารถตรวจสอบผู้ให้บริการโมเดลได้ โปรดลองอีกครั้ง',
+    retrySetup: 'ลองอีกครั้ง',
+    preparing: 'กำลังเตรียมผู้ช่วย…',
+    thinking: 'โมเดลกำลังตอบ · รอบ {{round}}',
+    executing: 'กำลังใช้เครื่องมือ…',
+    toolRunning: 'กำลังทำงาน',
+    recommendationFailed: 'โหลดโมเดลแนะนำไม่สำเร็จ โปรดเลือกโมเดล',
+    expand: 'ขยายผู้ช่วยพื้นที่ทำงาน',
+    collapse: 'ย่อผู้ช่วยพื้นที่ทำงาน',
     modelHint: 'สลับโมเดลผู้ช่วยสำหรับข้อความถัดไป',
     details: 'ดูข้อมูลดิบ',
     found: 'พบ {{count}} รายการ',
@@ -21,7 +187,7 @@ const thTH = {
       create_knowledge_base: 'สร้างฐานความรู้',
       get_pipeline: 'ดู Pipeline',
       get_knowledge_schema: 'ดูสคีมาของ knowledge engine',
-      list_operation_logs: 'ดูการติดตามการดำเนินการ',
+      list_operation_logs: 'ดูบันทึกการดำเนินการ',
     },
     resources: {
       models: 'ค้นหาโมเดลแชต',
@@ -35,7 +201,7 @@ const thTH = {
     newChat: 'บทสนทนาใหม่',
     close: 'ปิด',
     welcome:
-      'บอกฉันว่าคุณอยากสร้างอะไร ฉันตรวจสอบพื้นที่ทำงานนี้และเสนอการตั้งค่าฐานความรู้และ Pipeline ให้คุณยืนยันก่อนดำเนินการ',
+      'บอกฉันว่าคุณต้องการสร้างแอปอะไร ฉันสามารถตรวจสอบพื้นที่ทำงานนี้และเสนอการตั้งค่าสำหรับบอต, Agent, ฐานความรู้, Pipeline และอื่น ๆ แล้วดำเนินการหลังจากคุณยืนยันได้',
     discover: 'มีโมเดล ฐานความรู้ และ Pipeline อะไรที่ใช้ได้บ้าง?',
     build: 'ช่วยสร้างแอปตอบคำถามจากฐานความรู้ให้หน่อย',
     toolResult: 'ผลลัพธ์การดำเนินการ',
@@ -55,6 +221,14 @@ const thTH = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: '{{days}} วันที่ผ่านมา',
+        customRange: 'กำหนดช่วงเวลาเอง',
+        rangeStart: 'เวลาเริ่มต้น',
+        rangeEnd: 'เวลาสิ้นสุด',
+        rangeHint: 'เวลาเริ่มต้องก่อนเวลาสิ้นสุด สูงสุด 365 วัน',
+        autoRefreshLabel: 'รีเฟรชอัตโนมัติ',
+        autoRefreshOff: 'รีเฟรชอัตโนมัติ: ปิด',
+        autoRefreshEvery: 'รีเฟรชทุก {{seconds}} วินาที',
         title: 'ดูภาพรวมการทำงาน',
         description:
           'ดูการทำงานของบอท การใช้โมเดล ปริมาณข้อความ และประสิทธิภาพระบบ',
@@ -135,6 +309,14 @@ const thTH = {
           'ส่งข้อความทดสอบเพื่อตรวจเงื่อนไข คำตอบ AI และการประมวลผลผลลัพธ์ การเปลี่ยนแปลงจะถูกบันทึกก่อนทดสอบ',
       },
       monitoring: {
+        rangeDays: '{{days}} วันที่ผ่านมา',
+        customRange: 'กำหนดช่วงเวลาเอง',
+        rangeStart: 'เวลาเริ่มต้น',
+        rangeEnd: 'เวลาสิ้นสุด',
+        rangeHint: 'เวลาเริ่มต้องก่อนเวลาสิ้นสุด สูงสุด 365 วัน',
+        autoRefreshLabel: 'รีเฟรชอัตโนมัติ',
+        autoRefreshOff: 'รีเฟรชอัตโนมัติ: ปิด',
+        autoRefreshEvery: 'รีเฟรชทุก {{seconds}} วินาที',
         title: 'ดูประวัติการทำงาน',
         description:
           'ใช้แท็บข้างชื่อเพื่อดูบทสนทนา ขั้นตอนการประมวลผล และข้อผิดพลาด เมื่อไม่มีคำตอบหรือได้คำตอบที่ไม่คาดคิด',
@@ -299,6 +481,11 @@ const thTH = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'ไม่ได้บันทึกสาเหตุของข้อผิดพลาด โปรดตรวจสอบบันทึกของบอตหรือส่งข้อความอีกครั้งเพื่อดูรายละเอียด',
+    loadFailedTitle: 'ไม่สามารถโหลดเนื้อหาได้',
+    loadFailedHint:
+      'โปรดลองอีกครั้ง หากยังโหลดไม่ได้ ให้ตรวจสอบการเชื่อมต่อและสถานะบริการ',
     customValue: 'กำหนดเอง',
     loadFailed: 'โหลดไม่สำเร็จ โปรดลองอีกครั้ง',
     search: 'ค้นหา',
@@ -493,6 +680,9 @@ const thTH = {
     help: 'ขอความช่วยเหลือ',
   },
   models: {
+    starModel: 'ติดดาวเพื่อใช้เป็นค่าเริ่มต้นสำหรับตัวประมวลผลใหม่',
+    unstarModel: 'ยกเลิกดาว',
+    starModelFailed: 'อัปเดตโมเดลที่ติดดาวไม่สำเร็จ',
     codex: {
       account: 'การสมัครสมาชิก ChatGPT',
       description:
@@ -748,10 +938,13 @@ const thTH = {
     deleteConfirmation: 'คุณแน่ใจหรือไม่ว่าต้องการลบ Bot นี้?',
     platformAdapter: 'การเลือกแพลตฟอร์ม/อะแดปเตอร์',
     selectAdapter: 'เลือกอะแดปเตอร์',
+    saveAndEnable: 'บันทึกและเปิดใช้งาน',
+    searchAdapters: 'ค้นหาอะแดปเตอร์',
+    noMatchingAdapters: 'ไม่พบอะแดปเตอร์ที่ตรงกัน',
     legacyAdapters: 'อะแดปเตอร์รุ่นเก่า',
     legacyAdapterBadge: 'เก่า',
     legacyAdaptersHint:
-      'อะแดปเตอร์เหล่านี้มีรุ่นใหม่แบบขับเคลื่อนด้วยเหตุการณ์แล้ว\nและเก็บไว้เพื่อความเข้ากันได้กับการตั้งค่าเดิมเท่านั้น ไม่แนะนำสำหรับ Bot ใหม่',
+      'อะแดปเตอร์รุ่นเก่าเน้นรับส่งข้อความ และคงไว้เพื่อรองรับการตั้งค่าบอทเดิม\nรุ่นใหม่เชื่อมต่อแพลตฟอร์มเดียวกันผ่านเหตุการณ์ ทั้งข้อความ คำขอเป็นเพื่อนหรือการเปลี่ยนแปลงกลุ่มที่รองรับ แล้วส่งให้ตัวประมวลผล\nแนะนำรุ่นใหม่สำหรับบอทใหม่ บอทเดิมยังใช้รุ่นเก่าได้ เมื่อเปลี่ยนให้ตรวจการเชื่อมต่อและเส้นทางเหตุการณ์ ไม่ใช่การอัปเกรดบัญชีแพลตฟอร์ม',
     adapterConfig: 'การกำหนดค่าอะแดปเตอร์',
     viewAdapterDocs: 'ดูเอกสาร',
     bindPipeline: 'ผูก Pipeline',
@@ -764,6 +957,10 @@ const thTH = {
     earlier: 'ก่อนหน้า',
     dateFormat: '{{day}}/{{month}}',
     setBotEnableError: 'ไม่สามารถตั้งค่าสถานะเปิดใช้งาน Bot ได้',
+    enabling: 'กำลังเปิดใช้งาน…',
+    disabling: 'กำลังปิดใช้งาน…',
+    enableConfirmed: 'เปิดใช้งานแล้ว',
+    disableConfirmed: 'ปิดใช้งานแล้ว',
     log: 'บันทึก',
     configuration: 'การกำหนดค่า',
     logs: 'บันทึก',
@@ -772,6 +969,27 @@ const thTH = {
     routingConnection: 'การกำหนดเส้นทางและการเชื่อมต่อ',
     routingConnectionDescription:
       'ผูก Pipeline ที่ประมวลผลข้อความสำหรับ Bot นี้',
+    routingHelp: {
+      title: 'การกำหนดเส้นทางเหตุการณ์คืออะไร?',
+      intro: 'บอทรับได้ทั้งข้อความ คำขอเป็นเพื่อน และเหตุการณ์เข้าร่วมกลุ่ม',
+      events: 'เหตุการณ์ที่ได้รับ',
+      agents: 'Agent แต่ละตัว',
+      message: 'ได้รับข้อความ',
+      friend: 'คำขอเป็นเพื่อน',
+      group: 'เข้าร่วมกลุ่ม',
+      messageAction: 'Agent ตอบคำถาม',
+      friendAction: 'Agent ตรวจคำขอ',
+      groupAction: 'Agent ต้อนรับ',
+      rules: 'ตามกฎที่คุณตั้ง · ตัวอย่าง',
+      instructions:
+        'Agent ทำตามคำสั่งของคุณและใช้เครื่องมือเพื่อตอบข้อความหรือตรวจคำขอ',
+      plugin:
+        'สำหรับงานขั้นสูง ตัวประมวลผลปลั๊กอินลงทะเบียนหลายเหตุการณ์และจัดการด้วยโค้ดได้ เพิ่มการตั้งค่าตัวประมวลผลปลั๊กอินด้านล่าง',
+      docs: 'พัฒนาตัวประมวลผลปลั๊กอินของคุณเอง',
+      runnerDocs: 'พัฒนารันเนอร์ Agent ของคุณเอง',
+      pause: 'หยุดชั่วคราว',
+      play: 'เล่น',
+    },
     eventRouting: 'การกำหนดเส้นทางเหตุการณ์',
     eventRoutingDescription:
       'เลือกตัวประมวลผลสำหรับแต่ละเหตุการณ์ของ Bot และแก้ไขตรรกะในการตั้งค่า Agent หรือ Pipeline ที่เกี่ยวข้อง โดย Pipeline รองรับเฉพาะเหตุการณ์ข้อความ',
@@ -810,7 +1028,8 @@ const thTH = {
     enabled: 'เปิดใช้งาน',
     eventBindingDescriptionPlaceholder: 'คำอธิบายกฎ',
     noEventBindings: 'ไม่มีเส้นทางเหตุการณ์',
-    unsupportedPipelineEvent: 'ไปป์ไลน์ใช้ได้กับเหตุการณ์ message.* เท่านั้น',
+    unsupportedPipelineEvent:
+      'ไปป์ไลน์ใช้ได้กับเหตุการณ์ message.received เท่านั้น',
     disable: 'ปิดใช้งาน',
     enable: 'เปิดใช้งาน',
     disabledBindings: 'ปิดใช้งาน',
@@ -820,6 +1039,7 @@ const thTH = {
     advancedEventValues: 'ดูทั้งหมด',
     eventGroup: 'กลุ่ม',
     eventGroupNames: {
+      wecomcs: 'บริการลูกค้า WeCom',
       bot: 'สถานะบอท',
       feedback: 'ข้อเสนอแนะ',
       friend: 'เพื่อน',
@@ -834,6 +1054,9 @@ const thTH = {
     routeFallbackCatchAll: '{{route}} เป็นเส้นทางรับทุกเหตุการณ์',
     routeFallbackIgnored:
       'เหตุการณ์ที่ไม่ตรงกับเส้นทางจะถูกละเว้น เพิ่มเส้นทางรับทั้งหมดเฉพาะเมื่อต้องกำหนดผลลัพธ์ให้ทุกเหตุการณ์',
+    matchRecords: 'บันทึกการจับคู่',
+    matchRecordsDescription:
+      'แสดงผลล่าสุดของแต่ละกฎและเหตุการณ์ล่าสุดที่ไม่ตรงกัน บันทึกในหน่วยความจำจะถูกล้างเมื่อรีสตาร์ท',
     testRoute: 'ตรวจเส้นทาง',
     adapterEventDebugAction: 'ทดสอบการรับเหตุการณ์',
     adapterEventDebugTitle: 'ดีบักเหตุการณ์แพลตฟอร์ม',
@@ -915,6 +1138,15 @@ const thTH = {
     eventWildcard: 'ทุกเหตุการณ์',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'ผู้ใช้เข้าสู่แชตบอต',
+      wecombot_template_card_event: 'การโต้ตอบกับการ์ด',
+      dingtalk_card_action: 'การโต้ตอบกับการ์ด DingTalk',
+      wecomcs_enter_session: 'ลูกค้าเข้าสู่การสนทนา',
+      wecomcs_msg_send_fail: 'ส่งข้อความบริการลูกค้าไม่สำเร็จ',
+      wecomcs_servicer_status_change: 'สถานะเจ้าหน้าที่เปลี่ยนแปลง',
+      wecomcs_session_status_change: 'สถานะการสนทนาบริการลูกค้าเปลี่ยนแปลง',
+      wecomcs_reject_customer_msg_switch_change:
+        'การปฏิเสธข้อความลูกค้าเปลี่ยนแปลง',
       message_received: 'ได้รับข้อความ',
       message_edited: 'แก้ไขข้อความ',
       message_deleted: 'ลบข้อความ',
@@ -924,6 +1156,7 @@ const thTH = {
       friend_added: 'เพิ่มเพื่อน',
       group_member_joined: 'สมาชิกเข้าร่วมกลุ่ม',
       group_member_left: 'สมาชิกออกจากกลุ่ม',
+      group_info_updated: 'อัปเดตข้อมูลกลุ่ม',
       group_member_banned: 'สมาชิกถูกแบน',
       bot_invited_to_group: 'บอตได้รับเชิญเข้ากลุ่ม',
       bot_removed_from_group: 'บอตถูกนำออกจากกลุ่ม',
@@ -932,14 +1165,28 @@ const thTH = {
       platform_specific: 'เหตุการณ์เฉพาะแพลตฟอร์ม',
     },
     eventDescriptions: {
+      wecombot_enter_chat: 'ผู้ใช้เปิดแชตบอต ใช้สำหรับขั้นตอนต้อนรับและแนะนำ',
+      wecombot_template_card_event:
+        'ผู้ใช้คลิกหรือส่งการ์ด การตอบกลับแบบฟอร์มที่ระบบจัดการจะประมวลผลแยกกัน',
+      dingtalk_card_action:
+        'การคลิกและส่งการ์ด ความคิดเห็นและแบบฟอร์มที่ระบบจัดการจะประมวลผลแยกกัน',
+      wecomcs_enter_session:
+        'ลูกค้าเปิดลิงก์บริการ พร้อมข้อมูลแหล่งที่มาและรหัสต้อนรับ',
+      wecomcs_msg_send_fail:
+        'WeChat แจ้งการส่งล้มเหลว พร้อม ID ข้อความและประเภทข้อผิดพลาด',
+      wecomcs_servicer_status_change:
+        'เจ้าหน้าที่เริ่ม หยุด หรือพักการให้บริการ',
+      wecomcs_session_status_change:
+        'เจ้าหน้าที่รับ โอน ปิด หรือเปิดการสนทนาอีกครั้ง',
+      wecomcs_reject_customer_msg_switch_change:
+        'เจ้าหน้าที่เปิดหรือปิดการปฏิเสธข้อความลูกค้า',
       all: 'ตรงกับทุกเหตุการณ์ที่อะแดปเตอร์นี้ได้รับ',
       namespace: 'ตรงกับทุกเหตุการณ์ของ {{group}}',
       namespace_bot:
         'ตรงกับการเชิญ การนำบอทออก การปิดเสียง และเหตุการณ์สถานะอื่น',
       namespace_feedback: 'ตรงกับเหตุการณ์ข้อเสนอแนะจากแพลตฟอร์มหรือผู้ใช้',
       namespace_friend: 'ตรงกับคำขอเป็นเพื่อนและการเปลี่ยนแปลงความสัมพันธ์',
-      namespace_group:
-        'ตรงกับการเข้ากลุ่ม ออกจากกลุ่ม ถูกนำออก และเหตุการณ์กลุ่มอื่น',
+      namespace_group: 'ตรงกับการเปลี่ยนแปลงสมาชิกและการอัปเดตข้อมูลกลุ่ม',
       namespace_message: 'ตรงกับการรับ แก้ไข ลบข้อความ และแสดงปฏิกิริยา',
       namespace_platform: 'ตรงกับเหตุการณ์เฉพาะแพลตฟอร์มที่อะแดปเตอร์ให้มา',
       custom: 'เหตุการณ์กำหนดเองหรือยังไม่มีคำอธิบาย',
@@ -952,6 +1199,8 @@ const thTH = {
       friend_added: 'สร้างความสัมพันธ์เพื่อนแล้ว',
       group_member_joined: 'สมาชิกเข้าร่วมกลุ่มที่มีบอท',
       group_member_left: 'สมาชิกออกจากกลุ่มที่มีบอท',
+      group_info_updated:
+        'ข้อมูลกลุ่ม เช่น ชื่อหรือเจ้าของกลุ่ม มีการเปลี่ยนแปลง',
       group_member_banned: 'สมาชิกกลุ่มถูกแบนหรือนำออก',
       bot_invited_to_group: 'บอทได้รับเชิญเข้ากลุ่ม',
       bot_removed_from_group: 'บอทถูกนำออกจากกลุ่ม',
@@ -1038,6 +1287,13 @@ const thTH = {
     imagesAttached: 'รูปภาพที่แนบ',
     noLogs: 'ยังไม่มีบันทึก',
     sessionMonitor: {
+      resetContext: 'รีเซ็ตบริบท',
+      resetDescription:
+        'ข้อความใหม่จะเริ่มด้วยบริบทว่าง ประวัติการตรวจสอบ ไฟล์ และความจำระยะยาวจะยังคงอยู่ โปรดรอให้งานที่กำลังทำอยู่เสร็จก่อน',
+      resetSuccess: 'รีเซ็ตบริบทแล้ว',
+      resetError:
+        'รีเซ็ตไม่สำเร็จ โปรดรอให้งานที่กำลังทำอยู่เสร็จแล้วลองอีกครั้ง',
+
       title: 'เซสชัน',
       sessions: 'เซสชัน',
       noSessions: 'ไม่พบเซสชัน',
@@ -1085,7 +1341,26 @@ const thTH = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'เลือกทั้งหมด',
     monitoring: {
+      rangeDays: '{{days}} วันที่ผ่านมา',
+      customRange: 'กำหนดช่วงเวลาเอง',
+      rangeStart: 'เวลาเริ่มต้น',
+      rangeEnd: 'เวลาสิ้นสุด',
+      rangeHint: 'เวลาเริ่มต้องก่อนเวลาสิ้นสุด สูงสุด 365 วัน',
+      autoRefreshLabel: 'รีเฟรชอัตโนมัติ',
+      autoRefreshOff: 'รีเฟรชอัตโนมัติ: ปิด',
+      autoRefreshEvery: 'รีเฟรชทุก {{seconds}} วินาที',
+      runStatus: 'สถานะการทำงาน',
+      toolCalls: 'การเรียกเครื่องมือ',
+      delivery: 'การส่งข้อความ',
+      sendFailed: 'ส่งไม่สำเร็จ',
+      sendCount: 'เรียกส่งสำเร็จ {{count}} ครั้ง',
+      noSend: 'ไม่มีบันทึกการส่ง',
+      noSteps: 'ยังไม่มีบันทึกขั้นตอนการทำงาน',
+      recipient: 'ผู้รับ',
+      actionCompleted: 'เครื่องมือทำงานสำเร็จ',
+      simulatedAction: 'การจำลอง ไม่ได้ส่งข้อความจริง',
       description:
         'ดูเหตุการณ์เริ่มต้น ผลลัพธ์โมเดล และการเรียกเครื่องมือของแต่ละงาน',
       empty: 'ยังไม่มีการทำงาน เริ่มเหตุการณ์หรือทดสอบการดีบักเพื่อดูบันทึก',
@@ -1187,6 +1462,9 @@ const thTH = {
       'ทำงานตามขั้นตอนที่กำหนด: รับข้อความ เรียก AI และตอบกลับผู้ใช้ พร้อมตั้งค่าฐานความรู้และปลั๊กอินได้ รองรับเฉพาะเหตุการณ์ข้อความ เหมาะกับงานที่มีขั้นตอนชัดเจนและต้องการควบคุมกระบวนการประมวลผล',
     allEvents: 'รองรับทุกเหตุการณ์',
     messageEventsOnly: 'เฉพาะเหตุการณ์ข้อความ',
+    selectAgentRunner: 'เลือกรันเนอร์ Agent',
+    selectProcessorPlugin: 'เลือกปลั๊กอินประมวลผล',
+    createBasicInfoDescription: 'ตั้งชื่อ ไอคอน และเลือกปลั๊กอิน',
     chooseType: 'เลือกวิธีทำงาน',
     chooseTypeDescription:
       'เลือกประเภทตัวประมวลผลก่อน แล้วตั้งค่าเพิ่มเติมหลังสร้าง',
@@ -1593,6 +1871,11 @@ const thTH = {
     },
   },
   market: {
+    certifiedPlugin: 'ปลั๊กอินที่ผ่านการรับรอง',
+    certificationTooltip:
+      'เวอร์ชันนี้ผ่านการตรวจสอบการแยกข้อมูลและความเข้ากันได้กับสภาพแวดล้อมรันร่วมกัน และแพ็กเกจได้รับการลงนามแล้ว',
+    certificationLearnMore: 'เกี่ยวกับการรับรองปลั๊กอิน',
+
     runnerUsage: 'การใช้งานรันเนอร์',
     runnerUsageAll: 'ทั้งหมด',
     runnerUsageAgent: 'Agent / ไปป์ไลน์',
@@ -1985,6 +2268,14 @@ const thTH = {
       uploading: 'กำลังอัปโหลด...',
     },
     monitoring: {
+      rangeDays: '{{days}} วันที่ผ่านมา',
+      customRange: 'กำหนดช่วงเวลาเอง',
+      rangeStart: 'เวลาเริ่มต้น',
+      rangeEnd: 'เวลาสิ้นสุด',
+      rangeHint: 'เวลาเริ่มต้องก่อนเวลาสิ้นสุด สูงสุด 365 วัน',
+      autoRefreshLabel: 'รีเฟรชอัตโนมัติ',
+      autoRefreshOff: 'รีเฟรชอัตโนมัติ: ปิด',
+      autoRefreshEvery: 'รีเฟรชทุก {{seconds}} วินาที',
       title: 'บันทึกการทำงาน',
       workbench: 'กำหนดค่าและดีบัก',
       description:
@@ -2268,7 +2559,7 @@ const thTH = {
     totpStatusDisabled: 'ยังไม่เปิดใช้',
     totpCodesRemaining: 'เหลือรหัสกู้คืน {{count}} รหัส',
     totpManagerSectionDesc:
-      'เจ้าของและผู้ดูแลสามารถตรวจสอบและรีเซ็ตการยืนยันสองขั้นตอนของบัญชีใดก็ได้',
+      'เจ้าของและผู้ดูแลสามารถตรวจสอบและรีเซ็ตการยืนยันสองขั้นตอนของบัญชีในพื้นที่ทำงานนี้',
     revokeTotp: 'ผูกใหม่',
     totpAdminResetTitle: 'ผูกการยืนยันสองขั้นตอนใหม่ให้ {{user}}',
     totpAdminResetDesc:
@@ -2332,8 +2623,7 @@ const thTH = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2402,6 +2692,124 @@ const thTH = {
     },
   },
   monitoring: {
+    rangeDays: '{{days}} วันที่ผ่านมา',
+    customRange: 'กำหนดช่วงเวลาเอง',
+    rangeStart: 'เวลาเริ่มต้น',
+    rangeEnd: 'เวลาสิ้นสุด',
+    rangeHint: 'เวลาเริ่มต้องก่อนเวลาสิ้นสุด สูงสุด 365 วัน',
+    autoRefreshLabel: 'รีเฟรชอัตโนมัติ',
+    autoRefreshOff: 'รีเฟรชอัตโนมัติ: ปิด',
+    autoRefreshEvery: 'รีเฟรชทุก {{seconds}} วินาที',
+    unified: {
+      inputs: 'เหตุการณ์และข้อมูลเข้า',
+      outputs: 'ผลลัพธ์ที่สร้าง',
+      deliveries: 'บันทึกการส่ง',
+      conversation: 'บริบทการสนทนา',
+      related: 'การทำงานที่เกี่ยวข้อง',
+      events: 'เหตุการณ์การทำงาน',
+      llm_calls: 'การเรียกโมเดล',
+      tool_calls: 'การเรียกเครื่องมือ',
+      errors: 'ข้อผิดพลาด',
+      payload: 'รายละเอียด',
+      metadata: 'ข้อมูลทางเทคนิค',
+      loadMore: 'โหลดเพิ่มเติม',
+      noRecords: 'ไม่มีบันทึก',
+      contextHint: 'ข้อความอื่นอาจไม่ใช่ส่วนหนึ่งของการทำงานนี้',
+      unhandled: 'ไม่ได้เริ่มตัวประมวลผล',
+      roles: {
+        user: 'ผู้ใช้',
+        assistant: 'ผู้ช่วย',
+        system: 'ระบบ',
+        tool: 'เครื่องมือ',
+      },
+    },
+    execution: {
+      title: 'การทำงาน',
+      subtitle:
+        'เหตุการณ์ขาเข้าแต่ละรายการและการทำงานที่ถูกเรียกใช้ ทั้งในเอเจนต์และไปป์ไลน์',
+      tabs: { executions: 'การทำงาน' },
+      filters: {
+        source: 'ประเภท',
+        sourceAll: 'ทุกประเภท',
+        sourceAgent: 'เอเจนต์',
+        sourcePipeline: 'ไปป์ไลน์',
+        mode: 'โหมดการทำงาน',
+        modeAll: 'ใช้งานจริง + ดีบัก',
+        modeLive: 'เฉพาะใช้งานจริง',
+        modeDebug: 'เฉพาะดีบัก',
+        status: 'สถานะ',
+        statusAll: 'ทุกสถานะ',
+      },
+      status: {
+        completed: 'เสร็จสิ้น',
+        failed: 'ล้มเหลว',
+        running: 'กำลังทำงาน',
+        queued: 'อยู่ในคิว',
+        cancelled: 'ยกเลิกแล้ว',
+        ignored: 'ละเว้นแล้ว',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'เอเจนต์',
+        pipeline: 'ไปป์ไลน์',
+        processor: 'ตัวประมวลผล',
+      },
+      columns: {
+        time: 'เวลา',
+        target: 'ตัวประมวลผล',
+        status: 'สถานะ',
+        duration: 'ระยะเวลา',
+        tokens: 'โทเคน',
+      },
+      debug: 'ดีบัก',
+      empty: 'ไม่มีการทำงานในช่วงเวลานี้',
+      loading: 'กำลังโหลดการทำงาน...',
+      triggeredBy: 'เรียกใช้โดย {{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95: ประมาณ 50%/95% ทำงานเสร็จภายในเวลานี้ ใช้เฉพาะ Agent ที่เสร็จล่าสุด ไม่รวมเวลารอคิว',
+        tokensHint:
+          'รวมเฉพาะการใช้ที่บันทึกไว้ ผลรวมอาจไม่ครบหากโมเดลไม่ส่งข้อมูลการใช้กลับมา',
+        totalExecutions: 'การทำงาน',
+        bySource: 'เอเจนต์ {{agent}} · ไปป์ไลน์ {{pipeline}}',
+        successRate: 'อัตราการทำงานสำเร็จ',
+        successRateHint:
+          'สำเร็จ ÷ (สำเร็จ + ล้มเหลว) ไม่รวมงานที่กำลังทำงาน ถูกละเว้น หรือยกเลิก',
+        denominator:
+          'สำเร็จ {{completed}} จาก {{denominator}} รายการที่สิ้นสุดแล้ว',
+        running: 'กำลังทำงาน',
+        queued: 'อยู่ในคิว',
+        waiting: 'รอการป้อนข้อมูล',
+        failed: 'ล้มเหลว',
+        ignored: 'ละเว้นแล้ว',
+        latency: 'เวลาทำงาน',
+        p50: 'P50',
+        p95: 'P95',
+        sample: 'ตัวอย่างการทำงาน {{count}} ครั้ง',
+        tokens: 'โทเคน',
+        coverage: 'มีข้อมูลการใช้งาน {{withUsage}}/{{calls}} ครั้ง',
+        debugRuns: 'รวมการทำงานดีบัก {{count}} ครั้ง',
+      },
+      detail: {
+        title: 'การทำงาน',
+        timeline: 'ลำดับเวลา',
+        executionId: 'ID การทำงาน',
+        eventId: 'เหตุการณ์',
+        queue: 'คิว',
+        runner: 'รันเนอร์',
+        created: 'สร้างเมื่อ',
+        duration: 'ระยะเวลา',
+        usage: 'การใช้งาน',
+        noEvents: 'ไม่มีการบันทึกเหตุการณ์ติดตาม',
+        query: 'ข้อความขาเข้า',
+        modelCalls: 'การเรียกใช้โมเดล',
+        toolCalls: 'การเรียกใช้เครื่องมือ',
+        errors: 'ข้อผิดพลาด',
+        loadError: 'โหลดการติดตามการทำงานไม่สำเร็จ',
+        openObject: 'เปิดรายการ',
+        a11yDescription: 'การติดตามทั้งหมดของการทำงานนี้',
+      },
+    },
     title: 'แดชบอร์ด',
     description: 'ตรวจสอบกิจกรรม Bot, การเรียก LLM และประสิทธิภาพของระบบ',
     overview: 'ภาพรวม',
@@ -2417,6 +2825,9 @@ const thTH = {
       pipeline: 'Pipeline',
       allBots: 'Bot ทั้งหมด',
       selectBot: 'เลือก Bot',
+      processor: 'ตัวประมวลผล',
+      allProcessors: 'ตัวประมวลผลทั้งหมด',
+      selectProcessor: 'เลือกตัวประมวลผล',
       allPipelines: 'Pipeline ทั้งหมด',
       selectPipeline: 'เลือก Pipeline',
       loading: 'กำลังโหลด...',
@@ -2442,6 +2853,15 @@ const thTH = {
       sessions: 'การวิเคราะห์เซสชัน',
       errors: 'บันทึกข้อผิดพลาด',
     },
+    pipelineTrace: {
+      received: 'ข้อความที่ได้รับ',
+      processing: 'ขั้นตอนการประมวลผล',
+      replies: 'ข้อความตอบกลับ',
+      noSteps: 'ไม่มีบันทึกการเรียกโมเดลหรือเครื่องมือ',
+      success: 'สำเร็จ',
+      error: 'ล้มเหลว',
+      pending: 'รอดำเนินการ',
+    },
     messageList: {
       timestamp: 'เวลา',
       bot: 'Bot',
@@ -2665,7 +3085,7 @@ const thTH = {
     title: 'การตั้งค่า',
     nav: {
       workspace: 'Workspace',
-      operationTrace: 'การติดตาม',
+      operationTrace: 'บันทึกการดำเนินการ',
       models: 'โมเดล',
       api: 'API',
       storage: 'พื้นที่จัดเก็บ',
@@ -2673,21 +3093,21 @@ const thTH = {
     },
   },
   operationTrace: {
-    title: 'การติดตามการดำเนินการ',
+    title: 'บันทึกการดำเนินการ',
     description: 'ติดตามการแก้ไขและการดูของผู้ดูแลและเจ้าของ',
+    reset: 'รีเซ็ต',
+    settingsUpdated: 'บันทึกการตั้งค่าแล้ว',
+    settingsUpdateFailed: 'บันทึกการตั้งค่าไม่สำเร็จ',
+    levelBadge: 'ระดับ: {{level}}',
+    collectionSettings: 'การตั้งค่าการเก็บข้อมูล',
     captureLevel: 'ระดับการบันทึก',
     retention: 'นโยบายการเก็บรักษา',
-    retentionSummary: 'เก็บ {{days}} วัน · สูงสุด {{rows}} รายการ',
     retentionDays: 'จำนวนวันเก็บรักษา',
     maxRows: 'จำนวนบันทึกสูงสุด',
     dedupeWindow: 'ช่วงเวลาลบรายการซ้ำ (วินาที)',
     records: 'บันทึกการดำเนินการ',
     changesCount: '{{count}} การเปลี่ยนแปลง',
     hideDetails: 'ซ่อนรายละเอียด',
-    levelUpdated: 'อัปเดตระดับการบันทึกแล้ว',
-    levelUpdateFailed: 'อัปเดตระดับการบันทึกไม่สำเร็จ',
-    retentionUpdated: 'อัปเดตนโยบายการเก็บรักษาแล้ว',
-    retentionUpdateFailed: 'อัปเดตนโยบายการเก็บรักษาไม่สำเร็จ',
     export: 'ดาวน์โหลดบันทึก',
     exportFailed: 'ดาวน์โหลดบันทึกการดำเนินการไม่สำเร็จ',
     empty: 'ยังไม่มีบันทึกการดำเนินการ',
@@ -2701,7 +3121,7 @@ const thTH = {
     filterAllResources: 'ทุกทรัพยากร',
     filterActor: 'ผู้ดำเนินการ',
     filterAllActors: 'ผู้ดำเนินการทั้งหมด',
-    loadFailed: 'โหลดข้อมูลการติดตามไม่สำเร็จ',
+    loadFailed: 'โหลดบันทึกการดำเนินการไม่สำเร็จ',
     refresh: 'รีเฟรช',
     save: 'บันทึก',
     redacted: 'ปิดบังแล้ว',
@@ -2720,7 +3140,7 @@ const thTH = {
     elidedValue: '{{count}} ตัวอักษร',
     levels: {
       off: {
-        label: 'ปิดการติดตาม',
+        label: 'ปิดบันทึก',
       },
       mutation: {
         label: 'เฉพาะการแก้ไข',
diff --git a/web/src/i18n/locales/vi-VN.ts b/web/src/i18n/locales/vi-VN.ts
index 3360aada7..636ea68e6 100644
--- a/web/src/i18n/locales/vi-VN.ts
+++ b/web/src/i18n/locales/vi-VN.ts
@@ -1,6 +1,176 @@
 import pipelineMigration from './pipeline-migration/vi-VN';
 const viVN = {
+  planRevision: {
+    subscribedOn: 'Ngày đăng ký: {{date}}',
+    latestQuota: 'Hạn mức mới nhất của gói này',
+    showLatest: 'Xem hạn mức mới nhất của gói này',
+    showCurrent: 'Xem gói đăng ký hiện tại',
+    latestTab: 'Mới nhất',
+    currentTab: 'Hiện tại',
+    cancelToSubscribe:
+      'Hủy gói hiện tại để đăng ký hạn mức mới nhất vào kỳ tiếp theo',
+
+    current: 'Current subscription: {{plan}} · revision {{revision}}',
+    owned: 'Purchased revision {{revision}}',
+    latest: 'Latest revision {{revision}}',
+    changed:
+      'Your purchased benefits and the latest version differ. Highlighted rows show the changes.',
+    same: 'Benefits match latest revision {{revision}}. Your purchased revision remains unchanged.',
+    unavailable:
+      'The latest plan is unavailable. Your purchased benefits remain in effect.',
+    benefit: 'Benefit',
+    change: 'Changed',
+    enabled: 'Enabled',
+    disabled: 'Disabled',
+    monthlyCredits: 'Monthly bonus credits',
+    retention: 'Chat history retention (days)',
+    unknown:
+      'Current subscription details are unavailable. Refresh or manage your subscription in Space; latest catalog benefits are not your purchased benefits.',
+    handoff:
+      'Manage your subscription in Space. Cancel automatic renewal before purchasing a new version. Prepaid subscriptions end automatically unless renewed before expiry.',
+    cancelFirst:
+      'Cancel automatic renewal in subscription management first. Cancellation preserves paid benefits until the end of the period, without a refund. After confirmation, you can purchase the latest version, effective when payment succeeds.',
+    canceled:
+      'Renewal is canceled. Paid benefits remain until the end of the period, without a refund. A new purchase applies its benefits when payment succeeds.',
+    prepaid:
+      'This prepaid subscription ends automatically. You can renew your current plan before expiry; after expiry, subscriptions use the latest quotas for each plan.',
+    renew: 'Gia hạn gói đăng ký hiện tại',
+    buying: 'Purchasing revision {{revision}}.',
+    fields: {
+      bots_max: 'Bots',
+      members_max: 'Members',
+      pipelines_max: 'Processors',
+      plugins_max: 'Extensions',
+      knowledge_bases_max: 'Knowledge bases',
+      managed_sandbox_sessions_max: 'Sandbox sessions',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: 'Chưa thể đăng ký',
+    subscriptionUnavailableHint:
+      'Gói này hiện không nhận đăng ký mới. Vui lòng chọn gói khác. Đăng ký hiện có và gia hạn tự động không bị ảnh hưởng.',
+    platformNames: {
+      qqofficial: 'QQ Official API',
+      dingtalk: 'DingTalk',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: 'Lark',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: 'WeCom',
+      wecombot: 'WeComBot',
+      wecomcs: 'WeComCustomerService',
+      officialaccount: 'Official Account',
+      wechatpad: 'WeChatPad',
+      openclaw_weixin: 'OpenClaw WeChat',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: 'Quản lý đăng ký',
+    downgradeUnavailable: 'Không thể hạ cấp xuống gói miễn phí',
+    features: {
+      multiUserWorkspace: 'Không gian làm việc nhiều người dùng',
+      remoteE2BSandbox: 'Cấu hình sandbox E2B từ xa',
+      communitySupport: 'Hỗ trợ từ cộng đồng',
+      everythingInFree: 'Tất cả tính năng của gói miễn phí',
+      alwaysOnSandbox: 'Một sandbox toàn cục luôn hoạt động',
+      higherWorkspaceLimits: 'Hạn mức không gian làm việc cao hơn',
+    },
+    title: 'So sánh gói',
+    description: 'Chọn gói cho {{workspace}}',
+    billing: 'Chu kỳ thanh toán',
+    monthly: 'Hàng tháng',
+    annual: 'Hàng năm',
+    currency: 'Tiền tệ',
+    loading: 'Đang tải gói…',
+    error: 'Không thể tải các gói',
+    retry: 'Thử lại',
+    empty: 'Không có gói khả dụng',
+    current: 'Gói hiện tại',
+    recommended: 'Đề xuất',
+    free: 'Miễn phí',
+    annualNote: 'Tương đương {{price}} / tháng',
+    ownerOnly: 'Chỉ chủ sở hữu hoặc quản trị viên',
+    choose: 'Tiếp tục thanh toán',
+    handoff:
+      'Đăng nhập và thanh toán tại Space. Lựa chọn của bạn sẽ được giữ lại.',
+    quotas: 'Hạn mức không gian',
+    unlimited: 'Không giới hạn',
+    retention: 'Lịch sử tin nhắn · {{count}} ngày',
+    retentionLabel: 'Thời gian lưu tin nhắn',
+    retentionHelp: 'Tin nhắn được tự động xóa sau thời gian này.',
+    credits: '{{count}} tín dụng mỗi tháng',
+    creditsLabel: 'Tín dụng hàng tháng',
+    creditsHelp: 'Tín dụng được cấp mỗi tháng đăng ký, kể cả gói năm.',
+    platforms: 'Hỗ trợ nền tảng phổ biến',
+    excludedPlatforms: 'Không hỗ trợ: {{platforms}}',
+    allPlatforms: 'Tất cả nền tảng hiện có',
+    limits: {
+      bots: 'Bot',
+      pipelines: 'Bộ xử lý',
+      knowledge_bases: 'Kho kiến thức',
+      plugins: 'Tiện ích mở rộng',
+      members: 'Thành viên',
+      managed_sandbox_sessions: 'Phiên sandbox được quản lý',
+    },
+  },
+  workspaceLanguage: {
+    label: 'Ngôn ngữ không gian làm việc',
+    description:
+      'Dùng cho thông báo máy chủ và phản hồi mặc định của bot. Không thay đổi ngôn ngữ giao diện cá nhân.',
+    saved: 'Đã lưu ngôn ngữ',
+    saveFailed: 'Không thể lưu ngôn ngữ',
+  },
+  inflight: {
+    autoCollapse: 'Tự thu gọn',
+    empty: 'Không có sự kiện đang xử lý',
+    move: 'Kéo lên xuống hoặc dùng phím mũi tên',
+    botLogs: 'Nhật ký bot',
+    processorLogs: 'Nhật ký bộ xử lý',
+
+    title: 'Đang thực hiện',
+    expand: 'Mở rộng',
+    collapse: 'Thu gọn',
+    reconnecting: 'Đang kết nối lại; hiển thị trạng thái nhận được gần nhất.',
+    details: 'Xem chi tiết',
+    debug: 'Sự kiện gỡ lỗi',
+    limited:
+      'Chỉ hiển thị một phần. Xem bảng điều khiển để biết tất cả lượt chạy.',
+    queued: 'Đang chờ',
+    running: 'Đang xử lý',
+    waiting: 'Chờ nhập liệu',
+    tool: 'Đang gọi công cụ',
+    thinking: 'Đang suy nghĩ',
+    replying: 'Chuẩn bị trả lời',
+    generating: 'Đang tạo câu trả lời',
+    completed: 'Hoàn tất',
+    failed: 'Thất bại',
+    cancelled: 'Đã hủy',
+    ignored: 'Đã bỏ qua',
+  },
   assistant: {
+    movePosition: 'Di chuyển trợ lý theo chiều ngang',
+    resizeHeight: 'Điều chỉnh chiều cao trợ lý',
+    sessions: 'Cuộc trò chuyện',
+    currentChat: 'Cuộc trò chuyện hiện tại',
+    stop: 'Dừng tác vụ',
+    stopped:
+      'Đã dừng. Công cụ có thể đã thực hiện thay đổi. Kiểm tra tài nguyên trước khi thử lại.',
+    setupRequired:
+      'Đăng nhập bằng tài khoản LangBot hoặc cấu hình mô hình để bắt đầu sử dụng trợ lý không gian làm việc.',
+    loginAccount: 'Đăng nhập LangBot Account',
+    configureModels: 'Cấu hình mô hình',
+    providerCheckFailed: 'Không thể kiểm tra nhà cung cấp. Vui lòng thử lại.',
+    retrySetup: 'Thử lại',
+    preparing: 'Đang chuẩn bị trợ lý…',
+    thinking: 'Mô hình đang trả lời · Vòng {{round}}',
+    executing: 'Đang chạy công cụ…',
+    toolRunning: 'Đang chạy',
+    recommendationFailed:
+      'Không tải được mô hình đề xuất. Vui lòng chọn mô hình.',
+    expand: 'Mở rộng trợ lý không gian làm việc',
+    collapse: 'Thu gọn trợ lý không gian làm việc',
     modelHint: 'Đổi mô hình trợ lý cho tin nhắn tiếp theo',
     details: 'Xem dữ liệu thô',
     found: 'Tìm thấy {{count}} mục',
@@ -22,7 +192,7 @@ const viVN = {
       create_knowledge_base: 'Tạo cơ sở tri thức',
       get_pipeline: 'Xem Pipeline',
       get_knowledge_schema: 'Xem lược đồ knowledge engine',
-      list_operation_logs: 'Truy vấn truy vết thao tác',
+      list_operation_logs: 'Truy vấn nhật ký thao tác',
     },
     resources: {
       models: 'Tìm mô hình trò chuyện',
@@ -36,7 +206,7 @@ const viVN = {
     newChat: 'Cuộc trò chuyện mới',
     close: 'Đóng',
     welcome:
-      'Hãy cho tôi biết bạn muốn xây dựng gì. Tôi có thể xem không gian làm việc này và đề xuất cấu hình cơ sở tri thức cùng Pipeline để bạn xác nhận.',
+      'Hãy cho tôi biết bạn muốn xây dựng gì. Tôi có thể xem không gian làm việc này, đề xuất cấu hình cho bot, Agent, cơ sở tri thức, Pipeline và các tài nguyên khác, rồi áp dụng sau khi bạn xác nhận.',
     discover: 'Có những mô hình, cơ sở tri thức và Pipeline nào khả dụng?',
     build: 'Giúp tôi tạo ứng dụng hỏi đáp dựa trên cơ sở tri thức.',
     toolResult: 'Kết quả thao tác',
@@ -56,6 +226,14 @@ const viVN = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: '{{days}} ngày qua',
+        customRange: 'Khoảng thời gian tùy chỉnh',
+        rangeStart: 'Bắt đầu',
+        rangeEnd: 'Kết thúc',
+        rangeHint: 'Bắt đầu phải trước kết thúc. Tối đa 365 ngày.',
+        autoRefreshLabel: 'Tự động làm mới',
+        autoRefreshOff: 'Tự động làm mới: tắt',
+        autoRefreshEvery: 'Làm mới mỗi {{seconds}} giây',
         title: 'Xem hoạt động',
         description:
           'Xem hoạt động bot, mức sử dụng mô hình, lượng tin nhắn và hiệu suất hệ thống.',
@@ -136,6 +314,14 @@ const viVN = {
           'Gửi tin nhắn thử để kiểm tra điều kiện kích hoạt, câu trả lời AI và xử lý đầu ra. Thay đổi cấu hình được lưu trước khi thử.',
       },
       monitoring: {
+        rangeDays: '{{days}} ngày qua',
+        customRange: 'Khoảng thời gian tùy chỉnh',
+        rangeStart: 'Bắt đầu',
+        rangeEnd: 'Kết thúc',
+        rangeHint: 'Bắt đầu phải trước kết thúc. Tối đa 365 ngày.',
+        autoRefreshLabel: 'Tự động làm mới',
+        autoRefreshOff: 'Tự động làm mới: tắt',
+        autoRefreshEvery: 'Làm mới mỗi {{seconds}} giây',
         title: 'Xem lịch sử chạy',
         description:
           'Dùng thẻ cạnh tiêu đề để xem cuộc trò chuyện, các bước xử lý và lỗi khi không có phản hồi hoặc phản hồi không đúng dự kiến.',
@@ -302,6 +488,11 @@ const viVN = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      'Chưa lưu nguyên nhân lỗi. Hãy kiểm tra nhật ký bot hoặc gửi lại tin nhắn để biết thêm chi tiết.',
+    loadFailedTitle: 'Không thể tải nội dung',
+    loadFailedHint:
+      'Vui lòng thử lại sau. Nếu vẫn không tải được, hãy kiểm tra kết nối và trạng thái dịch vụ.',
     customValue: 'Tùy chỉnh',
     loadFailed: 'Không tải được. Vui lòng thử lại.',
     search: 'Tìm kiếm',
@@ -498,6 +689,9 @@ const viVN = {
     help: 'Trợ giúp',
   },
   models: {
+    starModel: 'Đánh dấu làm mô hình mặc định cho bộ xử lý mới',
+    unstarModel: 'Bỏ đánh dấu',
+    starModelFailed: 'Không thể cập nhật mô hình được đánh dấu',
     codex: {
       account: 'Gói đăng ký ChatGPT',
       description:
@@ -761,10 +955,13 @@ const viVN = {
     deleteConfirmation: 'Bạn có chắc chắn muốn xóa Bot này không?',
     platformAdapter: 'Nền tảng/Lựa chọn Adapter',
     selectAdapter: 'Chọn Adapter',
+    saveAndEnable: 'Lưu và bật',
+    searchAdapters: 'Tìm kiếm adapter',
+    noMatchingAdapters: 'Không tìm thấy adapter phù hợp',
     legacyAdapters: 'Adapter cũ',
     legacyAdapterBadge: 'Cũ',
     legacyAdaptersHint:
-      'Các adapter này đã có phiên bản hướng sự kiện mới hơn.\nChúng chỉ được giữ lại để tương thích với cấu hình hiện có và không được khuyến nghị cho Bot mới.',
+      'Adapter cũ chủ yếu gửi và nhận tin nhắn, được giữ để tương thích với cấu hình bot hiện có.\nAdapter mới kết nối cùng nền tảng qua sự kiện, gồm tin nhắn và các yêu cầu kết bạn hoặc thay đổi nhóm được hỗ trợ, rồi chuyển đến bộ xử lý.\nƯu tiên bản mới khi tạo bot. Bot hiện có vẫn dùng được bản cũ. Khi chuyển đổi, kiểm tra kết nối và tuyến sự kiện; tài khoản nền tảng không được nâng cấp.',
     adapterConfig: 'Cấu hình Adapter',
     viewAdapterDocs: 'Xem tài liệu',
     bindPipeline: 'Liên kết Pipeline',
@@ -777,6 +974,10 @@ const viVN = {
     earlier: 'Trước đó',
     dateFormat: '{{day}}/{{month}}',
     setBotEnableError: 'Thiết lập trạng thái bật Bot thất bại',
+    enabling: 'Đang bật…',
+    disabling: 'Đang tắt…',
+    enableConfirmed: 'Đã bật',
+    disableConfirmed: 'Đã tắt',
     log: 'Nhật ký',
     configuration: 'Cấu hình',
     logs: 'Nhật ký',
@@ -785,6 +986,28 @@ const viVN = {
     routingConnection: 'Định tuyến & Kết nối',
     routingConnectionDescription:
       'Liên kết Pipeline xử lý tin nhắn cho Bot này',
+    routingHelp: {
+      title: 'Định tuyến sự kiện là gì?',
+      intro:
+        'Bot nhận tin nhắn, lời mời kết bạn và các sự kiện như tham gia nhóm.',
+      events: 'Sự kiện đến',
+      agents: 'Các Agent khác nhau',
+      message: 'Nhận tin nhắn',
+      friend: 'Lời mời kết bạn',
+      group: 'Tham gia nhóm',
+      messageAction: 'Agent trả lời',
+      friendAction: 'Agent xét duyệt',
+      groupAction: 'Agent chào mừng',
+      rules: 'Theo quy tắc của bạn · Ví dụ',
+      instructions:
+        'Agent làm theo chỉ dẫn của bạn và gọi công cụ để trả lời, xét duyệt yêu cầu và thực hiện thao tác khác.',
+      plugin:
+        'Với nhu cầu nâng cao, bộ xử lý plugin đăng ký hàng loạt sự kiện và xử lý bằng mã. Thêm cấu hình bộ xử lý plugin bên dưới.',
+      docs: 'Phát triển bộ xử lý plugin của riêng bạn',
+      runnerDocs: 'Phát triển trình chạy Agent của riêng bạn',
+      pause: 'Tạm dừng',
+      play: 'Phát',
+    },
     eventRouting: 'Định tuyến sự kiện',
     eventRoutingDescription:
       'Chọn bộ xử lý cho từng sự kiện của Bot. Chỉnh sửa logic trong cấu hình Agent hoặc Pipeline tương ứng. Pipeline chỉ hỗ trợ sự kiện tin nhắn.',
@@ -824,7 +1047,7 @@ const viVN = {
     enabled: 'Đã bật',
     eventBindingDescriptionPlaceholder: 'Mô tả quy tắc',
     noEventBindings: 'Chưa có tuyến sự kiện',
-    unsupportedPipelineEvent: 'Pipeline chỉ hỗ trợ sự kiện message.*',
+    unsupportedPipelineEvent: 'Pipeline chỉ hỗ trợ sự kiện message.received',
     disable: 'Tắt',
     enable: 'Bật',
     disabledBindings: 'Đã tắt',
@@ -834,6 +1057,7 @@ const viVN = {
     advancedEventValues: 'Xem tất cả',
     eventGroup: 'Nhóm',
     eventGroupNames: {
+      wecomcs: 'Dịch vụ khách hàng WeCom',
       bot: 'Trạng thái bot',
       feedback: 'Phản hồi',
       friend: 'Bạn bè',
@@ -848,6 +1072,9 @@ const viVN = {
     routeFallbackCatchAll: '{{route}} là tuyến nhận mọi sự kiện.',
     routeFallbackIgnored:
       'Sự kiện không khớp tuyến nào sẽ bị bỏ qua. Chỉ thêm tuyến nhận tất cả khi mọi sự kiện cần kết quả rõ ràng.',
+    matchRecords: 'Lịch sử khớp',
+    matchRecordsDescription:
+      'Kết quả gần nhất của mỗi quy tắc và các sự kiện không khớp gần đây. Nhật ký trong bộ nhớ được xóa khi khởi động lại.',
     testRoute: 'Kiểm tra tuyến',
     adapterEventDebugAction: 'Kiểm tra lắng nghe',
     adapterEventDebugTitle: 'Gỡ lỗi sự kiện nền tảng',
@@ -931,6 +1158,15 @@ const viVN = {
     eventWildcard: 'Tất cả sự kiện',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: 'Người dùng vào cuộc trò chuyện với bot',
+      wecombot_template_card_event: 'Tương tác thẻ mẫu',
+      dingtalk_card_action: 'Tương tác thẻ DingTalk',
+      wecomcs_enter_session: 'Khách hàng vào hội thoại',
+      wecomcs_msg_send_fail: 'Gửi tin nhắn hỗ trợ thất bại',
+      wecomcs_servicer_status_change: 'Trạng thái nhân viên thay đổi',
+      wecomcs_session_status_change: 'Trạng thái hội thoại hỗ trợ thay đổi',
+      wecomcs_reject_customer_msg_switch_change:
+        'Cài đặt từ chối tin nhắn thay đổi',
       message_received: 'Nhận tin nhắn',
       message_edited: 'Tin nhắn được chỉnh sửa',
       message_deleted: 'Tin nhắn bị xóa',
@@ -940,6 +1176,7 @@ const viVN = {
       friend_added: 'Đã thêm bạn',
       group_member_joined: 'Thành viên tham gia nhóm',
       group_member_left: 'Thành viên rời nhóm',
+      group_info_updated: 'Thông tin nhóm được cập nhật',
       group_member_banned: 'Thành viên bị cấm tham gia nhóm',
       bot_invited_to_group: 'Bot được mời vào nhóm',
       bot_removed_from_group: 'Bot bị xóa khỏi nhóm',
@@ -948,6 +1185,22 @@ const viVN = {
       platform_specific: 'Sự kiện riêng của nền tảng',
     },
     eventDescriptions: {
+      wecombot_enter_chat:
+        'Người dùng mở cuộc trò chuyện với bot. Dùng cho quy trình chào mừng.',
+      wecombot_template_card_event:
+        'Người dùng nhấp hoặc gửi thẻ. Phản hồi biểu mẫu do hệ thống quản lý được xử lý riêng.',
+      dingtalk_card_action:
+        'Nhấp và gửi thẻ; phản hồi và biểu mẫu được quản lý được xử lý riêng.',
+      wecomcs_enter_session:
+        'Khách hàng mở liên kết hỗ trợ; gồm nguồn truy cập và mã chào mừng.',
+      wecomcs_msg_send_fail:
+        'WeChat thông báo gửi thất bại, gồm ID tin nhắn và loại lỗi.',
+      wecomcs_servicer_status_change:
+        'Nhân viên bắt đầu, dừng hoặc tạm dừng tiếp nhận.',
+      wecomcs_session_status_change:
+        'Nhân viên tiếp nhận, chuyển, kết thúc hoặc mở lại hội thoại.',
+      wecomcs_reject_customer_msg_switch_change:
+        'Nhân viên bật hoặc tắt từ chối tin nhắn khách hàng.',
       all: 'Khớp mọi sự kiện bộ chuyển đổi này nhận được.',
       namespace: 'Khớp mọi sự kiện {{group}}.',
       namespace_bot:
@@ -955,7 +1208,7 @@ const viVN = {
       namespace_feedback: 'Khớp sự kiện phản hồi từ nền tảng hoặc người dùng.',
       namespace_friend: 'Khớp yêu cầu kết bạn và thay đổi quan hệ bạn bè.',
       namespace_group:
-        'Khớp sự kiện tham gia, rời, bị xóa khỏi nhóm và sự kiện nhóm khác.',
+        'Khớp các thay đổi thành viên và cập nhật thông tin nhóm.',
       namespace_message: 'Khớp sự kiện nhận, sửa, xóa và thả cảm xúc tin nhắn.',
       namespace_platform:
         'Khớp sự kiện riêng của nền tảng do bộ chuyển đổi cung cấp.',
@@ -969,6 +1222,7 @@ const viVN = {
       friend_added: 'Đã tạo quan hệ bạn bè.',
       group_member_joined: 'Thành viên tham gia nhóm có bot.',
       group_member_left: 'Thành viên rời nhóm có bot.',
+      group_info_updated: 'Thông tin nhóm như tên hoặc chủ sở hữu đã thay đổi.',
       group_member_banned: 'Thành viên nhóm bị cấm hoặc xóa.',
       bot_invited_to_group: 'Bot được mời vào nhóm.',
       bot_removed_from_group: 'Bot bị xóa khỏi nhóm.',
@@ -1055,6 +1309,13 @@ const viVN = {
     imagesAttached: 'hình ảnh đính kèm',
     noLogs: 'Chưa có nhật ký',
     sessionMonitor: {
+      resetContext: 'Đặt lại ngữ cảnh',
+      resetDescription:
+        'Tin nhắn mới sẽ bắt đầu với ngữ cảnh trống. Lịch sử giám sát, tệp và bộ nhớ dài hạn được giữ lại. Hãy đợi các tác vụ đang chạy hoàn tất.',
+      resetSuccess: 'Đã đặt lại ngữ cảnh.',
+      resetError:
+        'Không thể đặt lại. Hãy đợi các tác vụ đang chạy hoàn tất rồi thử lại.',
+
       title: 'Phiên',
       sessions: 'Phiên',
       noSessions: 'Không tìm thấy phiên nào',
@@ -1102,7 +1363,26 @@ const viVN = {
     },
   },
   agents: {
+    apiToolsSelectAll: 'Chọn tất cả',
     monitoring: {
+      rangeDays: '{{days}} ngày qua',
+      customRange: 'Khoảng thời gian tùy chỉnh',
+      rangeStart: 'Bắt đầu',
+      rangeEnd: 'Kết thúc',
+      rangeHint: 'Bắt đầu phải trước kết thúc. Tối đa 365 ngày.',
+      autoRefreshLabel: 'Tự động làm mới',
+      autoRefreshOff: 'Tự động làm mới: tắt',
+      autoRefreshEvery: 'Làm mới mỗi {{seconds}} giây',
+      runStatus: 'Trạng thái thực thi',
+      toolCalls: 'Lượt gọi công cụ',
+      delivery: 'Gửi tin nhắn',
+      sendFailed: 'Gửi thất bại',
+      sendCount: '{{count}} lượt gọi gửi thành công',
+      noSend: 'Chưa ghi nhận lượt gửi',
+      noSteps: 'Chưa ghi nhận bước thực thi.',
+      recipient: 'Người nhận',
+      actionCompleted: 'Công cụ đã thực thi thành công.',
+      simulatedAction: 'Thao tác mô phỏng; không gửi tin nhắn thật.',
       description:
         'Xem sự kiện kích hoạt, đầu ra mô hình và quá trình gọi công cụ của mỗi tác vụ.',
       empty: 'Chưa có lượt chạy. Kích hoạt sự kiện hoặc chạy thử gỡ lỗi.',
@@ -1205,6 +1485,9 @@ const viVN = {
       'Chạy theo quy trình cố định: nhận tin nhắn, gọi AI và trả lời người dùng, với cơ sở tri thức và plugin có thể cấu hình. Chỉ xử lý sự kiện tin nhắn, phù hợp với tác vụ có các bước rõ ràng và cần kiểm soát quá trình xử lý.',
     allEvents: 'Hỗ trợ tất cả sự kiện',
     messageEventsOnly: 'Chỉ sự kiện tin nhắn',
+    selectAgentRunner: 'Chọn trình chạy Agent',
+    selectProcessorPlugin: 'Chọn plugin xử lý',
+    createBasicInfoDescription: 'Đặt tên, biểu tượng và chọn plugin',
     chooseType: 'Chọn cách hoạt động',
     chooseTypeDescription:
       'Chọn loại bộ xử lý trước. Bạn có thể cấu hình thêm sau khi tạo.',
@@ -1619,6 +1902,11 @@ const viVN = {
     },
   },
   market: {
+    certifiedPlugin: 'Plugin được chứng nhận',
+    certificationTooltip:
+      'Phiên bản này đã vượt qua đánh giá về cách ly dữ liệu và khả năng tương thích với môi trường chạy dùng chung. Gói đã được ký.',
+    certificationLearnMore: 'Tìm hiểu về chứng nhận',
+
     runnerUsage: 'Mục đích trình chạy',
     runnerUsageAll: 'Tất cả',
     runnerUsageAgent: 'Agent / Pipeline',
@@ -2014,6 +2302,14 @@ const viVN = {
       uploading: 'Đang tải lên...',
     },
     monitoring: {
+      rangeDays: '{{days}} ngày qua',
+      customRange: 'Khoảng thời gian tùy chỉnh',
+      rangeStart: 'Bắt đầu',
+      rangeEnd: 'Kết thúc',
+      rangeHint: 'Bắt đầu phải trước kết thúc. Tối đa 365 ngày.',
+      autoRefreshLabel: 'Tự động làm mới',
+      autoRefreshOff: 'Tự động làm mới: tắt',
+      autoRefreshEvery: 'Làm mới mỗi {{seconds}} giây',
       title: 'Nhật ký chạy',
       workbench: 'Cấu hình và gỡ lỗi',
       description: 'Xem nhật ký thực thi và lỗi của Pipeline này (24 giờ qua)',
@@ -2302,7 +2598,7 @@ const viVN = {
     totpStatusDisabled: 'Chưa bật',
     totpCodesRemaining: 'Còn {{count}} mã khôi phục',
     totpManagerSectionDesc:
-      'Chủ sở hữu và quản trị viên có thể xem và đặt lại yếu tố thứ hai của bất kỳ tài khoản nào.',
+      'Chủ sở hữu và quản trị viên có thể xem và đặt lại yếu tố thứ hai của các tài khoản trong không gian làm việc này.',
     revokeTotp: 'Liên kết lại',
     totpAdminResetTitle: 'Liên kết lại xác thực hai bước cho {{user}}',
     totpAdminResetDesc:
@@ -2367,8 +2663,7 @@ const viVN = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2437,6 +2732,124 @@ const viVN = {
     },
   },
   monitoring: {
+    rangeDays: '{{days}} ngày qua',
+    customRange: 'Khoảng thời gian tùy chỉnh',
+    rangeStart: 'Bắt đầu',
+    rangeEnd: 'Kết thúc',
+    rangeHint: 'Bắt đầu phải trước kết thúc. Tối đa 365 ngày.',
+    autoRefreshLabel: 'Tự động làm mới',
+    autoRefreshOff: 'Tự động làm mới: tắt',
+    autoRefreshEvery: 'Làm mới mỗi {{seconds}} giây',
+    unified: {
+      inputs: 'Sự kiện và đầu vào',
+      outputs: 'Kết quả tạo ra',
+      deliveries: 'Bản ghi gửi',
+      conversation: 'Ngữ cảnh hội thoại',
+      related: 'Lượt chạy liên quan',
+      events: 'Sự kiện thực thi',
+      llm_calls: 'Lần gọi mô hình',
+      tool_calls: 'Lần gọi công cụ',
+      errors: 'Lỗi',
+      payload: 'Chi tiết',
+      metadata: 'Thông tin kỹ thuật',
+      loadMore: 'Tải thêm',
+      noRecords: 'Chưa có bản ghi',
+      contextHint: 'Các tin nhắn khác không nhất thiết thuộc lượt chạy này.',
+      unhandled: 'Chưa chạy bộ xử lý',
+      roles: {
+        user: 'Người dùng',
+        assistant: 'Trợ lý',
+        system: 'Hệ thống',
+        tool: 'Công cụ',
+      },
+    },
+    execution: {
+      title: 'Lượt thực thi',
+      subtitle:
+        'Mọi sự kiện đầu vào và lượt thực thi được kích hoạt trên các tác nhân và pipeline.',
+      tabs: { executions: 'Lượt thực thi' },
+      filters: {
+        source: 'Loại',
+        sourceAll: 'Tất cả loại',
+        sourceAgent: 'Tác nhân',
+        sourcePipeline: 'Pipeline',
+        mode: 'Chế độ thực thi',
+        modeAll: 'Thực tế + gỡ lỗi',
+        modeLive: 'Chỉ thực tế',
+        modeDebug: 'Chỉ gỡ lỗi',
+        status: 'Trạng thái',
+        statusAll: 'Tất cả trạng thái',
+      },
+      status: {
+        completed: 'Hoàn thành',
+        failed: 'Thất bại',
+        running: 'Đang chạy',
+        queued: 'Trong hàng đợi',
+        cancelled: 'Đã hủy',
+        ignored: 'Đã bỏ qua',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Tác nhân',
+        pipeline: 'Pipeline',
+        processor: 'Bộ xử lý',
+      },
+      columns: {
+        time: 'Thời gian',
+        target: 'Bộ xử lý',
+        status: 'Trạng thái',
+        duration: 'Thời lượng',
+        tokens: 'Token',
+      },
+      debug: 'Gỡ lỗi',
+      empty: 'Không có lượt thực thi trong khoảng này',
+      loading: 'Đang tải lượt thực thi...',
+      triggeredBy: 'Được kích hoạt bởi {{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95: khoảng 50%/95% lượt chạy hoàn tất trong thời gian này. Chỉ tính Agent đã hoàn tất gần đây, không gồm thời gian xếp hàng.',
+        tokensHint:
+          'Chỉ cộng mức dùng đã ghi. Tổng có thể thiếu nếu mô hình không trả về dữ liệu sử dụng.',
+        totalExecutions: 'Lượt thực thi',
+        bySource: 'Tác nhân {{agent}} · Pipeline {{pipeline}}',
+        successRate: 'Tỷ lệ thực thi thành công',
+        successRateHint:
+          'Thành công ÷ (thành công + thất bại). Không tính lượt đang chạy, bỏ qua hoặc đã hủy.',
+        denominator:
+          'Thành công {{completed}} trong {{denominator}} lượt đã kết thúc',
+        running: 'Đang chạy',
+        queued: 'Trong hàng đợi',
+        waiting: 'Đang chờ nhập liệu',
+        failed: 'Thất bại',
+        ignored: 'Đã bỏ qua',
+        latency: 'Thời gian thực thi',
+        p50: 'P50',
+        p95: 'P95',
+        sample: 'Mẫu gồm {{count}} lượt thực thi',
+        tokens: 'Token',
+        coverage: 'Có dữ liệu sử dụng cho {{withUsage}}/{{calls}} lệnh gọi',
+        debugRuns: 'Bao gồm {{count}} lượt gỡ lỗi',
+      },
+      detail: {
+        title: 'Lượt thực thi',
+        timeline: 'Dòng thời gian',
+        executionId: 'ID thực thi',
+        eventId: 'Sự kiện',
+        queue: 'Hàng đợi',
+        runner: 'Trình thực thi',
+        created: 'Được tạo',
+        duration: 'Thời lượng',
+        usage: 'Mức sử dụng',
+        noEvents: 'Chưa ghi nhận sự kiện truy vết',
+        query: 'Tin nhắn đầu vào',
+        modelCalls: 'Lệnh gọi mô hình',
+        toolCalls: 'Lệnh gọi công cụ',
+        errors: 'Lỗi',
+        loadError: 'Không thể tải truy vết thực thi',
+        openObject: 'Mở đối tượng',
+        a11yDescription: 'Toàn bộ truy vết của lượt thực thi này',
+      },
+    },
     title: 'Bảng điều khiển',
     description: 'Giám sát hoạt động Bot, cuộc gọi LLM và hiệu suất hệ thống',
     overview: 'Tổng quan',
@@ -2452,6 +2865,9 @@ const viVN = {
       pipeline: 'Pipeline',
       allBots: 'Tất cả Bot',
       selectBot: 'Chọn Bot',
+      processor: 'Bộ xử lý',
+      allProcessors: 'Tất cả bộ xử lý',
+      selectProcessor: 'Chọn bộ xử lý',
       allPipelines: 'Tất cả Pipeline',
       selectPipeline: 'Chọn Pipeline',
       loading: 'Đang tải...',
@@ -2477,6 +2893,15 @@ const viVN = {
       sessions: 'Phân tích phiên',
       errors: 'Nhật ký lỗi',
     },
+    pipelineTrace: {
+      received: 'Tin nhắn nhận được',
+      processing: 'Quá trình xử lý',
+      replies: 'Phản hồi',
+      noSteps: 'Không có bản ghi gọi mô hình hoặc công cụ',
+      success: 'Thành công',
+      error: 'Thất bại',
+      pending: 'Đang chờ',
+    },
     messageList: {
       timestamp: 'Thời gian',
       bot: 'Bot',
@@ -2700,7 +3125,7 @@ const viVN = {
     title: 'Cài đặt',
     nav: {
       workspace: 'Workspace',
-      operationTrace: 'Truy vết',
+      operationTrace: 'Nhật ký',
       models: 'Mô hình',
       api: 'API',
       storage: 'Lưu trữ',
@@ -2708,21 +3133,21 @@ const viVN = {
     },
   },
   operationTrace: {
-    title: 'Truy vết thao tác',
+    title: 'Nhật ký thao tác',
     description: 'Theo dõi thao tác sửa và xem của quản trị viên',
+    reset: 'Đặt lại',
+    settingsUpdated: 'Đã lưu cài đặt',
+    settingsUpdateFailed: 'Lưu cài đặt thất bại',
+    levelBadge: 'Cấp độ: {{level}}',
+    collectionSettings: 'Cài đặt thu thập',
     captureLevel: 'Mức ghi nhận',
     retention: 'Chính sách lưu trữ',
-    retentionSummary: 'Lưu {{days}} ngày · tối đa {{rows}} bản ghi',
     retentionDays: 'Số ngày lưu',
     maxRows: 'Số bản ghi tối đa',
     dedupeWindow: 'Khoảng khử trùng lặp (giây)',
     records: 'Bản ghi thao tác',
     changesCount: '{{count}} thay đổi',
     hideDetails: 'Ẩn chi tiết',
-    levelUpdated: 'Đã cập nhật mức ghi nhận',
-    levelUpdateFailed: 'Cập nhật mức ghi nhận thất bại',
-    retentionUpdated: 'Đã cập nhật chính sách lưu trữ',
-    retentionUpdateFailed: 'Cập nhật chính sách lưu trữ thất bại',
     export: 'Tải nhật ký',
     exportFailed: 'Tải nhật ký thao tác thất bại',
     empty: 'Chưa có bản ghi thao tác',
@@ -2736,7 +3161,7 @@ const viVN = {
     filterAllResources: 'Tất cả tài nguyên',
     filterActor: 'Người thực hiện',
     filterAllActors: 'Tất cả người thực hiện',
-    loadFailed: 'Tải dữ liệu truy vết thất bại',
+    loadFailed: 'Tải nhật ký thao tác thất bại',
     refresh: 'Làm mới',
     save: 'Lưu',
     redacted: 'Đã ẩn',
@@ -2755,7 +3180,7 @@ const viVN = {
     elidedValue: '{{count}} ký tự',
     levels: {
       off: {
-        label: 'Tắt truy vết',
+        label: 'Tắt nhật ký',
       },
       mutation: {
         label: 'Chỉ thao tác sửa',
diff --git a/web/src/i18n/locales/zh-Hans.ts b/web/src/i18n/locales/zh-Hans.ts
index cebb8d2e6..6b1e82046 100644
--- a/web/src/i18n/locales/zh-Hans.ts
+++ b/web/src/i18n/locales/zh-Hans.ts
@@ -1,6 +1,169 @@
+import slackSetup from './slack-setup/zh-Hans';
 import pipelineMigration from './pipeline-migration/zh-Hans';
 const zhHans = {
+  planRevision: {
+    subscribedOn: '订阅日期：{{date}}',
+    latestQuota: '该计划最新配额',
+    showLatest: '查看该计划最新配额',
+    showCurrent: '查看当前订阅',
+    latestTab: '最新配额',
+    currentTab: '当前订阅',
+    cancelToSubscribe: '取消本期订阅，下一期即可订阅最新配额',
+
+    current: '当前订阅：{{plan}} · revision {{revision}}',
+    owned: '已购买 revision {{revision}}',
+    latest: '最新 revision {{revision}}',
+    changed: '已购买权益与最新版本存在差异，高亮行标示变动。',
+    same: '权益与最新 revision {{revision}} 相同，已购买版本保持不变。',
+    unavailable: '最新计划暂不可用，已购买权益继续有效。',
+    benefit: '权益',
+    change: '有变化',
+    enabled: '启用',
+    disabled: '禁用',
+    monthlyCredits: '每月赠送额度',
+    retention: '聊天记录保留天数',
+    unknown:
+      '当前订阅详情不可用。请刷新或前往 Space 管理订阅；最新目录权益不代表已购买权益。',
+    handoff:
+      '请在 Space 管理订阅。自动续费需先取消后才能购买新版本；预付订阅到期自动结束，除非到期前手动续订。',
+    cancelFirst:
+      '请先在订阅管理中取消自动续费。取消后保留本期已付权益，不退款。服务端确认取消后可重新购买最新版本，付款成功后新权益生效。',
+    canceled:
+      '续费已取消。本期已付权益保留至到期，不退款；重新购买的权益在付款成功后生效。',
+    prepaid:
+      '本预付订阅到期自动结束。到期前可续订当前方案；到期后只能按各计划的最新配额订阅。',
+    renew: '续订当前订阅',
+    buying: '正在购买 revision {{revision}}。',
+    fields: {
+      bots_max: '机器人',
+      members_max: '成员',
+      pipelines_max: '处理器',
+      plugins_max: '扩展',
+      knowledge_bases_max: '知识库',
+      managed_sandbox_sessions_max: '沙箱会话',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: '暂不可订阅',
+    subscriptionUnavailableHint:
+      '此计划暂不接受新的订阅，请选择其他计划。已有订阅和已签约的自动续费不受影响。',
+    platformNames: {
+      qqofficial: 'QQ 官方 API',
+      dingtalk: '钉钉',
+      discord: 'Discord',
+      telegram: '电报',
+      lark: '飞书',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: '企业微信',
+      wecombot: '企业微信智能机器人',
+      wecomcs: '企业微信客服',
+      officialaccount: '微信公众号',
