Commit Graph
574 Commits
Author SHA1 Message Date
BiFangKNT 807de8d24a fix(provider): sanitize replayed Groq message metadata
Remove response-only finish_reason metadata when converting stored messages
into subsequent model requests.

Strip provider_specific_fields from Groq messages and tool calls to avoid
strict request validation failures after a tool response. Preserve metadata
required by other providers, including Gemini thought signatures, without
mutating the original message entities.

Cover streaming and non-streaming request construction with regression tests.
2026-10-17 19:48:25 +08:00
BiFangKNT 2ca9c2c7fb fix(pipeline): enforce group response rules after routing
Stop treating an explicit Pipeline route as permission to bypass group
response rules. Routed group messages must still satisfy the configured
mention or prefix matchers.

Preserve the bypass for verified interaction submissions that resume an
existing run. Cover ordinary messages, mentions, prefixes, and interaction
callbacks with regression tests.
2026-10-17 19:48:15 +08:00
BiFangKNT 4d90358da0 perf(monitoring): exclude payloads from session summaries
Exclude message_content and variables from the message source used by
conversation session summaries. These fields were carried through window
sorting even though the resulting summaries did not use them.

Preserve complete payloads for message detail queries. Add regression
coverage for summary counts, session metadata, and full message retrieval.
2026-10-17 19:46:39 +08:00
BiFangKNT d663331889 fix(monitoring): bound inflight snapshot queries
Require a valid owner and unexpired lease for pending Pipeline messages
to appear as active tasks. Exclude historical ingress records without
rewriting their stored status.

Project only the fields needed by snapshots, skip linkage queries for empty
results, and resolve linked runs from selected message IDs instead of a
cross-table OR join that scans historical messages.

Cover lease filtering, terminal transitions, Workspace isolation, payload
exclusion, and indexed lookup plans with regression tests.
2026-10-17 19:45:19 +08:00
BiFangKNT 1318efc248 perf(persistence): index workspace execution leases
Add composite indexes on Workspace, status, and lease expiry for AgentRun
and MonitoringMessage so lease maintenance can locate active executions
without scanning completed history.

