huanghuoguoguo
ce6465094e
test(provider): fold method presence check into invocation tests
2026-10-17 15:18:01 +08:00
huanghuoguoguo
bcfe78aa0e
test(pipeline): verify message and provider behavior
2026-10-17 15:12:25 +08:00
huanghuoguoguo
d54e533985
fix(skill): pin revisions per run and preserve shared Box state
2026-10-17 09:32:41 +08:00
RockChinQ and Claude Opus 5.5
653f9e731d
Merge origin/master into feat/cloud-v2-subscriptions
...
Keep i18n'd requester error messages while adopting master's structured
RequesterError metadata.
Co-Authored-By: Claude Opus 5.5 <[email protected]
2026-10-17 01:37:14 +08:00
RockChinQ and Hyu
616981b0cf
perf(llm): harden request policy and optimize connections, reasoning and prompt caching ( #2661 )
...
* fix(llm): centralize request policy and preserve stream diagnostics
* test(llm): dispose SQLite engine after migration regression
* perf(llm): reuse connections and unify reasoning and prompt cache policy
* fix(llm): preserve explicit native prompt cache options
* perf(llm): fix measured TLS setup and streaming reuse overhead
* fix(llm): resolve plugin request-chain regressions found by E2E
---------
Co-authored-by: Hyu <[email protected]
2026-10-15 19:48:35 +08:00
RockChinQ
68a3f64cef
test(core): update bot fixtures for platform policy checks
2026-10-15 05:31:04 +00:00
Hyu
e29fd27f02
feat(i18n): localize backend messages by workspace language
2026-10-15 05:31:04 +00:00
Hyu
09a84f6642
fix(provider): preserve streaming when removing thinking content
2026-10-12 13:30:17 +08:00
Hyu
1c70218383
feat: reset bot session context while retaining monitoring history
2026-10-10 22:35:27 +08:00
Hyu
40a3a941a3
fix(monitoring): unify execution records and stabilize streaming cleanup
2026-10-08 03:02:37 +08:00
Hyu
9f76037e64
feat(monitoring): unify execution views and trace tool calls
2026-10-07 22:36:10 +08:00
Hyu
4482e8d876
feat(telemetry)!: upload one record per execution node and one closure per chain
...
Core no longer assembles a whole trace locally. Every workflow node is uploaded
the moment it completes as its own `execution_node` record, and each execution
closes with exactly one `execution_chain` record; all records of one execution
carry the same `event_id`, so Space groups them back into a single chain.
- every execution starts at an inbound event: event-less Pipeline lanes and
WebUI/API debug runs synthesize their virtual inbound event at the boundary
and are marked debug=true, so a Runner is never a trace origin
- sampling is decided once per event from a stable hash of the event id (the
previous `int(id[:8], 16)` raised on every `platform:`-prefixed id and so
dropped every successful inbound-event chain) and is reported on each record
as sample{mode,denominator}; the default mode is now `all`
- the local trace buffer, the 120s TTL, the per-process trace cap and the shared
open-node mirror are gone, which also removes the truncated-chain upload and
the re-buffered-closed-trace leak they caused
- a broken chain is always uploaded, and its reason falls back to the first
failed node so the chain explains itself
- flush keeps records until the sender accepts them and counts what it drops;
telemetry health replaces the unused send-task queue
- heartbeat reports bound Agent / plugin-processor / event-binding counts
- bound Pipeline and Agent resources (plugins, MCP servers, knowledge bases,
tools with their Host source, skills, models and the Runner schema selectors
that picked them) are reported on the chain record
- failure text is bounded and redacted instead of passing exception text through
- the dead per-query feature counters and their call sites are removed
Wire contract: schema 2 (`execution_node` / `execution_chain`).
2026-10-07 09:41:41 +08:00
Hyu
5a31ac75fc
style(tools): format the sandbox interpreter and skill tool changes with ruff
2026-09-30 18:45:00 +08:00
Hyu
4fc700b7e0
fix(tools): hide register_skill when Cloud sandboxes cannot scan host directories
...
scan_skill_directory is disabled for Cloud-managed sandboxes, but the
register_skill tool was still advertised to the model whenever a backend was
available, so every call failed closed without a usable explanation. Advertise
register_skill only when the runtime may scan sandbox directories, and keep
activate available in both modes.
2026-09-30 18:44:16 +08:00
Hyu
ff3f321120
fix(tools): resolve the sandbox interpreter for workspace file scripts
...
The workspace file tools (read/write/edit/glob/grep) ran a script through
`python`, which a sandbox built from a host rootfs may not provide: those hosts
ship `python3` only, and a read-only /usr prevents creating a `python` shim
inside the jail. Every file tool then failed with `/bin/sh: 1: python: not
found`. Resolve the interpreter with `command -v python3 || command -v python`,
matching the interpreter the Box service already uses for its own scripts.