+      wechatpad: 'WeChatPad（个人微信iPad）',
+      openclaw_weixin: '个人微信机器人',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: '管理订阅',
+    downgradeUnavailable: '不支持降级到免费版',
+    features: {
+      multiUserWorkspace: '多人协作工作区',
+      remoteE2BSandbox: '远程 E2B 沙箱配置',
+      communitySupport: '社区支持',
+      everythingInFree: '包含免费版全部功能',
+      alwaysOnSandbox: '一个常驻全局沙箱',
+      higherWorkspaceLimits: '更高的工作区配额',
+    },
+    title: '对比订阅计划',
+    description: '为 {{workspace}} 选择订阅计划',
+    billing: '计费周期',
+    monthly: '月付',
+    annual: '年付',
+    currency: '币种',
+    loading: '正在加载计划…',
+    error: '无法加载订阅计划',
+    retry: '重试',
+    empty: '暂无可用计划',
+    current: '当前计划',
+    recommended: '推荐',
+    free: '免费',
+    annualNote: '相当于 {{price}} / 月',
+    ownerOnly: '仅所有者或管理员可操作',
+    choose: '前往付款',
+    handoff: '将在 Space 中登录并付款，所选计划、周期和币种会保留。',
+    quotas: '工作区配额',
+    unlimited: '不限',
+    retention: '消息记录保留 {{count}} 天',
+    retentionLabel: '消息记录保留时长',
+    retentionHelp: '超过保留时长的消息记录会自动清理。',
+    credits: '每月赠送 {{count}} 积分',
+    creditsLabel: '每月赠送积分',
+    creditsHelp: '每个订阅月自动发放，年付计划同样按月发放。',
+    platforms: '支持主流平台',
+    excludedPlatforms: '不支持：{{platforms}}',
+    allPlatforms: '支持当前所有可用平台',
+    limits: {
+      bots: '机器人',
+      pipelines: '处理器',
+      knowledge_bases: '知识库',
+      plugins: '扩展',
+      members: '成员',
+      managed_sandbox_sessions: '托管沙盒会话',
+    },
+  },
+  workspaceLanguage: {
+    label: '工作区语言',
+    description: '用于后端提示和机器人的默认回复，不影响个人界面语言。',
+    saved: '语言设置已保存',
+    saveFailed: '语言设置保存失败',
+  },
+  slackSetup,
+  inflight: {
+    autoCollapse: '自动折叠',
+    empty: '当前没有正在处理的事件',
+    move: '上下拖动；也可用方向键移动',
+    botLogs: '机器人日志',
+    processorLogs: '处理器日志',
+
+    title: '处理中',
+    expand: '展开',
+    collapse: '收起',
+    reconnecting: '连接中断，正在重连；显示的是最后收到的状态。',
+    details: '查看详情',
+    debug: '调试事件',
+    limited: '仅展示部分执行，请在仪表盘查看全部记录。',
+    queued: '排队中',
+    running: '处理中',
+    waiting: '等待用户输入',
+    tool: '调用工具',
+    thinking: '思考中',
+    replying: '准备回复',
+    generating: '生成回复',
+    completed: '已完成',
+    failed: '失败',
+    cancelled: '已取消',
+    ignored: '已忽略',
+  },
   assistant: {
+    movePosition: '左右移动助手',
+    resizeHeight: '调整助手高度',
+    sessions: '切换会话',
+    currentChat: '当前会话',
+    stop: '停止任务',
+    stopped: '已停止。执行中的工具可能已经生效，请检查相关资源后再尝试。',
+    setupRequired: '登录 LangBot 账户或配置模型，即可开始使用工作区助手。',
+    loginAccount: '登录 LangBot 账户',
+    configureModels: '配置模型',
+    providerCheckFailed: '无法检查模型供应商，请重试。',
+    retrySetup: '重试',
+    preparing: '正在准备助手…',
+    thinking: '模型响应中 · 第 {{round}} 轮',
+    executing: '正在执行工具…',
+    toolRunning: '执行中',
+    recommendationFailed: '推荐模型加载失败，请手动选择模型。',
+    expand: '展开工作区助手',
+    collapse: '收起工作区助手',
     modelHint: '切换助手模型，从下一条消息生效',
     details: '查看原始数据',
     found: '找到 {{count}} 项',
@@ -14,17 +177,17 @@ const zhHans = {
     sendUnconfirmed: '发送状态未确认，请检查连接后重新打开会话。',
     draftPlaceholder: '可以先写下一条，回复完成后发送…',
     operations: {
-      create_pipeline: '创建 Pipeline',
-      configure_pipeline: '配置 Pipeline',
+      create_pipeline: '创建 流水线',
+      configure_pipeline: '配置 流水线',
       create_knowledge_base: '创建知识库',
-      get_pipeline: '查看 Pipeline',
+      get_pipeline: '查看 流水线',
       get_knowledge_schema: '查看知识引擎配置',
-      list_operation_logs: '查询操作溯源',
+      list_operation_logs: '查询操作日志',
     },
     resources: {
       models: '查询聊天模型',
       embedding_models: '查询嵌入模型',
-      pipelines: '查询 Pipeline',
+      pipelines: '查询 流水线',
       knowledge_bases: '查询知识库',
       knowledge_engines: '查询知识引擎',
     },
@@ -33,8 +196,8 @@ const zhHans = {
     newChat: '新对话',
     close: '关闭',
     welcome:
-      '告诉我你想构建什么应用。我可以查看当前工作区，并提出知识库和 Pipeline 配置方案，由你确认后执行。',
-    discover: '当前有哪些可用的模型、知识库和 Pipeline？',
+      '告诉我你想构建什么应用。我可以查看当前工作区，为机器人、Agent、知识库和 流水线 等提供配置方案，由你确认后执行。',
+    discover: '当前有哪些可用的模型、知识库和 流水线？',
     build: '帮我创建一个知识库问答应用。',
     toolResult: '操作结果',
     review: '请检查具体操作及参数，确认后执行',
@@ -53,6 +216,14 @@ const zhHans = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: '最近 {{days}} 天',
+        customRange: '自定义时间范围',
+        rangeStart: '开始时间',
+        rangeEnd: '结束时间',
+        rangeHint: '开始时间需早于结束时间，最多 365 天。',
+        autoRefreshLabel: '定时刷新',
+        autoRefreshOff: '定时刷新：关闭',
+        autoRefreshEvery: '每 {{seconds}} 秒刷新',
         title: '查看运行概况',
         description: '集中查看机器人活动、模型调用、消息量和系统运行情况。',
       },
@@ -128,6 +299,14 @@ const zhHans = {
           '发送测试消息，检查触发规则、AI 回复和输出处理是否符合预期；修改配置后，测试前会先保存。',
       },
       monitoring: {
+        rangeDays: '最近 {{days}} 天',
+        customRange: '自定义时间范围',
+        rangeStart: '开始时间',
+        rangeEnd: '结束时间',
+        rangeHint: '开始时间需早于结束时间，最多 365 天。',
+        autoRefreshLabel: '定时刷新',
+        autoRefreshOff: '定时刷新：关闭',
+        autoRefreshEvery: '每 {{seconds}} 秒刷新',
         title: '查看运行日志',
         description:
           '打开标题旁的“运行日志”，查看历史对话、处理过程和错误，排查回复异常。',
@@ -285,6 +464,10 @@ const zhHans = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      '这条历史记录未保存失败原因，请查看机器人日志，或重新发送消息获取详细错误。',
+    loadFailedTitle: '暂时无法加载',
+    loadFailedHint: '请稍后重试。如果仍无法加载，请检查网络连接与服务状态。',
     customValue: '自定义',
     loadFailed: '加载失败，请重试。',
     search: '搜索',
@@ -471,6 +654,9 @@ const zhHans = {
     help: '查看帮助文档',
   },
   models: {
+    starModel: '标星模型（用作新处理器的默认模型）',
+    unstarModel: '取消标星',
+    starModelFailed: '无法更新标星模型',
     codex: {
       account: 'ChatGPT 订阅',
       description:
@@ -609,7 +795,7 @@ const zhHans = {
     searchProviders: '搜索供应商...',
     langbotModelsDescription: 'LangBot Space 提供的云端模型',
     credits: '积分',
-    loginWithSpace: '使用 LangBot 账号登录',
+    loginWithSpace: '使用 LangBot 账户登录',
     loginToUseModels: '通过 LangBot 账号登录以使用云端模型',
     ownerMustBindSpace:
       '工作区所有者需要绑定 LangBot 账号才能使用 LangBot 模型。',
@@ -716,10 +902,13 @@ const zhHans = {
     deleteConfirmation: '你确定要删除这个机器人吗？',
     platformAdapter: '平台/适配器选择',
     selectAdapter: '选择适配器',
+    saveAndEnable: '保存并启用',
+    searchAdapters: '搜索适配器',
+    noMatchingAdapters: '未找到匹配的适配器',
     legacyAdapters: '旧版适配器',
     legacyAdapterBadge: '过时',
     legacyAdaptersHint:
-      '这些适配器已有对应的事件驱动接入版本，\n仅为兼容存量配置保留，不建议新建机器人时使用。',
+      '旧版主要用于消息收发，保留用于兼容已有机器人配置。\n新版是同一平台的事件驱动接入，除消息外，还可处理该平台支持的好友请求、群成员变化等事件，并通过事件路由交给处理器。\n新建机器人优先选择新版。已有机器人可继续使用旧版；切换时请按新版接入要求重新检查连接参数和事件路由，并非直接升级平台账号。',
     adapterConfig: '适配器配置',
     viewAdapterDocs: '查看文档',
     bindPipeline: '绑定流水线',
@@ -732,6 +921,10 @@ const zhHans = {
     earlier: '更久之前',
     dateFormat: '{{month}}月{{day}}日',
     setBotEnableError: '设置机器人启用状态失败',
+    enabling: '正在启用…',
+    disabling: '正在禁用…',
+    enableConfirmed: '已启用',
+    disableConfirmed: '已禁用',
     log: '日志',
     configuration: '配置',
     logs: '日志',
@@ -739,6 +932,26 @@ const zhHans = {
     basicInfoDescription: '设置机器人名称和描述',
     routingConnection: '路由与连接',
     routingConnectionDescription: '绑定处理此机器人消息的流水线',
+    routingHelp: {
+      title: '什么是事件路由？',
+      intro: '机器人不只会收到消息，也会收到好友请求、加入群聊等事件。',
+      events: '机器人收到事件',
+      agents: '交给不同 Agent',
+      message: '收到消息',
+      friend: '好友请求',
+      group: '加入群聊',
+      messageAction: '问答 Agent → 回复',
+      friendAction: '审核 Agent → 处理申请',
+      groupAction: '欢迎 Agent → 打招呼',
+      rules: '按您设置的规则分配 · 示例',
+      instructions: 'Agent 按您的指令调用工具，完成回复、审核等操作。',
+      plugin:
+        '专业场景可用插件处理器：批量注册事件，用代码处理，再在下方添加插件处理器配置。',
+      docs: '开发自己的插件处理器',
+      runnerDocs: '开发自己的 Agent 运行器',
+      pause: '暂停',
+      play: '播放',
+    },
     eventRouting: '事件路由',
     eventRoutingDescription: '设置收到事件后交给哪个处理器。',
     eventBindings: '事件路由',
@@ -763,7 +976,7 @@ const zhHans = {
     targetType: '目标类型',
     target: '处理器',
     targetAgent: 'Agent',
-    targetPipeline: 'Pipeline',
+    targetPipeline: '流水线',
     targetDiscard: '丢弃',
     selectTarget: '选择处理器',
     searchTarget: '搜索处理器…',
@@ -772,7 +985,7 @@ const zhHans = {
     enabled: '启用',
     eventBindingDescriptionPlaceholder: '规则说明',
     noEventBindings: '暂无事件路由',
-    unsupportedPipelineEvent: 'Pipeline 仅可用于 message.* 事件',
+    unsupportedPipelineEvent: '流水线 仅可用于 message.received 事件',
     disable: '禁用',
     enable: '启用',
     disabledBindings: '已禁用',
@@ -782,6 +995,7 @@ const zhHans = {
     advancedEventValues: '查看全部',
     eventGroup: '事件组',
     eventGroupNames: {
+      wecomcs: '企业微信客服',
       bot: '机器人状态',
       feedback: '反馈',
       friend: '好友',
@@ -796,6 +1010,9 @@ const zhHans = {
     routeFallbackCatchAll: '{{route}} 是兜底路由。',
     routeFallbackIgnored:
       '未命中任何路由的事件会被忽略。只有需要为每个事件指定结果时，才添加全局兜底路由。',
+    matchRecords: '匹配记录',
+    matchRecordsDescription:
+      '查看各规则最近一次匹配结果及近期未匹配事件。记录来自机器人内存日志，重启后清空。',
     testRoute: '检查路由',
     adapterEventDebugAction: '测试监听',
     adapterEventDebugTitle: '平台事件调试',
@@ -877,6 +1094,14 @@ const zhHans = {
     eventWildcard: '全部事件',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: '用户进入机器人会话',
+      wecombot_template_card_event: '模板卡片交互',
+      dingtalk_card_action: '钉钉卡片交互',
+      wecomcs_enter_session: '客户进入会话',
+      wecomcs_msg_send_fail: '客服消息发送失败',
+      wecomcs_servicer_status_change: '接待人员状态变更',
+      wecomcs_session_status_change: '客服会话状态变更',
+      wecomcs_reject_customer_msg_switch_change: '拒收客户消息设置变更',
       message_received: '收到消息',
       message_edited: '消息被编辑',
       message_deleted: '消息被删除',
@@ -886,6 +1111,7 @@ const zhHans = {
       friend_added: '好友添加成功',
       group_member_joined: '成员加入群组',
       group_member_left: '成员离开群组',
+      group_info_updated: '群信息更新',
       group_member_banned: '成员被封禁',
       bot_invited_to_group: '机器人被邀入群',
       bot_removed_from_group: '机器人被移出群',
@@ -894,12 +1120,25 @@ const zhHans = {
       platform_specific: '平台特定事件',
     },
     eventDescriptions: {
+      wecombot_enter_chat: '用户进入机器人会话，可用于欢迎和引导流程。',
+      wecombot_template_card_event:
+        '用户点击或提交模板卡片；已有交互表单的回调单独处理。',
+      dingtalk_card_action:
+        '用户点击或提交钉钉卡片；反馈和受管理的表单回调单独处理。',
+      wecomcs_enter_session:
+        '客户通过客服链接进入会话，可获取来源和欢迎语凭据。',
+      wecomcs_msg_send_fail:
+        '微信通知客服消息发送失败，包含消息 ID 和失败类型。',
+      wecomcs_servicer_status_change: '接待人员开始、停止接待或暂时挂起。',
+      wecomcs_session_status_change: '人工客服接入、转接、结束或重新接入会话。',
+      wecomcs_reject_customer_msg_switch_change:
+        '接待人员开启或取消拒收客户消息。',
       all: '匹配此适配器收到的全部事件。',
       namespace: '匹配所有{{group}}事件。',
       namespace_bot: '匹配机器人入群、退群、禁言和解除禁言等状态事件。',
       namespace_feedback: '匹配平台或用户反馈事件。',
       namespace_friend: '匹配好友请求、好友添加成功等好友关系事件。',
-      namespace_group: '匹配成员加入、离开或被移出群组等群组成员事件。',
+      namespace_group: '匹配群成员变动、群信息更新等群组事件。',
       namespace_message: '匹配消息接收、编辑、删除和表态事件。',
       namespace_platform: '匹配适配器提供的平台专属事件。',
       custom: '自定义或暂未提供说明的事件。',
@@ -912,6 +1151,7 @@ const zhHans = {
       friend_added: '好友关系建立成功。',
       group_member_joined: '有成员加入机器人所在群组。',
       group_member_left: '有成员离开机器人所在群组。',
+      group_info_updated: '群名称、群主等群资料发生变更。',
       group_member_banned: '群成员被封禁或移出。',
       bot_invited_to_group: '机器人被邀请加入群组。',
       bot_removed_from_group: '机器人被移出群组。',
@@ -995,6 +1235,12 @@ const zhHans = {
     imagesAttached: '张图片',
     noLogs: '暂无日志',
     sessionMonitor: {
+      resetContext: '重置上下文',
+      resetDescription:
+        '重置后，新消息将从空上下文开始。监控历史会保留，文件和长期记忆不受影响。请等待当前任务结束后再重置。',
+      resetSuccess: '上下文已重置，新消息将开启新的对话。',
+      resetError: '重置失败，请等待当前任务结束后重试。',
+
       title: '会话监控',
       sessions: '会话列表',
       noSessions: '暂无会话',
@@ -1041,7 +1287,26 @@ const zhHans = {
     },
   },
   agents: {
+    apiToolsSelectAll: '全选',
     monitoring: {
+      rangeDays: '最近 {{days}} 天',
+      customRange: '自定义时间范围',
+      rangeStart: '开始时间',
+      rangeEnd: '结束时间',
+      rangeHint: '开始时间需早于结束时间，最多 365 天。',
+      autoRefreshLabel: '定时刷新',
+      autoRefreshOff: '定时刷新：关闭',
+      autoRefreshEvery: '每 {{seconds}} 秒刷新',
+      runStatus: '执行状态',
+      toolCalls: '工具调用',
+      delivery: '消息发送',
+      sendFailed: '发送失败',
+      sendCount: '{{count}} 次发送调用成功',
+      noSend: '暂无发送记录',
+      noSteps: '暂未记录执行步骤。',
+      recipient: '接收目标',
+      actionCompleted: '工具已执行完成。',
+      simulatedAction: '模拟执行，未发送真实消息。',
       description: '查看每次任务的触发事件、模型输出和工具执行过程。',
       empty: '暂无运行记录。触发平台事件或运行调试后，可在这里查看。',
       input: '触发输入',
@@ -1141,15 +1406,18 @@ const zhHans = {
       '只处理消息事件，由 AI 直接生成回复，并提供知识库、插件等实用功能。',
     allEvents: '支持全部事件',
     messageEventsOnly: '仅支持消息事件',
+    selectAgentRunner: '选择 Agent 运行器',
+    selectProcessorPlugin: '选择处理器插件',
+    createBasicInfoDescription: '设置名称、图标并选择插件',
     chooseType: '选择处理方式',
     chooseTypeDescription: '先选择处理器的工作方式，创建后可继续完善配置。',
     diagramHint: '工作方式示意',
     agentDiagramTitle: 'Agent：自主判断并行动',
     agentDiagramDescription:
       'Agent 可以接收多种平台事件，结合模型和工具理解当前情况，并自主决定下一步动作。',
-    pipelineDiagramTitle: 'Pipeline：按流程逐步处理',
+    pipelineDiagramTitle: '流水线：按流程逐步处理',
     pipelineDiagramDescription:
-      'Pipeline 只处理消息事件，并严格按照你编排的固定步骤依次执行。',
+      '流水线 只处理消息事件，并严格按照你编排的固定步骤依次执行。',
     pipelineDiagramFlow: '消息沿固定流程依次处理',
     diagramEvents: '平台事件',
     diagramAgentCanUse: 'Agent 可使用',
@@ -1531,6 +1799,11 @@ const zhHans = {
     },
   },
   market: {
+    certifiedPlugin: '认证插件',
+    certificationTooltip:
+      '此版本已通过数据隔离与共享运行环境兼容性审核，插件包已签名。',
+    certificationLearnMore: '了解插件认证',
+
     runnerUsage: '运行器用途',
     runnerUsageAll: '全部',
     runnerUsageAgent: 'Agent / 流水线',
@@ -1914,6 +2187,14 @@ const zhHans = {
       uploading: '上传中...',
     },
     monitoring: {
+      rangeDays: '最近 {{days}} 天',
+      customRange: '自定义时间范围',
+      rangeStart: '开始时间',
+      rangeEnd: '结束时间',
+      rangeHint: '开始时间需早于结束时间，最多 365 天。',
+      autoRefreshLabel: '定时刷新',
+      autoRefreshOff: '定时刷新：关闭',
+      autoRefreshEvery: '每 {{seconds}} 秒刷新',
       title: '运行日志',
       workbench: '配置与调试',
       description: '查看此流水线的运行记录和错误信息（最近24小时）',
@@ -2175,7 +2456,8 @@ const zhHans = {
     totpRecoveryCodesRegenerated: '已生成新的恢复代码',
     totpStatusDisabled: '未启用',
     totpCodesRemaining: '剩余 {{count}} 个恢复代码',
-    totpManagerSectionDesc: '所有者和管理员可以查看并重置任意账户的两步验证。',
+    totpManagerSectionDesc:
+      '所有者和管理员可以查看并重置本工作区账户的两步验证。',
     revokeTotp: '重新绑定',
     totpAdminResetTitle: '重新绑定 {{user}} 的两步验证',
     totpAdminResetDesc:
@@ -2235,8 +2517,7 @@ const zhHans = {
     upgradePlan: '切换或升级计划',
     ossSingletonDescription:
       '当前自托管实例只有一个工作区，但可以包含多个用户。',
-    cloudManagedDescription:
-      '此工作区托管于 LangBot Cloud。成员在此管理，计费在 Cloud 中打开。',
+    cloudManagedDescription: '此工作区托管于 LangBot Cloud。',
     loadFailed: '加载工作区信息失败',
     members: '成员',
     you: '你',
@@ -2302,6 +2583,122 @@ const zhHans = {
     },
   },
   monitoring: {
+    rangeDays: '最近 {{days}} 天',
+    customRange: '自定义时间范围',
+    rangeStart: '开始时间',
+    rangeEnd: '结束时间',
+    rangeHint: '开始时间需早于结束时间，最多 365 天。',
+    autoRefreshLabel: '定时刷新',
+    autoRefreshOff: '定时刷新：关闭',
+    autoRefreshEvery: '每 {{seconds}} 秒刷新',
+    unified: {
+      inputs: '事件与输入',
+      outputs: '生成结果',
+      deliveries: '发送记录',
+      conversation: '会话上下文',
+      related: '关联执行',
+      events: '执行事件',
+      llm_calls: '模型调用',
+      tool_calls: '工具调用',
+      errors: '错误',
+      payload: '查看详情',
+      metadata: '技术信息',
+      loadMore: '加载更多',
+      noRecords: '暂无记录',
+      contextHint: '会话中的其他消息不一定属于本次执行。',
+      unhandled: '未启动处理器',
+      roles: {
+        user: '用户',
+        assistant: '助手',
+        system: '系统',
+        tool: '工具',
+      },
+    },
+    execution: {
+      title: '执行记录',
+      subtitle: '每条入站事件及其触发的执行，覆盖 Agent 与 流水线。',
+      tabs: { executions: '执行记录' },
+      filters: {
+        source: '类型',
+        sourceAll: '全部类型',
+        sourceAgent: 'Agent',
+        sourcePipeline: '流水线',
+        mode: '流量',
+        modeAll: '真实 + 调试',
+        modeLive: '仅真实',
+        modeDebug: '仅调试',
+        status: '状态',
+        statusAll: '全部状态',
+      },
+      status: {
+        completed: '已完成',
+        failed: '失败',
+        running: '运行中',
+        queued: '排队中',
+        cancelled: '已取消',
+        ignored: '已忽略',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Agent',
+        pipeline: '流水线',
+        processor: '处理器',
+      },
+      columns: {
+        time: '时间',
+        target: '处理器',
+        status: '状态',
+        duration: '耗时',
+        tokens: 'Token',
+      },
+      debug: '调试',
+      empty: '该时间范围内暂无执行记录',
+      loading: '加载执行记录中...',
+      triggeredBy: '触发：{{trigger}}',
+      cards: {
+        latencyHint:
+          'P50/P95：约 50%/95% 的执行能在此时间内完成。仅统计最近已完成的 Agent 执行，不含排队时间。',
+        tokensHint:
+          '仅汇总已记录的用量；部分模型未返回用量时，总数可能不完整。',
+        totalExecutions: '执行总数',
+        bySource: 'Agent {{agent}} · 流水线 {{pipeline}}',
+        successRate: '执行成功率',
+        successRateHint:
+          '成功次数 ÷（成功 + 失败）。不计进行中、已忽略和已取消的执行。',
+        denominator: '{{denominator}} 次已结束，{{completed}} 次成功',
+        running: '运行中',
+        queued: '排队中',
+        waiting: '等待输入',
+        failed: '失败',
+        ignored: '已忽略',
+        latency: '执行耗时',
+        p50: 'P50',
+        p95: 'P95',
+        sample: '根据 {{count}} 次执行计算',
+        tokens: 'Token',
+        coverage: '{{withUsage}}/{{calls}} 次调用有用量',
+        debugRuns: '其中 {{count}} 条为调试执行',
+      },
+      detail: {
+        title: '执行详情',
+        timeline: '执行时间线',
+        executionId: '执行 ID',
+        eventId: '事件',
+        queue: '队列',
+        runner: 'Runner',
+        created: '创建时间',
+        duration: '耗时',
+        usage: '用量',
+        noEvents: '暂无追踪事件',
+        query: '入站消息',
+        modelCalls: '模型调用',
+        toolCalls: '工具调用',
+        errors: '错误',
+        loadError: '加载执行轨迹失败',
+        openObject: '打开对象',
+        a11yDescription: '此次执行的完整追踪',
+      },
+    },
     title: '仪表盘',
     description: '查看机器人活动、LLM调用和系统性能',
     overview: '概览',
@@ -2317,6 +2714,9 @@ const zhHans = {
       pipeline: '流水线',
       allBots: '全部机器人',
       selectBot: '选择机器人',
+      processor: '处理器',
+      allProcessors: '全部处理器',
+      selectProcessor: '选择处理器',
       allPipelines: '全部流水线',
       selectPipeline: '选择流水线',
       loading: '加载中...',
@@ -2342,6 +2742,15 @@ const zhHans = {
       sessions: '会话分析',
       errors: '错误日志',
     },
+    pipelineTrace: {
+      received: '收到消息',
+      processing: '处理过程',
+      replies: '回复消息',
+      noSteps: '未记录模型或工具调用',
+      success: '成功',
+      error: '失败',
+      pending: '处理中',
+    },
     messageList: {
       timestamp: '时间戳',
       bot: '机器人',
@@ -2561,7 +2970,7 @@ const zhHans = {
     title: '设置',
     nav: {
       workspace: '工作区',
-      operationTrace: '操作溯源',
+      operationTrace: '操作日志',
       models: '模型',
       api: 'API',
       storage: '存储',
@@ -2569,21 +2978,21 @@ const zhHans = {
     },
   },
   operationTrace: {
-    title: '操作溯源',
+    title: '操作日志',
     description: '追踪管理员与拥有者的修改和查看操作，记录“什么改成了什么”。',
+    reset: '重置',
+    settingsUpdated: '设置已保存',
+    settingsUpdateFailed: '保存设置失败',
+    levelBadge: '等级：{{level}}',
+    collectionSettings: '收集设置',
     captureLevel: '操作等级',
     retention: '保留策略',
-    retentionSummary: '保留 {{days}} 天 · 最多 {{rows}} 条',
     retentionDays: '保留天数',
     maxRows: '最大记录数',
     dedupeWindow: '去重窗口（秒）',
     records: '操作记录',
     changesCount: '{{count}} 项变更',
     hideDetails: '收起详情',
-    levelUpdated: '操作等级已更新',
-    levelUpdateFailed: '操作等级更新失败',
-    retentionUpdated: '保留策略已更新',
-    retentionUpdateFailed: '保留策略更新失败',
     export: '下载日志',
     exportFailed: '下载操作日志失败',
     empty: '暂无操作记录',
@@ -2597,7 +3006,7 @@ const zhHans = {
     filterAllResources: '全部资源',
     filterActor: '操作者',
     filterAllActors: '全部操作者',
-    loadFailed: '加载操作溯源数据失败',
+    loadFailed: '加载操作日志数据失败',
     refresh: '刷新',
     save: '保存',
     redacted: '已脱敏',
@@ -2616,7 +3025,7 @@ const zhHans = {
     elidedValue: '{{count}} 字符',
     levels: {
       off: {
-        label: '关闭溯源',
+        label: '关闭日志',
       },
       mutation: {
         label: '仅修改操作',
@@ -2635,7 +3044,7 @@ const zhHans = {
       adapter: '适配器',
       model_provider: '模型提供商',
       llm_model: '模型',
-      pipeline: '管道',
+      pipeline: '流水线',
       user: '用户',
       workspace: '工作区',
       monitoring: '监控',
@@ -2664,8 +3073,8 @@ const zhHans = {
     },
     actions: {
       audit_log_view: '查看操作日志',
-      settings_update: '修改溯源设置',
-      settings_view: '查看溯源设置',
+      settings_update: '修改日志设置',
+      settings_view: '查看日志设置',
       member_invite: '邀请成员',
       member_role_update: '修改成员角色',
       member_remove: '移除成员',
@@ -2693,7 +3102,7 @@ const zhHans = {
       codex_view: '查看 Codex 授权',
       codex_authorize: '授权 Codex',
       embed: '嵌入式会话',
-      pipeline_extensions_update: '修改管道扩展绑定',
+      pipeline_extensions_update: '修改流水线扩展绑定',
       export: '导出数据',
       execute: '执行任务',
       debug: '调试任务',
diff --git a/web/src/i18n/locales/zh-Hant.ts b/web/src/i18n/locales/zh-Hant.ts
index 029a73ba9..8a8b749c2 100644
--- a/web/src/i18n/locales/zh-Hant.ts
+++ b/web/src/i18n/locales/zh-Hant.ts
@@ -1,6 +1,169 @@
+import slackSetup from './slack-setup/zh-Hant';
 import pipelineMigration from './pipeline-migration/zh-Hant';
 const zhHant = {
+  planRevision: {
+    subscribedOn: '訂閱日期：{{date}}',
+    latestQuota: '該方案最新配額',
+    showLatest: '查看該計畫最新配額',
+    showCurrent: '查看目前訂閱',
+    latestTab: '最新配額',
+    currentTab: '目前訂閱',
+    cancelToSubscribe: '取消本期訂閱，下一期即可訂閱最新配額',
+
+    current: '当前订阅：{{plan}} · revision {{revision}}',
+    owned: '已购买 revision {{revision}}',
+    latest: '最新 revision {{revision}}',
+    changed: '已购买权益与最新版本存在差异，高亮行标示变动。',
+    same: '权益与最新 revision {{revision}} 相同，已购买版本保持不变。',
+    unavailable: '最新计划暂不可用，已购买权益继续有效。',
+    benefit: '权益',
+    change: '有变化',
+    enabled: '启用',
+    disabled: '禁用',
+    monthlyCredits: '每月赠送额度',
+    retention: '聊天记录保留天数',
+    unknown:
+      '当前订阅详情不可用。请刷新或前往 Space 管理订阅；最新目录权益不代表已购买权益。',
+    handoff:
+      '请在 Space 管理订阅。自动续费需先取消后才能购买新版本；预付订阅到期自动结束，除非到期前手动续订。',
+    cancelFirst:
+      '请先在订阅管理中取消自动续费。取消后保留本期已付权益，不退款。服务端确认取消后可重新购买最新版本，付款成功后新权益生效。',
+    canceled:
+      '续费已取消。本期已付权益保留至到期，不退款；重新购买的权益在付款成功后生效。',
+    prepaid:
+      '本預付訂閱到期自動結束。到期前可續訂目前方案；到期後只能按各方案的最新配額訂閱。',
+    renew: '續訂目前訂閱',
+    buying: '正在购买 revision {{revision}}。',
+    fields: {
+      bots_max: '机器人',
+      members_max: '成员',
+      pipelines_max: '处理器',
+      plugins_max: '扩展',
+      knowledge_bases_max: '知识库',
+      managed_sandbox_sessions_max: '沙箱会话',
+    },
+  },
+  cloudPlans: {
+    subscriptionUnavailable: '暫不可訂閱',
+    subscriptionUnavailableHint:
+      '此方案暫不接受新訂閱，請選擇其他方案。既有訂閱和已簽約的自動續費不受影響。',
+    platformNames: {
+      qqofficial: 'QQ 官方 API',
+      dingtalk: '釘釘',
+      discord: 'Discord',
+      telegram: 'Telegram',
+      lark: '飛書',
+      slack: 'Slack',
+      kook: 'KOOK',
+      line: 'LINE',
+      wecom: '企業微信',
+      wecombot: '企業微信智慧機器人',
+      wecomcs: '企業微信客服',
+      officialaccount: '微信公眾號',
+      wechatpad: 'WeChatPad（個人微信iPad）',
+      openclaw_weixin: '個人微信機器人',
+      aiocqhttp: 'OneBot v11',
+      satori: 'Satori',
+    },
+    manageSubscription: '管理訂閱',
+    downgradeUnavailable: '不支援降級至免費版',
+    features: {
+      multiUserWorkspace: '多人協作工作區',
+      remoteE2BSandbox: '遠端 E2B 沙箱設定',
+      communitySupport: '社群支援',
+      everythingInFree: '包含免費版全部功能',
+      alwaysOnSandbox: '一個常駐全域沙箱',
+      higherWorkspaceLimits: '更高的工作區配額',
+    },
+    title: '比較訂閱方案',
+    description: '為 {{workspace}} 選擇訂閱方案',
+    billing: '計費週期',
+    monthly: '月付',
+    annual: '年付',
+    currency: '幣別',
+    loading: '正在載入方案…',
+    error: '無法載入訂閱方案',
+    retry: '重試',
+    empty: '暫無可用方案',
+    current: '目前方案',
+    recommended: '推薦',
+    free: '免費',
+    annualNote: '相當於 {{price}} / 月',
+    ownerOnly: '僅擁有者或管理員可操作',
+    choose: '前往付款',
+    handoff: '將在 Space 中登入並付款，所選方案、週期和幣別會保留。',
+    quotas: '工作區配額',
+    unlimited: '不限',
+    retention: '訊息記錄保留 {{count}} 天',
+    retentionLabel: '訊息記錄保留時間',
+    retentionHelp: '超過保留時間的訊息記錄會自動清理。',
+    credits: '每月贈送 {{count}} 積分',
+    creditsLabel: '每月贈送積分',
+    creditsHelp: '每個訂閱月自動發放，年付方案同樣按月發放。',
+    platforms: '支援主流平台',
+    excludedPlatforms: '不支援：{{platforms}}',
+    allPlatforms: '支援目前所有可用平台',
+    limits: {
+      bots: '機器人',
+      pipelines: '處理器',
+      knowledge_bases: '知識庫',
+      plugins: '擴充',
+      members: '成員',
+      managed_sandbox_sessions: '託管沙盒工作階段',
+    },
+  },
+  workspaceLanguage: {
+    label: '工作區語言',
+    description: '用於後端提示和機器人的預設回覆，不影響個人介面語言。',
+    saved: '語言設定已儲存',
+    saveFailed: '語言設定儲存失敗',
+  },
+  slackSetup,
+  inflight: {
+    autoCollapse: '自動摺疊',
+    empty: '目前沒有正在處理的事件',
+    move: '上下拖動；也可用方向鍵移動',
+    botLogs: '機器人日誌',
+    processorLogs: '處理器日誌',
+
+    title: '處理中',
+    expand: '展開',
+    collapse: '收起',
+    reconnecting: '連線中斷，正在重新連線；顯示最後收到的狀態。',
+    details: '查看詳情',
+    debug: '除錯事件',
+    limited: '僅顯示部分執行，請在儀表板查看全部記錄。',
+    queued: '排隊中',
+    running: '處理中',
+    waiting: '等待使用者輸入',
+    tool: '呼叫工具',
+    thinking: '思考中',
+    replying: '準備回覆',
+    generating: '產生回覆',
+    completed: '已完成',
+    failed: '失敗',
+    cancelled: '已取消',
+    ignored: '已忽略',
+  },
   assistant: {
+    movePosition: '左右移動助手',
+    resizeHeight: '調整助手高度',
+    sessions: '切換對話',
+    currentChat: '目前對話',
+    stop: '停止任務',
+    stopped: '已停止。執行中的工具可能已經生效，請檢查相關資源後再嘗試。',
+    setupRequired: '登入 LangBot 帳戶或設定模型，即可開始使用工作區助手。',
+    loginAccount: '登入 LangBot Account',
+    configureModels: '設定模型',
+    providerCheckFailed: '無法檢查模型供應商，請重試。',
+    retrySetup: '重試',
+    preparing: '正在準備助手…',
+    thinking: '模型回應中 · 第 {{round}} 輪',
+    executing: '正在執行工具…',
+    toolRunning: '執行中',
+    recommendationFailed: '推薦模型載入失敗，請手動選擇模型。',
+    expand: '展開工作區助手',
+    collapse: '收起工作區助手',
     modelHint: '切換助手模型，從下一條訊息生效',
     details: '檢視原始資料',
     found: '找到 {{count}} 項',
@@ -14,17 +177,17 @@ const zhHant = {
     sendUnconfirmed: '傳送狀態未確認，請檢查連線後重新開啟對話。',
     draftPlaceholder: '可以先寫下一條，回覆完成後傳送…',
     operations: {
-      create_pipeline: '建立 Pipeline',
-      configure_pipeline: '設定 Pipeline',
+      create_pipeline: '建立 流水線',
+      configure_pipeline: '設定 流水線',
       create_knowledge_base: '建立知識庫',
-      get_pipeline: '查看 Pipeline',
+      get_pipeline: '查看 流水線',
       get_knowledge_schema: '查看知識引擎設定',
-      list_operation_logs: '查詢操作溯源',
+      list_operation_logs: '查詢操作日誌',
     },
     resources: {
       models: '查詢聊天模型',
       embedding_models: '查詢嵌入模型',
-      pipelines: '查詢 Pipeline',
+      pipelines: '查詢 流水線',
       knowledge_bases: '查詢知識庫',
       knowledge_engines: '查詢知識引擎',
     },
@@ -33,8 +196,8 @@ const zhHant = {
     newChat: '新對話',
     close: '關閉',
     welcome:
-      '告訴我你想建構什麼應用。我可以查看目前工作區，並提出知識庫和 Pipeline 設定方案，由你確認後執行。',
-    discover: '目前有哪些可用的模型、知識庫和 Pipeline？',
+      '告訴我你想建構什麼應用。我可以查看目前工作區，為機器人、Agent、知識庫和 流水線 等提供設定方案，由你確認後執行。',
+    discover: '目前有哪些可用的模型、知識庫和 流水線？',
     build: '幫我建立一個知識庫問答應用。',
     toolResult: '操作結果',
     review: '請檢查具體操作及參數，確認後執行',
@@ -53,6 +216,14 @@ const zhHant = {
   sidebarGuide: {
     steps: {
       monitoring: {
+        rangeDays: '最近 {{days}} 天',
+        customRange: '自訂時間範圍',
+        rangeStart: '開始時間',
+        rangeEnd: '結束時間',
+        rangeHint: '開始時間須早於結束時間，最多 365 天。',
+        autoRefreshLabel: '定時重新整理',
+        autoRefreshOff: '定時重新整理：關閉',
+        autoRefreshEvery: '每 {{seconds}} 秒重新整理',
         title: '查看執行概況',
         description: '集中查看機器人活動、模型呼叫、訊息量和系統運行情況。',
       },
@@ -128,6 +299,14 @@ const zhHant = {
           '傳送測試訊息，檢查觸發規則、AI 回覆和輸出處理是否符合預期；修改設定後，測試前會先儲存。',
       },
       monitoring: {
+        rangeDays: '最近 {{days}} 天',
+        customRange: '自訂時間範圍',
+        rangeStart: '開始時間',
+        rangeEnd: '結束時間',
+        rangeHint: '開始時間須早於結束時間，最多 365 天。',
+        autoRefreshLabel: '定時重新整理',
+        autoRefreshOff: '定時重新整理：關閉',
+        autoRefreshEvery: '每 {{seconds}} 秒重新整理',
         title: '查看執行日誌',
         description:
           '開啟標題旁的「執行日誌」，查看歷史對話、處理過程和錯誤，排查回覆異常。',
@@ -281,6 +460,10 @@ const zhHant = {
     editionCloud: 'Cloud',
   },
   common: {
+    failureReasonNotRecorded:
+      '這筆歷史記錄未儲存失敗原因，請查看機器人日誌，或重新傳送訊息取得詳細錯誤。',
+    loadFailedTitle: '暫時無法載入',
+    loadFailedHint: '請稍後重試。如果仍無法載入，請檢查網路連線與服務狀態。',
     customValue: '自訂',
     loadFailed: '載入失敗，請重試。',
     search: '搜尋',
@@ -467,6 +650,9 @@ const zhHant = {
     help: '查看說明文件',
   },
   models: {
+    starModel: '標星模型（用作新處理器的預設模型）',
+    unstarModel: '取消標星',
+    starModelFailed: '無法更新標星模型',
     codex: {
       account: 'ChatGPT 訂閱',
       description:
@@ -499,7 +685,7 @@ const zhHant = {
       confirmDisconnect: '確認中斷',
     },
     title: '模型設定',
-    description: '設定和管理可在流程線中使用的模型',
+    description: '設定和管理可在流水線中使用的模型',
     createModel: '建立模型',
     editModel: '編輯模型',
     getModelListError: '取得模型清單失敗：',
@@ -713,14 +899,17 @@ const zhHant = {
     deleteConfirmation: '您確定要刪除這個機器人嗎？',
     platformAdapter: '平台/適配器選擇',
     selectAdapter: '選擇適配器',
+    saveAndEnable: '儲存並啟用',
+    searchAdapters: '搜尋適配器',
+    noMatchingAdapters: '找不到符合條件的適配器',
     legacyAdapters: '舊版適配器',
     legacyAdapterBadge: '過時',
     legacyAdaptersHint:
-      '這些適配器已有對應的事件驅動接入版本，\n僅為相容既有設定保留，不建議新建機器人時使用。',
+      '舊版主要用於訊息收發，保留以相容現有機器人設定。\n新版是同一平台的事件驅動接入，除訊息外，也可處理平台支援的好友請求、群成員變化等事件，透過事件路由交給處理器。\n新建機器人優先選新版。現有機器人可繼續使用舊版；切換時請依新版要求檢查連線參數和事件路由，並非直接升級平台帳號。',
     adapterConfig: '適配器設定',
     viewAdapterDocs: '查看文檔',
-    bindPipeline: '綁定流程線',
-    selectPipeline: '選擇流程線',
+    bindPipeline: '綁定流水線',
+    selectPipeline: '選擇流水線',
     selectBot: '請選擇機器人',
     botLogTitle: '機器人日誌',
     enableAutoRefresh: '開啟自動重新整理',
@@ -729,16 +918,40 @@ const zhHant = {
     earlier: '更久之前',
     dateFormat: '{{month}}月{{day}}日',
     setBotEnableError: '設定機器人啟用狀態失敗',
+    enabling: '正在啟用…',
+    disabling: '正在停用…',
+    enableConfirmed: '已啟用',
+    disableConfirmed: '已停用',
     log: '日誌',
     configuration: '設定',
     logs: '日誌',
     basicInfo: '基礎資訊',
     basicInfoDescription: '設定機器人名稱和描述',
     routingConnection: '路由與連接',
-    routingConnectionDescription: '綁定處理此機器人訊息的流程線',
+    routingConnectionDescription: '綁定處理此機器人訊息的流水線',
+    routingHelp: {
+      title: '什麼是事件路由？',
+      intro: '機器人不只會收到訊息，也會收到好友請求、加入群聊等事件。',
+      events: '機器人收到事件',
+      agents: '交給不同 Agent',
+      message: '收到訊息',
+      friend: '好友請求',
+      group: '加入群聊',
+      messageAction: '問答 Agent → 回覆',
+      friendAction: '審核 Agent → 處理申請',
+      groupAction: '歡迎 Agent → 打招呼',
+      rules: '依您設定的規則分配 · 範例',
+      instructions: 'Agent 依您的指令呼叫工具，完成回覆、審核等操作。',
+      plugin:
+        '專業場景可用插件處理器：批次註冊事件，用程式碼處理，再於下方新增插件處理器設定。',
+      docs: '開發自己的插件處理器',
+      runnerDocs: '開發自己的 Agent 運行器',
+      pause: '暫停',
+      play: '播放',
+    },
     eventRouting: '事件路由',
     eventRoutingDescription:
-      '選擇此機器人收到不同事件時交給哪個處理器。在對應的 Agent 或 Pipeline 設定中編輯處理邏輯；Pipeline 僅支援訊息事件。',
+      '選擇此機器人收到不同事件時交給哪個處理器。在對應的 Agent 或 流水線 設定中編輯處理邏輯；流水線 僅支援訊息事件。',
     eventBindings: '事件路由',
     addEventBinding: '添加路由',
     addBehavior: '添加行爲',
@@ -761,7 +974,7 @@ const zhHant = {
     targetType: '目標類型',
     target: '處理器',
     targetAgent: 'Agent',
-    targetPipeline: 'Pipeline',
+    targetPipeline: '流水線',
     targetDiscard: '丟棄',
     selectTarget: '選擇處理器',
     searchTarget: '搜索處理器…',
@@ -770,7 +983,7 @@ const zhHant = {
     enabled: '啟用',
     eventBindingDescriptionPlaceholder: '規則說明',
     noEventBindings: '暫無事件路由',
-    unsupportedPipelineEvent: 'Pipeline 僅可用於 message.* 事件',
+    unsupportedPipelineEvent: '流水線 僅可用於 message.received 事件',
     disable: '禁用',
     enable: '啟用',
     disabledBindings: '已禁用',
@@ -780,6 +993,7 @@ const zhHant = {
     advancedEventValues: '查看全部',
     eventGroup: '事件組',
     eventGroupNames: {
+      wecomcs: '企業微信客服',
       bot: '機器人狀態',
       feedback: '反饋',
       friend: '好友',
@@ -794,6 +1008,9 @@ const zhHant = {
     routeFallbackCatchAll: '{{route}} 是兜底路由。',
     routeFallbackIgnored:
       '未命中任何路由的事件會被忽略。只有需要爲每個事件指定結果時，才添加全局兜底路由。',
+    matchRecords: '匹配記錄',
+    matchRecordsDescription:
+      '查看各規則最近一次匹配結果及近期未匹配事件。記錄來自機器人記憶體日誌，重新啟動後清空。',
     testRoute: '檢查路由',
     adapterEventDebugAction: '測試監聽',
     adapterEventDebugTitle: '平臺事件調試',
@@ -875,6 +1092,14 @@ const zhHant = {
     eventWildcard: '全部事件',
     eventNamespaceWildcard: '{{namespace}}.*',
     eventNames: {
+      wecombot_enter_chat: '使用者進入機器人會話',
+      wecombot_template_card_event: '範本卡片互動',
+      dingtalk_card_action: '釘釘卡片互動',
+      wecomcs_enter_session: '客戶進入會話',
+      wecomcs_msg_send_fail: '客服訊息傳送失敗',
+      wecomcs_servicer_status_change: '接待人員狀態變更',
+      wecomcs_session_status_change: '客服會話狀態變更',
+      wecomcs_reject_customer_msg_switch_change: '拒收客戶訊息設定變更',
       message_received: '收到訊息',
       message_edited: '訊息已編輯',
       message_deleted: '訊息已刪除',
@@ -884,6 +1109,7 @@ const zhHant = {
       friend_added: '新增好友',
       group_member_joined: '成員加入群組',
       group_member_left: '成員離開群組',
+      group_info_updated: '群組資訊更新',
       group_member_banned: '成員被封禁',
       bot_invited_to_group: '機器人被邀請加入群組',
       bot_removed_from_group: '機器人被移出群組',
@@ -892,12 +1118,25 @@ const zhHant = {
       platform_specific: '平台特定事件',
     },
     eventDescriptions: {
+      wecombot_enter_chat: '使用者進入機器人會話，可用於歡迎和引導流程。',
+      wecombot_template_card_event:
+        '使用者點擊或提交範本卡片；已有互動表單的回呼單獨處理。',
+      dingtalk_card_action:
+        '使用者點擊或提交釘釘卡片；回饋和受管理的表單回呼單獨處理。',
+      wecomcs_enter_session:
+        '客戶透過客服連結進入會話，可取得來源和歡迎語憑證。',
+      wecomcs_msg_send_fail:
+        '微信通知客服訊息傳送失敗，包含訊息 ID 和失敗類型。',
+      wecomcs_servicer_status_change: '接待人員開始、停止接待或暫時掛起。',
+      wecomcs_session_status_change: '人工客服接入、轉接、結束或重新接入會話。',
+      wecomcs_reject_customer_msg_switch_change:
+        '接待人員開啟或取消拒收客戶訊息。',
       all: '匹配此適配器收到的全部事件。',
       namespace: '匹配所有{{group}}事件。',
       namespace_bot: '匹配機器人入羣、退羣、禁言和解除禁言等狀態事件。',
       namespace_feedback: '匹配平臺或使用者反饋事件。',
       namespace_friend: '匹配好友請求、好友添加成功等好友關係事件。',
-      namespace_group: '匹配成員加入、離開或被移出羣組等羣組成員事件。',
+      namespace_group: '匹配群組成員變動、群組資訊更新等事件。',
       namespace_message: '匹配訊息接收、編輯、刪除和表態事件。',
       namespace_platform: '匹配適配器提供的平臺專屬事件。',
       custom: '自定義或暫未提供說明的事件。',
@@ -910,6 +1149,7 @@ const zhHant = {
       friend_added: '好友關係建立成功。',
       group_member_joined: '有成員加入機器人所在羣組。',
       group_member_left: '有成員離開機器人所在羣組。',
+      group_info_updated: '群組名稱、群主等群組資料發生變更。',
       group_member_banned: '羣成員被封禁或移出。',
       bot_invited_to_group: '機器人被邀請加入羣組。',
       bot_removed_from_group: '機器人被移出羣組。',
@@ -933,7 +1173,7 @@ const zhHant = {
     operator_regex: '正則匹配',
     routingRules: '條件路由規則',
     routingRulesDescription:
-      '按順序匹配，命中第一條規則後路由到對應流程線；都不匹配時使用上方預設流程線',
+      '按順序匹配，命中第一條規則後路由到對應流水線；都不匹配時使用上方預設流水線',
     addRoutingRule: '新增規則',
     ruleTypeLauncherType: '會話類型',
     ruleTypeLauncherId: '會話 ID',
@@ -993,6 +1233,12 @@ const zhHant = {
     imagesAttached: '張圖片已附加',
     noLogs: '暫無日誌',
     sessionMonitor: {
+      resetContext: '重設上下文',
+      resetDescription:
+        '重設後，新訊息將從空上下文開始。監控歷史會保留，檔案和長期記憶不受影響。請等待目前任務結束後再重設。',
+      resetSuccess: '上下文已重設，新訊息將開始新的對話。',
+      resetError: '重設失敗，請等待目前任務結束後重試。',
+
       title: '會話監控',
       sessions: '會話列表',
       noSessions: '暫無會話',
@@ -1039,7 +1285,26 @@ const zhHant = {
     },
   },
   agents: {
+    apiToolsSelectAll: '全選',
     monitoring: {
+      rangeDays: '最近 {{days}} 天',
+      customRange: '自訂時間範圍',
+      rangeStart: '開始時間',
+      rangeEnd: '結束時間',
+      rangeHint: '開始時間須早於結束時間，最多 365 天。',
+      autoRefreshLabel: '定時重新整理',
+      autoRefreshOff: '定時重新整理：關閉',
+      autoRefreshEvery: '每 {{seconds}} 秒重新整理',
+      runStatus: '執行狀態',
+      toolCalls: '工具呼叫',
+      delivery: '訊息傳送',
+      sendFailed: '傳送失敗',
+      sendCount: '{{count}} 次傳送呼叫成功',
+      noSend: '暫無傳送記錄',
+      noSteps: '尚未記錄執行步驟。',
+      recipient: '接收目標',
+      actionCompleted: '工具已執行完成。',
+      simulatedAction: '模擬執行，未傳送真實訊息。',
       description: '查看每次任務的觸發事件、模型輸出和工具執行過程。',
       empty: '尚無執行紀錄。觸發平台事件或執行除錯後，可在此查看。',
       input: '觸發輸入',
@@ -1129,7 +1394,7 @@ const zhHant = {
     selectFromSidebar: '從側邊欄選擇一個處理器',
     agentType: 'Agent',
     agentTypeDescription: '透過執行器處理訊息、群成員、好友、回饋等平台事件。',
-    pipelineType: '流程線',
+    pipelineType: '流水線',
     kindBadgeAgent: 'Agent',
     kindBadgePipeline: '流水線',
     groupByKind: '依類型分組',
@@ -1138,15 +1403,18 @@ const zhHant = {
       '依「接收訊息、呼叫 AI、回覆使用者」的固定流程執行，可設定知識庫與外掛擴充。僅處理訊息事件，適合步驟明確、需要控制處理過程的場景。',
     allEvents: '支援全部事件',
     messageEventsOnly: '僅支援訊息事件',
+    selectAgentRunner: '選擇 Agent 運行器',
+    selectProcessorPlugin: '選擇處理器插件',
+    createBasicInfoDescription: '設定名稱、圖示並選擇插件',
     chooseType: '選擇處理方式',
     chooseTypeDescription: '先選擇處理器的工作方式，創建後可繼續完善配置。',
     diagramHint: '工作方式示意',
     agentDiagramTitle: 'Agent：自主判斷並行動',
     agentDiagramDescription:
       'Agent 可以接收多種平臺事件，結合模型和工具理解當前情況，並自主決定下一步動作。',
-    pipelineDiagramTitle: 'Pipeline：按流程逐步處理',
+    pipelineDiagramTitle: '流水線：按流程逐步處理',
     pipelineDiagramDescription:
-      'Pipeline 只處理訊息事件，並嚴格按照你編排的固定步驟依次執行。',
+      '流水線 只處理訊息事件，並嚴格按照你編排的固定步驟依次執行。',
     pipelineDiagramFlow: '訊息沿固定流程依次處理',
     diagramEvents: '平臺事件',
     diagramAgentCanUse: 'Agent 可使用',
@@ -1322,7 +1590,7 @@ const zhHant = {
   },
   plugins: {
     title: '外掛擴展',
-    description: '安裝和設定用於擴展功能的外掛，請在流程線配置中選用',
+    description: '安裝和設定用於擴展功能的外掛，請在流水線配置中選用',
     createPlugin: '建立外掛',
     editPlugin: '編輯外掛',
     installed: '已安裝',
@@ -1529,6 +1797,11 @@ const zhHant = {
     },
   },
   market: {
+    certifiedPlugin: '認證外掛',
+    certificationTooltip:
+      '此版本已通過資料隔離與共享執行環境相容性審核，外掛套件已簽章。',
+    certificationLearnMore: '瞭解外掛認證',
+
     runnerUsage: '運行器用途',
     runnerUsageAll: '全部',
     runnerUsageAgent: 'Agent / 流水線',
@@ -1758,15 +2031,15 @@ const zhHant = {
     deleteMCPHint: '刪除後，此 MCP 伺服器設定將無法恢復。',
   },
   pipelines: {
-    title: '流程線',
-    description: '流程線定義了對訊息事件的處理流程，用於綁定到機器人',
-    createPipeline: '建立流程線',
-    selectFromSidebar: '從側邊欄選擇一個流程線',
-    editPipeline: '編輯流程線',
+    title: '流水線',
+    description: '流水線定義了對訊息事件的處理流程，用於綁定到機器人',
+    createPipeline: '建立流水線',
+    selectFromSidebar: '從側邊欄選擇一個流水線',
+    editPipeline: '編輯流水線',
     chat: '對話',
     configuration: '設定',
     debugChat: '對話除錯',
-    getPipelineListError: '取得流程線清單失敗：',
+    getPipelineListError: '取得流水線清單失敗：',
     daysAgo: '天前',
     today: '今天',
     updateTime: '更新於',
@@ -1777,7 +2050,7 @@ const zhHant = {
     recentlyEdited: '最近編輯',
     earliestEdited: '最早編輯',
     basicInfo: '基本資訊',
-    basicInfoDescription: '設定流程線名稱、圖示和描述',
+    basicInfoDescription: '設定流水線名稱、圖示和描述',
     managementDescription: '複製或刪除此流水線。',
     aiCapabilities: 'AI 能力',
     triggerConditions: '觸發條件',
@@ -1785,26 +2058,26 @@ const zhHant = {
     outputProcessing: '輸出處理',
     nameRequired: '名稱不能為空',
     descriptionRequired: '描述不能為空',
-    createSuccess: '建立成功 請編輯流程線詳細參數',
+    createSuccess: '建立成功 請編輯流水線詳細參數',
     createError: '建立失敗：',
     saveSuccess: '儲存成功',
     saveError: '儲存失敗：',
     copySuffix: ' Copy',
     deleteConfirmation:
-      '您確定要刪除這個流程線嗎？已綁定此流程線的機器人將無法使用。',
-    defaultPipelineCannotDelete: '預設流程線不可刪除',
+      '您確定要刪除這個流水線嗎？已綁定此流水線的機器人將無法使用。',
+    defaultPipelineCannotDelete: '預設流水線不可刪除',
     deleteSuccess: '刪除成功',
     deleteError: '刪除失敗：',
     copyConfirmTitle: '確認複製',
     copyConfirmation:
-      '確定要複製這個流程線嗎？複製將創建一個包含完整配置的新流程線。',
+      '確定要複製這個流水線嗎？複製將創建一個包含完整配置的新流水線。',
     unsavedChanges: '有未儲存的變更',
     dangerZone: '危險區域',
     dangerZoneDescription: '不可逆的操作',
-    deletePipelineAction: '刪除此流程線',
-    deletePipelineHint: '刪除後，綁定此流程線的機器人將無法正常運作。',
-    copyPipelineAction: '複製此流程線',
-    copyPipelineHint: '建立一條新的流程線，並複製所有設定。',
+    deletePipelineAction: '刪除此流水線',
+    deletePipelineHint: '刪除後，綁定此流水線的機器人將無法正常運作。',
+    copyPipelineAction: '複製此流水線',
+    copyPipelineHint: '建立一條新的流水線，並複製所有設定。',
     extensions: {
       title: '擴展集成',
       loadError: '載入插件清單失敗',
@@ -1839,7 +2112,7 @@ const zhHant = {
       selectSkills: '選擇技能',
       noSkillsAvailable: '暫無可用技能',
       mcpServersScopeTooltip:
-        '這裡僅控制此流程線綁定哪些 MCP 伺服器；具體 MCP 工具和資源在 AI 能力的內建 Agent 表單中選擇。',
+        '這裡僅控制此流水線綁定哪些 MCP 伺服器；具體 MCP 工具和資源在 AI 能力的內建 Agent 表單中選擇。',
       enableAllMCPServersTooltip:
         '啟用後，所有已配置且啟用的 MCP 伺服器都會進入 AI 能力中的 MCP 工具和資源候選範圍。',
     },
@@ -1874,8 +2147,8 @@ const zhHant = {
       selectKnowledgeBases: '選擇知識庫',
     },
     debugDialog: {
-      title: '流程線對話',
-      selectPipeline: '選擇流程線',
+      title: '流水線對話',
+      selectPipeline: '選擇流水線',
       sessionType: '對話類型',
       privateChat: '私聊',
       groupChat: '群聊',
@@ -1890,7 +2163,7 @@ const zhHant = {
       resetSuccess: '對話已重設',
       resetFailed: '重設失敗',
       loadMessagesFailed: '載入訊息失敗',
-      loadPipelinesFailed: '載入流程線失敗',
+      loadPipelinesFailed: '載入流水線失敗',
       atTips: '提及機器人',
       streaming: '串流傳輸',
       streamOutput: '串流',
@@ -1911,9 +2184,17 @@ const zhHant = {
       uploading: '上傳中...',
     },
     monitoring: {
+      rangeDays: '最近 {{days}} 天',
+      customRange: '自訂時間範圍',
+      rangeStart: '開始時間',
+      rangeEnd: '結束時間',
+      rangeHint: '開始時間須早於結束時間，最多 365 天。',
+      autoRefreshLabel: '定時重新整理',
+      autoRefreshOff: '定時重新整理：關閉',
+      autoRefreshEvery: '每 {{seconds}} 秒重新整理',
       title: '執行日誌',
       workbench: '設定與除錯',
-      description: '檢視此流程線的執行記錄和錯誤資訊（最近24小時）',
+      description: '檢視此流水線的執行記錄和錯誤資訊（最近24小時）',
       detailedLogs: '詳細日誌',
     },
   },
@@ -2172,7 +2453,8 @@ const zhHant = {
     totpRecoveryCodesRegenerated: '已產生新的恢復代碼',
     totpStatusDisabled: '未啟用',
     totpCodesRemaining: '剩餘 {{count}} 個恢復代碼',
-    totpManagerSectionDesc: '擁有者與管理員可以檢視並重設任意帳號的兩步驗證。',
+    totpManagerSectionDesc:
+      '擁有者與管理員可以檢視並重設本工作區帳號的兩步驗證。',
     revokeTotp: '重新綁定',
     totpAdminResetTitle: '重新綁定 {{user}} 的兩步驗證',
     totpAdminResetDesc:
@@ -2233,8 +2515,7 @@ const zhHant = {
     upgradePlan: 'Change or upgrade plan',
     ossSingletonDescription:
       'This self-hosted instance has one Workspace and can include multiple users.',
-    cloudManagedDescription:
-      'This Workspace is hosted by LangBot Cloud. Manage members here; billing opens in Cloud.',
+    cloudManagedDescription: 'This Workspace is hosted by LangBot Cloud.',
     loadFailed: 'Failed to load Workspace information',
     members: 'Members',
     you: 'You',
@@ -2303,6 +2584,122 @@ const zhHant = {
     },
   },
   monitoring: {
+    rangeDays: '最近 {{days}} 天',
+    customRange: '自訂時間範圍',
+    rangeStart: '開始時間',
+    rangeEnd: '結束時間',
+    rangeHint: '開始時間須早於結束時間，最多 365 天。',
+    autoRefreshLabel: '定時重新整理',
+    autoRefreshOff: '定時重新整理：關閉',
+    autoRefreshEvery: '每 {{seconds}} 秒重新整理',
+    unified: {
+      inputs: '事件與輸入',
+      outputs: '生成結果',
+      deliveries: '傳送記錄',
+      conversation: '對話上下文',
+      related: '關聯執行',
+      events: '執行事件',
+      llm_calls: '模型呼叫',
+      tool_calls: '工具呼叫',
+      errors: '錯誤',
+      payload: '查看詳情',
+      metadata: '技術資訊',
+      loadMore: '載入更多',
+      noRecords: '暫無記錄',
+      contextHint: '對話中的其他訊息不一定屬於本次執行。',
+      unhandled: '未啟動處理器',
+      roles: {
+        user: '使用者',
+        assistant: '助手',
+        system: '系統',
+        tool: '工具',
+      },
+    },
+    execution: {
+      title: '執行記錄',
+      subtitle: '每條入站事件及其觸發的執行，涵蓋 Agent 與 流水線。',
+      tabs: { executions: '執行記錄' },
+      filters: {
+        source: '類型',
+        sourceAll: '全部類型',
+        sourceAgent: 'Agent',
+        sourcePipeline: '流水線',
+        mode: '流量',
+        modeAll: '真實 + 偵錯',
+        modeLive: '僅真實',
+        modeDebug: '僅偵錯',
+        status: '狀態',
+        statusAll: '全部狀態',
+      },
+      status: {
+        completed: '已完成',
+        failed: '失敗',
+        running: '執行中',
+        queued: '排隊中',
+        cancelled: '已取消',
+        ignored: '已忽略',
+      },
+      // prettier-ignore
+      kind: {
+        agent: 'Agent',
+        pipeline: '流水線',
+        processor: '處理器',
+      },
+      columns: {
+        time: '時間',
+        target: '處理器',
+        status: '狀態',
+        duration: '耗時',
+        tokens: 'Token',
+      },
+      debug: '偵錯',
+      empty: '此時間範圍內暫無執行記錄',
+      triggeredBy: '觸發：{{trigger}}',
+      loading: '載入執行記錄中...',
+      cards: {
+        latencyHint:
+          'P50/P95：約 50%/95% 的執行能在此時間內完成。僅統計最近已完成的 Agent 執行，不含排隊時間。',
+        tokensHint:
+          '僅加總已記錄的用量；部分模型未回傳用量時，總數可能不完整。',
+        totalExecutions: '執行總數',
+        bySource: 'Agent {{agent}} · 流水線 {{pipeline}}',
+        successRate: '執行成功率',
+        successRateHint:
+          '成功次數 ÷（成功 + 失敗）。不計進行中、已忽略及已取消的執行。',
+        denominator: '{{completed}} / {{denominator}} 已結束',
+        running: '執行中',
+        queued: '排隊中',
+        waiting: '等待互動',
+        failed: '失敗',
+        ignored: '已忽略',
+        latency: '執行耗時',
+        p50: 'P50',
+        p95: 'P95',
+        sample: '樣本 {{count}} 次',
+        tokens: 'Token',
+        coverage: '{{withUsage}}/{{calls}} 次呼叫有用量',
+        debugRuns: '其中 {{count}} 筆為偵錯執行',
+      },
+      detail: {
+        title: '執行詳情',
+        timeline: '執行時間軸',
+        executionId: '執行 ID',
+        eventId: '事件',
+        queue: '佇列',
+        runner: 'Runner',
+        created: '建立時間',
+        duration: '耗時',
+        usage: '用量',
+        noEvents: '暫無追蹤事件',
+        query: '入站訊息',
+        modelCalls: '模型呼叫',
+        toolCalls: '工具呼叫',
+        errors: '錯誤',
+        loadError: '載入執行軌跡失敗',
+        openObject: '開啟物件',
+        a11yDescription: '此次執行的完整追蹤',
+      },
+    },
     title: '儀表盤',
     description: '監控機器人活動、LLM調用和系統效能',
     overview: '概覽',
@@ -2318,6 +2715,9 @@ const zhHant = {
       pipeline: '流水線',
       allBots: '全部機器人',
       selectBot: '選擇機器人',
+      processor: '處理器',
+      allProcessors: '全部處理器',
+      selectProcessor: '選擇處理器',
       allPipelines: '全部流水線',
       selectPipeline: '選擇流水線',
       loading: '載入中...',
@@ -2343,6 +2743,15 @@ const zhHant = {
       sessions: '會話分析',
       errors: '錯誤日誌',
     },
+    pipelineTrace: {
+      received: '收到訊息',
+      processing: '處理過程',
+      replies: '回覆訊息',
+      noSteps: '未記錄模型或工具呼叫',
+      success: '成功',
+      error: '失敗',
+      pending: '處理中',
+    },
     messageList: {
       timestamp: '時間戳記',
       bot: '機器人',
@@ -2562,7 +2971,7 @@ const zhHant = {
     title: '設定',
     nav: {
       workspace: 'Workspace',
-      operationTrace: '操作溯源',
+      operationTrace: '操作日誌',
       models: '模型',
       api: 'API',
       storage: '儲存',
@@ -2570,21 +2979,21 @@ const zhHant = {
     },
   },
   operationTrace: {
-    title: '操作溯源',
+    title: '操作日誌',
     description: '追蹤管理員與擁有者的修改與查看操作，記錄「什麼改成了什麼」。',
+    reset: '重設',
+    settingsUpdated: '設定已儲存',
+    settingsUpdateFailed: '儲存設定失敗',
+    levelBadge: '等級：{{level}}',
+    collectionSettings: '收集設定',
     captureLevel: '操作等級',
     retention: '保留策略',
-    retentionSummary: '保留 {{days}} 天 · 最多 {{rows}} 條',
     retentionDays: '保留天數',
     maxRows: '最大記錄數',
     dedupeWindow: '去重視窗（秒）',
     records: '操作記錄',
     changesCount: '{{count}} 項變更',
     hideDetails: '收合詳情',
-    levelUpdated: '操作等級已更新',
-    levelUpdateFailed: '操作等級更新失敗',
-    retentionUpdated: '保留策略已更新',
-    retentionUpdateFailed: '保留策略更新失敗',
     export: '下載日誌',
     exportFailed: '下載操作日誌失敗',
     empty: '暫無操作記錄',
@@ -2598,7 +3007,7 @@ const zhHant = {
     filterAllResources: '全部資源',
     filterActor: '操作者',
     filterAllActors: '全部操作者',
-    loadFailed: '載入操作溯源資料失敗',
+    loadFailed: '載入操作日誌資料失敗',
     refresh: '重新整理',
     save: '儲存',
     redacted: '已遮罩',
@@ -2617,7 +3026,7 @@ const zhHant = {
     elidedValue: '{{count}} 字元',
     levels: {
       off: {
-        label: '關閉溯源',
+        label: '關閉日誌',
       },
       mutation: {
         label: '僅修改操作',
@@ -2636,7 +3045,7 @@ const zhHant = {
       adapter: '適配器',
       model_provider: '模型提供商',
       llm_model: '模型',
-      pipeline: '管道',
+      pipeline: '流水線',
       user: '使用者',
       workspace: '工作區',
       monitoring: '監控',
@@ -2665,8 +3074,8 @@ const zhHant = {
     },
     actions: {
       audit_log_view: '查看操作日誌',
-      settings_update: '修改溯源設定',
-      settings_view: '查看溯源設定',
+      settings_update: '修改日誌設定',
+      settings_view: '查看日誌設定',
       member_invite: '邀請成員',
       member_role_update: '修改成員角色',
       member_remove: '移除成員',
@@ -2694,7 +3103,7 @@ const zhHant = {
       codex_view: '查看 Codex 授權',
       codex_authorize: '授權 Codex',
       embed: '嵌入式工作階段',
-      pipeline_extensions_update: '修改管道擴展綁定',
+      pipeline_extensions_update: '修改流水線擴展綁定',
       export: '匯出資料',
       execute: '執行任務',
       debug: '偵錯任務',
diff --git a/web/tests/e2e/bot-session-tool-timeline.spec.ts b/web/tests/e2e/bot-session-tool-timeline.spec.ts
index 5f580f4a0..ae397007e 100644
--- a/web/tests/e2e/bot-session-tool-timeline.spec.ts
+++ b/web/tests/e2e/bot-session-tool-timeline.spec.ts
@@ -152,17 +152,20 @@ test.describe('bot session request recovery', () => {
             await page.getByRole('button', { name: /Recovery user 1/ }).click();
           else
             await page
-              .getByRole('button', { name: 'Next', exact: true })
-              .last()
+              .getByRole('button', { name: 'Next page', exact: true })
               .click();
         }
       }
-      await expect(page.getByRole('alert')).toBeVisible();
+      await expect(
+        page.getByRole('tabpanel', { name: /Sessions/ }).getByRole('alert'),
+      ).toBeVisible();
       await expect(
         page.getByText('No sessions found', { exact: true }),
       ).toHaveCount(0);
       if (failure === 'analysis') {
-        await expect(page.getByRole('alert')).toContainText(/Tool/i);
+        await expect(
+          page.getByRole('tabpanel', { name: /Sessions/ }).getByRole('alert'),
+        ).toContainText(/Tool/i);
         await expect(
           page.getByText('Successful message', { exact: true }),
         ).toBeVisible();
@@ -177,10 +180,13 @@ test.describe('bot session request recovery', () => {
         ).toHaveCount(0);
       failing = false;
       await page
+        .getByRole('tabpanel', { name: /Sessions/ })
         .getByRole('alert')
         .getByRole('button', { name: 'Retry', exact: true })
         .click();
-      await expect(page.getByRole('alert')).toHaveCount(0);
+      await expect(
+        page.getByRole('tabpanel', { name: /Sessions/ }).getByRole('alert'),
+      ).toHaveCount(0);
       if (failure === 'initial list' || failure === 'list page') {
         await expect(
           page.getByRole('button', {
@@ -313,7 +319,9 @@ test.describe('bot session request races', () => {
         await expect.poll(() => released).toBe(true);
         // Allow the released HTTP response and React's queued update to settle.
         await page.waitForTimeout(200);
-        await expect(page.getByRole('alert')).toHaveCount(0);
+        await expect(
+          page.getByRole('tabpanel', { name: /Sessions/ }).getByRole('alert'),
+        ).toHaveCount(0);
         await expect(page.getByText('Stale list', { exact: true })).toHaveCount(
           0,
         );
diff --git a/web/tests/e2e/codex-subscription.spec.ts b/web/tests/e2e/codex-subscription.spec.ts
index 5a67867f9..6a8ad70c4 100644
--- a/web/tests/e2e/codex-subscription.spec.ts
+++ b/web/tests/e2e/codex-subscription.spec.ts
@@ -196,7 +196,7 @@ for (const width of [1280, 390, 320]) {
     await page.getByRole('button', { name: 'Done', exact: true }).click();
     await expect(page.getByText('My Codex', { exact: true })).toBeVisible();
     await expect(
-      page.getByRole('button', { name: 'Add Model', exact: true }),
+      page.getByRole('button', { name: 'Add', exact: true }),
     ).toBeVisible();
     expect(state.creates).toBe(1);
     expect(
@@ -309,7 +309,7 @@ test('model test retains the connected provider identity', async ({ page }) => {
   });
   await page.goto('/home/bots');
   await page.getByRole('button', { name: 'Models', exact: true }).click();
-  await page.getByRole('button', { name: 'Add Model', exact: true }).click();
+  await page.getByRole('button', { name: 'Add', exact: true }).click();
   await page
     .getByPlaceholder('Model Name', { exact: true })
     .fill('fixture-codex-model');
diff --git a/web/tests/e2e/contextual-guides.spec.ts b/web/tests/e2e/contextual-guides.spec.ts
index ad6d3c808..b4da33b88 100644
--- a/web/tests/e2e/contextual-guides.spec.ts
+++ b/web/tests/e2e/contextual-guides.spec.ts
@@ -203,7 +203,8 @@ test('bot guide starts on the detail page after adapter-only creation', async ({
   const adapterCard = page.locator('[data-slot="card"]').filter({
     has: page.getByText('Adapter Configuration', { exact: true }),
   });
-  await expect(adapterCard.getByRole('switch')).toBeChecked();
+  await expect(adapterCard.getByRole('switch')).toHaveCount(0);
+  await expect(page.getByRole('radio', { name: /^Webhook/ })).toBeChecked();
   await guide.getByRole('button', { name: 'Next' }).click();
 
