Pre-compaction assistant messages carry thinkingSignature values bound to the
original conversation prefix. After compaction the prefix changes (summarized
content is replaced by the compaction summary), so Anthropic rejects those
signatures with "Invalid signature in thinking block", permanently stalling the
session through gateway restarts.
stripInvalidThinkingSignatures only catches absent/blank signatures; this adds
stripStaleThinkingSignaturesForCompactionReplay (thinking.ts) which identifies
pre-compaction assistant messages by timestamp comparison against the latest
compaction summary and strips their signature fields. Called in
sanitizeSessionHistory (replay-history.ts) before stripInvalidThinkingSignatures
for all signed-thinking providers (Anthropic, Bedrock, Vertex). Also fixes
buildSuccessorEntries (compaction-successor-transcript.ts) to strip only
pre-compaction kept entries when writing the rotation successor JSONL; uses
strict < timestamp boundary so same-instant post-compaction messages are not
affected.
Docs: update transcript-hygiene.md Anthropic and Bedrock sections.
Tests: 8 new cases for stripStaleThinkingSignaturesForCompactionReplay; 1 new
case for buildSuccessorEntries verifying pre/post-compaction signature boundary.
Fixes#90108
Fixes the OpenAI-compatible stream transport regression where a valid ChatGPT Codex HTTP 200 stream could arrive without a `content-type` header and be rejected before the OpenAI SDK consumed it.
Prepared head SHA: 0d7f8abb17
Co-authored-by: jalehman <550978+jalehman@users.noreply.github.com>
* fix(docker): qualify base image refs for podman short-name mode
Podman with short-name-mode=enforcing (the Fedora/RHEL default) blocked
the build: `FROM oven/bun:1.3.13...` is an ambiguous short name with no
alias, so Podman prompted interactively for a registry (the apparent
"hang") or, headless, failed with "short-name resolution enforced but
cannot prompt without a TTY". `node:*` only resolved because a `node`
short-name alias ships in registries.conf.d.
Fully-qualify the node and bun base images with docker.io/ so registry
resolution is deterministic. Pinned digests are unchanged, so resolved
image content is identical, and Docker/Buildx builds are unaffected.
Also qualify the docker.io/ prefix in the digest-refresh maintenance
comments so the documented update path matches the defaults and does not
reintroduce the same short-name ambiguity for Podman users.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* test(docker): expect qualified base image refs
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: sallyom <somalley@redhat.com>
Fixes#88213.
Browser tab routes now use the configured action timeout for Chrome MCP existing-session reachability checks instead of failing through the old 300ms probe. Non-Chrome-MCP profiles keep the short probe, and configured timeout values are clamped to the safe timer range.
Proof: autoreview clean; node scripts/run-vitest.mjs extensions/browser/src/browser/routes/tabs.test.ts; merge-base git diff --check; PR CI green.
Co-authored-by: Ted Li <tl2493@columbia.edu>
Keep ACP sessions_spawn runs that request streamTo=parent in the subagent registry so completion handoffs can announce back to users while inline delivery suppression still prevents double delivery.
Fixes stream-to-parent child completions dropping in direct-session channels.
Thanks @scotthuang.
Classify clearly read-only exec/bash shell probes as non-mutating so failed inspection commands no longer add misleading final tool-warning messages after a useful assistant reply. Ambiguous or mutating shell forms still fail closed as mutating, including redirects, pipes, heredocs, mutating git/gh forms, and gh web-launch flags.
Verification:
- pnpm test src/agents/tool-mutation.test.ts
- pnpm test src/agents/embedded-agent-runner/run/payloads.test.ts src/agents/embedded-agent-runner/run/payloads.errors.test.ts
- CI/check rollup for head 346853fb07 had no pending or failing checks; historical cancelled/skipped Mantis proof jobs were non-blocking.
Co-authored-by: Markus <markuscontasul@gmail.com>
Consolidate repeated gateway test setup into shared helpers and keep the preauth WebSocket fixture bounded with maxPayload.\n\nVerification: focused gateway Vitest passed, autoreview clean, and ready-state GitHub Actions CI passed on c6f6957e55.
Summary:
- The PR wraps outbound post-delivery transcript mirroring in warning-only error handling and adds regression tests for thrown and not-ok mirror append failures.
- PR surface: Source +16, Tests +61. Total +77 across 2 files.
- Reproducibility: yes. A high-confidence source reproduction is to make appendAssistantMessageToSessionTransc ... a/outbound/deliver.ts:1970 and the caller retry path treats that exception as a failed direct announcement.
Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(outbound): keep channel send durable when transcript mirror fails…
Validation:
- ClawSweeper review passed for head dfe0fd7119.
- Required merge gates passed before the squash merge.