2026-09-30 18:44:16 +08:00
huanghuoguoguo
c56fa0ba2e
fix(skill): reconcile rebase with runner contracts
2026-09-28 10:54:31 +08:00
huanghuoguoguo
9b8d1b4034
feat(skill): pin run revision bindings
2026-09-28 10:22:01 +08:00
huanghuoguoguo
4e64e0cfc2
refactor(skill): compose generic box mounts in core
2026-09-28 10:20:00 +08:00
huanghuoguoguo
7bf20490f4
feat(skill): use SDK store without box execution
2026-09-28 10:18:16 +08:00
RockChinQ
e77acfa3ab
fix(box): align integration tests with runner-owned bindings
2026-09-18 01:07:43 +08:00
RockChinQ
b264d46d77
feat: delegate sandbox policy to runners and simplify pipeline migration
2026-09-17 23:05:27 +08:00
RockChinQ
8c119bc4b6
feat: add verified manual pipeline migration for plugin runners
2026-09-16 18:13:11 +00:00
RockChinQ
273b1ea3cf
Merge master into dev/4.11.x and preserve plugin runner architecture
...
Reconcile migration branches without rewriting published revisions; retain additive Codex, monitoring, provider and platform fixes. Keep dynamic runner schemas and Host ownership, restore compatibility regressions, and preserve safe model-test error handling.
2026-09-16 08:13:07 +00:00
RockChinQ
f24a7c9bb2
feat(runner): unify plugin execution across agents and event processors
2026-09-10 18:04:38 +08:00
huige66631 and RockChinQ
ce6b647fe7
feat(mcp): detect OAuth-protected remote MCP servers ( #2363 )
...
* feat(mcp): surface OAuth-required server tests
* fix(mcp): show connection failure details in status cards
---------
Co-authored-by: RockChinQ <[email protected]
2026-09-09 16:13:33 +08:00
Hyu and dadachann
0f216a0d4d
feat(provider): support Codex subscriptions with ChatGPT sign-in ( #2513 )
...
* feat(provider): support Codex subscriptions with ChatGPT sign-in
* style: format Codex live integration test
* fix(provider): preserve Codex identity in temporary model tests
* fix(web): portal provider selector without dialog overflow
* fix(web): allow native scrolling in provider dropdown
* fix(provider): surface safe Codex quota and upstream errors
* fix(web): provide reliable Codex copy feedback in dialogs
* feat(provider): confirm cascade deletion from edit dialog
* fix(persistence): discard connections after failed commit
* fix(web): polish provider loading and confirmation motion
---------
Co-authored-by: dadachann <[email protected]
2026-09-06 23:31:02 +08:00
RockChinQ
75f117ae50
fix(provider): preserve explicit streaming finish reasons
2026-09-06 00:33:20 +08:00
Hyu
8f0a55a1f4
fix(agent-debug): stream execution traces with platform mocks and coverage
2026-09-05 15:34:29 +08:00
Hyu
fde04e64f6
Merge remote-tracking branch 'origin/master' into dev/4.11.x
...
# Conflicts:
# src/langbot/pkg/api/http/controller/groups/pipelines/pipelines.py
# src/langbot/pkg/api/http/service/bot.py
# src/langbot/pkg/provider/runners/localagent.py
# src/langbot/templates/metadata/pipeline/ai.yaml
# tests/unit_tests/api/service/test_bot_service.py
# tests/unit_tests/provider/runners/test_difysvapi_runner.py
# tests/unit_tests/utils/test_safe_regex.py
# web/src/app/infra/entities/adapter-categories.ts
# web/src/app/wizard/page.tsx
# web/src/i18n/locales/en-US.ts
# web/src/i18n/locales/ja-JP.ts
# web/src/i18n/locales/zh-Hans.ts
# web/tests/e2e/plugin-page-auth.spec.ts
2026-08-31 17:17:47 +08:00
Neos and Hyu
e69a80f5e9
Windows 兼容性修复 + Dify API 适配 + 企微 WS 加载优化 ( #2470 )
...
* fix: Windows fsync requires write access to file descriptor
- _fsync_file: os.O_RDONLY -> os.O_RDWR (Windows _commit() requires write access)
- _fsync_directory: tolerate OSError on fsync and os.open (chmod 0o700 blocks access on Windows)
* fix: Dify SSE empty data lines and upload response format mismatch
- Skip malformed/flushed SSE data lines (JSONDecodeError)
- Accept HTTP 200 besides 201 for file upload response
- Unwrap data wrapper in upload response (resp.get('data', resp))
* feat: WeCom WS mode sends empty initial stream frame for loading spinner
Send an empty reply_stream frame immediately after stream session
creation so the WeCom client shows its built-in loading indicator
while the pipeline processes the message (e.g. RAG retrieval).