   await expect(
@@ -318,7 +319,7 @@ test('bot adapter gallery keeps categories and fits desktop and mobile', async (
     )
     .toBe(true);
 
-  await gallery.getByRole('button', { name: /Legacy adapters/ }).click();
+  await gallery.getByRole('button', { name: /^Legacy adapters \d+$/ }).click();
   await expect(
     gallery.getByRole('button', { name: /Legacy Adapter/ }),
   ).toBeVisible();
@@ -641,11 +642,11 @@ for (const available of [true, false]) {
       guide.getByRole('heading', { name: 'Review execution results' }),
     ).toBeVisible();
     await expect(
-      page.locator('[data-guide="event-processor-form-tab-logs"]'),
+      page.getByRole('tab', { name: 'Run logs', exact: true }),
     ).toHaveAttribute('data-state', 'active');
     await guide.getByRole('button', { name: 'Previous', exact: true }).click();
     await expect(
-      page.locator('[data-guide="event-processor-form-tab-config"]'),
+      page.getByRole('tab', { name: 'Configure & debug', exact: true }),
     ).toHaveAttribute('data-state', 'active');
     await guide.getByRole('button', { name: 'Next', exact: true }).click();
     if (available) {
@@ -655,7 +656,7 @@ for (const available of [true, false]) {
       ).toBeVisible();
       await expect(guide).toContainText('no per-event routing needed');
       await expect(
-        page.locator('[data-guide="event-processor-form-tab-config"]'),
+        page.getByRole('tab', { name: 'Configure & debug', exact: true }),
       ).toHaveAttribute('data-state', 'active');
     }
     await guide.getByRole('button', { name: 'Finish', exact: true }).click();
diff --git a/web/tests/e2e/crud-smoke.spec.ts b/web/tests/e2e/crud-smoke.spec.ts
index 5fa7f155f..ba0d3b176 100644
--- a/web/tests/e2e/crud-smoke.spec.ts
+++ b/web/tests/e2e/crud-smoke.spec.ts
@@ -6,7 +6,7 @@ import {
 } from './fixtures/langbot-api';
 