Prepared head SHA: dfe0fd7119
Review: https://github.com/openclaw/openclaw/pull/89812#issuecomment-4611974387
Co-authored-by: harjoth <harjoth.khara@gmail.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Summary:
- The PR adds a narrow Feishu runtime-setter entrypoint, wires it into the Feishu setup entry, and adds regression coverage for setup-only runtime registration.
- PR surface: Source +7, Tests +22. Total +29 across 4 files.
- Reproducibility: yes. source inspection gives a high-confidence reproduction path: current Feishu setup-only ... ate when that setter is present. I did not run a live Feishu tenant message repro in this read-only review.
Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(feishu): wire setup runtime setter
Validation:
- ClawSweeper review passed for head befd074ca6.
- Required merge gates passed before the squash merge.
Prepared head SHA: befd074ca6
Review: https://github.com/openclaw/openclaw/pull/89814#issuecomment-4612032021
Co-authored-by: Glenn-Agent <glenn_agent@163.com>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Summary:
- The branch updates Telegram dispatch so a verbose/status final arriving after a streamed final answer uses a fresh answer-lane message, with default and progress-mode regression tests.
- PR surface: Source +14, Tests +52. Total +66 across 2 files.
- Reproducibility: yes. The linked bug report gives a concrete Telegram `/reset`, `/v on`, short-prompt path, and source inspection shows current main can route a second final payload through the finalized answer lane.
Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(telegram): isolate verbose status after streamed finals
Validation:
- ClawSweeper review passed for head 4d476a957f.
- Required merge gates passed before the squash merge.
Prepared head SHA: 4d476a957f
Review: https://github.com/openclaw/openclaw/pull/89813#issuecomment-4612006920
Co-authored-by: kesslerio <martin@kessler.io>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Add operator-only Control UI chat send timing milestones across gateway dispatch, model selection, agent-run start, dispatch completion, and post-dispatch completion. The Control UI records these server phases into the existing chat send timing buffer, and the gateway broadcast guard now scopes the new timing event with other read-visible chat events.
Summary:
- The PR makes prompt-release fence bookkeeping exception-safe so the session write lock is released even when fence reads throw, and adds a regression test for that path.
- PR surface: Source +6, Tests +27. Total +33 across 2 files.
- Reproducibility: yes. source-reproducible with provided real-output proof: current main clears `heldLock` be ... ire timing out after an injected `EIO`. I did not run the harness locally because this review is read-only.
Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(agents): release session write lock if fence read throws on promp…
Validation:
- ClawSweeper review passed for head 394d978437.
- Required merge gates passed before the squash merge.
Prepared head SHA: 394d978437
Review: https://github.com/openclaw/openclaw/pull/89811#issuecomment-4611966479
Co-authored-by: Spencer Fuller <spencer.p.fuller@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Harden Workboard modal and drawer accessibility.
Summary:
- Add Workboard dialog focus lifecycle handling for initial focus, Tab/Shift+Tab containment, Escape close, and opener restore.
- Mark Workboard background content inert/aria-hidden while modal or drawer dialogs are active.
- Add focused unit and Chromium browser smoke coverage for the audited modal/drawer accessibility requirements.
- Keep UI browser test aliases able to resolve shared workspace packages used by the Workboard view.
Verification:
- node scripts/run-vitest.mjs ui/src/ui/views/workboard.test.ts
- node scripts/run-vitest.mjs ui/src/ui/views/workboard.browser.test.ts
- (cd ui && pnpm exec vitest run --config vitest.config.ts --project browser src/ui/views/workboard.browser.test.ts)
- GitHub checks green at 6557012430
Start the active Control UI chat refresh after Gateway hello without waiting for the slower bootstrap fetch. Keep startup canvas embeds fail-closed until bootstrap config arrives, and recreate preview iframes when sandbox policy changes.
Fix WebChat stream/history reconciliation so visible assistant text survives stale history reloads, tool-history catch-up, and terminal final/error/abort events.\n\nRefactors the UI path into stream reconciliation, stream text, and typed tool-message helpers so persisted history and live stream state use the same matching rules.\n\nCloses #67035.
Fix the Control UI WebChat race where terminal assistant messages could be committed while chatStream was still live, causing history and active stream to render the same reply twice. Terminal final/aborted handling now snapshots fallback text, clears the active run/stream through the lifecycle owner, then appends the visible assistant message.\n\nFixes #71992.\n\nVerification: node scripts/run-vitest.mjs run ui/src/ui/controllers/chat.test.ts ui/src/ui/chat/run-lifecycle.test.ts ui/src/ui/chat/build-chat-items.test.ts; node scripts/run-vitest.mjs run ui/src/ui/app-chat.test.ts ui/src/ui/controllers/sessions.test.ts; node scripts/run-vitest.mjs run --config test/vitest/vitest.ui-e2e.config.ts --configLoader runner ui/src/ui/e2e/chat-flow.e2e.test.ts; Blacksmith Testbox tbx_01kt6a4zn7awkdy12d6b0q2d1q / run 26873514898; autoreview clean; PR CI 121 pass / 10 skipped.
