Commit Graph

11375 Commits

Author SHA1 Message Date
Peter Steinberger
842e6f1643 fix(slack): preserve assistant thread context 2026-05-16 22:31:22 +01:00
Peter Steinberger
80eeb688c1 feat(slack): add assistant thread lifecycle 2026-05-16 22:31:22 +01:00
Peter Steinberger
c4fb12ee8d fix: preserve xAI Grok 4.3 default reasoning (#81227) 2026-05-16 22:25:43 +01:00
Jason O'Neal
4f886e7334 fix(xai): gate reasoning effort to supported models 2026-05-16 22:25:43 +01:00
Marcus Castro
6e70d9e4b6 fix: accept WhatsApp group-prefixed targets (#82738) 2026-05-16 18:18:18 -03:00
Vincent Koc
55edadf86f fix(qa-lab): ignore heartbeat parity transcripts 2026-05-17 04:24:17 +08:00
Vincent Koc
640735cebe fix(qa): serialize runtime parity cells 2026-05-17 04:19:05 +08:00
wuyangfan
35e1c7ac41 fix(webchat): keep trustedLocalMedia internal to reply payloads
Restore Omit on public plugin-sdk ReplyPayload; set trustedLocalMedia via
runtime assertion in speech-core and explicitly on dispatch TTS-only finals.

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-16 21:18:13 +01:00
wuyangfan
eec18fccb4 fix(webchat): forward trustedLocalMedia on accumulated block TTS tail
Avoid per-block final-mode synthesis (duplicate with dispatch tail). Mark
TTS output as trusted local media and pass the flag through the TTS-only
final payload WebChat consumes after block streaming.

Fixes #82628

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-16 21:18:13 +01:00
wuyangfan
14117c303d fix(webchat): enable auto-TTS for block replies with trusted local media
WebChat streaming uses kind=block for assistant text; final-mode TTS skipped
those payloads. Mark synthesized audio as trustedLocalMedia and export the
full ReplyPayload type so the gateway can serve local TTS files.

Fixes #82628

Co-authored-by: Cursor <cursoragent@cursor.com>
2026-05-16 21:18:13 +01:00
KateWilkins
03012ac5a1 fix: update xai image generation model (#81399)
Updates the xAI image model catalog and docs to use `grok-imagine-image-quality` after `grok-imagine-image-pro` retirement.

Co-authored-by: Kate <kate@trantor.dev>
2026-05-16 21:09:21 +01:00
Manzojunior
a9e0a897a1 fix: handle xai video pending status (#82610)
Treats xAI video `pending` poll status as in-flight processing and keeps polling until `done`.

Co-authored-by: Chase Young <manzo0924@gmail.com>
2026-05-16 21:03:34 +01:00
Vincent Koc
5db30ab47d fix(extensions): satisfy runtime boundary checks 2026-05-17 03:48:46 +08:00
Vincent Koc
f345b54d04 test(qa-lab): add runtime parity axis 2026-05-17 03:32:50 +08:00
Peter Steinberger
6e4cc222cb fix(xai): refresh oauth and model catalog 2026-05-16 20:25:07 +01:00
Neerav Makwana
ccdcdc7d1b fix(memory-core): retry disabled dreaming cron cleanup 2026-05-16 19:58:01 +01:00
Vincent Koc
440333125c test(qa-lab): add personal agent scenarios 2026-05-17 02:56:53 +08:00
Vincent Koc
e1061a8b46 test(live): tolerate provider drift in release checks 2026-05-17 02:36:48 +08:00
Vincent Koc
8330582493 test: repair broad qa surface regressions 2026-05-17 02:32:14 +08:00
Jason O'Neal
e8b4003933 fix(google): keep first login identity errors 2026-05-17 02:10:11 +08:00
Jason O'Neal
995c702b07 fix(google): wrap Gemini CLI refresh credentials
# Conflicts:
#	CHANGELOG.md
2026-05-17 02:10:11 +08:00
Jason O'Neal
b34454f5b3 fix(google): refresh Gemini CLI OAuth tokens 2026-05-17 02:10:11 +08:00
Gio Della-Libera
4003a955ee fix(telegram): normalize announce group targets (#81229) 2026-05-16 10:32:58 -07:00
Gio Della-Libera
37cd82913f fix(discord): bind delayed identify to socket generation (#82225)
* fix(discord): bind delayed identify to socket generation
* chore: refresh CI after main repairs
2026-05-16 10:19:35 -07:00
Vincent Koc
d13749b2fc test(codex): keep dynamic tool helper tests fast 2026-05-17 01:12:20 +08:00
Pavan Kumar Gondhi
6a12c6f799 fix(gateway): scope session data lookups by agent [AI] (#81386)
* fix: scope gateway session lookups by agent