 async function save(page: Page) {
-  const button = page.getByRole('button', { name: /^Save$/ });
+  const button = page.getByRole('button', { name: /^Save(?: and enable)?$/ });
   await expect(button).toBeEnabled();
   await button.click();
 }
@@ -127,10 +127,10 @@ test.describe('frontend CRUD smoke flows', () => {
 
     await page.goto('/home/agents?id=new');
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
     await page.locator('input[name="name"]').fill('Viewer Pipeline');
-    await page
-      .locator('input[name="description"]')
-      .fill('Viewer monitoring permission regression.');
+
     await submit(page);
     await expect(page).toHaveURL(/\/home\/agents\?id=pipeline-1$/);
 
@@ -218,6 +218,7 @@ test.describe('frontend CRUD smoke flows', () => {
       'data-slot',
       'toggle-group-item',
     );
+    await agentTypeCard.click();
     await expect(agentTypeCard).toHaveAttribute('aria-checked', 'true');
     await expect(page.getByTestId('agent-diagram')).toBeVisible();
     await expect(
@@ -238,6 +239,8 @@ test.describe('frontend CRUD smoke flows', () => {
       '0.5',
     );
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
     await expect(pipelineTypeCard).toHaveAttribute('aria-checked', 'true');
     await expect(page.getByTestId('pipeline-diagram')).toBeVisible();
     await expect(
@@ -246,9 +249,7 @@ test.describe('frontend CRUD smoke flows', () => {
     await expect(page.getByTestId('agent-diagram')).toHaveCount(0);
     await expect(page.locator('input[name="name"]')).toBeVisible();
     await page.locator('input[name="name"]').fill('Escalation Pipeline');
-    await page
-      .locator('input[name="description"]')
-      .fill('Routes urgent customer issues.');
+
     await submit(page);
 
     await expect(page).toHaveURL(/\/home\/agents\?id=pipeline-1$/);
@@ -550,10 +551,13 @@ test.describe('bot advanced flows', () => {
     );
     await page.keyboard.press('Escape');
 
-    await page.getByRole('button', { name: 'Refresh status' }).hover();
+    await page
+      .getByRole('button', { name: 'Match records', exact: true })
+      .click();
     await expect(
       page.getByText('Failed to refresh route status.'),
     ).toBeVisible();
+    await page.keyboard.press('Escape');
 
     await page.getByRole('button', { name: 'Check route' }).click();
     const routeDialog = page.getByRole('dialog');
@@ -612,7 +616,7 @@ test.describe('bot advanced flows', () => {
     });
     const listenBox = await listenButton.boundingBox();
     const saveBox = await page
-      .getByRole('button', { name: /^Save$/ })
+      .getByRole('button', { name: /^Save(?: and enable)?$/ })
       .boundingBox();
     expect(listenBox).not.toBeNull();
     expect(saveBox).not.toBeNull();
@@ -741,6 +745,7 @@ test.describe('bot advanced flows', () => {
 
     // Select an adapter but leave the required name empty.
     await selectPlaywrightAdapter(page);
+    await page.locator('input[name="name"]').fill('');
     await submit(page);
 
     await expect(page.getByText(/cannot be empty/i)).toBeVisible();
@@ -790,6 +795,8 @@ test.describe('pipeline advanced flows', () => {
     // Create a pipeline
     await page.goto('/home/agents?id=new');
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
     await page.locator('input[name="name"]').fill('Tab Test Pipeline');
     await submit(page);
 
@@ -820,6 +827,8 @@ test.describe('pipeline advanced flows', () => {
     // Create a pipeline
     await page.goto('/home/agents?id=new');
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
     await page.locator('input[name="name"]').fill('Dirty Form Pipeline');
     await submit(page);
 
@@ -842,8 +851,11 @@ test.describe('pipeline advanced flows', () => {
 
     await page.goto('/home/agents?id=new');
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
 
-    // Submit without filling name
+    // Clear the name suggested by the selected runner.
+    await page.locator('input[name="name"]').fill('');
     await submit(page);
 