* fix(compat): harden Windows, Dify, and WeCom changes
* fix(migration): clean interrupted SQLite temp files
---------
Co-authored-by: Hyu <[email protected]
2026-08-29 22:34:10 +08:00
fishzjp
b66db86bff
fix(provider): stringify MCP tool results for OpenAI-compatible APIs ( #2476 )
...
execute_func_call returns list[ContentElement] for MCP tools, but the
runner assigned that list directly to the tool-message content. The
OpenAI chat-completions spec requires tool-message content to be a
string, so OpenAI-compatible endpoints return HTTP 500 when the raw
list is sent.
Serialize the list to a string before building the tool message, using
ContentElement.__str__ which returns the text payload for text elements
and a human-readable placeholder for images and files. Fixes #2457 .
2026-08-27 18:19:04 +08:00
fishzjp
95b8736e93
fix(provider): tolerate trimmed image parts in litellm message conversion ( #2475 )
...
SessionManager clears image_base64 on past turns to save memory, and
exclude_none serialization drops the hollowed field entirely, so a
replayed history part can arrive as {'type': 'image_base64'} with no
payload. The converter accessed the missing key unconditionally and
raised KeyError on every turn after an image was sent.
Prefer the base64 payload when present, fall back to an image_url that
survived on the same element, and drop hollow parts otherwise (same
strategy as the existing file-part handling). Fixes #2469 .
2026-08-27 18:07:43 +08:00
Junyan Qin
f94243c725
chore: merge master into dev/4.11.x
2026-08-14 16:04:52 +08:00
Dongchuan Fu
e37987215e
feat(provider): add pipeline reasoning controls ( #2373 )
...
* feat(provider): add pipeline reasoning controls
* fix(provider): preserve local agent model compatibility
* refactor(web): use shadcn reasoning slider
* fix(runtime): stabilize reasoning chat delivery
* fix(provider): route reasoning controls by model family
* fix(provider): handle hosted Kimi reasoning protocols
* fix(provider): map qwen reasoning levels to budgets
* fix(provider): preserve think tags in streamed reasoning
* fix(provider): preserve reasoning tool metadata
* style(provider): satisfy ruff checks after merge
* fix(persistence): preserve reasoning migration compatibility
2026-08-09 17:38:01 +08:00
huanghuoguoguo
79611a5513
fix(agent): harden runner integration and QA
2026-08-01 09:23:41 +08:00
huanghuoguoguo
55f42a4ebc
chore(merge): sync master into dev/4.11.x
2026-07-31 19:29:38 +08:00
RockChinQ and dadachann
e1ac5e0fc8
feat(tenancy): add Workspace multi-tenant foundation ( #2353 )
...
* Document multi-tenant workspace architecture
* Add OSS and commercial workspace boundaries
* docs: redesign multi-tenant workspace architecture
* feat(tenancy): implement workspace isolation
* docs(tenancy): record verification evidence
* docs(tenancy): revise single-instance SaaS topology
* docs(tenancy): refine architecture options
* docs: finalize cloud v2 multi-tenant decisions
* feat(tenancy): establish cloud isolation foundations
* feat(tenancy): harden shared cloud runtime boundaries
* docs(tenancy): record final isolation verification
* fix(tenancy): close isolation and permission gaps
* docs(tenancy): record final isolation verification
* feat(tenancy): connect cloud workspace control plane
* fix(build): install git for pinned SDK
* docs(cloud): update control plane verification
* chore: update multi-tenant SDK pin
* fix(cloud): skip legacy model sync during startup
* test(cloud): preserve minimal model manager fixtures
* fix(cloud): preserve authenticated account context
* fix(cloud): reuse authenticated account for user info
* feat(cloud): complete Workspace settings navigation
* test(web): cover Workspace dropdown menu
* feat(web): place workspace controls in sidebar
* refactor(web): streamline workspace controls
* style(web): format workspace layout test
* fix(cloud): surface runtime and workspace plan status
* fix(plugin): keep runtime identity stable across restarts
* fix(ui): widen and center workspace switcher
* fix(ui): hide roles from workspace switcher
* fix(ui): align workspace switcher with sidebar entries
* feat(workspace): add in-product collaboration and direct Cloud launch
* style: format collaboration changes
* fix(workspace): bind collaboration APIs to tenant UoW
* fix(cloud): preserve Core-owned collaboration state
* test(cloud): require Space identity for invite registration
* feat(cloud): complete secure invitation experience
* style(web): format invitation flows
* fix(cloud): recover box runtime without unscoped skill reload
* feat(oss): enforce invitation account and owner billing flows