* addressing review-skill

* addressing review-skill

* addressing review-skill

* addressing review-skill

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing review-skill

* addressing review-skill

* addressing review-skill

* addressing review-skill

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing ci

* addressing ci

* fix: complete root-cause handling

* addressing review-skill

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* addressing codex review

* Fix Swift protocol optional initializer defaults

* Stabilize node command lookup in approval test

* Fix browser proxy approval test node lookup

* Trim unrelated changes from issue 642 fix

* Remove unrelated formatting churn from issue 642 fix

* Fix Swift protocol generator lint

* docs: add changelog entry for PR merge
2026-05-16 22:31:02 +05:30
Vincent Koc
9a204008ba test(extensions): stabilize plugin prerelease shards 2026-05-17 01:00:43 +08:00
Peter Steinberger
bdfc078487 test(matrix): add state-after E2EE QA coverage
Adds Matrix QA coverage for the state_after E2EE regression fixed by #82631.
2026-05-16 17:59:30 +01:00
Peter Steinberger
a3e7fc7de7 test(telegram): fix cache invalidation test contexts 2026-05-16 17:56:01 +01:00
Peter Steinberger
b11f67964c test(telegram): cover bot info cache invalidation 2026-05-16 17:56:01 +01:00
Peter Steinberger
95741daeb4 fix(telegram): cache startup bot info 2026-05-16 17:56:01 +01:00
Vincent Koc
6ed16d9356 test(extensions): harden plugin prerelease shards 2026-05-17 00:40:42 +08:00
Vincent Koc
fddac1c507 test(extensions): align mocks with runtime contracts 2026-05-17 00:23:23 +08:00
Peter Steinberger
777d289979 fix(matrix): avoid state-after sync opt-in 2026-05-16 16:57:24 +01:00
Peter Steinberger
2fcaab0010 fix: clean up approval handler PR landing (#82482) 2026-05-16 16:41:07 +01:00
Feelw00
ea9793b2e1 fix(approvals): release Matrix reaction target on mid-flight cancel
Address the ClawSweeper R2 finding that the pre-bind stopped guard
introduced in this PR drops a delivered entry without any cleanup. The
prior PR comment block was correct only for adapters whose deliverPending
has no in-process side effects; Matrix registers a reaction target in
both an in-memory Map and a persistent store inside deliverPending, so
the entry would leak until the 24h TTL (or process restart) every time
stop() landed between deliverPending and bindPending.

Add an optional cancelDelivered interaction hook on the runtime types,
forward it through both the spec-to-adapter wrapper
(createChannelApprovalNativeRuntimeAdapter) and the lazy adapter wrapper
(createLazyChannelApprovalNativeRuntimeAdapter), and invoke it from the
two stopped guards in deliverTarget: the pre-bind guard always calls it,
and the post-bind guard calls it on the branch where bindPending
returned no handle (so unbindPending cannot run). Matrix implements the
hook by calling unregisterMatrixApprovalReactionTarget on the entry's
roomId + reactionEventId, which is the exact key
registerMatrixApprovalReactionTarget uses inside deliverPending.

The other native runtime adapters (Slack, Discord, Telegram, qqbot)
leave the hook unimplemented because their deliverPending paths only
emit remote messages and keep no in-process state to drop.

Regression coverage:
- invokes cancelDelivered when stop() fires between deliverPending and
  bindPending (Deferred-gated deliverPending, asserts bindPending /
  unbindPending never run and cancelDelivered receives the entry)
- invokes cancelDelivered when stop() fires after bindPending returned
  null (asserts unbindPending stays uncalled while cancelDelivered fires)

AI-assisted: drafted with claude code (claude-opus-4-7).
2026-05-16 16:41:07 +01:00
Kagura
ffdc7aa7a6 fix(slack): route DM thread replies to main session instead of thread-scoped session (#82418)
* fix(slack): route DM thread replies to main session instead of thread-scoped session

DM thread replies (user replies inside a thread under a bot message in a
DM) were routed to a thread-specific session key instead of the user's
main DM session.  This caused the agent to never receive the inbound on
the expected session, making the bot appear unresponsive.