     // Should show validation error for name (zod validation)
@@ -1415,6 +1427,8 @@ test.describe('cross-resource flows', () => {
     // Create a pipeline first
     await page.goto('/home/agents?id=new');
     await page.getByRole('radio', { name: /^Pipeline/ }).click();
+    await page.locator('#agent-create-form').getByRole('combobox').click();
+    await page.getByRole('option', { name: /Local Agent/ }).click();
     await page.locator('input[name="name"]').fill('Production Pipeline');
     await submit(page);
     await expect(page).toHaveURL(/\/home\/agents\?id=pipeline-1$/);
diff --git a/web/tests/e2e/dashboard-mobile.spec.ts b/web/tests/e2e/dashboard-mobile.spec.ts
new file mode 100644
index 000000000..55b7371cb
--- /dev/null
+++ b/web/tests/e2e/dashboard-mobile.spec.ts
@@ -0,0 +1,28 @@
+import { expect, test } from '@playwright/test';
+import { installLangBotApiMocks } from './fixtures/langbot-api';
+
+for (const width of [375, 430]) {
+  test(`dashboard filters fit mobile ${width}`, async ({ page }) => {
+    await page.setViewportSize({ width, height: 812 });
+    await installLangBotApiMocks(page, { authenticated: true });
+    await page.goto('/home/monitoring');
+    const bot = page.getByRole('combobox', { name: 'Bot', exact: true });
+    await expect(bot).toBeVisible();
+    const grid = bot.locator('..').locator('..');
+    expect(
+      await grid.evaluate(
+        (el) => getComputedStyle(el).gridTemplateColumns.split(' ').length,
+      ),
+    ).toBe(2);
+    const box = await grid.boundingBox();
+    expect(box!.x).toBeGreaterThanOrEqual(0);
+    expect(box!.x + box!.width).toBeLessThanOrEqual(width);
+    await expect(
+      page.getByRole('button', { name: /Reset Filters/i }),
+    ).toBeVisible();
+    await page.screenshot({
+      path: `test-results/dashboard-${width}.png`,
+      fullPage: true,
+    });
+  });
+}
diff --git a/web/tests/e2e/entity-loading.spec.ts b/web/tests/e2e/entity-loading.spec.ts
index f3e681200..27a7f190e 100644
--- a/web/tests/e2e/entity-loading.spec.ts
+++ b/web/tests/e2e/entity-loading.spec.ts
@@ -240,7 +240,7 @@ for (const scenario of cases) {
     await page.goto(scenario.url);
     const error = page
       .getByRole('alert')
-      .filter({ hasText: 'Failed to load. Please try again.' });
+      .filter({ hasText: 'Unable to load this content' });
     await expect(error).toBeVisible();
     fail = false;
     await error.getByRole('button', { name: 'Retry', exact: true }).click();
diff --git a/web/tests/e2e/event-processor-marketplace.spec.ts b/web/tests/e2e/event-processor-marketplace.spec.ts
new file mode 100644
index 000000000..cbf73e369
--- /dev/null
+++ b/web/tests/e2e/event-processor-marketplace.spec.ts
@@ -0,0 +1,179 @@
+import { test, expect } from '@playwright/test';
+import { installLangBotApiMocks } from './fixtures/langbot-api';
+
+for (const scenario of ['install', 'retry', 'reload'] as const) {
+  test(`event processor marketplace ${scenario}: filters usage and selects registered component`, async ({
+    page,
+  }) => {
+    await installLangBotApiMocks(page, { authenticated: true });
+    const ref = 'plugin:qa/events/default';
+    let installed = false;
+    let finish = scenario !== 'reload';
+    let installs = 0;
+    let registrations = 0;
+    const requests: Record<string, unknown>[] = [];
+    const descriptor = {
+      id: ref,
+      plugin_author: 'qa',
+      plugin_name: 'events',
+      label: { en_US: 'Event handler' },
+      supported_event_patterns: ['group.member_joined'],
+      config_schema: [
+        {
+          name: 'greeting',
+          type: 'string',
+          label: { en_US: 'Greeting' },
+          default: 'Welcome',
+          required: true,
+        },
+      ],
+    };
+    let processor = {
+      uuid: 'event-market',
+      kind: 'event_processor',
+      name: 'Event marketplace',
+      component_ref: '',
+      config: {},
+      supported_event_patterns: [],
+    };
+    await page.route('**/api/v1/agents**', async (route) => {
+      const path = new URL(route.request().url()).pathname;
+      let data: unknown;
+      if (path.endsWith('/_/metadata')) {
+        if (installed) registrations++;
+        data = {
+          event_processors: installed && registrations > 1 ? [descriptor] : [],
+          kinds: [],
+          platform_tools: [],
+        };
+      } else if (path.endsWith('/runs')) data = { items: [], has_more: false };
+      else if (path.endsWith('/event-market')) {
+        if (route.request().method() === 'PUT')
+          processor = { ...processor, ...route.request().postDataJSON() };
+        data = { agent: processor };
+      } else return route.fallback();
+      await route.fulfill({ json: { code: 0, data } });
+    });
+    await page.route('**/api/v1/marketplace/**', async (route) => {
+      const path = new URL(route.request().url()).pathname;
+      if (!path.endsWith('/search')) return route.fallback();
+      requests.push(route.request().postDataJSON());
+      const plugin = (
+        name: string,
+        usages?: string[],
+        components = { Runner: 1 },
+      ) => ({
+        author: 'qa',
+        name,
+        label: { en_US: name },
+        description: { en_US: name + ' description' },
+        latest_version: '1.0.0',
+        type: 'plugin',
+        components,
+        runner_usages: usages,
+        install_count: 0,
+      });
+      await route.fulfill({
+        json: {
+          code: 0,
+          data: {
+            total: 5,
+            [path.includes('/extensions/') ? 'extensions' : 'plugins']: [
+              plugin('events', ['event']),
+              plugin('agent-only', ['agent']),
+              plugin('mixed', ['agent', 'event']),
+              plugin('unknown'),
+              plugin('knowledge-only', ['event'], {
+                KnowledgeEngine: 1,
+              } as any),
+            ],
+          },
+        },
+      });
+    });
+    await page.route('**/api/v1/plugins/install/marketplace', async (route) => {
+      installs++;
+      expect(route.request().postDataJSON()).toEqual({
+        plugin_author: 'qa',
+        plugin_name: 'events',
+        plugin_version: '1.0.0',
+      });
+      await route.fulfill({ json: { code: 0, data: { task_id: installs } } });
+    });
+    await page.route('**/api/v1/system/tasks/*', async (route) => {
+      const failed = scenario === 'retry' && installs === 1;
+      installed = finish && !failed;
+      await route.fulfill({
+        json: {
+          code: 0,
+          data: {
+            id: installs,
+            name: 'plugin-install-marketplace',
+            runtime: {
+              done: finish,
+              exception: failed ? 'Download failed' : null,
+            },
+            task_context: {
+              current_action: 'Downloading',
+              metadata: { progress_percent: 25 },
+            },
+          },
+        },
+      });
+    });
+    await page.goto('/home/agents?id=event-market');
+    const select = page.getByRole('combobox', {
+      name: 'Plugin processor',
+      includeHidden: true,
+    });
+    await select.click();
+    const install = page.getByRole('button', {
+      name: 'Install events',
+      exact: true,
+    });
+    await expect(install).toBeVisible();
+    await expect(
+      page.getByRole('button', { name: 'Install mixed', exact: true }),
+    ).toBeVisible();
+    for (const excluded of ['agent-only', 'unknown', 'knowledge-only'])
+      await expect(
+        page.getByRole('button', { name: `Install ${excluded}`, exact: true }),
+      ).toHaveCount(0);
+    expect(requests.length).toBeGreaterThan(0);
+    for (const request of requests)
+      expect(request).toMatchObject({
+        component_filter: 'Runner',
+        runner_usage: 'event',
+        type_filter: 'plugin',
+      });
+    await install.click();
+    if (scenario === 'retry') {
+      await expect(
+        page
+          .getByRole('alert', { includeHidden: true })
+          .filter({ hasText: 'Download failed' }),
+      ).toBeVisible();
+      await install.click();
+    }
+    if (scenario === 'reload') {
+      await expect.poll(() => installs).toBe(1);
+      await expect
+        .poll(() =>
+          page.evaluate(() =>
+            Object.keys(sessionStorage).some((key) =>
+              key.includes('event-processor:event-market'),
+            ),
+          ),
+        )
+        .toBe(true);
+      finish = true;
+      await page.reload();
+    }
+    await expect(select).toContainText('Event handler', { timeout: 20000 });
+    await page.keyboard.press('Escape');
+    await expect(page.locator('input[name="greeting"]')).toHaveValue('Welcome');
+    expect(installs).toBe(scenario === 'retry' ? 2 : 1);
+    await page.getByRole('button', { name: 'Save', exact: true }).click();
+    await expect.poll(() => processor.component_ref).toBe(ref);
+  });
+}
diff --git a/web/tests/e2e/event-processor.spec.ts b/web/tests/e2e/event-processor.spec.ts
index 9a4f6d1f1..6e095d274 100644
--- a/web/tests/e2e/event-processor.spec.ts
+++ b/web/tests/e2e/event-processor.spec.ts
@@ -1,7 +1,7 @@
 import { expect, test } from '@playwright/test';
 import { installLangBotApiMocks } from './fixtures/langbot-api';
 
-test('create first, select a plugin in the header, debug beside scrollable logs', async ({
+test('select a plugin before creation and retain debug state while browsing run logs', async ({
   page,
 }) => {
   await installLangBotApiMocks(page, { authenticated: true });
@@ -45,11 +45,7 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
     ) {
       const payload = route.request().postDataJSON();
       creations.push(payload);
-      Object.assign(processor, payload, {
-        component_ref: null,
-        config: {},
-        supported_event_patterns: [],
-      });
+      Object.assign(processor, payload);
       await route.fulfill({
         json: { code: 0, data: { uuid: processor.uuid, kind: processor.kind } },
       });
@@ -222,6 +218,8 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
   await page
     .getByRole('textbox', { name: 'Name', exact: false })
     .fill('Welcome processor');
+  await page.locator('#agent-create-form').getByRole('combobox').click();
+  await page.getByRole('option').filter({ hasText: 'Welcome' }).click();
   await page.getByRole('button', { name: 'Submit', exact: true }).click();
   await expect(page).toHaveURL(/id=processor-qa/);
   expect(creations).toHaveLength(1);
@@ -229,35 +227,33 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
     kind: 'event_processor',
     name: 'Welcome processor',
   });
-  expect(creations[0]).not.toHaveProperty('component_ref');
+  expect(creations[0]).toHaveProperty('component_ref', ref);
   expect(creations[0]).not.toHaveProperty('config');
   const panel = page.getByRole('region', { name: 'Event Debug' });
-  const logs = page.getByRole('region', { name: 'Plugin processor' });
+  const settings = page.getByRole('region', { name: 'Plugin processor' });
+  const logs = page.getByRole('region', { name: 'Run logs', exact: true });
+  const configTab = page.getByRole('tab', {
+    name: /^Configure & debug/,
+  });
+  const logsTab = page.getByRole('tab', { name: 'Run logs', exact: true });
   await expect(panel).toBeVisible();
-  await expect(logs).toBeVisible();
-  await expect(logs.getByRole('tab')).toHaveCount(2);
-  await expect(
-    logs.getByRole('tab', { name: 'Configuration', exact: true }),
-  ).toHaveAttribute('data-state', 'active');
+  await expect(settings).toBeVisible();
+  await expect(configTab).toHaveAttribute('data-state', 'active');
   await expect(page.getByRole('dialog')).toHaveCount(0);
-  await expect(panel.getByRole('button', { name: 'Run test' })).toHaveCount(0);
+  await expect(panel.getByRole('button', { name: 'Run test' })).toBeVisible();
   await page.getByRole('combobox', { name: 'Plugin processor' }).click();
   await page.getByRole('option').filter({ hasText: 'Welcome' }).click();
   await expect(
     panel.getByRole('combobox', { name: 'Event type' }),
   ).toContainText('group.member_joined');
-  const settings = logs.getByRole('tabpanel', {
-    name: 'Configuration',
-    exact: true,
-  });
   await expect(settings.getByText('Greeting *', { exact: true })).toBeVisible();
   await expect(
     page.getByRole('button', { name: 'Plugin settings', exact: true }),
   ).toHaveCount(0);
   await settings.getByRole('textbox').fill('Welcome');
-  await logs.getByRole('tab', { name: 'Logs', exact: true }).click();
+  await logsTab.click();
   await expect(settings).toBeHidden();
-  await logs.getByRole('tab', { name: 'Configuration', exact: true }).click();
+  await configTab.click();
   await expect(settings.getByRole('textbox')).toHaveValue('Welcome');
   await panel
     .getByRole('textbox', { name: 'Member ID' })
@@ -265,9 +261,9 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
   await panel.getByRole('button', { name: 'Save and run' }).click();
   await expect(panel.getByText('Debug handler invoked once')).toBeVisible();
   expect(operations).toEqual(['save', 'debug']);
-  await expect(
-    logs.getByRole('tab', { name: 'Logs', exact: true }),
-  ).toHaveAttribute('data-state', 'active');
+  await expect(configTab).toHaveAttribute('data-state', 'active');
+  await logsTab.click();
+  await expect(logs).toBeVisible();
   expect(debugRequests).toHaveLength(1);
   expect(debugRequests[0]).toMatchObject({
     event_type: 'group.member_joined',
@@ -298,10 +294,8 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
   await logs.getByText('Final log after pagination').scrollIntoViewIfNeeded();
   await expect(logs.getByText('Final log after pagination')).toBeInViewport();
   expect(cursors).toContain('100');
+  await configTab.click();
   await expect(panel.getByText('Debug handler invoked once')).toBeVisible();
-  const debugBox = await panel.boundingBox();
-  const logBox = await logs.boundingBox();
-  expect(debugBox!.x).toBeLessThan(logBox!.x);
   await page.setViewportSize({ width: 390, height: 700 });
   await panel
     .getByRole('button', { name: 'Run test' })
@@ -309,6 +303,7 @@ test('create first, select a plugin in the header, debug beside scrollable logs'
   await expect(
     panel.getByRole('button', { name: 'Run test' }),
   ).toBeInViewport();
+  await logsTab.click();
   await logs.getByText('Final log after pagination').scrollIntoViewIfNeeded();
   await expect(logs.getByText('Final log after pagination')).toBeInViewport();
 });
diff --git a/web/tests/e2e/fixtures/langbot-api.ts b/web/tests/e2e/fixtures/langbot-api.ts
index 254f8178a..f73ce5675 100644
--- a/web/tests/e2e/fixtures/langbot-api.ts
+++ b/web/tests/e2e/fixtures/langbot-api.ts
@@ -98,6 +98,7 @@ export interface WorkspaceEntryMock {
 
 interface LangBotApiMockState {
   authenticated: boolean;
+  withAssistant: boolean;
   bots: BotMock[];
   counters: Record<string, number>;
   knowledgeBases: KnowledgeBaseMock[];
@@ -227,6 +228,127 @@ function emptyMonitoringData() {
   };
 }
 
+type RawMonitoringMessage = Record<string, unknown>;
+
+/** Mirror of the backend execution status groups so fixtures stay realistic. */
+function executionStatusGroup(status: string, source: 'agent' | 'pipeline') {
+  if (source === 'agent') {
+    if (status === 'failed' || status === 'timeout') return 'failed';
+    if (status === 'created' || status === 'queued') return 'queued';
+    if (status === 'claimed' || status === 'running') return 'running';
+    if (status === 'cancelled') return 'cancelled';
+    return status;
+  }
+  if (status === 'success') return 'completed';
+  if (status === 'error') return 'failed';
+  if (status === 'pending') return 'running';
+  if (status === 'discarded' || status === 'not_matched') return 'ignored';
+  return status;
+}
+
+/** Best-effort text extraction from a serialized message chain (mirrors backend). */
+function executionText(raw: unknown): string {
+  const source = String(raw ?? '');
+  try {
+    const parts: string[] = [];
+    const walk = (node: unknown, depth = 0): void => {
+      if (depth > 6) return;
+      if (Array.isArray(node)) {
+        node.forEach((item) => walk(item, depth + 1));
+      } else if (node && typeof node === 'object') {
+        for (const [key, value] of Object.entries(node)) {
+          if (key === 'text' && typeof value === 'string') parts.push(value);
+          else walk(value, depth + 1);
+        }
+      }
+    };
+    walk(JSON.parse(source));
+    return parts.join(' ').trim() || source;
+  } catch {
+    return source;
+  }
+}
+
+/** Derive the unified execution list from a mocked monitoring payload. */
+export function buildExecutionList(monitoringData: unknown) {
+  const payload = monitoringData as
+    { messages?: RawMonitoringMessage[] } | undefined;
+  const items = (payload?.messages ?? [])
+    .filter((message) => message.role !== 'assistant')
+    .map((message) => {
+      const status = String(message.status ?? 'success');
+      return {
+        source: 'pipeline' as const,
+        id: String(message.id),
+        event_id: null,
+        status,
+        status_group: executionStatusGroup(status, 'pipeline'),
+        title: executionText(message.message_content),
+        input_preview: executionText(message.message_content),
+        target_kind: 'pipeline' as const,
+        target_id: (message.pipeline_id as string) ?? null,
+        target_name: (message.pipeline_name as string) ?? null,
+        bot_id: (message.bot_id as string) ?? null,
+        bot_name: (message.bot_name as string) ?? null,
+        pipeline_id: (message.pipeline_id as string) ?? null,
+        pipeline_name: (message.pipeline_name as string) ?? null,
+        runner_id: (message.runner_name as string) ?? null,
+        conversation_id: null,
+        session_id: (message.session_id as string) ?? null,
+        created_at_ms: Date.parse(String(message.timestamp ?? '')) || null,
+        started_at_ms: null,
+        finished_at_ms: null,
+        duration_ms: null,
+        usage: null,
+        cost: null,
+        queue_name: null,
+        debug: false,
+        has_error: status === 'error',
+      };
+    });
+
+  const byStatus: Record<string, number> = {};
+  for (const item of items) {
+    byStatus[item.status_group] = (byStatus[item.status_group] ?? 0) + 1;
+  }
+  const completed = byStatus.completed ?? 0;
+  const failed = byStatus.failed ?? 0;
+  const denominator = completed + failed;
+
+  return {
+    items,
+    total: items.length,
+    limit: items.length,
+    offset: 0,
+    has_more: false,
+    summary: {
+      executions: {
+        total: items.length,
+        by_source: { agent: 0, pipeline: items.length },
+        by_status: byStatus,
+        completed,
+        failed,
+        running: byStatus.running ?? 0,
+        queued: byStatus.queued ?? 0,
+        cancelled: byStatus.cancelled ?? 0,
+        ignored: byStatus.ignored ?? 0,
+        waiting: 0,
+        success_rate:
+          denominator > 0
+            ? Math.round((completed / denominator) * 10000) / 100
+            : null,
+        denominator,
+        p50_duration_ms: null,
+        p95_duration_ms: null,
+        duration_sample: 0,
+        debug: 0,
+        real: items.length,
+      },
+      tokens: { total_tokens: 0, calls: 0, calls_with_usage: 0 },
+    },
+  };
+}
+
 function emptyTokenStatistics() {
   return {
     summary: {
@@ -610,7 +732,9 @@ async function handleBackendApi(route: Route, state: LangBotApiMockState) {
 
   if (path === '/api/v1/user/info') {
     return fulfillJson(route, {
-      account_uuid: 'account-playwright',
+      // Most tests exercise an unlinked local user. Assistant tests opt into
+      // an account identity and mock its additional provider/conversation APIs.
+      account_uuid: state.withAssistant ? 'account-playwright' : null,
       user: 'admin@example.com',
       account_type: 'local',
       has_password: true,
@@ -625,6 +749,22 @@ async function handleBackendApi(route: Route, state: LangBotApiMockState) {
     return fulfillJson(route, { workspaces: state.workspaces });
   }
 
+  if (path === '/api/v1/workspaces/current/language') {
+    return fulfillJson(route, {
+      language: 'en_US',
+      supported_languages: [
+        'en_US',
+        'zh_Hans',
+        'zh_Hant',
+        'ja_JP',
+        'es_ES',
+        'ru_RU',
+        'th_TH',
+        'vi_VN',
+      ],
+    });
+  }
+
   if (path === '/api/v1/workspaces/current') {
     const selectedWorkspaceUuid = request.headers()['x-workspace-id'];
     const entry = state.workspaces.find(
@@ -1163,6 +1303,76 @@ async function handleBackendApi(route: Route, state: LangBotApiMockState) {
     return fulfillJson(route, data.overview || emptyMonitoringData().overview);
   }
 
+  if (path === '/api/v1/monitoring/executions') {
+    return fulfillJson(route, buildExecutionList(state.monitoringData));
+  }
+
+  const executionDetail = path.match(
+    /^\/api\/v1\/monitoring\/executions\/(agent|pipeline|event|auto)\/(.+)$/,
+  );
+  if (executionDetail) {
+    const [, , executionId] = executionDetail;
+    const payload = state.monitoringData as {
+      messages?: RawMonitoringMessage[];
+      llmCalls?: RawMonitoringMessage[];
+      toolCalls?: RawMonitoringMessage[];
+      errors?: RawMonitoringMessage[];
+    };
+    const row = buildExecutionList(payload).items.find(
+      (item) => item.id === executionId,
+    );
+    if (!row)
+      return route.fulfill({
+        status: 404,
+        json: { code: 404, msg: 'Execution not found' },
+      });
+    const root = (payload.messages ?? []).find(
+      (item) => item.id === executionId,
+    )!;
+    const content = (item: RawMonitoringMessage) => ({
+      ...item,
+      content: item.message_content,
+      timestamp_ms: Date.parse(String(item.timestamp)),
+      origin: item.role === 'assistant' ? 'delivery' : 'input',
+    });
+    const pages = Object.fromEntries(
+      Object.entries({
+        inputs: [content(root)],
+        outputs: [],
+        deliveries: (payload.messages ?? [])
+          .filter((item) => item.parent_message_id === executionId)
+          .map(content),
+        conversation: (payload.messages ?? [])
+          .filter(
+            (item) =>
+              item.session_id === root.session_id &&
+              item.bot_id === root.bot_id,
+          )
+          .map(content),
+        related: [],
+        events: [],
+        llm_calls: (payload.llmCalls ?? []).filter(
+          (item) => item.message_id === executionId,
+        ),
+        tool_calls: (payload.toolCalls ?? []).filter(
+          (item) => item.message_id === executionId,
+        ),
+        errors: (payload.errors ?? []).filter(
+          (item) => item.message_id === executionId,
+        ),
+      }).map(([key, items]) => [
+        key,
+        { items, next_offset: items.length, has_more: false },
+      ]),
+    );
+    return fulfillJson(route, {
+      source: row.source,
+      row,
+      pages,
+      legacy_context: true,
+    });
+  }
+
   if (path === '/api/v1/monitoring/token-statistics') {
     return fulfillJson(route, emptyTokenStatistics());
   }
@@ -1267,6 +1477,7 @@ export async function installLangBotApiMocks(
   page: Page,
   options: {
     authenticated?: boolean;
+    withAssistant?: boolean;
     language?: string;
     monitoringData?: unknown;
     monitoringSessions?: unknown[];
@@ -1280,6 +1491,7 @@ export async function installLangBotApiMocks(
 ) {
   const {
     authenticated = false,
+    withAssistant = false,
     language = 'en-US',
     monitoringData,
     monitoringSessions,
@@ -1292,6 +1504,7 @@ export async function installLangBotApiMocks(
   } = options;
   const state: LangBotApiMockState = {
     authenticated,
+    withAssistant,
     bots: [],
     counters: {},
     knowledgeBases: [],
diff --git a/web/tests/e2e/fixtures/runner-migration-contract.json b/web/tests/e2e/fixtures/runner-migration-contract.json
index 5acd29ce2..1c066a061 100644
--- a/web/tests/e2e/fixtures/runner-migration-contract.json
+++ b/web/tests/e2e/fixtures/runner-migration-contract.json
@@ -1,5 +1,5 @@
 {
-  "planner_version": "4",
+  "planner_version": "5",
   "fixture_only": true,
   "legacy_sections": {
     "local-agent": {
@@ -86,7 +86,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "LocalAgent",
-          "version": "0.1.10",
+          "version": "0.2.1",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/LocalAgent",
           "label": {
             "en_US": "Local Agent",
@@ -124,6 +124,8 @@
           "pages": []
         },
         "execution": {
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1",
           "python": {
             "path": "main.py",
             "attr": "LocalAgentPlugin"
@@ -152,8 +154,8 @@
             {
               "name": "model",
               "description": {
-                "en_US": "Primary/fallback model UUIDs and per-model reasoning levels. Host validates and applies reasoning for every request; this is per-pipeline configuration.",
-                "zh_Hans": "主模型/备用模型 UUID 与各模型推理等级。由 Host 对每次请求校验并应用，属于流水线独立配置。"
+                "en_US": "Choose a primary model and optional fallback models, and set the reasoning level for each model.",
+                "zh_Hans": "选择主模型和备用模型，并设置各模型的推理等级。"
               },
               "label": {
                 "en_US": "Model",
@@ -217,14 +219,9 @@
                 "zh_Hans": "选择沙箱文件和环境的共享方式。自定义模板使用 {变量名} 插值，如 {launcher_type}_{launcher_id}_{sender_id}；插值结果相同就复用同一个沙箱，{global} 表示在工作区内全局共享。可用变量：{launcher_type}（会话类型）、{launcher_id}（会话 ID）、{sender_id}（用户 ID）、{conversation_id}（对话上下文 ID）、{bot_id}（机器人 ID）、{query_id}（请求 ID）、{run_id}（运行 ID）、{event_id}（事件 ID）、{event_type}（事件类型），也可引用公开请求变量。仅支持变量名，不支持运算或属性访问；变量不可用时会报错。"
               },
               "type": "select",
+              "allow_custom": true,
               "required": true,
               "default": "{launcher_type}_{launcher_id}",
-              "show_if": {
-                "field": "box-enabled",
-                "operator": "eq",
-                "value": true
-              },
-              "allow_custom": true,
               "options": [
                 {
                   "name": "{global}",
@@ -291,7 +288,12 @@
                     "ru_RU": "По сообщению (изолированно)"
                   }
                 }
-              ]
+              ],
+              "show_if": {
+                "field": "box-enabled",
+                "operator": "eq",
+                "value": true
+              }
             },
             {
               "name": "advanced-settings",
@@ -578,8 +580,8 @@
           }
         }
       },
-      "package_sha256": "52780785960722cbc0f2a9335c48a852cf631a32e4d04ce1e1ff601af7c0b599",
-      "schema_sha256": "cdb336b8655b53dec482cfe9e191d0ed1d409929423584e5848255aabe499338"
+      "package_sha256": "2d2ce17055d70dd7736e6902595be3262992fd1598a00bd9d79d19fef1c16329",
+      "schema_sha256": "c1364cf0f211162bb701e27ed92b351fd9ff4cfa0b4469cfeb83ad58136cdf47"
     },
     "dify-agent": {
       "manifest": {
@@ -588,7 +590,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "DifyAgent",
-          "version": "0.1.10",
+          "version": "0.2.3",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/dify-agent",
           "label": {
             "en_US": "Dify Agent",
@@ -629,7 +631,9 @@
           "python": {
             "path": "main.py",
             "attr": "DifyAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -948,7 +952,7 @@
           }
         }
       },
-      "package_sha256": "6fc89a8d127b338b6f8d2deca78001b7dbaccfe9862c88c0d2ce95bd1df1ebb2",
+      "package_sha256": "6fb3c388e713449ff0db4c4491cf4bf89173b2a8de69d5c4c2e0c1fb21a3f4a8",
       "schema_sha256": "dccc4b7df7730c4e5e2cfc2ca2e67db5ea9195ad9720a79a329bc11fdcaeb840"
     },
     "coze-agent": {
@@ -958,7 +962,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "CozeAgent",
-          "version": "0.1.10",
+          "version": "0.2.0",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/coze-agent",
           "label": {
             "en_US": "Coze Agent",
@@ -999,7 +1003,9 @@
           "python": {
             "path": "main.py",
             "attr": "CozeAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -1259,7 +1265,7 @@
           }
         }
       },
-      "package_sha256": "c8e48659a0358bd8291137379f6159e61f2b1dbe72766772d81c9af49f7a8317",
+      "package_sha256": "d17917c8fcdbb5e044cdae7ba4c7c966835ade4350c97f9b38184a17de9d79b2",
       "schema_sha256": "3f7c4d4e5c8a7ace3e804f0ca02739c48439abcc8f360bfc17394b9cf6660674"
     },
     "dashscope-agent": {
@@ -1269,7 +1275,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "DashScopeAgent",
-          "version": "0.1.10",
+          "version": "0.2.2",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/dashscope-agent",
           "label": {
             "en_US": "Dashscope Agent",
@@ -1310,7 +1316,9 @@
           "python": {
             "path": "main.py",
             "attr": "DashScopeAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -1552,7 +1560,7 @@
           }
         }
       },
-      "package_sha256": "cd4e28e53eb986b906d6aa92d26b49d8ad4a9514fcc764f359bbdb84ad1a3f00",
+      "package_sha256": "200450d94da2173acd58239e4565c3d4a9cd4830a09173ff35bb24aa31bdd8a4",
       "schema_sha256": "aadfce3fe0157b124235437f251c24fbfea0e498ece4be8f9de6028b97573937"
     },
     "n8n-agent": {
@@ -1562,7 +1570,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "N8nAgent",
-          "version": "0.1.10",
+          "version": "0.2.0",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/n8n-agent",
           "label": {
             "en_US": "N8n Agent",
@@ -1603,7 +1611,9 @@
           "python": {
             "path": "main.py",
             "attr": "N8nAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -2028,7 +2038,7 @@
           }
         }
       },
-      "package_sha256": "507af9773690ff90c8ce2cdf8c3660ca01050e195eeb954cb631ea0116fe680c",
+      "package_sha256": "a83807decf3f62d2bd9a8c5515849c867140e22b3377ba248cb1e594757a5961",
       "schema_sha256": "9e492577b0e8d197fd792cd1456771e5ff1731ebf897a128c70f4850b4dfece0"
     },
     "langflow-agent": {
@@ -2038,7 +2048,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "LangflowAgent",
-          "version": "0.1.10",
+          "version": "0.2.2",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/langflow-agent",
           "label": {
             "en_US": "Langflow Agent",
@@ -2079,7 +2089,9 @@
           "python": {
             "path": "main.py",
             "attr": "LangflowAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -2306,7 +2318,7 @@
           }
         }
       },
-      "package_sha256": "93fee3064505d0a69d8b16d8aceb4a171ffdffed63a36278eef635267681f5a1",
+      "package_sha256": "f82975125ec6916fdf88876e93a3020d23728f96fec36c4980dd0670cfaa6608",
       "schema_sha256": "c7d940674bae9acab993801c6957da778bbdb3fa8734bfad9cda96dbb57f5fee"
     },
     "deerflow-agent": {
@@ -2316,7 +2328,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "DeerFlowAgent",
-          "version": "0.1.10",
+          "version": "0.2.1",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/deerflow-agent",
           "label": {
             "en_US": "DeerFlow Agent",
@@ -2357,7 +2369,9 @@
           "python": {
             "path": "main.py",
             "attr": "DeerFlowAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -2588,7 +2602,7 @@
           }
         }
       },
-      "package_sha256": "10c93fec92e2d34c791baa541b71debf060e80aaa5fb64c386dc25fbc6b510a5",
+      "package_sha256": "6413462042008652f4a0fd4ed0514c4943ceefbbdcea36c9a58a9cbf080168d9",
       "schema_sha256": "8a865320f37e367095a1f332bcc490cad8c0af273bc82baf768015242457c8f6"
     },
     "tbox-agent": {
@@ -2598,7 +2612,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "TboxAgent",
-          "version": "0.1.8",
+          "version": "0.2.1",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/tbox-agent",
           "label": {
             "en_US": "Tbox Agent",
@@ -2639,7 +2653,9 @@
           "python": {
             "path": "main.py",
             "attr": "TboxAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -2756,7 +2772,7 @@
           }
         }
       },
-      "package_sha256": "ed9c7a3f7c904a1fc6a95e8e7de0253f85282bf159250e355480fdcc91b4f629",
+      "package_sha256": "d88a3d14f2996383fab27a6fa1adf089f3bd575c2aa7a2c76c09f4eef682a9ff",
       "schema_sha256": "ac34c6a84859728328aa8d554aadbd1bef9b92c35ce4524812a541bac58c5088"
     },
     "weknora-agent": {
@@ -2766,7 +2782,7 @@
         "metadata": {
           "author": "langbot-team",
           "name": "WeKnoraAgent",
-          "version": "0.1.10",
+          "version": "0.2.1",
           "repository": "https://github.com/langbot-app/langbot-plugins/tree/main/Runner/weknora-agent",
           "label": {
             "en_US": "WeKnora Agent",
@@ -2807,7 +2823,9 @@
           "python": {
             "path": "main.py",
             "attr": "WeKnoraAgentPlugin"
-          }
+          },
+          "sharedRuntime": "shared-runtime-v1",
+          "componentModel": "stateless-v1"
         }
       },
       "component": {
@@ -3002,7 +3020,7 @@
           }
         }
       },
-      "package_sha256": "63841513ad9a0272f1e9863ab70a280989651a26db421b3f0df634972978135b",
+      "package_sha256": "41cb9bd32b3f794fad95a7d20cf8d9f52f81da1e1e4b3049f4b5f1c51581b980",
       "schema_sha256": "78d6d36d38f3991cc1a0bfc057d09a18a43c55e6af45b6ed61908d66bf475aa2"
     }
   },
@@ -3155,7 +3173,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "LocalAgent",
-          "version": "0.1.10"
+          "version": "0.2.1"
         },
         "config": {
           "trigger": {
@@ -3197,8 +3215,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/LocalAgent/default"
+              "id": "plugin:langbot-team/LocalAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/LocalAgent/default": {
@@ -3220,9 +3238,9 @@
                 "tools": [],
                 "knowledge-bases": [],
                 "model": {
+                  "primary": "synthetic-model",
                   "fallbacks": [],
-                  "reasoning": {},
-                  "primary": "synthetic-model"
+                  "reasoning": {}
                 },
                 "prompt": [
                   {
@@ -3256,6 +3274,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -3459,7 +3478,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "DifyAgent",
-          "version": "0.1.10"
+          "version": "0.2.3"
         },
         "config": {
           "trigger": {
@@ -3501,17 +3520,17 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/DifyAgent/default"
+              "id": "plugin:langbot-team/DifyAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/DifyAgent/default": {
                 "timeout": 30,
                 "advanced-settings": false,
-                "base-url": "https://dify.invalid/v1",
                 "api-key": "synthetic-key",
                 "app-type": "chat",
                 "base-prompt": "Synthetic prompt",
+                "base-url": "https://dify.invalid/v1",
                 "user-id-source": "legacy-session",
                 "enable-all-tools": false,
                 "tools": [],
@@ -3543,6 +3562,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -3742,7 +3762,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "CozeAgent",
-          "version": "0.1.10"
+          "version": "0.2.0"
         },
         "config": {
           "trigger": {
@@ -3784,16 +3804,16 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/CozeAgent/default"
+              "id": "plugin:langbot-team/CozeAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/CozeAgent/default": {
                 "auto-save-history": false,
                 "advanced-settings": false,
+                "api-base": "https://api.coze.cn",
                 "api-key": "synthetic-key",
                 "bot-id": "synthetic-bot",
-                "api-base": "https://api.coze.cn",
                 "timeout": 120,
                 "user-id-source": "legacy-session",
                 "enable-all-tools": false,
@@ -3826,6 +3846,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -4025,7 +4046,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "DashScopeAgent",
-          "version": "0.1.10"
+          "version": "0.2.2"
         },
         "config": {
           "trigger": {
@@ -4067,8 +4088,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/DashScopeAgent/default"
+              "id": "plugin:langbot-team/DashScopeAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/DashScopeAgent/default": {
@@ -4108,6 +4129,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -4299,7 +4321,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "N8nAgent",
-          "version": "0.1.10"
+          "version": "0.2.0"
         },
         "config": {
           "trigger": {
@@ -4341,8 +4363,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/N8nAgent/default"
+              "id": "plugin:langbot-team/N8nAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/N8nAgent/default": {
@@ -4389,6 +4411,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -4584,7 +4607,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "LangflowAgent",
-          "version": "0.1.10"
+          "version": "0.2.2"
         },
         "config": {
           "trigger": {
@@ -4626,8 +4649,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/LangflowAgent/default"
+              "id": "plugin:langbot-team/LangflowAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/LangflowAgent/default": {
@@ -4635,8 +4658,8 @@
                 "output-type": "chat",
                 "tweaks": {},
                 "advanced-settings": false,
-                "base-url": "https://langflow.invalid",
                 "api-key": "synthetic-key",
+                "base-url": "https://langflow.invalid",
                 "flow-id": "synthetic-flow",
                 "enable-all-tools": false,
                 "tools": [],
@@ -4667,6 +4690,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -4862,7 +4886,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "DeerFlowAgent",
-          "version": "0.1.10"
+          "version": "0.2.1"
         },
         "config": {
           "trigger": {
@@ -4904,8 +4928,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/DeerFlowAgent/default"
+              "id": "plugin:langbot-team/DeerFlowAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/DeerFlowAgent/default": {
@@ -4948,6 +4972,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -5131,7 +5156,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "TboxAgent",
-          "version": "0.1.8"
+          "version": "0.2.1"
         },
         "config": {
           "trigger": {
@@ -5173,8 +5198,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/TboxAgent/default"
+              "id": "plugin:langbot-team/TboxAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/TboxAgent/default": {
@@ -5211,6 +5236,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -5402,7 +5428,7 @@
         "target_plugin": {
           "author": "langbot-team",
           "name": "WeKnoraAgent",
-          "version": "0.1.10"
+          "version": "0.2.1"
         },
         "config": {
           "trigger": {
@@ -5444,8 +5470,8 @@
           },
           "ai": {
             "runner": {
-              "expire-time": 0,
-              "id": "plugin:langbot-team/WeKnoraAgent/default"
+              "id": "plugin:langbot-team/WeKnoraAgent/default",
+              "expire-time": 0
             },
             "runner_config": {
               "plugin:langbot-team/WeKnoraAgent/default": {
@@ -5456,9 +5482,9 @@
                 "timeout": 120,
                 "base-prompt": "请回答用户的问题。",
                 "advanced-settings": false,
-                "base-url": "https://weknora.invalid/api/v1",
                 "api-key": "synthetic-key",
                 "app-type": "agent",
+                "base-url": "https://weknora.invalid/api/v1",
                 "enable-all-tools": false,
                 "tools": [],
                 "knowledge-bases": [],
@@ -5488,6 +5514,7 @@
             }
           }
         },
+        "extensions_preferences": null,
         "changed_paths": [
           "ai.runner.runner",
           "ai.runner.id",
@@ -5738,12 +5765,110 @@
               "zh_Hans": "失败提示文本"
             },
             "description": {
-              "en_US": "The text to display when a request fails. Only effective when Exception Handling Strategy is set to \"Show Failure Hint\"",
-              "zh_Hans": "请求失败时显示的提示文本，仅在异常处理策略设置为\"仅文字提示\"时生效"
+              "en_US": "Leave empty to use the Workspace language. Used for request and runner execution failures; custom text is not translated.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。用于请求和运行器执行失败；自定义文本不会自动翻译。"
             },
             "type": "string",
             "required": false,
-            "default": "Request failed."
+            "default": ""
+          },
+          {
+            "name": "rate-limit-hint",
+            "label": {
+              "en_US": "Rate Limit Hint",
+              "zh_Hans": "限速提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "content-filter-hint",
+            "label": {
+              "en_US": "Content Filter Hint",
+              "zh_Hans": "内容过滤提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "content-filter-error-hint",
+            "label": {
+              "en_US": "Content Check Failure Hint",
+              "zh_Hans": "内容检查失败提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "tool-call-hint",
+            "label": {
+              "en_US": "Tool Call Hint",
+              "zh_Hans": "工具调用提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language. Use {tools} for tool names. Requires Track Function Calls.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。 可用 {tools} 插入工具名称，需开启跟踪函数调用。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "runner-not-found-hint",
+            "label": {
+              "en_US": "Runner Missing Hint",
+              "zh_Hans": "运行器不存在提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language. Use {runner} for the runner ID.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。 可用 {runner} 插入运行器 ID。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "runner-not-authorized-hint",
+            "label": {
+              "en_US": "Runner Authorization Hint",
+              "zh_Hans": "运行器未授权提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
+          },
+          {
+            "name": "runner-unavailable-hint",
+            "label": {
+              "en_US": "Runner Unavailable Hint",
+              "zh_Hans": "运行器不可用提示"
+            },
+            "description": {
+              "en_US": "Leave empty to use the Workspace language.",
+              "zh_Hans": "留空时使用工作区语言的默认提示。"
+            },
+            "type": "string",
+            "required": false,
+            "default": ""
           },
           {
             "name": "at-sender",
diff --git a/web/tests/e2e/form-dirty-regressions.spec.ts b/web/tests/e2e/form-dirty-regressions.spec.ts
new file mode 100644
index 000000000..930c165f3
--- /dev/null
+++ b/web/tests/e2e/form-dirty-regressions.spec.ts
@@ -0,0 +1,317 @@
+import { expect, test, type Page } from '@playwright/test';
+import {
+  installLangBotApiMocks,
+  pipelineMetadata,
+} from './fixtures/langbot-api';
+
+const runner = 'plugin:langbot-team/LocalAgent/default';
+
+async function setup(page: Page, kind: 'agent' | 'pipeline' = 'agent') {
+  await installLangBotApiMocks(page, { authenticated: true });
+  // Exercise the self-hosted form, without the unrelated account assistant overlay.
+  await page.route('**/api/v1/user/info', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: { account_uuid: null, user_id: 1, username: 'Form tester' },
+      },
+    }),
+  );
+  const metadata = pipelineMetadata();
+  let pipelineSaved = false;
+  page.on('request', (r) => {
+    if (r.method() === 'PUT' && r.url().endsWith('/pipelines/pipeline-draft'))
+      pipelineSaved = true;
+  });
+  await page.route('**/api/v1/pipelines/pipeline-draft', async (route) => {
+    if (route.request().method() !== 'GET' || pipelineSaved)
+      return route.fallback();
+    await route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          pipeline: {
+            uuid: 'pipeline-draft',
+            name: 'Pipeline draft',
+            emoji: '⚙️',
+            description: '',
+            config: {
+              ai: {
+                runner: { id: runner },
+                runner_config: {
+                  [runner]: {
+                    model: { primary: 'llm-valid', fallbacks: [] },
+                    instruction: 'Original instruction',
+                    advanced: false,
+                    detail: 'Original detail',
+                  },
+                },
+              },
+              trigger: {},
+              safety: {},
+              output: {},
+            },
+          },
+        },
+      },
+    });
+  });
+  metadata.configs[0].stages[1].config.push(
+    ...([
+      {
+        name: 'instruction',
+        label: { en_US: 'Instruction' },
+        type: 'text',
+        required: true,
+        default: 'Original instruction',
+      },
+      {
+        name: 'advanced',
+        label: { en_US: 'Advanced' },
+        type: 'boolean',
+        default: false,
+      },
+      {
+        name: 'detail',
+        label: { en_US: 'Detail' },
+        type: 'text',
+        default: 'Original detail',
+        show_if: { field: 'advanced', operator: 'eq', value: true },
+      },
+    ] as any),
+  );
+  await page.route(
+    `**/api/v1/${kind === 'agent' ? 'agents' : 'pipelines'}/_/metadata`,
+    (route) =>
+      route.fulfill({
+        json: {
+          code: 0,
+          data:
+            kind === 'agent'
+              ? {
+                  runner_config: metadata.configs[0],
+                  kinds: [],
+                  event_processors: [],
+                }
+              : metadata,
+        },
+      }),
+  );
+  await page.goto(
+    `/home/${kind === 'agent' ? 'agents' : 'pipelines'}?id=${kind}-draft`,
+  );
+  if (kind === 'pipeline')
+    await page.getByRole('tab', { name: /^AI$/ }).click();
+  const input = page.locator('textarea[name="instruction"]');
+  await expect(input).toHaveValue('Original instruction');
+  return {
+    input,
+    save: page.getByRole('button', { name: 'Save', exact: true }),
+    endpoint: `/api/v1/${kind === 'agent' ? 'agents' : 'pipelines'}/${kind}-draft`,
+  };
+}
+
+for (const kind of ['agent', 'pipeline'] as const) {
+  test(`${kind}: revert, conditional fields, immediate save and reload retain the latest draft`, async ({
+    page,
+  }) => {
+    const { input, save, endpoint } = await setup(page, kind);
+    await expect(save).toBeDisabled();
+    await input.fill('Changed');
+    await expect(save).toBeEnabled();
+    await input.fill('Original instruction');
+    await expect(save).toBeDisabled();
+    await page
+      .locator('label')
+      .filter({ hasText: /^Advanced$/ })
+      .locator('../..')
+      .getByRole('switch')
+      .click();
+    const detail = page.locator('textarea[name="detail"]');
+    await detail.fill('Retained while hidden');
+    await page
+      .locator('label')
+      .filter({ hasText: /^Advanced$/ })
+      .locator('../..')
+      .getByRole('switch')
+      .click();
+    await expect(detail).toHaveCount(0);
+    await input.pressSequentially(' final characters');
+    const request = page.waitForRequest(
+      (r) => new URL(r.url()).pathname === endpoint && r.method() === 'PUT',
+    );
+    await save.click();
+    const payload = (await request).postDataJSON();
+    const config = kind === 'agent' ? payload.config : payload.config.ai;
+    expect(config.runner_config[runner]).toMatchObject({
+      instruction: 'Original instruction final characters',
+      detail: 'Retained while hidden',
+      advanced: false,
+    });
+    await expect(save).toBeDisabled();
+    await page.reload();
+    if (kind === 'pipeline')
+      await page.getByRole('tab', { name: /^AI$/ }).click();
+    await expect(input).toHaveValue('Original instruction final characters');
+    await expect(save).toBeDisabled();
+    await page
+      .locator('label')
+      .filter({ hasText: /^Advanced$/ })
+      .locator('../..')
+      .getByRole('switch')
+      .click();
+    await expect(detail).toHaveValue('Retained while hidden');
+  });
+}
+
+test('agent: save completion does not absorb edits made while the request is pending', async ({
+  page,
+}) => {
+  const { input, save, endpoint } = await setup(page);
+  let release!: () => void;
+  const gate = new Promise<void>((resolve) => {
+    release = resolve;
+  });
+  await page.route(`**${endpoint}`, async (route) => {
+    if (route.request().method() === 'PUT') await gate;
+    await route.fallback();
+  });
+  await input.fill('First draft');
+  const request = page.waitForRequest(
+    (r) => r.method() === 'PUT' && r.url().endsWith(endpoint),
+  );
+  await save.click();
+  await request;
+  await input.fill('Second draft typed during save');
+  release();
+  await expect(save).toBeEnabled();
+  const next = page.waitForRequest(
+    (r) => r.method() === 'PUT' && r.url().endsWith(endpoint),
+  );
+  await save.click();
+  expect(
+    (await next).postDataJSON().config.runner_config[runner].instruction,
+  ).toBe('Second draft typed during save');
+  await expect(save).toBeDisabled();
+});
+
+test('agent: failed save keeps the draft, and debugging saves the latest value first', async ({
+  page,
+}) => {
+  const { input, save, endpoint } = await setup(page);
+  let fail = true;
+  const operations: string[] = [];
+  await page.route(`**${endpoint}`, async (route) => {
+    if (route.request().method() !== 'PUT') return route.fallback();
+    if (fail)
+      return route.fulfill({
+        status: 500,
+        json: { code: 500, msg: 'Deliberate save failure' },
+      });
+    operations.push('save');
+    expect(
+      route.request().postDataJSON().config.runner_config[runner].instruction,
+    ).toBe('Latest before debug');
+    await route.fallback();
+  });
+  await input.fill('Unsaved draft');
+  await save.click();
+  await expect(page.getByText(/Deliberate save failure/)).toBeVisible();
+  await expect(save).toBeEnabled();
+  await expect(input).toHaveValue('Unsaved draft');
+  fail = false;
+  page.on('request', (request) => {
+    if (request.url().includes('/debug/stream')) operations.push('debug');
+  });
+  await page.getByRole('textbox', { name: 'Message content' }).fill('Hello');
+  await input.fill('Latest before debug');
+  await page.getByRole('button', { name: 'Save and run', exact: true }).click();
+  await expect(page.getByText('Mock Agent response')).toBeVisible();
+  expect(operations).toEqual(['save', 'debug']);
+  await expect(save).toBeDisabled();
+});
+
+test('agent: required-field status still reacts after unrelated keystrokes are isolated', async ({
+  page,
+}) => {
+  const { input } = await setup(page);
+  await expect(page.getByText('Runner ready', { exact: true })).toBeVisible();
+  await input.fill('');
+  await expect(page.getByText('Runner ready', { exact: true })).toHaveCount(0);
+  await input.fill('Valid again');
+  await expect(page.getByText('Runner ready', { exact: true })).toBeVisible();
+});
+
+test('knowledge: early edits are not absorbed into initialization, and save uses a submitted snapshot', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  const engine = {
+    plugin_id: 'builtin/minimal-knowledge',
+    name: { en_US: 'Test engine' },
+    capabilities: ['text_retrieval'],
+    creation_schema: [],
+    retrieval_schema: [
+      {
+        name: 'instructions',
+        type: 'text',
+        label: { en_US: 'Retrieval instructions' },
+        default: 'Original',
+      },
+    ],
+  };
+  let base = {
+    uuid: 'kb-draft',
+    name: 'Knowledge draft',
+    description: '',
+    emoji: '📚',
+    initialized: true,
+    knowledge_engine_plugin_id: engine.plugin_id,
+    knowledge_engine: engine,
+    creation_settings: {},
+    retrieval_settings: { instructions: 'Original' },
+  };
+  await page.route('**/api/v1/knowledge/engines', (route) =>
+    route.fulfill({ json: { code: 0, data: { engines: [engine] } } }),
+  );
+  let release!: () => void;
+  const gate = new Promise<void>((resolve) => {
+    release = resolve;
+  });
+  let saving = false;
+  await page.route('**/api/v1/knowledge/bases/kb-draft', async (route) => {
+    if (route.request().method() === 'PUT') {
+      const payload = route.request().postDataJSON();
+      saving = true;
+      await gate;
+      base = { ...base, ...payload };
+      return route.fulfill({ json: { code: 0, data: { uuid: base.uuid } } });
+    }
+    return route.fulfill({ json: { code: 0, data: { base } } });
+  });
+  await page.goto('/home/knowledge?id=kb-draft');
+  const input = page.locator('textarea[name="instructions"]');
+  const save = page.getByRole('button', { name: 'Save', exact: true });
+  await expect(input).toHaveValue('Original');
+  await input.fill('Early edit');
+  await expect(save).toBeEnabled();
+  await input.fill('Original');
+  await expect(save).toBeDisabled();
+  await input.fill('Submitted value');
+  await save.click();
+  await expect.poll(() => saving).toBe(true);
+  await input.fill('Typed while saving');
+  release();
+  await expect
+    .poll(() => base.retrieval_settings.instructions)
+    .toBe('Submitted value');
+  await expect(save).toBeEnabled();
+  await save.click();
+  await expect
+    .poll(() => base.retrieval_settings.instructions)
+    .toBe('Typed while saving');
+  await expect(save).toBeDisabled();
+  await page.reload();
+  await expect(input).toHaveValue('Typed while saving');
+  await expect(save).toBeDisabled();
+});
diff --git a/web/tests/e2e/home-smoke.spec.ts b/web/tests/e2e/home-smoke.spec.ts
index 34a69670d..3f7b808ed 100644
--- a/web/tests/e2e/home-smoke.spec.ts
+++ b/web/tests/e2e/home-smoke.spec.ts
@@ -59,10 +59,17 @@ test.describe('authenticated app shell', () => {
     await page.goto('/home/monitoring');
 