* style: format OSS account service
* test(oss): cover invitation logout handoff
* fix(oss): resolve workspace owner in scoped session
* feat(cloud): harden multi-tenant runtime resources
* fix(cloud): bound runtime restart storms
* fix(cloud): eliminate periodic runtime CPU spikes
* fix(cloud): enforce instance capacity ceilings
* fix(cloud): scope public login capability discovery
* fix(cloud): bound tenant maintenance and monitoring work
* fix(runtime): bound tenant resource amplification
* fix(deps): pin green multi-tenant plugin SDK
* fix(cloud): handle unavailable skill capability
* fix(security): require authentication for image file endpoint (H-2)
- Changed /api/v1/files/image from AuthType.NONE to USER_TOKEN_OR_API_KEY
- Added Permission.RESOURCE_VIEW requirement
- Prevents unauthenticated cross-tenant file access via leaked keys
- Fixes HIGH severity finding from multi-tenant security review
docs: add comprehensive database migration guide
- Complete migration steps for OSS → multi-tenant
- Backup, execution, verification procedures
- Rollback scenarios and recovery plans
- Performance tuning recommendations
* test: add comprehensive cross-tenant isolation tests
Added 7 critical test scenarios for multi-tenant boundaries:
- Cross-tenant bot access prevention
- Viewer role read-only enforcement
- Removed member immediate access revocation
- Model provider credential isolation
- WebSocket message isolation
- Invitation token workspace scoping
- Multi-workspace context validation
These tests address P0-2 coverage gaps for:
- workspaces.py (membership & invitation flows)
- user.py (authentication & authorization)
- websocket_chat.py (real-time isolation)
- plugins.py (resource access control)
docs: finalize database migration guide
* fix(security): resolve M-1, M-2, M-3 security findings
M-1: WebSocket authorization TOCTOU race (FIXED)
- Changed _revalidate_websocket_authorization to return RequestContext
- Ensures validated context is used immediately without race window
- Prevents removed members from sending messages during revalidation gap
M-2: Model Manager cache workspace isolation (VERIFIED)
- Confirmed _CacheKey already uses 4-tuple: (instance, workspace, generation, resource)
- Cache is properly scoped per workspace, no cross-tenant leakage possible
- No code change needed, documented as working correctly
M-3: Invitation lock workspace scoping (FIXED)
- Changed lock key from token_digest to workspace_uuid:token_digest
- Prevents DoS where attacker locks token in Workspace A to block Workspace B
- Locks now isolated per workspace
All MEDIUM severity findings from security review now resolved.
* fix(cloud): unblock tenant CI and enforce knowledge quotas
* fix(tenancy): scope rerank model sync
---------
Co-authored-by: dadachann <[email protected]
2026-07-30 21:43:35 +08:00
Hyu and chan
dd8d1007a1
feat: support Space rerank models ( #2358 )
...
Co-authored-by: chan <[email protected]
2026-07-26 16:37:34 +08:00
Junyan Qin
283c6949f4
fix(mcp): validate tool call timeout config
2026-07-23 18:26:41 +08:00
Junyan Qin
0dfae76e39
fix(mcp): make tool call timeout configurable
2026-07-23 18:26:37 +08:00
Junyan Qin
73e47ea2d4
fix(mcp): keep waiting for box recovery
2026-07-23 17:41:27 +08:00
Junyan Qin
06e03af994
fix(runtime): complete reconnect recovery paths
2026-07-23 17:28:31 +08:00
Junyan Qin
998b76d53a
fix(runtime): make plugin and box connectors resilient
2026-07-21 18:43:24 +08:00
Junyan Qin
1765c43262
Revert "fix(runtime): make plugin and box connectors resilient"
...
This reverts commit 0b461e5830
2026-07-21 18:43:01 +08:00
Junyan Qin
0b461e5830
fix(runtime): make plugin and box connectors resilient
2026-07-21 18:41:22 +08:00
DongXiaoming
76c5003c21
fix(localagent): stop re-seeding stream accumulator with previous round content ( #2329 )
...
When the LLM (e.g. MiniMax-M3) returns multiple rounds after tool calls, the _StreamAccumulator was initialized with initial_content=first_content, causing every subsequent round to repeat the entire first message. Remove the re-seeding so each round starts with a clean accumulator.
Add regression test verifying multi-round tool call content is not duplicated.
2026-07-20 16:38:11 +08:00
huanghuoguoguo
8afab5fa49
fix(mcp): recover box stdio sessions after disconnect
2026-07-19 22:47:09 +08:00
huanghuoguoguo
cfa7c7b336
fix(tools): support skill mount root discovery
2026-07-19 22:46:01 +08:00
huanghuoguoguo
54c872b74f
fix(agent-runner): close sandbox skill authoring flow
2026-07-19 22:46:01 +08:00
huanghuoguoguo
c55be79f9f
fix(agent-runner): prepare local runner functional tests
2026-07-19 22:46:01 +08:00
huanghuoguoguo