The root cause was in prepare-routing.ts: canonicalThreadId for
isDirectMessage was set to threadTs when isThreadReply was true, creating
a session key like agent:main:slack:direct:u3🧵<ts>.  DM threads
are a UI affordance — not a session boundary — so all DM messages should
route to the main DM session regardless of thread_ts.

Also adds a diagnostic logVerbose warning when assistant_app_thread
message_changed events fail sender resolution (Case 2 of #82390),
which was previously completely silent.

Fixes #82390

* chore(slack): polish DM thread routing PR

* test(slack): update DM thread routing contract

* test(slack): flatten non-main DM thread expectations

* fix(slack): preserve bound DM thread routes

* test(slack): align DM thread session fixtures

* fix(slack): keep flattened DM thread metadata scoped

* fix(slack): preserve DM thread delivery routes

---------

Co-authored-by: Peter Steinberger <steipete@gmail.com>
2026-05-16 16:24:36 +01:00
Peter Steinberger
2074cde6cf test: stabilize codex app-server waits 2026-05-16 16:21:34 +01:00
Peter Steinberger
21c5f8dc6d fix(codex): keep run lane timeout progress-aware 2026-05-16 16:21:34 +01:00
Peter Steinberger
a641a27bd4 fix(codex): harden app-server progress watchdog 2026-05-16 16:21:34 +01:00
Peter Steinberger
efe3790dd3 fix(codex): preserve queued turn-start progress 2026-05-16 16:21:34 +01:00
Peter Steinberger
6778da05d6 fix(codex): scope app-server request watchdog progress 2026-05-16 16:21:34 +01:00
Eva (agent)
d7d597cfd8 fix: scope codex attempt watchdog to turn progress 2026-05-16 16:21:34 +01:00
Eva (agent)
722161271e fix: make codex app-server timeout progress-aware 2026-05-16 16:21:34 +01:00
Peter Steinberger
1769e6a2f0 fix(agents): log detail-less responses failures (#82593)
* fix(agents): log detail-less responses failures

* fix(github-copilot): guard device login fetches

* test(ci): refresh stale cron and session expectations

* test(ci): keep cron legacy string fixture

* fix(agents): redact array response failure ids

* fix(agents): classify empty response failures
2026-05-16 15:57:40 +01:00
Gio Della-Libera
0b708a2574 OC Path: restore YAML support (#81436)
* OC Path: restore YAML support
* fix(oc-path): guard yaml writes and empty sequences
* fix(oc-path): guard yaml insertion keys
* fix(oc-path): guard yaml object key
* fix(oc-path): classify yaml root insertions
* style(oc-path): format yaml branch after rebase
* fix(oc-path): reject malformed yaml edits
* docs(oc-path): clarify yaml file support
* fix(ci): refresh yaml branch after rebase
* fix(ci): clean shared blockers for yaml path PR
* fix(changelog): keep yaml path note scoped
* fix(ci): preserve current shared contracts


---------

Co-authored-by: Gio Della-Libera <giodl73@gmail.com
2026-05-16 07:52:08 -07:00
Peter Steinberger
16e5d6692d fix(gateway): bound traced channel startup handoff (#82592)
* fix(gateway): bound traced channel startup handoff

* fix(github-copilot): guard device login fetches

* fix(gateway): skip stopped traced channel handoffs

* test(net): keep guarded fetch mocks hermetic
2026-05-16 15:15:57 +01:00
吴杨帆
eebdbabae9 fix: omit Ollama think for non-reasoning models
Preserve native Ollama thinking controls for supported models and explicit think=false, but avoid sending truthy think payloads for models marked reasoning=false.\n\nCo-authored-by: 吴杨帆 <85487201+leno23@users.noreply.github.com>
2026-05-16 15:10:12 +01:00
Peter Steinberger
58083866d0 fix: sync codex app-server protocol drift 2026-05-16 15:03:51 +01:00
Kyzcreig
8092761d7b fix: mark codex compaction events completed 2026-05-16 15:03:51 +01:00