     await expect(page).toHaveURL(/\/home\/monitoring$/);
-    await expect(page.getByText('Total Messages').first()).toBeVisible();
+    // The unified execution view is the default entry point.
+    await expect(page.getByRole('tab', { name: 'Executions' })).toHaveAttribute(
+      'data-state',
+      'active',
+    );
+    await expect(
+      page.getByText('Execution Success Rate').first(),
+    ).toBeVisible();
     await expect(
       page.getByRole('tab', { name: 'Message Records' }),
-    ).toBeVisible();
+    ).toHaveCount(0);
     await expect(
       page.getByRole('tab', { name: 'Token Monitoring' }),
     ).toBeVisible();
diff --git a/web/tests/e2e/login.spec.ts b/web/tests/e2e/login.spec.ts
index 11ae6108f..e60c78574 100644
--- a/web/tests/e2e/login.spec.ts
+++ b/web/tests/e2e/login.spec.ts
@@ -20,7 +20,7 @@ test('local account login reaches the authenticated home shell', async ({
   await expect(page).toHaveURL(/\/home(?:\/monitoring)?$/);
   await expect(page.getByText('Home').first()).toBeVisible();
   await expect(page.getByRole('button', { name: 'Dashboard' })).toBeVisible();
-  await expect(page.getByText('Total Messages').first()).toBeVisible();
+  await expect(page.getByText('Execution Success Rate').first()).toBeVisible();
   await expect(page.getByText('Unable to connect to server')).toHaveCount(0);
 });
 
diff --git a/web/tests/e2e/monitoring-executions.spec.ts b/web/tests/e2e/monitoring-executions.spec.ts
new file mode 100644
index 000000000..6e253bad6
--- /dev/null
+++ b/web/tests/e2e/monitoring-executions.spec.ts
@@ -0,0 +1,496 @@
+import { expect, test } from '@playwright/test';
+import {
+  installLangBotApiMocks,
+  buildExecutionList,
+} from './fixtures/langbot-api';
+
+const monitoring = {
+  messages: [
+    {
+      id: 'input-one',
+      role: 'user',
+      message_content: 'Open this execution',
+      timestamp: '2026-07-02T10:00:00Z',
+      bot_id: 'bot',
+      pipeline_id: 'pipe',
+      pipeline_name: 'Processor A',
+    },
+  ],
+};
+const row = buildExecutionList(monitoring).items[0];
+const pageOf = (
+  items: unknown[],
+  has_more = false,
+  next_offset = items.length,
+) => ({ items, has_more, next_offset });
+
+test('renders typed and custom events without duplicate text or empty attachment messages', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    monitoringData: monitoring,
+  });
+  await page.route('**/api/v1/monitoring/executions/*/*?*', async (route) => {
+    await route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          source: 'pipeline',
+          row,
+          pages: {
+            inputs: pageOf([
+              {
+                id: 'message',
+                content: {
+                  text: '你好~',
+                  contents: [{ type: 'text', text: '你好~', image_url: null }],
+                  attachments: [],
+                },
+              },
+              {
+                id: 'member',
+                event_type: 'group.member_banned',
+                content: {
+                  member: { id: 'member-42', nickname: 'Alice' },
+                  group: { id: 'group-9' },
+                  duration: 0,
+                  permanent: false,
+                  custom: { reason: 'Moderation reason' },
+                },
+              },
+              {
+                id: 'edited',
+                content: {
+                  message_chain: [{ type: 'Plain', text: 'Typed chain text' }],
+                  editor: { id: 'editor-7' },
+                  new_content: [{ type: 'Plain', text: 'Updated text' }],
+                },
+              },
+              {
+                id: 'custom',
+                content: {
+                  type: 'CustomEvent',
+                  text: 'Distinct text',
+                  content: 'Different content',
+                  properties: { nested: ['custom-value'] },
+                },
+              },
+              {
+                id: 'same-again',
+                content: {
+                  text: '你好~',
+                  contents: [{ type: 'text', text: '你好~' }],
+                },
+              },
+              {
+                id: 'image',
+                content: {
+                  contents: [
+                    {
+                      type: 'image_url',
+                      image_url: {
+                        url: 'data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7',
+                      },
+                    },
+                  ],
+                },
+              },
+            ]),
+            outputs: pageOf([
+              {
+                id: 'assistant',
+                content: {
+                  content: 'Agent answer',
+                  tool_calls: [
+                    {
+                      id: 'tool-1',
+                      function: {
+                        name: 'event_reply',
+                        arguments: '{"text":"sent"}',
+                      },
+                    },
+                  ],
+                },
+                attachments: [],
+              },
+            ]),
+            deliveries: pageOf([]),
+          },
+        },
+      },
+    });
+  });
+  await page.goto('/home/monitoring');
+  await page
+    .getByRole('button')
+    .filter({ hasText: 'Open this execution' })
+    .click();
+  const sheet = page.getByRole('dialog');
+  await expect(
+    sheet.getByText('你好~', { exact: true }).filter({ visible: true }),
+  ).toHaveCount(2);
+  await expect(sheet.getByText('[Empty message]', { exact: true })).toHaveCount(
+    0,
+  );
+  for (const details of await sheet
+    .locator('summary')
+    .filter({ hasText: /^Details$/ })
+    .all())
+    await details.click();
+  for (const text of [
+    'Agent answer',
+    'member-42',
+    'Alice',
+    'group-9',
+    'Moderation reason',
+    '0',
+    'false',
+    'Typed chain text',
+    'editor-7',
+    'Updated text',
+    'Distinct text',
+    'Different content',
+    'custom-value',
+    'event_reply',
+  ]) {
+    await expect(
+      sheet.getByText(text, { exact: true }).filter({ visible: true }).first(),
+    ).toBeVisible();
+  }
+  await expect(
+    sheet.getByRole('img', { name: 'Message attachment' }),
+  ).toBeVisible();
+});
+
+test('pages a long trace without dropping inputs and opens related executions', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    monitoringData: monitoring,
+  });
+  const requests: string[] = [];
+  await page.route('**/api/v1/monitoring/executions/*/*?*', async (route) => {
+    const url = new URL(route.request().url());
+    requests.push(url.search);
+    const sibling = url.pathname.endsWith('/sibling');
+    const more = url.searchParams.get('section') === 'events';
+    await route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          source: 'pipeline',
+          row: sibling
+            ? { ...row, id: 'sibling', target_name: 'Processor B' }
+            : row,
+          pages: more
+            ? {
+                events: pageOf(
+                  [
+                    {
+                      id: 101,
+                      sequence: 101,
+                      type: 'run.completed',
+                      data: { final: 'Last trace event' },
+                    },
+                  ],
+                  false,
+                  101,
+                ),
+              }
+            : {
+                inputs: pageOf([
+                  {
+                    id: 'in',
+                    content: sibling ? 'Sibling input' : 'Original input',
+                  },
+                ]),
+                outputs: pageOf([
+                  { id: 'out', content: 'Generated but not sent' },
+                ]),
+                deliveries: pageOf([]),
+                events: pageOf(
+                  [{ id: 1, sequence: 1, type: 'run.started', data: {} }],
+                  true,
+                  100,
+                ),
+                related: pageOf(
+                  sibling
+                    ? []
+                    : [{ ...row, id: 'sibling', target_name: 'Processor B' }],
+                ),
+              },
+        },
+      },
+    });
+  });
+  await page.goto('/home/monitoring');
+  await page
+    .getByRole('button')
+    .filter({ hasText: 'Open this execution' })
+    .click();
+  const sheet = page.getByRole('dialog');
+  await expect(
+    sheet.getByText('Generated but not sent').filter({ visible: true }),
+  ).toBeVisible();
+  await expect(sheet.getByText('Delivery records (0)')).toBeVisible();
+  await sheet.getByText('Execution events (1+)', { exact: true }).click();
+  await sheet.getByRole('button', { name: 'Load more' }).click();
+  await expect(sheet.getByText('run.completed', { exact: true })).toBeVisible();
+  await expect(
+    sheet.getByText('Original input').filter({ visible: true }),
+  ).toBeVisible();
+  expect(
+    requests.some(
+      (query) =>
+        query.includes('section=events') && query.includes('offset=100'),
+    ),
+  ).toBe(true);
+  await sheet.getByText('Related executions (1)', { exact: true }).click();
+  await sheet.getByRole('button', { name: /Processor B/ }).click();
+  await expect(
+    sheet.getByText('Sibling input').filter({ visible: true }),
+  ).toBeVisible();
+  await expect(
+    sheet.getByText('Original input').filter({ visible: true }),
+  ).toHaveCount(0);
+});
+
+test('shows non-message events without fabricating a reply or a processor', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  const event = {
+    ...row,
+    source: 'event',
+    id: 'event-only',
+    event_id: 'event-only',
+    target_id: null,
+    target_kind: 'event',
+    target_name: null,
+    title: 'group.member.joined',
+    input_preview: 'Member joined',
+    status_group: 'ignored',
+    status: 'ignored',
+  };
+  await page.route('**/api/v1/monitoring/executions?*', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: { ...buildExecutionList(monitoring), items: [event] },
+      },
+    }),
+  );
+  await page.route('**/api/v1/monitoring/executions/event/*?*', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          source: 'event',
+          row: event,
+          pages: {
+            inputs: pageOf([
+              {
+                id: 'event-only',
+                event_type: 'group.member.joined',
+                content: { member_id: 'member-42' },
+                metadata: {
+                  routes: [
+                    { status: 'not_matched', reason: 'No matching route' },
+                  ],
+                },
+              },
+            ]),
+            outputs: pageOf([]),
+            deliveries: pageOf([]),
+          },
+        },
+      },
+    }),
+  );
+  await page.goto('/home/monitoring');
+  await page.getByRole('button').filter({ hasText: 'Member joined' }).click();
+  const sheet = page.getByRole('dialog');
+  await expect(
+    sheet.getByRole('heading', { name: 'No processor run' }),
+  ).toBeVisible();
+  await sheet
+    .locator('summary')
+    .filter({ hasText: /^Details$/ })
+    .first()
+    .click();
+  await expect(sheet.getByText(/member-42/)).toBeVisible();
+  await expect(sheet.getByText(/No matching route/)).toBeVisible();
+  await expect(sheet.getByText('Delivery records (0)')).toBeVisible();
+});
+
+test('filters all three processor kinds with the sidebar icons', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  const processors = [
+    { uuid: 'agent-one', name: 'Agent One', kind: 'agent', emoji: '🤖' },
+    {
+      uuid: 'pipeline-one',
+      name: 'Pipeline One',
+      kind: 'pipeline',
+      emoji: '⚙️',
+    },
+    {
+      uuid: 'event-one',
+      name: 'Event One',
+      kind: 'event_processor',
+      emoji: '🧩',
+    },
+  ];
+  await page.route('**/api/v1/agents', (route) =>
+    route.fulfill({ json: { code: 0, data: { agents: processors } } }),
+  );
+  await page.goto('/home/monitoring');
+  const filter = page.locator('#monitoring-filter-processor');
+  await filter.click();
+  const popup = page.getByRole('dialog');
+  for (const processor of processors) {
+    const option = popup.getByRole('checkbox', {
+      name: processor.name,
+      exact: true,
+    });
+    await expect(
+      option.locator('..').getByText(processor.emoji, { exact: true }),
+    ).toBeVisible();
+    const request = page.waitForRequest((req) => {
+      const url = new URL(req.url());
+      return (
+        url.pathname.endsWith('/monitoring/executions') &&
+        url.searchParams.getAll('pipelineId').includes(processor.uuid)
+      );
+    });
+    await option.check();
+    await request;
+    await expect(option).toBeChecked();
+  }
+  await expect(filter).toContainText('3');
+  await popup
+    .getByRole('button', { name: 'All Processors', exact: true })
+    .click();
+  await expect(popup.getByRole('checkbox', { checked: true })).toHaveCount(0);
+  await expect(filter).toContainText('All Processors');
+});
+
+test('shares mode and status filters with dashboard charts and records', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  await page.goto('/home/monitoring');
+  for (const [id, option, parameter, value] of [
+    ['monitoring-filter-mode', 'Debug only', 'mode', 'debug'],
+    ['monitoring-filter-status', 'Failed', 'status', 'failed'],
+  ]) {
+    await page.locator(`#${id}`).click();
+    const requests = ['/monitoring/data', '/monitoring/executions'].map(
+      (path) =>
+        page.waitForRequest((req) => {
+          const url = new URL(req.url());
+          return (
+            url.pathname.endsWith(path) &&
+            url.searchParams.get(parameter) === value
+          );
+        }),
+    );
+    await page.getByRole('option', { name: option, exact: true }).click();
+    await Promise.all(requests);
+  }
+});
+
+test('remembers all filters on reload and persists resetting them', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  await page.route('**/api/v1/platform/bots', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          bots: [{ uuid: 'bot-one', name: 'Saved Bot', adapter: 'aiocqhttp' }],
+        },
+      },
+    }),
+  );
+  await page.route('**/api/v1/agents', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          agents: [
+            {
+              uuid: 'processor-one',
+              name: 'Saved Processor',
+              kind: 'agent',
+              emoji: '🤖',
+            },
+          ],
+        },
+      },
+    }),
+  );
+  await page.goto('/home/monitoring');
+  const selections = [
+    ['bot', 'Saved Bot'],
+    ['processor', 'Saved Processor'],
+    ['mode', 'Debug only'],
+    ['status', 'Failed'],
+    ['time', 'Last 7 days'],
+  ];
+  for (const [id, label] of selections) {
+    await page.locator(`#monitoring-filter-${id}`).click();
+    if (id === 'processor') {
+      await page.getByRole('checkbox', { name: label, exact: true }).check();
+      await page.keyboard.press('Escape');
+    } else await page.getByRole('option', { name: label, exact: true }).click();
+  }
+  await page.reload();
+  for (const [id, label] of selections) {
+    if (id === 'processor') {
+      await page.locator('#monitoring-filter-processor').click();
+      await expect(
+        page.getByRole('checkbox', { name: label, exact: true }),
+      ).toBeChecked();
+      await page.keyboard.press('Escape');
+    } else
+      await expect(page.locator(`#monitoring-filter-${id}`)).toContainText(
+        label,
+      );
+  }
+  await page
+    .getByRole('button', { name: 'Reset Filters', exact: true })
+    .click();
+  await page.reload();
+  for (const [id, label] of [
+    ['bot', 'All Bots'],
+    ['processor', 'All Processors'],
+    ['mode', 'Live + debug'],
+    ['status', 'All statuses'],
+    ['time', 'Last 24 hours'],
+  ]) {
+    await expect(page.locator(`#monitoring-filter-${id}`)).toContainText(label);
+  }
+});
+
+test('reset removes linked resource filters from the URL', async ({ page }) => {
+  await installLangBotApiMocks(page, { authenticated: true });
+  await page.goto(
+    '/home/monitoring?botId=linked-bot&pipelineId=linked-processor',
+  );
+  await page
+    .getByRole('button', { name: 'Reset Filters', exact: true })
+    .click();
+  await expect(page).toHaveURL(/\/home\/monitoring$/);
+  await page.reload();
+  await expect(page.locator('#monitoring-filter-bot')).toContainText(
+    'All Bots',
+  );
+  await expect(page.locator('#monitoring-filter-processor')).toContainText(
+    'All Processors',
+  );
+});
diff --git a/web/tests/e2e/monitoring-turns.spec.ts b/web/tests/e2e/monitoring-turns.spec.ts
index 2abc76eb6..f81b69d6a 100644
--- a/web/tests/e2e/monitoring-turns.spec.ts
+++ b/web/tests/e2e/monitoring-turns.spec.ts
@@ -1,6 +1,9 @@
 import { expect, test, Route } from '@playwright/test';
 
-import { installLangBotApiMocks } from './fixtures/langbot-api';
+import {
+  installLangBotApiMocks,
+  buildExecutionList,
+} from './fixtures/langbot-api';
 import { buildConversationTurns } from '../../src/app/home/monitoring/utils/conversationTurns';
 import {
   ErrorLog,
@@ -254,7 +257,15 @@ function rawMonitoringData() {
       success_rate: 100,
       active_sessions: 2,
     },
-    messages: scenario.messages.map(rawMessage),
+    messages: scenario.messages.map((item) => ({
+      ...rawMessage(item),
+      parent_message_id:
+        item.role === 'assistant'
+          ? item.id === 'agent-assistant-4'
+            ? 'agent-user-2'
+            : 'agent-user-1'
+          : undefined,
+    })),
     llmCalls: scenario.llmCalls.map(rawLlmCall),
     toolCalls: scenario.toolCalls.map(rawToolCall),
     embeddingCalls: [],
@@ -273,8 +284,13 @@ function rawMonitoringData() {
 
 async function respond(route: Route, label: string) {
   const data = rawMonitoringData();
-  data.messages = [rawMessage(message(label, 'user', 10, label))];
-  await route.fulfill({ json: { code: 0, data } });
+  data.messages = [
+    {
+      ...rawMessage(message(label, 'user', 10, label)),
+      parent_message_id: undefined,
+    },
+  ];
+  await route.fulfill({ json: { code: 0, data: buildExecutionList(data) } });
 }
 
 test.describe('monitoring request contracts', () => {
@@ -283,7 +299,7 @@ test.describe('monitoring request contracts', () => {
   }) => {
     await installLangBotApiMocks(page, { authenticated: true });
     let failing = true;
-    await page.route('**/api/v1/monitoring/data?*', async (route) => {
+    await page.route('**/api/v1/monitoring/executions?*', async (route) => {
       expect(route.request().headers().authorization).toBe(
         'Bearer playwright-token',
       );
@@ -298,14 +314,14 @@ test.describe('monitoring request contracts', () => {
       else await respond(route, 'Recovered monitoring');
     });
     await page.goto('/home/monitoring');
-    await expect(page.getByRole('alert')).toContainText(
-      'Failed to load monitoring data',
-    );
+    await expect(
+      page.getByText('Failed to load execution trace'),
+    ).toBeVisible();
     await expect(page.getByText('No message records')).toHaveCount(0);
     failing = false;
     await page.getByRole('button', { name: 'Retry', exact: true }).click();
     await expect(
-      page.getByText('Recovered monitoring', { exact: true }),
+      page.getByText('Recovered monitoring', { exact: true }).first(),
     ).toBeVisible();
     await expect(page.getByRole('alert')).toHaveCount(0);
   });
@@ -315,17 +331,18 @@ test.describe('monitoring request contracts', () => {
   }) => {
     await installLangBotApiMocks(page, { authenticated: true });
     const pending: Route[] = [];
-    await page.route('**/api/v1/monitoring/data?*', (route) => {
+    await page.route('**/api/v1/monitoring/executions?*', (route) => {
       pending.push(route);
     });
     await page.goto('/home/monitoring');
     await expect.poll(() => pending.length).toBe(2);
-    await page.getByRole('combobox').last().click();
+    await page.locator('#monitoring-filter-time').click();
     await page.getByRole('option', { name: /Last 7 days/i }).click();
     await expect.poll(() => pending.length).toBe(3);
     await respond(pending[2], 'Latest filter data');
+    // The message-turn view now lives behind its own tab.
     await expect(
-      page.getByText('Latest filter data', { exact: true }),
+      page.getByText('Latest filter data', { exact: true }).first(),
     ).toBeVisible();
     await respond(pending[0], 'Obsolete filter data');
     await respond(pending[1], 'Obsolete filter data');
@@ -336,7 +353,7 @@ test.describe('monitoring request contracts', () => {
         ),
     );
     await expect(
-      page.getByText('Latest filter data', { exact: true }),
+      page.getByText('Latest filter data', { exact: true }).first(),
     ).toBeVisible();
     await page
       .getByRole('button', { name: 'Refresh Data', exact: true })
@@ -345,19 +362,19 @@ test.describe('monitoring request contracts', () => {
     await expect(
       page.getByText('Obsolete filter data', { exact: true }),
     ).toHaveCount(0);
-    await page.getByRole('combobox').last().click();
+    await page.locator('#monitoring-filter-time').click();
     await page.getByRole('option', { name: /Last 24 hours/i }).click();
     await expect.poll(() => pending.length).toBe(5);
     await respond(pending[4], 'Current result');
     await expect(
-      page.getByText('Current result', { exact: true }),
+      page.getByText('Current result', { exact: true }).first(),
     ).toBeVisible();
     await pending[3].fulfill({
       status: 500,
       json: { code: 500, msg: 'old failure' },
     });
     await expect(
-      page.getByText('Current result', { exact: true }),
+      page.getByText('Current result', { exact: true }).first(),
     ).toBeVisible();
     await expect(page.getByRole('alert')).toHaveCount(0);
   });
@@ -386,7 +403,7 @@ test.describe('monitoring request contracts', () => {
       page.getByText(
         'Showing 7 of 125 messages. Conversation traces may be incomplete.',
       ),
-    ).toBeVisible();
+    ).toHaveCount(0);
     await expect(
       page.getByText('Traffic range truncated. Choose a shorter time range.'),
     ).toBeVisible();
@@ -589,56 +606,45 @@ test.describe('monitoring conversation turn grouping', () => {
     expect(turns[0].totalToolDuration).toBe(45);
   });
 
-  test('renders user-only, multi-agent, and multi-turn cases in the monitoring page', async ({
+  test('replaces the messages tab with full execution inputs, deliveries and calls', async ({
     page,
   }) => {
     await installLangBotApiMocks(page, {
       authenticated: true,
       monitoringData: rawMonitoringData(),
     });
-
     await page.goto('/home/monitoring');
-
-    await expect(page.getByText('3 conversation turns')).toBeVisible();
     await expect(
-      page.getByText('Standalone question with no reply'),
-    ).toBeVisible();
-    await expect(page.getByText('No assistant reply recorded')).toBeVisible();
-    await expect(page.getByText('Need deployment plan')).toBeVisible();
-    await expect(page.getByText('Agent step 1: inspect repo')).toBeVisible();
-    await expect(page.getByText('Assistant +2')).toBeVisible();
-    await expect(page.getByText('3 LLM')).toBeVisible();
-    await expect(page.getByText('2 tools')).toBeVisible();
-    await expect(page.getByText('790 tokens')).toBeVisible();
-    await expect(page.getByText('1 errors')).toBeVisible();
-    await expect(page.getByText('Continue with rollback plan')).toBeVisible();
-    await expect(page.getByText('Rollback plan ready')).toBeVisible();
-
-    const agentTurn = page
-      .locator('div[role="button"]')
-      .filter({ hasText: 'Need deployment plan' });
-    await expect(agentTurn).toHaveCount(1);
-    await agentTurn.click();
-
-    await expect(page.getByText('Conversation Trace')).toBeVisible();
-    await expect(page.getByText('Agent step 2: run tests')).toBeVisible();
+      page.getByRole('tab', { name: 'Message Records' }),
+    ).toHaveCount(0);
+    await page
+      .getByRole('button')
+      .filter({ hasText: 'Need deployment plan' })
+      .click();
+    const sheet = page.getByRole('dialog');
+    await expect(sheet.getByText('Event / input (1)')).toBeVisible();
     await expect(
-      page.getByText('Final answer: deployment plan ready'),
+      sheet.getByText('Need deployment plan', { exact: true }).first(),
+    ).toBeVisible();
+    await expect(sheet.getByText('Delivery records (3)')).toBeVisible();
+    await expect(
+      sheet.getByText('Final answer: deployment plan ready').first(),
+    ).toBeVisible();
+    await sheet.getByText('Model calls (3)', { exact: true }).click();
+    await expect(sheet.getByText('gpt-5.5').first()).toBeVisible();
+    await sheet.getByText('Tool calls (2)', { exact: true }).click();
+    await expect(
+      sheet.locator('span.break-all').filter({ hasText: /^repo_search$/ }),
+    ).toBeVisible();
+    await expect(
+      sheet.getByText('Errors (1)', { exact: true }).locator('..'),
+    ).toHaveAttribute('open', '');
+    await expect(
+      sheet.getByText('Tool retry failed').filter({ visible: true }).first(),
+    ).toBeVisible();
+    await sheet.getByText('Conversation context (6)', { exact: true }).click();
+    await expect(
+      sheet.getByText('Rollback plan ready').filter({ visible: true }),
     ).toBeVisible();
-    await expect(page.getByText('LLM Calls (3)')).toBeVisible();
-    await expect(page.getByText('#3 gpt-5.5')).toBeVisible();
-    await expect(page.getByText('In: 300')).toBeVisible();
-    await expect(page.getByText('Out: 90')).toBeVisible();
-    await expect(page.getByText('Total: 390')).toBeVisible();
-    await expect(page.getByText('Tool Calls (2)')).toBeVisible();
-    await expect(page.getByText('#1 repo_search')).toBeVisible();
-    await expect(page.getByText('#2 run_tests')).toBeVisible();
-    await expect(page.getByText('Arguments')).toHaveCount(0);
-    await expect(page.getByText('Result')).toHaveCount(0);
-
-    await page.getByText('#1 repo_search').click();
-    await expect(page.getByText('Arguments').first()).toBeVisible();
-    await expect(page.getByText('Result').first()).toBeVisible();
-    await expect(page.getByText('Tool retry failed')).toBeVisible();
   });
 });
diff --git a/web/tests/e2e/owner-space-billing.spec.ts b/web/tests/e2e/owner-space-billing.spec.ts
index 0c5f7c5e5..4db78e3f6 100644
--- a/web/tests/e2e/owner-space-billing.spec.ts
+++ b/web/tests/e2e/owner-space-billing.spec.ts
@@ -56,8 +56,9 @@ test('an OSS local-only owner is prompted to bind before using LangBot Models',
   await page.goto('/home?action=showModelSettings');
 