Keep ORM metadata and Alembic migration 0039 consistent, including fresh
installations where the indexes already exist. Cover migration round trips,
data preservation, and SQLite query plans with regression tests.
2026-10-17 19:41:55 +08:00
huanghuoguoguo 51b4de966d test: require pinned LocalAgent contracts in CI 2026-10-17 19:31:10 +08:00
huanghuoguoguo 774b633c26 fix(box): recover private outbox files without following links 2026-10-17 19:11:00 +08:00
huanghuoguoguo 837fa1d4fc test(web): verify session pagination and workspace quota contracts 2026-10-17 16:31:20 +08:00
huanghuoguoguo 0f522e5286 test(persistence): verify migration data and startup task bounds 2026-10-17 16:08:08 +08:00
huanghuoguoguo 6ea7fb9fc8 test(discord): assert attachment and guild delivery targets 2026-10-17 16:07:00 +08:00
huanghuoguoguo 44872101f6 Merge master and preserve multimodal test contracts 2026-10-17 16:03:19 +08:00
huanghuoguoguo 9b350132f4 test(api): cover send delegation and reuse the prior list contract 2026-10-17 15:57:35 +08:00
huanghuoguoguo c03f33d855 test(runtime): verify sandbox and platform contracts 2026-10-17 15:57:15 +08:00
huanghuoguoguo 4a2780b841 test(vector): verify real Valkey search and deletion results 2026-10-17 15:53:26 +08:00
huanghuoguoguo cbc049400c test(api): verify bot response and service delegation contracts 2026-10-17 15:53:00 +08:00
huanghuoguoguo 1c0fd7d0ba test: strengthen startup and persistence boundary oracles 2026-10-17 15:45:09 +08:00
huanghuoguoguo 5ad95e100e test(api): verify maintenance outcomes and upload context 2026-10-17 15:44:19 +08:00
huanghuoguoguo bac69e987a test(api): verify identity predicates and credential side effects 2026-10-17 15:43:50 +08:00
huanghuoguoguo 1adc2a6046 test(api): verify orchestration writes and scoped state transitions 2026-10-17 15:43:38 +08:00
huanghuoguoguo d5bc009426 test(api): verify extension persistence and connection task behavior 2026-10-17 15:42:54 +08:00
huanghuoguoguo e0c738557a test(agent): verify persistence and denied action side effects 2026-10-17 15:35:33 +08:00
huanghuoguoguo ca997e6faa test(api): verify provider persistence and returned service payloads 2026-10-17 15:29:54 +08:00
huanghuoguoguo efe1e4a758 test(auth): verify revocation and authorization ordering 2026-10-17 15:29:39 +08:00
huanghuoguoguo 69a415e562 test(models): verify model persistence and runtime boundaries 2026-10-17 15:26:38 +08:00
huanghuoguoguo 89ef102238 test(cloud): close directory observer database connections 2026-10-17 15:25:24 +08:00
huanghuoguoguo 569dd07024 fix(auth): reject malformed credential payloads before authentication 2026-10-17 15:25:16 +08:00
huanghuoguoguo ce6465094e test(provider): fold method presence check into invocation tests 2026-10-17 15:18:01 +08:00
huanghuoguoguo d25c1e0f0c test(api): verify monitoring payloads and embed side effects 2026-10-17 15:16:26 +08:00
huanghuoguoguo c1ab2d1110 test(services): verify persisted state and delivery boundaries 2026-10-17 15:13:08 +08:00
huanghuoguoguo bcfe78aa0e test(pipeline): verify message and provider behavior 2026-10-17 15:12:25 +08:00
RockChinQ c05aac538f fix(discord): complete streaming media and event delivery 2026-10-17 15:11:29 +08:00
huanghuoguoguo a2066c95e7 test(io): verify storage config and utility results 2026-10-17 15:07:52 +08:00
huanghuoguoguo 4afd86c1f6 test: consolidate duplicate backend contracts and strengthen stage checks 2026-10-17 15:03:50 +08:00
huanghuoguoguo 8bad1a14fe test(api): verify resource payloads and scoped service arguments 2026-10-17 15:02:41 +08:00
huanghuoguoguo 24300adc72 test(runtime): assert filter and message delivery semantics 2026-10-17 15:01:09 +08:00
huanghuoguoguo 8d3e99d0fb test(core): strengthen config and task behavior oracles 2026-10-17 14:58:17 +08:00
huanghuoguoguo bbf517edbc test(auth): exercise knowledge and event action boundaries (#2667) 2026-10-17 12:46:41 +08:00
huanghuoguoguo 5641637d3d Merge pull request #2665 from langbot-app/test/regression-boundary-cleanup-20261017
test: remove self-confirming tests and strengthen regression boundaries
2026-10-17 11:26:46 +08:00
huanghuoguoguo 3f2d52e557 Merge pull request #2512 from langbot-app/feat/2410-sdk-skill-store
feat(skill): decouple read-only skills from Box execution
2026-10-17 10:13:26 +08:00
huanghuoguoguo d54e533985 fix(skill): pin revisions per run and preserve shared Box state 2026-10-17 09:32:41 +08:00
huanghuoguoguo 438cbd41a9 test(core): provision real plugin, CLI and SeekDB contracts (#2636)
* test(core): provision official plugin CLI and SeekDB contracts

* test(core): trigger report guard tests and trim fixture checks
2026-10-17 09:13:44 +08:00
huanghuoguoguo 2b79b04d36 Merge pull request #2635 from langbot-app/test/core-contract-coverage-gaps
test(core): verify runtime behavior and close PostgreSQL coverage gaps
2026-10-17 09:13:07 +08:00
RockChinQandClaude 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
RockChinQandClaude Opus 5.5 16cfb6f8bd feat(cloud): show purchased plan revision alongside the latest revision
- Expose the workspace's purchased plan revision from the entitlement
  snapshot.
- Stack the purchased revision card in front of the latest revision in
  the plan comparison dialog; both cards share the same layout, and the
  purchased card's copy and prices come from the Space revision endpoint
  while benefits stay on the entitlement snapshot.
- Add a cancel-first tooltip for subscribing to the newer revision.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-10-17 00:25:41 +08:00
RockChinQandHyu 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 338b856eec fix(monitoring): restore Cloud in-flight snapshots and collapsed dock 2026-10-15 16:42:24 +08:00
RockChinQ 68a3f64cef test(core): update bot fixtures for platform policy checks 2026-10-15 05:31:04 +00:00
RockChinQ 625ee67113 fix(ci): align workspace fixtures and constrain mobile tours 2026-10-15 05:31:04 +00:00
RockChinQ b8d292c94d fix(websocket): recover debug connections and resolve backend URLs consistently 2026-10-15 05:31:04 +00:00
RockChinQ 528faf7851 test(cloud): include adapter names in runtime tenancy fixtures 2026-10-15 05:31:04 +00:00