   await expect(
-    page.getByRole('button', {
-      name: 'The Workspace owner must connect a LangBot Account for LangBot Models.',
-    }),
+    page.getByText(
+      'The Workspace owner must connect a LangBot Account for LangBot Models.',
+      { exact: true },
+    ),
   ).toBeVisible();
 });
diff --git a/web/tests/e2e/pipeline-monitoring-turns.spec.ts b/web/tests/e2e/pipeline-monitoring-turns.spec.ts
index 306398c62..2ebb71b76 100644
--- a/web/tests/e2e/pipeline-monitoring-turns.spec.ts
+++ b/web/tests/e2e/pipeline-monitoring-turns.spec.ts
@@ -169,27 +169,36 @@ test.describe('pipeline monitoring conversation turns', () => {
     await expect(
       page.getByText('Pipeline single user message without reply'),
     ).toBeVisible();
-    await expect(
-      page.getByText('Pipeline needs a deployment plan'),
-    ).toBeVisible();
-    await expect(
-      page.getByText('Pipeline agent step 1: inspect repository'),
-    ).toBeVisible();
-    await expect(page.getByText('Assistant +2')).toBeVisible();
-    await expect(page.getByText('2 tools')).toBeVisible();
-
-    const agentTurn = page
-      .locator('div[role="button"]')
-      .filter({ hasText: 'Pipeline needs a deployment plan' });
-    await expect(agentTurn).toHaveCount(1);
+    const agentTurn = page.getByRole('button', {
+      name: /Pipeline Bot Success Pipeline needs a deployment plan/,
+    });
     await agentTurn.click();
-
-    await expect(page.getByText('Tool Calls (2)')).toBeVisible();
-    await expect(page.getByText('#1 repo_search')).toBeVisible();
-    await expect(page.getByText('#2 run_tests')).toBeVisible();
-    await expect(page.getByText('Arguments')).toHaveCount(0);
-    await page.getByText('#1 repo_search').click();
-    await expect(page.getByText('Arguments')).toBeVisible();
-    await expect(page.getByText('Result')).toBeVisible();
+    const detail = page.getByRole('article');
+    await expect(
+      detail.getByText('Pipeline needs a deployment plan', { exact: true }),
+    ).toBeVisible();
+    for (const text of [
+      'Pipeline agent step 1: inspect repository',
+      'Pipeline agent step 2: run tests',
+      'Pipeline final answer: deployment ready',
+    ])
+      await expect(detail.getByText(text, { exact: true })).toBeVisible();
+    await expect(
+      detail.getByText('Tool Calls 2', { exact: true }),
+    ).toBeVisible();
+    const search = detail
+      .getByRole('listitem')
+      .filter({ hasText: 'repo_search' });
+    await expect(
+      search.getByText('repo_search', { exact: true }),
+    ).toBeVisible();
+    await expect(detail.getByText('run_tests', { exact: true })).toBeVisible();
+    await expect(search.getByText(/"query"/)).toHaveCount(0);
+    await search
+      .getByRole('button', { name: 'Arguments', exact: true })
+      .click();
+    await expect(search.getByText(/"query"/)).toBeVisible();
+    await search.getByRole('button', { name: 'Result', exact: true }).click();
+    await expect(search.getByText(/"ok"/)).toBeVisible();
   });
 });
diff --git a/web/tests/e2e/plugin-subscriptions.spec.ts b/web/tests/e2e/plugin-subscriptions.spec.ts
index 61a1bdb46..3992c69cd 100644
--- a/web/tests/e2e/plugin-subscriptions.spec.ts
+++ b/web/tests/e2e/plugin-subscriptions.spec.ts
@@ -84,9 +84,11 @@ test('creates a configured processor, persists subscriptions separately and reus
     const request = page.waitForRequest(
       (r) => r.method() === 'PUT' && r.url().endsWith('/platform/bots/bot-1'),
     );
-    await page.getByRole('button', { name: /^Save$/ }).click();
+    await page.getByRole('button', { name: /^Save(?: and enable)?$/ }).click();
     const body = (await request).postDataJSON();
-    await expect(page.getByRole('button', { name: /^Save$/ })).toBeDisabled();
+    await expect(
+      page.getByRole('button', { name: /^Save(?: and enable)?$/ }),
+    ).toBeDisabled();
     // Move away so hovering the next control does not pause an overlapping toast.
     await page.mouse.move(0, 0);
     await expect(
diff --git a/web/tests/e2e/wizard-platform-regressions.spec.ts b/web/tests/e2e/wizard-platform-regressions.spec.ts
index 2c1cca56f..7859e6acd 100644
--- a/web/tests/e2e/wizard-platform-regressions.spec.ts
+++ b/web/tests/e2e/wizard-platform-regressions.spec.ts
@@ -715,6 +715,8 @@ test.describe('wizard and QR platform regressions', () => {
     await expect(dialog.getByRole('button', { name: 'Retry' })).toBeEnabled();
     await dialog.getByRole('button', { name: 'Cancel' }).click();
     await expect(dialog).toHaveCount(0);
-    await expect(page.getByRole('button', { name: /^Save$/ })).toBeVisible();
+    await expect(
+      page.getByRole('button', { name: /^Save(?: and enable)?$/ }),
+    ).toBeVisible();
   });
 });
diff --git a/web/tests/e2e/workspace-assistant.spec.ts b/web/tests/e2e/workspace-assistant.spec.ts
new file mode 100644
index 000000000..1f0a4edcc
--- /dev/null
+++ b/web/tests/e2e/workspace-assistant.spec.ts
@@ -0,0 +1,811 @@
+import { expect, test } from '@playwright/test';
+import { installLangBotApiMocks } from './fixtures/langbot-api';
+
+const chat = {
+  uuid: 'assistant-test',
+  revision: 0,
+  status: 'ready',
+  messages: [],
+  pending: [],
+  error: null,
+  model_uuid: null,
+  model_name: null,
+};
+
+test('resizes from the top handle, keeps height on reopen, and fits small viewports', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Model' } },
+    }),
+  );
+  await page.emulateMedia({ reducedMotion: 'reduce' });
+  await page.goto('/home/bots');
+  const panel = page.getByRole('dialog', { name: 'Workspace assistant' });
+  const handle = panel.getByRole('separator', {
+    name: 'Resize assistant height',
+  });
+  await expect(handle).toBeVisible();
+  const initial = (await panel.boundingBox())!;
+  const grip = (await handle.boundingBox())!;
+  await page.mouse.move(grip.x + grip.width / 2, grip.y + grip.height / 2);
+  await page.mouse.down();
+  await page.mouse.move(
+    grip.x + grip.width / 2,
+    grip.y + grip.height / 2 - 140,
+    { steps: 8 },
+  );
+  await page.mouse.up();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.height)
+    .toBeCloseTo(initial.height + 140, 0);
+  const expanded = (await panel.boundingBox())!;
+  expect(expanded.y + expanded.height).toBeCloseTo(
+    initial.y + initial.height,
+    0,
+  );
+  await panel
+    .getByRole('button', { name: 'Collapse workspace assistant', exact: true })
+    .click();
+  const dock = page.getByRole('button', {
+    name: 'Expand workspace assistant',
+    exact: true,
+  });
+  await expect(dock).toBeVisible();
+  expect((await dock.boundingBox())!.height).toBe(21);
+  await dock.click();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.height)
+    .toBeCloseTo(expanded.height, 0);
+  await page.reload();
+  await expect(panel).toBeVisible();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.height)
+    .toBeCloseTo(expanded.height, 0);
+  await handle.focus();
+  await page.keyboard.press('Home');
+  await expect.poll(async () => (await panel.boundingBox())!.height).toBe(280);
+  await page.keyboard.press('ArrowUp');
+  await expect.poll(async () => (await panel.boundingBox())!.height).toBe(304);
+  await page.keyboard.press('End');
+  const preferredHeight = page.viewportSize()!.height - 48;
+  await expect
+    .poll(async () => (await panel.boundingBox())!.height)
+    .toBe(page.viewportSize()!.height - 48);
+  await page.setViewportSize({ width: 375, height: 580 });
+  await expect
+    .poll(async () => (await panel.boundingBox())!.y)
+    .toBeGreaterThanOrEqual(48);
+  await expect(panel.getByRole('textbox')).toBeVisible();
+  await expect.poll(async () => (await panel.boundingBox())!.height).toBe(532);
+  await page.reload();
+  await expect(panel).toBeVisible();
+  await expect.poll(async () => (await panel.boundingBox())!.height).toBe(532);
+  expect(
+    await page.evaluate(() =>
+      localStorage.getItem('langbot-assistant-panel-height'),
+    ),
+  ).toBe(String(preferredHeight));
+  await page.setViewportSize({ width: 1280, height: 1000 });
+  await expect
+    .poll(async () => (await panel.boundingBox())!.height)
+    .toBe(preferredHeight);
+  const moveHandle = panel.getByRole('slider', {
+    name: 'Move assistant horizontally',
+  });
+  const beforeMove = (await panel.boundingBox())!;
+  const title = (await moveHandle.boundingBox())!;
+  await page.mouse.move(title.x + title.width / 2, title.y + title.height / 2);
+  await page.mouse.down();
+  await page.mouse.move(
+    title.x + title.width / 2 + 180,
+    title.y + title.height / 2,
+    { steps: 8 },
+  );
+  await page.mouse.up();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.x)
+    .toBeCloseTo(beforeMove.x + 180, 0);
+  await page.reload();
+  await expect(panel).toBeVisible();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.x)
+    .toBeCloseTo(beforeMove.x + 180, 0);
+  await moveHandle.focus();
+  await page.keyboard.press('End');
+  await expect
+    .poll(async () => {
+      const box = (await panel.boundingBox())!;
+      return box.x + box.width;
+    })
+    .toBe(1268);
+  await page.keyboard.press('Home');
+  await expect.poll(async () => (await panel.boundingBox())!.x).toBe(12);
+  await page.setViewportSize({ width: 375, height: 580 });
+  await expect.poll(async () => (await panel.boundingBox())!.x).toBe(12);
+  expect((await panel.boundingBox())!.width).toBe(351);
+  await page.reload();
+  await expect(panel).toBeVisible();
+  await expect.poll(async () => (await panel.boundingBox())!.x).toBe(12);
+  await page.setViewportSize({ width: 1440, height: 1000 });
+  await expect.poll(async () => (await panel.boundingBox())!.x).toBe(12);
+  await panel
+    .getByRole('button', { name: 'Collapse workspace assistant', exact: true })
+    .click();
+  await expect(dock).toBeVisible();
+  const dockBox = (await dock.boundingBox())!;
+  const startX = dockBox.x + dockBox.width / 2;
+  const startY = dockBox.y + dockBox.height / 2;
+  await page.mouse.move(startX, startY);
+  await page.mouse.down();
+  await page.mouse.move(startX + 150, startY, { steps: 8 });
+  await page.mouse.up();
+  await expect(dock).toBeVisible();
+  await expect(panel).toHaveCount(0);
+  await expect
+    .poll(async () => (await dock.boundingBox())!.x)
+    .toBeCloseTo(dockBox.x + 150, 0);
+  await dock.click();
+  await expect(panel).toBeVisible();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.x)
+    .toBeCloseTo(162, 0);
+  await page.reload();
+  await expect(panel).toBeVisible();
+  await expect
+    .poll(async () => (await panel.boundingBox())!.x)
+    .toBeCloseTo(162, 0);
+});
+
+test('sending survives model labels appearing and disappearing in stream snapshots', async ({
+  page,
+}) => {
+  const errors: string[] = [];
+  page.on('pageerror', (error) => errors.push(error.message));
+  page.on('console', (message) => {
+    if (/createPortal|removeChild|createRoot/.test(message.text()))
+      errors.push(message.text());
+  });
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Recommended model' } },
+    }),
+  );
+  let current: Record<string, unknown> = chat;
+  await page.route('**/api/v1/assistant/conversations', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data:
+          route.request().method() === 'POST' ? chat : { conversations: [] },
+      },
+    }),
+  );
+  await page.route(
+    '**/api/v1/assistant/conversations/assistant-test',
+    (route) =>
+      route.fulfill({
+        json: { code: 0, data: current },
+      }),
+  );
+  let turn = 0;
+  await page.route(
+    '**/api/v1/assistant/conversations/*/turn/stream',
+    async (route) => {
+      turn += 1;
+      current = {
+        ...chat,
+        revision: turn,
+        model_uuid: 'llm-valid',
+        model_name: turn === 2 ? 'Snapshot model' : null,
+        messages: [{ role: 'assistant', content: `Reply ${turn}` }],
+      };
+      await route.fulfill({
+        contentType: 'application/x-ndjson',
+        body:
+          JSON.stringify({ kind: 'snapshot', data: current }) +
+          '\n' +
+          JSON.stringify({ kind: 'completed', data: current }) +
+          '\n',
+      });
+    },
+  );
+  await page.goto('/home/bots');
+  const assistant = page.getByRole('dialog', { name: 'Workspace assistant' });
+  const model = assistant.getByRole('combobox', { name: 'Select Model' });
+  await expect(model).toContainText('Valid Mock Model');
+  for (let i = 1; i <= 3; i++) {
+    await assistant.getByRole('textbox').fill(`Hello ${i}`);
+    await assistant.getByRole('button', { name: 'Send', exact: true }).click();
+    await expect(
+      assistant.getByText(`Reply ${i}`, { exact: true }),
+    ).toBeVisible();
+    await expect(model).toContainText(
+      i === 2 ? 'Snapshot model' : 'Valid Mock Model',
+    );
+  }
+  expect(errors).toEqual([]);
+});
+
+test('selects the available local model on entry and submits the visible model', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Recommended model' } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/conversations', (route) =>
+    route.fulfill({ json: { code: 0, data: chat } }),
+  );
+  await page.route(
+    '**/api/v1/assistant/conversations/*/turn/stream',
+    async (route) => {
+      expect(route.request().postDataJSON().model_uuid).toBe('llm-valid');
+      const result = {
+        ...chat,
+        revision: 1,
+        model_uuid: 'llm-valid',
+        model_name: 'Recommended model',
+        messages: [
+          { role: 'user', content: 'Hello' },
+          { role: 'assistant', content: 'Hello back' },
+        ],
+      };
+      await route.fulfill({
+        contentType: 'application/x-ndjson',
+        body: JSON.stringify({ kind: 'completed', data: result }) + '\n',
+      });
+    },
+  );
+  await page.goto('/home/bots');
+  const assistant = page.getByRole('dialog', { name: 'Workspace assistant' });
+  await expect(
+    assistant.getByRole('combobox', { name: 'Select Model' }),
+  ).toContainText('Valid Mock Model');
+  await assistant.getByRole('textbox').fill('Hello');
+  await assistant.getByRole('button', { name: 'Send', exact: true }).click();
+  await expect(
+    assistant.getByText('Hello back', { exact: true }),
+  ).toBeVisible();
+});
+
+test('late recommendation cannot overwrite manual selection, including request payload', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  let release!: () => void;
+  const gate = new Promise<void>((resolve) => {
+    release = resolve;
+  });
+  await page.route('**/api/v1/assistant/recommended-model', async (route) => {
+    await gate;
+    await route.fulfill({
+      json: {
+        code: 0,
+        data: { uuid: 'recommended-other', name: 'Other recommendation' },
+      },
+    });
+  });
+  await page.route('**/api/v1/assistant/conversations', (route) =>
+    route.fulfill({ json: { code: 0, data: chat } }),
+  );
+  await page.route(
+    '**/api/v1/assistant/conversations/*/turn/stream',
+    async (route) => {
+      expect(route.request().postDataJSON().model_uuid).toBe('llm-valid');
+      await route.fulfill({
+        contentType: 'application/x-ndjson',
+        body:
+          JSON.stringify({
+            kind: 'completed',
+            data: {
+              ...chat,
+              revision: 1,
+              model_uuid: 'llm-valid',
+              model_name: 'Valid Mock Model',
+              messages: [{ role: 'assistant', content: 'Manual model used' }],
+            },
+          }) + '\n',
+      });
+    },
+  );
+  await page.goto('/home/bots');
+  const assistant = page.getByRole('dialog', { name: 'Workspace assistant' });
+  await assistant.getByRole('combobox', { name: 'Select Model' }).click();
+  await page.getByRole('option', { name: /Valid Mock Model/ }).click();
+  release();
+  await expect(assistant.getByText('Preparing assistant…')).toHaveCount(0);
+  await expect(
+    assistant.getByRole('combobox', { name: 'Select Model' }),
+  ).toContainText('Valid Mock Model');
+  await assistant.getByRole('textbox').fill('Hello');
+  await assistant.getByRole('button', { name: 'Send', exact: true }).click();
+  await expect(
+    assistant.getByText('Manual model used', { exact: true }),
+  ).toBeVisible();
+});
+
+test('renders streaming text and tool progress before completion', async ({
+  page,
+}) => {
+  const { createServer } = await import('node:http');
+  let release!: () => void;
+  const gate = new Promise<void>((resolve) => {
+    release = resolve;
+  });
+  const started = {
+    ...chat,
+    revision: 1,
+    status: 'running',
+    model_uuid: 'llm-valid',
+    model_name: 'Recommended model',
+    messages: [{ role: 'user', content: 'Inspect resources' }],
+  };
+  let currentConversation: Record<string, unknown> = started;
+  const server = createServer(async (req, res) => {
+    res.setHeader('Access-Control-Allow-Origin', req.headers.origin || '*');
+    res.setHeader('Access-Control-Allow-Credentials', 'true');
+    res.setHeader(
+      'Access-Control-Allow-Headers',
+      req.headers['access-control-request-headers'] || '*',
+    );
+    if (req.method === 'OPTIONS') {
+      res.end();
+      return;
+    }
+    res.writeHead(200, { 'Content-Type': 'application/x-ndjson' });
+    const emit = (kind: string, data: unknown) =>
+      res.write(JSON.stringify({ kind, data }) + '\n');
+    emit('snapshot', started);
+    emit('phase', { phase: 'thinking', round: 1 });
+    emit('text', { text: 'Inspecting your resources' });
+    emit('phase', {
+      phase: 'tool',
+      tool: { name: 'list_resources', arguments: { kind: 'models' } },
+    });
+    await gate;
+    currentConversation = {
+      ...started,
+      status: 'ready',
+      messages: [
+        ...started.messages,
+        {
+          role: 'tool',
+          content: '{}',
+          tool: {
+            name: 'list_resources',
+            arguments: { kind: 'models' },
+            result: { total: 1, items: [] },
+          },
+        },
+        { role: 'assistant', content: 'Found your model' },
+      ],
+    };
+    emit('completed', currentConversation);
+    res.end();
+  });
+  await new Promise<void>((resolve) => server.listen(0, '127.0.0.1', resolve));
+  const address = server.address() as { port: number };
+  try {
+    await installLangBotApiMocks(page, {
+      authenticated: true,
+      withAssistant: true,
+    });
+    // The panel refreshes the persisted conversation after streaming finishes.
+    // Return the same snapshot as the stream instead of the generic API fallback.
+    await page.route(
+      '**/api/v1/assistant/conversations/assistant-test',
+      (route) =>
+        route.fulfill({ json: { code: 0, data: currentConversation } }),
+    );
+    await page.route('**/api/v1/provider/providers', (route) =>
+      route.fulfill({
+        json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+      }),
+    );
+    await page.route('**/api/v1/assistant/recommended-model', (route) =>
+      route.fulfill({
+        json: {
+          code: 0,
+          data: { uuid: 'llm-valid', name: 'Recommended model' },
+        },
+      }),
+    );
+    await page.route('**/api/v1/assistant/conversations', (route) =>
+      route.fulfill({ json: { code: 0, data: chat } }),
+    );
+    await page.route(
+      '**/api/v1/assistant/conversations/*/turn/stream',
+      (route) =>
+        route.continue({ url: `http://127.0.0.1:${address.port}/stream` }),
+    );
+    await page.goto('/home/bots');
+    const assistant = page.getByRole('dialog', { name: 'Workspace assistant' });
+    await expect(
+      assistant.getByRole('combobox', { name: 'Select Model' }),
+    ).toContainText('Valid Mock Model');
+    await assistant.getByRole('textbox').fill('Inspect resources');
+    await assistant.getByRole('button', { name: 'Send', exact: true }).click();
+    await expect(
+      assistant.getByText('Inspecting your resources', { exact: true }),
+    ).toBeVisible();
+    await expect(
+      assistant.getByRole('button', { name: /Find chat models Running/ }),
+    ).toBeVisible();
+    await expect(
+      assistant.getByRole('button', { name: 'Stop task', exact: true }),
+    ).toBeEnabled();
+    release();
+    await expect(
+      assistant.getByText('Found your model', { exact: true }),
+    ).toBeVisible();
+    await expect(
+      assistant.getByText('Inspecting your resources', { exact: true }),
+    ).toHaveCount(0);
+    const bubble = assistant
+      .locator('[class*="assistantMessage"]')
+      .filter({ hasText: 'Found your model' });
+    await expect(
+      bubble.getByRole('button', { name: /Find chat models Completed/ }),
+    ).toBeVisible();
+  } finally {
+    release();
+    server.closeAllConnections();
+    await new Promise<void>((resolve) => server.close(() => resolve()));
+  }
+});
+
+test('without usable models shows setup only and rechecks after model settings close', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/requesters', (route) =>
+    route.fulfill({ json: { code: 0, data: { requesters: [] } } }),
+  );
+  let configured = false;
+  let recommendations = 0;
+  await page.route('**/api/v1/provider/models/llm', (route) => {
+    if (configured) return route.fallback();
+    return route.fulfill({ json: { code: 0, data: { models: [] } } });
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          providers: configured
+            ? [
+                {
+                  uuid: 'provider',
+                  name: 'Custom provider',
+                  requester: 'openai-chat-completions',
+                  config: {},
+                },
+              ]
+            : [],
+        },
+      },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) => {
+    recommendations++;
+    return route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Recommended model' } },
+    });
+  });
+  await page.goto('/home/bots');
+  const setup = page.getByRole('dialog', {
+    name: 'Workspace assistant',
+  });
+  await expect(
+    setup.getByRole('button', { name: 'Sign in to LangBot Account' }),
+  ).toBeVisible();
+  await expect(setup.getByRole('textbox')).toHaveCount(0);
+  expect(recommendations).toBe(0);
+  await setup.getByRole('button', { name: 'Configure models' }).click();
+  await expect(
+    page.getByRole('dialog', { name: 'Settings', exact: true }),
+  ).toBeVisible();
+  configured = true;
+  await page.keyboard.press('Escape');
+  await expect(
+    page
+      .getByRole('dialog', { name: 'Workspace assistant' })
+      .getByRole('combobox', { name: 'Select Model' }),
+  ).toContainText('Valid Mock Model');
+});
+
+test('setup login starts the existing LangBot Account authorization flow', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [] } },
+    }),
+  );
+  await page.route('**/api/v1/provider/models/llm', (route) =>
+    route.fulfill({ json: { code: 0, data: { models: [] } } }),
+  );
+  await page.goto('/home/bots');
+  const request = page.waitForRequest(
+    (request) =>
+      request.url().includes('/space/') && request.url().includes('authorize'),
+  );
+  await page
+    .getByRole('dialog', { name: 'Workspace assistant' })
+    .getByRole('button', { name: 'Sign in to LangBot Account' })
+    .click();
+  expect((await request).url()).toContain('authorize');
+});
+
+test('switches conversations, restores background progress after reload and stops only the selected task', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Recommended model' } },
+    }),
+  );
+  const running = {
+    ...chat,
+    uuid: 'running-chat',
+    revision: 3,
+    status: 'running',
+    model_uuid: 'llm-valid',
+    model_name: 'Recommended model',
+    messages: [{ role: 'user', content: 'Background request' }],
+    progress: {
+      text: 'Working in the background',
+      phase: 'thinking',
+      round: 2,
+    },
+  };
+  const ready = {
+    ...chat,
+    uuid: 'ready-chat',
+    model_uuid: 'llm-valid',
+    model_name: 'Recommended model',
+    messages: [{ role: 'assistant', content: 'Previous answer' }],
+  };
+  let stopped = false;
+  await page.route('**/api/v1/assistant/conversations', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          conversations: [
+            {
+              uuid: running.uuid,
+              title: 'Background request',
+              status: stopped ? 'failed' : 'running',
+            },
+            {
+              uuid: ready.uuid,
+              title: 'Previous conversation',
+              status: 'ready',
+            },
+          ],
+        },
+      },
+    }),
+  );
+  await page.route('**/api/v1/assistant/conversations/*', (route) => {
+    const data = route.request().url().endsWith('ready-chat')
+      ? ready
+      : stopped
+        ? {
+            ...running,
+            status: 'failed',
+            error: 'stopped',
+            progress: undefined,
+          }
+        : running;
+    return route.fulfill({ json: { code: 0, data } });
+  });
+  await page.route(
+    '**/api/v1/assistant/conversations/running-chat/stop',
+    (route) => {
+      expect(route.request().postDataJSON()).toEqual({ revision: 3 });
+      stopped = true;
+      return route.fulfill({
+        json: {
+          code: 0,
+          data: {
+            ...running,
+            status: 'failed',
+            error: 'stopped',
+            progress: undefined,
+          },
+        },
+      });
+    },
+  );
+  await page.goto('/home/bots');
+  const assistant = page.getByRole('dialog', { name: 'Workspace assistant' });
+  await assistant
+    .getByRole('combobox', { name: 'Conversations', exact: true })
+    .click();
+  await page
+    .getByRole('option', { name: '◌ Background request', exact: true })
+    .click();
+  await expect(
+    assistant.getByText('Working in the background', { exact: true }),
+  ).toBeVisible();
+  await assistant
+    .getByRole('combobox', { name: 'Conversations', exact: true })
+    .click();
+  await page
+    .getByRole('option', { name: 'Previous conversation', exact: true })
+    .click();
+  await expect(
+    assistant.getByText('Previous answer', { exact: true }),
+  ).toBeVisible();
+  expect(stopped).toBe(false);
+  await assistant
+    .getByRole('combobox', { name: 'Conversations', exact: true })
+    .click();
+  await page
+    .getByRole('option', { name: '◌ Background request', exact: true })
+    .click();
+  await page.reload();
+  await expect(
+    assistant.getByText('Working in the background', { exact: true }),
+  ).toBeVisible();
+  await assistant
+    .getByRole('button', { name: 'Stop task', exact: true })
+    .click();
+  await expect(assistant.getByRole('alert')).toContainText('Stopped.');
+  await expect(
+    assistant.getByRole('button', { name: 'Stop task', exact: true }),
+  ).toHaveCount(0);
+  await assistant
+    .getByRole('button', { name: 'New conversation', exact: true })
+    .click();
+  await expect(assistant.getByRole('textbox')).toBeEmpty();
+});
+
+test('keeps the reading position through background updates and resumes following at the bottom', async ({
+  page,
+}) => {
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    withAssistant: true,
+  });
+  await page.route('**/api/v1/provider/providers', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { providers: [{ uuid: 'provider' }] } },
+    }),
+  );
+  await page.route('**/api/v1/assistant/recommended-model', (route) =>
+    route.fulfill({
+      json: { code: 0, data: { uuid: 'llm-valid', name: 'Model' } },
+    }),
+  );
+  let updates = 0;
+  await page.route('**/api/v1/assistant/conversations', (route) =>
+    route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          conversations: [
+            {
+              uuid: 'scroll-chat',
+              title: 'Long conversation',
+              status: 'running',
+            },
+          ],
+        },
+      },
+    }),
+  );
+  await page.route('**/api/v1/assistant/conversations/scroll-chat', (route) => {
+    updates++;
+    return route.fulfill({
+      json: {
+        code: 0,
+        data: {
+          ...chat,
+          uuid: 'scroll-chat',
+          status: 'running',
+          model_uuid: 'llm-valid',
+          model_name: 'Model',
+          messages: [
+            {
+              role: 'assistant',
+              content: Array.from(
+                { length: 40 },
+                (_, i) => `Paragraph ${i}`,
+              ).join('\n\n'),
+            },
+          ],
+          progress: {
+            text: Array.from({ length: updates }, (_, i) => `Update ${i}`).join(
+              '\n\n',
+            ),
+            phase: 'thinking',
+          },
+        },
+      },
+    });
+  });
+  await page.goto('/home/bots');
+  const panel = page.getByRole('dialog', { name: 'Workspace assistant' });
+  await panel
+    .getByRole('combobox', { name: 'Conversations', exact: true })
+    .click();
+  await page
+    .getByRole('option', { name: '◌ Long conversation', exact: true })
+    .click();
+  const messages = panel.locator('[aria-live="polite"]');
+  const gap = () =>
+    messages.evaluate((el) => el.scrollHeight - el.clientHeight - el.scrollTop);
+  await expect.poll(gap).toBeLessThan(2);
+  await messages.hover();
+  await page.mouse.wheel(0, -600);
+  await expect.poll(gap).toBeGreaterThan(200);
+  const position = await messages.evaluate((el) => el.scrollTop);
+  const previousUpdates = updates;
+  await expect.poll(() => updates).toBeGreaterThan(previousUpdates + 1);
+  await expect
+    .poll(() => messages.evaluate((el) => el.scrollTop))
+    .toBeCloseTo(position, 0);
+  await messages.hover();
+  await page.mouse.wheel(0, 10000);
+  await expect.poll(gap).toBeLessThan(2);
+  const beforeFollow = updates;
+  await expect.poll(() => updates).toBeGreaterThan(beforeFollow + 1);
+  await expect.poll(gap).toBeLessThan(2);
+});
diff --git a/web/tests/e2e/workspace-language.spec.ts b/web/tests/e2e/workspace-language.spec.ts
new file mode 100644
index 000000000..7ff0a5fc1
--- /dev/null
+++ b/web/tests/e2e/workspace-language.spec.ts
@@ -0,0 +1,80 @@
+import { expect, test } from '@playwright/test';
+import {
+  installLangBotApiMocks,
+  makeWorkspaceEntry,
+} from './fixtures/langbot-api';
+
+test('workspace language persists independently of personal UI language', async ({
+  page,
+}, testInfo) => {
+  const workspace = makeWorkspaceEntry(
+    'workspace-language',
+    'Language Workspace',
+    'local',
+  );
+  workspace.permissions = ['workspace.view', 'workspace.update'];
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    workspaces: [workspace],
+  });
+  let language = 'en_US';
+  await page.route('**/api/v1/workspaces/current/language', async (route) => {
+    if (route.request().method() === 'PUT') {
+      expect(route.request().headers()['x-workspace-id']).toBe(
+        'workspace-language',
+      );
+      language = route.request().postDataJSON().language;
+    }
+    await route.fulfill({
+      json: {
+        code: 0,
+        msg: 'ok',
+        data: {
+          language,
+          supported_languages: ['en_US', 'zh_Hans', 'ja_JP'],
+        },
+      },
+    });
+  });
+  await page.goto('/home?action=showWorkspaceSettings');
+  const select = page.getByRole('combobox', {
+    name: 'Workspace language',
+    exact: true,
+  });
+  await expect(select).toHaveText('English');
+  await select.click();
+  await page.getByRole('option', { name: '日本語', exact: true }).click();
+  await expect(select).toHaveText('日本語');
+  await expect(page.getByText('Language saved', { exact: true })).toBeVisible();
+  await page.reload();
+  await page.goto('/home?action=showWorkspaceSettings');
+  await expect(select).toHaveText('日本語');
+  await expect(
+    page.getByText('Workspace language', { exact: true }),
+  ).toBeVisible();
+  await page.screenshot({
+    path: testInfo.outputPath('workspace-language.png'),
+    fullPage: true,
+  });
+});
+
+test('viewer can read but cannot edit workspace language', async ({ page }) => {
+  const workspace = makeWorkspaceEntry(
+    'workspace-language',
+    'Language Workspace',
+    'local',
+  );
+  workspace.membership.role = 'viewer';
+  workspace.permissions = ['workspace.view'];
+  await installLangBotApiMocks(page, {
+    authenticated: true,
+    workspaces: [workspace],
+  });
+  await page.goto('/home?action=showWorkspaceSettings');
+  const select = page.getByRole('combobox', {
+    name: 'Workspace language',
+    exact: true,
+  });
+  await expect(select).toHaveText('English');
+  await expect(select).toBeDisabled();
+});
diff --git a/web/tests/revisions/index.html b/web/tests/revisions/index.html
new file mode 100644
index 000000000..90e9c1c0b
--- /dev/null
+++ b/web/tests/revisions/index.html
@@ -0,0 +1,9 @@
+<html>
+  <head>
+    <meta charset="utf-8" />
+  </head>
+  <body>
+    <div id="root"></div>
+    <script type="module" src="./preview.tsx"></script>
+  </body>
+</html>
diff --git a/web/tests/revisions/preview.tsx b/web/tests/revisions/preview.tsx
new file mode 100644
index 000000000..c40db14b6
--- /dev/null
+++ b/web/tests/revisions/preview.tsx
@@ -0,0 +1,42 @@
+import React from 'react';
+import { createRoot } from 'react-dom/client';
+import { createInstance } from 'i18next';
+import { I18nextProvider } from 'react-i18next';
+import enUS from '../../src/i18n/locales/en-US';
+import CloudPlansDialog from '../../src/app/home/components/workspace-settings/CloudPlansDialog';
+import '../../src/app/global.css';
+const mode = new URLSearchParams(location.search).get('mode') || 'changed';
+const current = {
+  name: 'pro',
+  subscription_started_at: 1790812800,
+  revision: 1,
+  limits: {
+    'bots.max': 2,
+    'members.max': 3,
+    'chat_history.retention_days': 30,
+  },
+  entitlement_features: { mcp_stdio: false, 'platform.telegram': false },
+  monthly_bonus_credits: 10,
+};
+const i18n = createInstance();
+await i18n.init({
+  lng: 'en-US',
+  resources: { 'en-US': { translation: enUS } },
+});
+createRoot(document.getElementById('root')!).render(
+  <I18nextProvider i18n={i18n}>
+    <CloudPlansDialog
+      open
+      onOpenChange={() => {}}
+      workspace={
+        {
+          workspace: { uuid: 'fixture', name: 'Review workspace' },
+          membership: { role: mode === 'member' ? 'viewer' : 'owner' },
+          plan_name: 'pro',
+          current_plan: mode === 'unknown' ? null : current,
+        } as any
+      }
+      spaceURL="https://space.example.test"
+    />
+  </I18nextProvider>,
+);
diff --git a/web/tests/revisions/verify.mjs b/web/tests/revisions/verify.mjs
new file mode 100644
index 000000000..6d4ba1efb
--- /dev/null
+++ b/web/tests/revisions/verify.mjs
@@ -0,0 +1,225 @@
+import { createServer } from 'vite';
+import { chromium } from '@playwright/test';
+import { fileURLToPath } from 'node:url';
+import path from 'node:path';
+import { mkdir, writeFile } from 'node:fs/promises';
+const workdir = path.resolve(
+  path.dirname(fileURLToPath(import.meta.url)),
+  '../../../..',
+);
+const evidence = path.join(workdir, 'evidence');
+await mkdir(evidence, { recursive: true });
+const servers = [];
+let browser;
+const results = [];
+const latest = (mode) => ({
+  uuid: 'latest',
+  name: 'pro',
+  revision: mode === 'same' ? 1 : 2,
+  display_name: { en_US: 'Pro' },
+  description: { en_US: 'Workspace plan' },
+  price_cny: '70',
+  price_usd: '10',
+  price_jpy: '1500',
+  features: [],
+  limits:
+    mode === 'equal' || mode === 'same'
+      ? { 'bots.max': 2, 'members.max': 3 }
+      : { 'bots.max': 8, 'members.max': 5 },
+  entitlement_features: { mcp_stdio: !['same', 'equal'].includes(mode) },
+  disabled_adapters: ['same', 'equal'].includes(mode) ? ['telegram'] : [],
+  chat_history_retention_days: ['same', 'equal'].includes(mode) ? 30 : 90,
+  monthly_bonus_credits: ['same', 'equal'].includes(mode) ? 10 : 20,
+  is_active: true,
+  subscription_disabled: mode === 'paused',
+  sort_order: 1,
+  highlight: false,
+});
+try {
+  for (const [repo] of [
+    ['space', 4185],
+    ['core', 4186],
+  ]) {
+    const server = await createServer({
+      configFile: path.join(
+        workdir,
+        repo,
+        'web/tests/revisions/vite.config.mjs',
+      ),
+    });
+    await server.listen();
+    servers.push(server);
+  }
+  browser = await chromium.launch({ headless: true });
+  for (const repo of ['space', 'core']) {
+    const modes =
+      repo === 'space'
+        ? [
+            'changed',
+            'same',
+            'equal',
+            'paused',
+            'removed',
+            'member',
+            'stripe',
+            'cancel-failure',
+          ]
+        : [
+            'changed',
+            'same',
+            'equal',
+            'paused',
+            'removed',
+            'member',
+            'unknown',
+            'error',
+          ];
+    for (const mode of modes) {
+      const page = await browser.newPage({
+        viewport: { width: 1360, height: 1100 },
+      });
+      page.setDefaultTimeout(120000);
+      const errors = [];
+      page.on('pageerror', (error) => {
+        errors.push(error.message);
+        console.error(error.message);
+      });
+      await page.route('**/*', async (route) => {
+        const url = new URL(route.request().url());
+        if (url.hostname === '127.0.0.1') return route.continue();
+        if (url.pathname === '/api/v1/cloud/v2/plans')
+          return route.fulfill({
+            status: mode === 'error' ? 500 : 200,
+            contentType: 'application/json',
+            body: JSON.stringify({
+              code: 0,
+              data: mode === 'removed' ? [] : [latest(mode)],
+            }),
+          });
+        return route.abort();
+      });
+      const port = repo === 'space' ? 4185 : 4186;
+      await page.goto(
+        `http://127.0.0.1:${port}/tests/revisions/index.html?mode=${mode}${mode === 'stripe' ? '&intent=1' : ''}`,
+        { waitUntil: 'domcontentloaded', timeout: 120000 },
+      );
+      if (repo === 'core' && mode === 'error')
+        await page.getByRole('alert').waitFor();
+      else if (mode === 'unknown')
+        await page
+          .getByText(/Current subscription details are unavailable/)
+          .waitFor();
+      else
+        await page
+          .locator(
+            repo === 'space'
+              ? '[data-plan-comparison]'
+              : '[data-cloud-plans] article',
+          )
+          .first()
+          .waitFor();
+      if (errors.length)
+        throw new Error(`${repo}/${mode}: ${errors.join(';')}`);
+      if (await page.locator('[data-revision-comparison]').count())
+        throw new Error('Standalone comparison remains');
+      if (
+        ['same', 'equal'].includes(mode) &&
+        (await page.locator('[data-owned-plan]').count())
+      )
+        throw new Error('Identical benefits duplicated');
+      const area = page.locator(
+        repo === 'space' ? '[data-plan-comparison]' : '[data-cloud-plans]',
+      );
+      if (!['unknown', 'error'].includes(mode)) {
+        const copy = await area.innerText();
+        if (
+          /&#|revision|mcp stdio|workspace multi user|platform telegram/.test(
+            copy,
+          )
+        )
+          throw new Error('Internal metadata leaked into plan presentation');
+        if (
+          repo === 'core' &&
+          (await area.locator('[data-changed], .bg-amber-500\\/10').count())
+        )
+          throw new Error('Diff highlighting remains');
+        if (
+          ['same', 'equal'].includes(mode) &&
+          /Subscribed on|Latest quotas for this plan/.test(copy)
+        )
+          throw new Error('Unnecessary comparison labels');
+      }
+      if (mode === 'changed') {
+        await area.getByText(/Subscribed on/).waitFor();
+        await area
+          .getByText('Latest quotas for this plan', { exact: true })
+          .waitFor();
+        const owned = page.locator('[data-owned-plan]');
+        if ((await owned.count()) !== 1)
+          throw new Error('Missing integrated owned version');
+        const horizontal = await owned.evaluate((el) => {
+          const next = el.nextElementSibling;
+          return (
+            !!next &&
+            Math.abs(
+              next.getBoundingClientRect().y - el.getBoundingClientRect().y,
+            ) < 5 &&
+            next.getBoundingClientRect().x > el.getBoundingClientRect().x
+          );
+        });
+        if (!horizontal)
+          throw new Error('Owned and latest are not adjacent horizontally');
+      }
+      if (repo === 'space' && mode === 'stripe') {
+        if (await page.getByRole('dialog').count())
+          throw new Error('Deep link bypassed renewal gate');
+        await page
+          .getByRole('button', { name: 'Manage subscription' })
+          .first()
+          .click();
+        await page.getByText('Server refresh confirmed cancellation').waitFor();
+        if (!(await page.getByRole('dialog').count()))
+          await page
+            .getByRole('button', { name: /Upgrade to|Subscribe to|Choose/ })
+            .last()
+            .click();
+        await page.getByRole('dialog').waitFor();
+      }
+      if (repo === 'space' && mode === 'cancel-failure') {
+        await page
+          .getByRole('button', { name: 'Manage subscription' })
+          .first()
+          .click();
+        await page.getByText('Cancellation failed').waitFor();
+        if (await page.getByRole('dialog').count())
+          throw new Error('Failed cancellation opened checkout');
+      }
+      if (
+        mode === 'member' &&
+        repo === 'core' &&
+        (await page.getByRole('link').count())
+      )
+        throw new Error('Member purchase link');
+      if (
+        mode === 'changed' ||
+        mode === 'equal' ||
+        mode === 'member' ||
+        mode === 'paused'
+      )
+        await page.screenshot({
+          path: path.join(evidence, `${repo}-${mode}.png`),
+          fullPage: true,
+        });
+      results.push(`${repo}/${mode}: PASS`);
+      console.log(results.at(-1));
+      await page.close();
+    }
+  }
+} finally {
+  await writeFile(
+    path.join(evidence, 'browser-results.txt'),
+    results.join('\n') + '\n',
+  );
+  if (browser) await browser.close();
+  for (const server of servers) await server.close();
+}
diff --git a/web/tests/revisions/vite.config.mjs b/web/tests/revisions/vite.config.mjs
new file mode 100644
index 000000000..5021b7626
--- /dev/null
+++ b/web/tests/revisions/vite.config.mjs
@@ -0,0 +1,14 @@
+import path from 'node:path';
+import tailwind from '@tailwindcss/postcss';
+import { fileURLToPath } from 'node:url';
+const root = path.resolve(
+  path.dirname(fileURLToPath(import.meta.url)),
+  '../..',
+);
+export default {
+  root,
+  resolve: { alias: { '@': path.join(root, 'src') } },
+  esbuild: { jsx: 'automatic' },
+  css: { postcss: { plugins: [tailwind({ base: root })] } },
+  server: { host: '127.0.0.1', port: 4186, strictPort: true },
+};
diff --git a/web/tests/unit/assistant-dock-rail.test.mjs b/web/tests/unit/assistant-dock-rail.test.mjs
deleted file mode 100644
index c1a6a9156..000000000
--- a/web/tests/unit/assistant-dock-rail.test.mjs
+++ /dev/null
@@ -1,175 +0,0 @@
-import assert from 'node:assert/strict';
-import fs from 'node:fs';
-import test from 'node:test';
-import ts from 'typescript';
-
-// The floating assistant button docks against a viewport edge, collapses into a
-// short blue rail, and re-expands on hover. The transition rules are pure so
-// they can be pinned here: the race this guards against is a button that drops
-// under the cursor and never becomes visible again.
-const sourcePath = new URL(
-  '../../src/app/home/components/assistant-dock.ts',
-  import.meta.url,
-);
-assert.ok(fs.existsSync(sourcePath), 'Missing assistant dock module');
-const compiled = ts.transpileModule(fs.readFileSync(sourcePath, 'utf8'), {
-  compilerOptions: { module: ts.ModuleKind.CommonJS },
-}).outputText;
-const loaded = { exports: {} };
-new Function('require', 'module', 'exports', compiled)(
-  () => {
-    throw new Error('assistant-dock must stay dependency free');
-  },
-  loaded,
-  loaded.exports,
-);
-const {
-  ASSISTANT_BUTTON_SIZE,
-  ASSISTANT_SNAP_THRESHOLD,
-  ASSISTANT_RAIL_WIDTH,
-  clampAssistantPosition,
-  resolveAssistantEdge,
-  shouldCollapseRail,
-  shouldExpandRail,
-  restingRailExpanded,
-} = loaded.exports;
-
-const viewport = { width: 1280, height: 800 };
-const maxX = viewport.width - ASSISTANT_BUTTON_SIZE;
-
-test('clamping keeps the button fully inside the viewport', () => {
-  assert.deepEqual(clampAssistantPosition(-50, -50, viewport), { x: 0, y: 0 });
-  assert.deepEqual(clampAssistantPosition(9999, 9999, viewport), {
-    x: maxX,
-    y: viewport.height - ASSISTANT_BUTTON_SIZE,
-  });
-  assert.deepEqual(clampAssistantPosition(400, 300, viewport), {
-    x: 400,
-    y: 300,
-  });
-});
-
-test('a narrow viewport never produces a negative travel range', () => {
-  assert.deepEqual(clampAssistantPosition(10, 10, { width: 20, height: 20 }), {
-    x: 0,
-    y: 0,
-  });
-});
-
-test('positions within the snap threshold dock to the nearest edge', () => {
-  assert.equal(resolveAssistantEdge(0, viewport), 'left');
-  assert.equal(
-    resolveAssistantEdge(ASSISTANT_SNAP_THRESHOLD, viewport),
-    'left',
-  );
-  assert.equal(resolveAssistantEdge(maxX, viewport), 'right');
-  assert.equal(
-    resolveAssistantEdge(maxX - ASSISTANT_SNAP_THRESHOLD, viewport),
-    'right',
-  );
-});
-
-test('a centred position stays free floating', () => {
-  assert.equal(resolveAssistantEdge(Math.round(maxX / 2), viewport), null);
-  assert.equal(
-    resolveAssistantEdge(ASSISTANT_SNAP_THRESHOLD + 1, viewport),
-    null,
-  );
-});
-
-test('the left edge wins ties so a centred drop is deterministic', () => {
-  // A viewport sized so both gaps are inside the snap threshold at x = 0.
-  const tight = { width: ASSISTANT_BUTTON_SIZE, height: 400 };
-  assert.equal(resolveAssistantEdge(0, tight), 'left');
-});
-
-test('hovering a docked rail expands it, hovering a free button does not', () => {
-  assert.equal(
-    shouldExpandRail({
-      dockedEdge: 'right',
-      railExpanded: false,
-      dragging: false,
-    }),
-    true,
-  );
-  assert.equal(
-    shouldExpandRail({
-      dockedEdge: null,
-      railExpanded: false,
-      dragging: false,
-    }),
-    false,
-  );
-  // Already expanded: nothing to do, so no redundant state churn on every move.
-  assert.equal(
-    shouldExpandRail({
-      dockedEdge: 'left',
-      railExpanded: true,
-      dragging: false,
-    }),
-    false,
-  );
-});
-
-test('a drag in progress never expands or collapses the rail', () => {
-  const dragging = { dockedEdge: 'left', railExpanded: false, dragging: true };
-  assert.equal(shouldExpandRail(dragging), false);
-  assert.equal(shouldCollapseRail(dragging), false);
-});
-
-test('only an *expanded* docked button collapses when the pointer leaves', () => {
-  // Regression: keying this on `!railExpanded` made a revealed button
-  // impossible to collapse again, so it stayed open forever after one hover.
-  assert.equal(
-    shouldCollapseRail({
-      dockedEdge: 'left',
-      railExpanded: true,
-      dragging: false,
-    }),
-    true,
-  );
-  assert.equal(
-    shouldCollapseRail({
-      dockedEdge: 'left',
-      railExpanded: false,
-      dragging: false,
-    }),
-    false,
-  );
-  assert.equal(
-    shouldCollapseRail({
-      dockedEdge: null,
-      railExpanded: true,
-      dragging: false,
-    }),
-    false,
-  );
-});
-
-test('expand then leave round-trips back to the collapsed rail', () => {
-  const docked = { dockedEdge: 'right', railExpanded: false, dragging: false };
-  // Hover: reveal the button.
-  assert.equal(shouldExpandRail(docked), true);
-  // Pointer leaves: the revealed button must collapse again.
-  assert.equal(shouldCollapseRail({ ...docked, railExpanded: true }), true);
-  // A second hover re-reveals it, so the cycle is repeatable.
-  assert.equal(shouldExpandRail(docked), true);
-});
-
-test('dropping always starts collapsed so the rail is never stuck open', () => {
-  // The pointer is still over the button on drop; the next pointermove expands.
-  assert.equal(restingRailExpanded(), false);
-  assert.equal(
-    shouldExpandRail({
-      dockedEdge: 'right',
-      railExpanded: restingRailExpanded(),
-      dragging: false,
-    }),
-    true,
-  );
-});
-
-test('the rail strip stays thinner than the button it replaces', () => {
-  assert.ok(ASSISTANT_RAIL_WIDTH > 0);
-  assert.ok(ASSISTANT_RAIL_WIDTH < ASSISTANT_BUTTON_SIZE);
-});
diff --git a/web/tests/unit/cloud-plans.test.mjs b/web/tests/unit/cloud-plans.test.mjs
new file mode 100644
index 000000000..d8208ac43
--- /dev/null
+++ b/web/tests/unit/cloud-plans.test.mjs
@@ -0,0 +1,194 @@
+import assert from 'node:assert/strict';
+import test from 'node:test';
+import { readFileSync } from 'node:fs';
+import ts from 'typescript';
+const source = readFileSync(
+  new URL(
+    '../../src/app/home/components/workspace-settings/cloud-plans.ts',
+    import.meta.url,
+  ),
+  'utf8',
+);
+const m = { exports: {} };
+new Function(
+  'module',
+  'exports',
+  ts.transpileModule(source, {
+    compilerOptions: {
+      module: ts.ModuleKind.CommonJS,
+      target: ts.ScriptTarget.ES2022,
+    },
+  }).outputText,
+)(m, m.exports);
+const {
+  cloudPlanCheckoutURL,
+  cloudSubscriptionManagementURL,
+  cloudPlanPrice,
+  localizedPlanText,
+  fetchPublicCloudPlans,
+} = m.exports;
+
+test('checkout links preserve the workspace and selection, never source credentials', () => {
+  const url = new URL(
+    cloudPlanCheckoutURL(
+      'https://space.example/prefix?token=secret#access_token=secret',
+      'workspace-1',
+      'pro & team',
+      'monthly',
+      'JPY',
+    ),
+  );
+  assert.equal(url.pathname, '/prefix/cloud');
+  assert.equal(url.hash, '');
+  assert.deepEqual(Object.fromEntries(url.searchParams), {
+    environment: 'new',
+    step: 'plan',
+    workspace: 'workspace-1',
+    plan: 'pro & team',
+    billing: 'monthly',
+    currency: 'JPY',
+  });
+  assert.throws(() =>
+    cloudPlanCheckoutURL(
+      'https://user:secret@space.example',
+      'w',
+      'pro',
+      'annual',
+      'USD',
+    ),
+  );
+  assert.throws(() =>
+    cloudPlanCheckoutURL('javascript:alert(1)', 'w', 'pro', 'annual', 'USD'),
+  );
+});
+test('prices match server annual discount and locale fallback', () => {
+  const plan = { price_cny: '29', price_usd: '5.5', price_jpy: '800' };
+  assert.equal(cloudPlanPrice(plan, 'monthly', 'USD'), 5.5);
+  assert.equal(cloudPlanPrice(plan, 'annual', 'CNY'), 290);
+  assert.equal(cloudPlanPrice(plan, 'annual', 'JPY'), 8000);
+  assert.equal(
+    localizedPlanText({ zh_Hans: '专业版', en_US: 'Pro' }, 'zh-Hans'),
+    '专业版',
+  );
+  assert.equal(localizedPlanText({ en_US: 'Pro' }, 'ja-JP'), 'Pro');
+  assert.throws(() =>
+    cloudPlanPrice({ ...plan, price_usd: 'NaN' }, 'monthly', 'USD'),
+  );
+});
+test('public catalog fetch omits cookies and authorization, hides inactive plans, validates response', async (t) => {
+  let options, url;
+  const plan = {
+    uuid: '1',
+    name: 'pro',
+    limits: {},
+    price_cny: '29',
+    price_usd: '5',
+    price_jpy: '800',
+    is_active: true,
+    subscription_disabled: true,
+  };
+  t.mock.method(globalThis, 'fetch', async (u, o) => {
+    url = u;
+    options = o;
+    return {
+      ok: true,
+      json: async () => ({
+        code: 0,
+        data: [plan, { ...plan, uuid: 'hidden', is_active: false }],
+      }),
+    };
+  });
+  assert.deepEqual(
+    await fetchPublicCloudPlans(
+      'https://space.example',
+      new AbortController().signal,
+    ),
+    [plan],
+  );
+  assert.equal(url.pathname, '/api/v1/cloud/v2/plans');
+  assert.equal(options.credentials, 'omit');
+  assert.equal(options.referrerPolicy, 'no-referrer');
+  assert.equal(options.headers, undefined);
+  globalThis.fetch = async () => ({ ok: false });
+  await assert.rejects(fetchPublicCloudPlans('https://space.example'));
+  globalThis.fetch = async () => ({
+    ok: true,
+    json: async () => ({ code: 0, data: [{ ...plan, price_cny: 'broken' }] }),
+  });
+  await assert.rejects(fetchPublicCloudPlans('https://space.example'));
+});
+
+const platformsModule = { exports: {} };
+new Function(
+  'module',
+  'exports',
+  ts.transpileModule(
+    readFileSync(
+      new URL(
+        '../../src/app/home/components/workspace-settings/cloud-plan-platforms.ts',
+        import.meta.url,
+      ),
+      'utf8',
+    ),
+    {
+      compilerOptions: {
+        module: ts.ModuleKind.CommonJS,
+        target: ts.ScriptTarget.ES2022,
+      },
+    },
+  ).outputText,
+)(platformsModule, platformsModule.exports);
+const { supportedPlanPlatforms, KNOWN_ADAPTERS } = platformsModule.exports;
+test('supported platforms hide blocked families and merge Omni variants', () => {
+  const platforms = supportedPlanPlatforms([
+    'aiocqhttp',
+    'line',
+    'discord-eba',
+  ]);
+  const names = platforms.map((platform) => platform.name);
+  assert.equal(names.includes('aiocqhttp'), false);
+  assert.equal(names.includes('LINE'), false);
+  assert.equal(names.includes('discord'), true);
+  assert.equal(
+    names.some((name) => name.endsWith('-omni')),
+    false,
+  );
+  assert.equal(new Set(names).size, names.length);
+  assert.ok(platforms.every((platform) => platform.label && platform.icon));
+  assert.deepEqual(
+    supportedPlanPlatforms(KNOWN_ADAPTERS.map((adapter) => adapter.name)),
+    [],
+  );
+  assert.deepEqual(
+    supportedPlanPlatforms(['unknown-adapter']),
+    supportedPlanPlatforms(),
+  );
+});
+
+test('plan currency follows the current UI language', () => {
+  for (const locale of ['zh-Hans', 'zh_Hans', 'zh-Hans-CN'])
+    assert.equal(m.exports.cloudPlanCurrencyForLanguage(locale), 'CNY');
+  for (const locale of ['ja-JP', 'ja_JP', 'ja'])
+    assert.equal(m.exports.cloudPlanCurrencyForLanguage(locale), 'JPY');
+  for (const locale of [
+    'zh-Hant',
+    'zh_Hant',
+    'en-US',
+    'es-ES',
+    'ru-RU',
+    'th-TH',
+    'vi-VN',
+  ])
+    assert.equal(m.exports.cloudPlanCurrencyForLanguage(locale), 'USD');
+});
+
+test('subscription management links use the production environment', () => {
+  const url = new URL(
+    cloudSubscriptionManagementURL('https://space.example', 'w'),
+  );
+  assert.deepEqual(Object.fromEntries(url.searchParams), {
+    environment: 'new',
+    step: 'subscription',
+    workspace: 'w',
+  });
+});
diff --git a/web/tests/unit/codex-subscription.test.mjs b/web/tests/unit/codex-subscription.test.mjs
index dfafce695..03399248c 100644
--- a/web/tests/unit/codex-subscription.test.mjs
+++ b/web/tests/unit/codex-subscription.test.mjs
@@ -24,7 +24,7 @@ test('all locale catalogs cover Codex states and preserve the expiry placeholder
   const directory = new URL('../../src/i18n/locales/', import.meta.url);
   const files = fs
     .readdirSync(directory)
-    .filter((file) => file.endsWith('.ts'));
+    .filter((file) => /^[a-z]{2}-[A-Za-z]+\.ts$/.test(file));
   assert.equal(files.length, 8);
   let expected;
   for (const file of files) {
diff --git a/web/tests/unit/embed-widget-assistant-dedupe.test.mjs b/web/tests/unit/embed-widget-assistant-dedupe.test.mjs
index 42f172d3d..51f93aee4 100644
--- a/web/tests/unit/embed-widget-assistant-dedupe.test.mjs
+++ b/web/tests/unit/embed-widget-assistant-dedupe.test.mjs
@@ -129,11 +129,13 @@ function launchWidget() {
   FakeWebSocket.instances = [];
   const document = new FakeDocument();
   const window = {
+    location: { href: 'https://embed.test/' },
     crypto: { randomUUID: () => '00000000-0000-4000-8000-000000000000' },
     sessionStorage: { getItem: () => null, setItem: () => {} },
   };
   const context = vm.createContext({
     document,
+    URL,
     window,
     navigator: { clipboard: { writeText: () => Promise.resolve() } },
     WebSocket: FakeWebSocket,
diff --git a/web/tests/unit/installed-plugin-icon-cache.test.mjs b/web/tests/unit/installed-plugin-icon-cache.test.mjs
new file mode 100644
index 000000000..81e72046e
--- /dev/null
+++ b/web/tests/unit/installed-plugin-icon-cache.test.mjs
@@ -0,0 +1,179 @@
+import assert from 'node:assert/strict';
+import fs from 'node:fs';
+import path from 'node:path';
+import test from 'node:test';
+import { createRequire } from 'node:module';
+import { fileURLToPath } from 'node:url';
+import ts from 'typescript';
+
+const currentDirectory = path.dirname(fileURLToPath(import.meta.url));
+const sourcePath = path.resolve(
+  currentDirectory,
+  '../../src/app/infra/hooks/useInstalledPluginIcon.ts',
+);
+const source = fs.readFileSync(sourcePath, 'utf8');
+const compiled = ts.transpileModule(source, {
+  compilerOptions: { module: ts.ModuleKind.CommonJS },
+}).outputText;
+
+// --- minimal React + backend mocks so the hook can run under node:test ---
+let stateSlots = [];
+let lastCleanup;
+
+const reactMock = {
+  useState(initial) {
+    const value = typeof initial === 'function' ? initial() : initial;
+    const index = stateSlots.length;
+    stateSlots.push(value);
+    return [
+      value,
+      (next) => {
+        stateSlots[index] =
+          typeof next === 'function' ? next(stateSlots[index]) : next;
+      },
+    ];
+  },
+  useEffect(effect) {
+    const cleanup = effect();
+    if (typeof cleanup === 'function') lastCleanup = cleanup;
+  },
+};
+
+const httpMock = {
+  httpClient: {
+    getAuthenticatedPluginIconURL: () => {
+      throw new Error('fetch not configured for this test');
+    },
+  },
+};
+
+const sourceRequire = createRequire(sourcePath);
+
+function loadHook() {
+  const loadedModule = { exports: {} };
+  const requireShim = (id) => {
+    if (id === 'react') return reactMock;
+    if (id === '@/app/infra/http') return httpMock;
+    return sourceRequire(id);
+  };
+  new Function('require', 'module', 'exports', compiled)(
+    requireShim,
+    loadedModule,
+    loadedModule.exports,
+  );
+  return loadedModule.exports.useInstalledPluginIcon;
+}
+
+const revoked = [];
+globalThis.URL.revokeObjectURL = (url) => revoked.push(url);
+
+function reset() {
+  stateSlots = [];
+  lastCleanup = undefined;
+  revoked.length = 0;
+}
+
+const flush = () => new Promise((resolve) => setTimeout(resolve, 0));
+
+test('caches per author/name and revokes once the last subscriber leaves', async () => {
+  let fetches = 0;
+  let resolveFetch;
+  httpMock.httpClient.getAuthenticatedPluginIconURL = () => {
+    fetches += 1;
+    return new Promise((resolve) => {
+      resolveFetch = resolve;
+    });
+  };
+  reset();
+  const useInstalledPluginIcon = loadHook();
+
+  const first = useInstalledPluginIcon('acme', 'demo');
+  const firstCleanup = lastCleanup;
+  assert.equal(first, null);
+  assert.equal(fetches, 1);
+
+  resolveFetch('blob:acme/demo');
+  await flush();
+
+  // Second subscriber reuses the cached object URL synchronously.
+  const second = useInstalledPluginIcon('acme', 'demo');
+  const secondCleanup = lastCleanup;
+  assert.equal(second, 'blob:acme/demo');
+  assert.equal(fetches, 1);
+
+  firstCleanup();
+  assert.deepEqual(revoked, []);
+  secondCleanup();
+  assert.deepEqual(revoked, ['blob:acme/demo']);
+
+  // Cache was dropped with the last subscriber, so a new mount refetches.
+  useInstalledPluginIcon('acme', 'demo');
+  assert.equal(fetches, 2);
+});
+
+test('de-duplicates concurrent subscribers and revokes a shared icon once', async () => {
+  let fetches = 0;
+  let resolveFetch;
+  httpMock.httpClient.getAuthenticatedPluginIconURL = () => {
+    fetches += 1;
+    return new Promise((resolve) => {
+      resolveFetch = resolve;
+    });
+  };
+  reset();
+  const useInstalledPluginIcon = loadHook();
+
+  useInstalledPluginIcon('acme', 'demo');
+  const firstCleanup = lastCleanup;
+  useInstalledPluginIcon('acme', 'demo');
+  const secondCleanup = lastCleanup;
+  assert.equal(fetches, 1);
+
+  resolveFetch('blob:shared');
+  await flush();
+
+  firstCleanup();
+  assert.deepEqual(revoked, []);
+  secondCleanup();
+  assert.deepEqual(revoked, ['blob:shared']);
+});
+
+test('caches nothing and revokes nothing when the authenticated fetch fails', async () => {
+  let fetches = 0;
+  httpMock.httpClient.getAuthenticatedPluginIconURL = () => {
+    fetches += 1;
+    return Promise.reject(new Error('not found'));
+  };
+  reset();
+  const useInstalledPluginIcon = loadHook();
+
+  const value = useInstalledPluginIcon('acme', 'missing');
+  const cleanup = lastCleanup;
+  assert.equal(value, null);
+
+  await flush();
+  assert.equal(useInstalledPluginIcon('acme', 'missing'), null);
+  assert.equal(fetches, 2); // failures are not negatively cached
+
+  cleanup();
+  assert.deepEqual(revoked, []);
+});
+
+test('returns a marketplace URL verbatim without fetching', () => {
+  let fetches = 0;
+  httpMock.httpClient.getAuthenticatedPluginIconURL = () => {
+    fetches += 1;
+    return Promise.resolve('blob:unused');
+  };
+  reset();
+  const useInstalledPluginIcon = loadHook();
+
+  const value = useInstalledPluginIcon(
+    'acme',
+    'demo',
+    'https://space.langbot.app/icon',
+  );
+  assert.equal(value, 'https://space.langbot.app/icon');
+  assert.equal(fetches, 0);
+  assert.equal(lastCleanup, undefined);
+});
diff --git a/web/tests/unit/monitoring-call-links.test.mjs b/web/tests/unit/monitoring-call-links.test.mjs
new file mode 100644
index 000000000..6d443d722
--- /dev/null
+++ b/web/tests/unit/monitoring-call-links.test.mjs
@@ -0,0 +1,131 @@
+import assert from 'node:assert/strict';
+import fs from 'node:fs';
+import test from 'node:test';
+import ts from 'typescript';
+
+// The resolver is pure: recorded calls carry a time and a session, the execution
+// list carries the runs' windows, so the link is decided without any re-query.
+const source = fs.readFileSync(
+  new URL('../../src/app/home/monitoring/utils/callLinks.ts', import.meta.url),
+  'utf8',
+);
+const module = { exports: {} };
+new Function(
+  'exports',
+  ts.transpileModule(source, {
+    compilerOptions: {
+      module: ts.ModuleKind.CommonJS,
+      target: ts.ScriptTarget.ES2022,
+    },
+  }).outputText,
+)(module.exports);
+const { findCallExecution, resolveCallTarget } = module.exports;
+
+const at = (iso) => new Date(iso);
+
+function run(overrides) {
+  return {
+    source: 'agent',
+    id: 'run-1',
+    session_id: 'session-a',
+    started_at_ms: at('2026-10-07T10:00:00Z').getTime(),
+    finished_at_ms: at('2026-10-07T10:00:30Z').getTime(),
+    ...overrides,
+  };
+}
+
+function call(overrides) {
+  return {
+    id: 'call-1',
+    timestamp: at('2026-10-07T10:00:10Z'),
+    sessionId: 'session-a',
+    ...overrides,
+  };
+}
+
+test('links a pipeline call to the execution carrying its message id', () => {
+  const execution = run({
+    source: 'pipeline',
+    id: 'message-9',
+    session_id: null,
+  });
+  assert.equal(
+    findCallExecution(call({ messageId: 'message-9', sessionId: undefined }), [
+      execution,
+    ]),
+    execution,
+  );
+});
+
+test('links an agent call to the run whose window contains it', () => {
+  const execution = run({});
+  assert.equal(findCallExecution(call({}), [execution]), execution);
+});
+
+test('prefers the run that was live when a session has consecutive runs', () => {
+  const first = run({ id: 'run-1' });
+  const second = run({
+    id: 'run-2',
+    started_at_ms: at('2026-10-07T10:05:00Z').getTime(),
+    finished_at_ms: at('2026-10-07T10:05:30Z').getTime(),
+  });
+  assert.equal(
+    findCallExecution(call({ timestamp: at('2026-10-07T10:05:10Z') }), [
+      first,
+      second,
+    ]).id,
+    'run-2',
+  );
+});
+
+test('an open run owns every later call of its session', () => {
+  const execution = run({ finished_at_ms: null });
+  assert.equal(
+    findCallExecution(call({ timestamp: at('2026-10-07T10:20:00Z') }), [
+      execution,
+    ]),
+    execution,
+  );
+});
+
+test('a call outside every window links to nothing', () => {
+  const execution = run({});
+  assert.equal(
+    findCallExecution(call({ timestamp: at('2026-10-07T11:00:00Z') }), [
+      execution,
+    ]),
+    null,
+  );
+  assert.equal(
+    findCallExecution(call({ timestamp: at('2026-10-07T09:00:00Z') }), [
+      execution,
+    ]),
+    null,
+  );
+  assert.equal(
+    findCallExecution(call({ sessionId: 'other' }), [execution]),
+    null,
+  );
+});
+
+test('a call keeps its message target when the execution list cannot resolve it', () => {
+  assert.deepEqual(
+    resolveCallTarget(
+      call({ messageId: 'message-9', sessionId: 'other' }),
+      [run({})],
+      new Set(['message-9']),
+    ),
+    { kind: 'message', id: 'message-9' },
+  );
+  // An Agent call carries its run id in the same field: an id the message list
+  // does not hold must not become a jump to the message tab.
+  assert.equal(
+    resolveCallTarget(
+      call({ messageId: 'run-9', sessionId: 'other' }),
+      [run({})],
+      new Set(['message-9']),
+    ),
+    null,
+  );
+  assert.equal(resolveCallTarget(call({}), []), null);
+});
diff --git a/web/tests/unit/oss-cloud-ui-privacy.test.mjs b/web/tests/unit/oss-cloud-ui-privacy.test.mjs
index 4481aa18f..41a81e375 100644
--- a/web/tests/unit/oss-cloud-ui-privacy.test.mjs
+++ b/web/tests/unit/oss-cloud-ui-privacy.test.mjs
@@ -17,6 +17,9 @@ const homeSidebarSource = readSource(
 const botFormSource = readSource(
   'src/app/home/bots/components/bot-form/BotForm.tsx',
 );
+const dynamicFormSource = readSource(
+  'src/app/home/components/dynamic-form/DynamicFormComponent.tsx',
+);
 const kbFormSource = readSource(
   'src/app/home/knowledge/components/kb-form/KBForm.tsx',
 );
@@ -52,6 +55,27 @@ test('keeps bot forms spaced with responsive setup and detail columns', () => {
   assert.match(kbFormSource, /<form[\s\S]*?className="space-y-6"/);
 });
 
+test('hides the manifest connection switch behind the dedicated selector', () => {
+  assert.match(
+    botFormSource,
+    /webhookField\?\.show_if\?\.field\s*\?\s*\[webhookField\.show_if\.field\]/,
+  );
+  assert.match(
+    botFormSource,
+    /hiddenItemNames=\{connectionModeControllerNames\}/,
+  );
+  assert.match(
+    dynamicFormSource,
+    /if \(hiddenItemNames\?\.includes\(config\.name\)\) return null/,
+  );
+  assert.doesNotMatch(
+    dynamicFormSource.match(
+      /const editableItems = useMemo\([\s\S]*?\n  \);/,
+    )?.[0] ?? '',
+    /hiddenItemNames/,
+  );
+});
+
 test('does not expose storage analysis in Cloud settings or via a deep link', () => {
   assert.match(
     homeSidebarSource,
diff --git a/web/tests/unit/plan-revisions.test.mjs b/web/tests/unit/plan-revisions.test.mjs
new file mode 100644
index 000000000..d5c57f805
--- /dev/null
+++ b/web/tests/unit/plan-revisions.test.mjs
@@ -0,0 +1,82 @@
+import assert from 'node:assert/strict';
+import test from 'node:test';
+import { readFileSync } from 'node:fs';
+import ts from 'typescript';
+const m = { exports: {} };
+new Function(
+  'module',
+  'exports',
+  ts.transpileModule(
+    readFileSync(
+      new URL(
+        '../../src/app/home/components/workspace-settings/plan-revisions.ts',
+        import.meta.url,
+      ),
+      'utf8',
+    ),
+    {
+      compilerOptions: {
+        module: ts.ModuleKind.CommonJS,
+        target: ts.ScriptTarget.ES2022,
+      },
+    },
+  ).outputText,
+)(m, m.exports);
+const { revisionRows } = m.exports;
+const current = {
+  name: 'pro',
+  revision: 1,
+  limits: { 'bots.max': 2, 'chat_history.retention_days': 30 },
+  monthly_bonus_credits: 10,
+  entitlement_features: { 'platform.telegram': false, mcp_stdio: false },
+};
+test('same version and revision-only changes do not duplicate comparisons', () => {
+  assert.ok(revisionRows(current, current).every((row) => !row.changed));
+  assert.ok(
+    revisionRows(current, { ...current, revision: 2 }).every(
+      (row) => !row.changed,
+    ),
+  );
+});
+test('compares effective quotas, credits, retention, platforms and feature switches', () => {
+  const latest = {
+    name: 'pro',
+    revision: 2,
+    limits: { 'bots.max': 5 },
+    chat_history_retention_days: 60,
+    monthly_bonus_credits: 20,
+    entitlement_features: { mcp_stdio: true },
+    disabled_adapters: [],
+  };
+  const changed = revisionRows(current, latest).filter((row) => row.changed);
+  assert.deepEqual(
+    changed.map((row) => row.key).sort(),
+    [
+      'feature:mcp_stdio',
+      'feature:platform.telegram',
+      'limit:bots.max',
+      'limit:chat_history.retention_days',
+      'monthly_credits',
+    ].sort(),
+  );
+  assert.equal(
+    changed.find((row) => row.key === 'feature:platform.telegram').after,
+    true,
+  );
+});
+test('catalog representations normalize to snapshot values', () => {
+  const latest = {
+    name: 'pro',
+    revision: 2,
+    limits: { 'bots.max': 2 },
+    chat_history_retention_days: 30,
+    monthly_bonus_credits: 10,
+    disabled_adapters: ['telegram'],
+  };
+  assert.ok(revisionRows(current, latest).every((row) => !row.changed));
+});
+test('missing latest keeps current values without invented changes', () => {
+  const rows = revisionRows(current);
+  assert.equal(rows.find((row) => row.key === 'limit:bots.max').before, 2);
+  assert.ok(rows.every((row) => !row.changed));
+});
diff --git a/web/tests/unit/websocket-client.test.mjs b/web/tests/unit/websocket-client.test.mjs
new file mode 100644
index 000000000..c89e7d7f4
--- /dev/null
+++ b/web/tests/unit/websocket-client.test.mjs
@@ -0,0 +1,175 @@
+import assert from 'node:assert/strict';
+import test from 'node:test';
+import fs from 'node:fs';
+import ts from 'typescript';
+function load(file, deps = {}, globals = {}) {
+  const source = fs
+    .readFileSync(
+      new URL('../../src/app/infra/websocket/' + file, import.meta.url),
+      'utf8',
+    )
+    .replaceAll('import.meta.env.VITE_API_BASE_URL', 'undefined');
+  const code = ts.transpileModule(source, {
+    compilerOptions: {
+      module: ts.ModuleKind.CommonJS,
+      target: ts.ScriptTarget.ES2022,
+    },
+  }).outputText;
+  const m = { exports: {} };
+  new Function('require', 'module', 'exports', ...Object.keys(globals), code)(
+    (id) => deps[id],
+    m,
+    m.exports,
+    ...Object.values(globals),
+  );
+  return m.exports;
+}
+const urls = load('url.ts');
+test('OSS and Cloud URLs follow HTTP backend scheme, port and prefix', () => {
+  for (const [base, page, expected] of [
+    [undefined, 'http://localhost:3000/home', 'ws://localhost:3000/api/ws'],
+    [undefined, 'https://cloud.test/home', 'wss://cloud.test/api/ws'],
+    [
+      'http://localhost:5300',
+      'http://localhost:3000',
+      'ws://localhost:5300/api/ws',
+    ],
+    ['https://api.test/root/', 'https://ui.test', 'wss://api.test/root/api/ws'],
+    ['/backend/', 'https://ui.test/home', 'wss://ui.test/backend/api/ws'],
+    ['https://api.test', 'http://ui.test', 'wss://api.test/api/ws'],
+  ])
+    assert.equal(
+      urls.backendWebSocketURL(base, page, '/api/ws').toString(),
+      expected,
+    );
+});
+function setup() {
+  let workspace = 'one',
+    next = 0;
+  const timers = new Map(),
+    win = new EventTarget(),
+    doc = new EventTarget();
+  win.location = { href: 'https://cloud.test/home' };
+  doc.visibilityState = 'visible';
+  class WS {
+    static CONNECTING = 0;
+    static OPEN = 1;
+    static instances = [];
+    readyState = 0;
+    constructor(url) {
+      this.url = url;
+      WS.instances.push(this);
+    }
+    send(p) {
+      this.sent = JSON.parse(p);
+    }
+    close() {
+      this.readyState = 3;
+    }
+    closed() {
+      this.readyState = 3;
+      this.onclose?.();
+    }
+    open() {
+      this.readyState = 1;
+      this.onopen?.();
+    }
+    receive(data) {
+      this.onmessage?.({ data: JSON.stringify(data) });
+    }
+  }
+  const { WebSocketClient } = load(
+    'WebSocketClient.ts',
+    {
+      './url': urls,
+      '@/app/infra/http/workspaceContext': {
+        getActiveWorkspaceUuid: () => workspace,
+      },
+    },
+    {
+      window: win,
+      document: doc,
+      WebSocket: WS,
+      localStorage: { getItem: () => 'token' },
+      setTimeout: (fn) => {
+        timers.set(++next, fn);
+        return next;
+      },
+      clearTimeout: (id) => timers.delete(id),
+      setInterval: () => 1,
+      clearInterval: () => {},
+      console: { warn() {}, error() {}, debug() {} },
+    },
+  );
+  return {
+    client: new WebSocketClient('pipeline'),
+    WS,
+    win,
+    timers,
+    setWorkspace: (v) => (workspace = v),
+  };
+}
+test('only authenticated sockets are connected; disposal removes recovery', async () => {
+  const { client, WS, win } = setup(),
+    pending = client.connect(),
+    ws = WS.instances[0];
+  ws.open();
+  assert.equal(client.isConnected(), false);
+  assert.equal(ws.sent.workspace_uuid, 'one');
+  ws.receive({ type: 'connected', connection_id: 'id' });
+  assert.equal(await pending, 'id');
+  assert(client.isConnected());
+  client.disconnect();
+  ws.closed();
+  win.dispatchEvent(new Event('focus'));
+  assert.equal(WS.instances.length, 1);
+});
+test('closed handshakes settle and transient retries survive more than five failures', async () => {
+  const { client, WS, timers } = setup();
+  const first = client.connect().catch(() => {});
+  WS.instances[0].closed();
+  await first;
+  for (let i = 0; i < 7; i++) {
+    const [id, fn] = [...timers][0];
+    timers.delete(id);
+    fn();
+    WS.instances.at(-1).closed();
+    await Promise.resolve();
+  }
+  assert.equal(WS.instances.length, 8);
+  assert.equal(timers.size, 1);
+  client.disconnect();
+  assert.equal(timers.size, 0);
+});
+test('stale closes do not affect replacement sockets and workspace changes block recovery', async () => {
+  const { client, WS, win, setWorkspace } = setup();
+  const pending = client.connect().catch(() => {}),
+    old = WS.instances[0];
+  client.disconnect();
+  const current = client.connect(),
+    ws = WS.instances[1];
+  old.closed();
+  await pending;
+  ws.open();
+  ws.receive({ type: 'connected', connection_id: 'new' });
+  await current;
+  assert(client.isConnected());
+  setWorkspace('two');
+  ws.closed();
+  win.dispatchEvent(new Event('focus'));
+  assert.equal(WS.instances.length, 2);
+  client.disconnect();
+});
+test('authentication rejection does not reconnect', async () => {
+  const { client, WS, timers, win } = setup();
+  const pending = client.connect().catch(() => {}),
+    ws = WS.instances[0];
+  ws.open();
+  ws.receive({ type: 'error', message: 'Unauthorized' });
+  ws.closed();
+  await pending;
+  win.dispatchEvent(new Event('focus'));
+  assert.equal(timers.size, 0);
+  assert.equal(WS.instances.length, 1);
+  client.disconnect();
+});
diff --git a/web/tests/unit/workspace-switcher-layout.test.mjs b/web/tests/unit/workspace-switcher-layout.test.mjs
index 7b7e241f5..09e63aefa 100644
--- a/web/tests/unit/workspace-switcher-layout.test.mjs
+++ b/web/tests/unit/workspace-switcher-layout.test.mjs
@@ -90,9 +90,8 @@ test('keeps workspace plan and settings controls on the workspace row', () => {
 });
 
 test('moves Cloud plan upgrades into Workspace Settings', () => {
-  assert.match(workspaceSettingsPanelSource, /workspace\.upgradePlan/);
-  assert.match(workspaceSettingsPanelSource, /cloudPortalURL/);
-  assert.match(workspaceSettingsPanelSource, /step=plan/);
+  assert.match(workspaceSettingsPanelSource, /cloudPlans\.title/);
+  assert.match(workspaceSettingsPanelSource, /<CloudPlansDialog/);
   assert.match(workspaceSettingsPanelSource, /systemInfo\.cloud_service_url/);
 });