Fixes #87699.\n\nRoutes ACK-completed Control UI chat sends through the existing run lifecycle reconciliation path so stale selected-session rows cannot re-enable the composer/Stop state after the conversation has already completed.\n\nVerification: focused UI/unit tests, Control UI E2E chat-flow test, autoreview clean, Testbox changed gate tbx_01kt68xvz17fcnmd3wj6f7pk6f, and PR CI run 26872484363 green after failed-job rerun for transient runner setup failures.
Add a coalesced chat.metadata Gateway method so the Control UI can fetch model and command metadata without blocking a clean first message path. Reuses existing models/commands builders, keeps compatibility fallback for older gateways, updates protocol artifacts, and adds focused gateway/UI/e2e coverage.
Summary:
- The PR changes shared poll-intent detection so `pollDurationHours` and `pollMulti` alone no longer make `send` actions fail, with focused unit and outbound validation coverage.
- PR surface: Source -2, Tests +40. Total +38 across 3 files.
- Reproducibility: yes. Source inspection shows current main and `v2026.5.28` expose `pollDurationHours` throu ... d message schema, classify non-zero shared duration as poll intent, and throw before a `send` can dispatch.
Automerge notes:
- No ClawSweeper repair was needed after automerge opt-in.
Validation:
- ClawSweeper review passed for head 0fd95756cd.
- Required merge gates passed before the squash merge.
Prepared head SHA: 0fd95756cd
Review: https://github.com/openclaw/openclaw/pull/89601#issuecomment-4606487310
Co-authored-by: Gabriel Fratica <gabriel@codez.ro>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Split Anthropic system prompts at the cache boundary so only stable prefixes get cache_control, strip the internal marker when cache control is disabled, and keep OpenAI-compatible Anthropic cache-control routes from caching dynamic suffixes.\n\nFixes #89386.
Forward configured stop sequences to Gemini generationConfig.stopSequences in the bundled Google transport, matching the shared Google provider behavior and the @google/genai request contract.\n\nThanks @coder999999999.
Keep startup non-breaking for existing installs when hooks.token reuses Gateway auth, but surface a startup warning, critical security audit finding, and doctor --fix repair that rotates persisted hooks.token.
Closes#87376.
Co-authored-by: Coy Geek <65363919+coygeek@users.noreply.github.com>
Enable Mistral prompt cache keys without long-retention forwarding. Update cached-read pricing and doctor migration for existing Mistral provider config. Fixes#83709.
Run `before_agent_finalize` for embedded agents before terminal delivery so revise decisions can retry without leaking a final assistant reply.
The embedded subscription now defers terminal assistant events, block replies, and lifecycle delivery until the pre-terminal gate resolves; accepted revise decisions suppress delivery, while hook failures and continue decisions finalize normally. It also preserves existing replay-invalid liveness behavior while still preventing revise after side-effecting turns.
Closes#87585
Co-authored-by: ai-hpc <mail.speedy.hpc@hotmail.com>
Move QQBot credential backups, gateway sessions, known-user records, and ref-index rows into plugin SQLite KV stores. Import shipped JSON/JSONL state files on first use and keep auxiliary known-user/ref-index state best-effort so message delivery is not blocked by cache persistence failures.
Surface stale Codex OAuth sidecar references as unresolved auth failures in auth health, model status, and gateway status instead of hiding them as generic missing auth.
Also refresh the running gateway after doctor auth-profile repairs by reloading secrets/runtime auth snapshots and then refreshing the model auth-status cache.
Thanks @TurboTheTurtle.
Fixes#84252.
* fix(providers): use native reasoning mode for direct Gemini API, keep CLI tagged
Gemini 2.5+ delivers reasoning via native thinkingParts (thinkingConfig.
includeThoughts). Having tagged mode active at the same time injects a
<think>…</think>/<final>…</final> directive into the system prompt; the
model opens a <think> block before a tool call, never closes it, and
returns an empty post-tool turn (content:[], payloads=0 error, #69220).
Fix: override resolveReasoningOutputMode in buildGoogleProvider() only —
not in the shared GOOGLE_GEMINI_PROVIDER_HOOKS. The Gemini CLI backend
(google-gemini-cli) runs gemini --output-format json and parses a text
response field, not native thought parts; it must stay on tagged mode.
A regression test confirms google-gemini-cli remains "tagged".
Also remove the dead BUILTIN_REASONING_OUTPUT_MODES entry keyed on
"google-generative-ai" from provider-utils.ts — that string is only
ever the transport model.api value, never the provider id passed to
resolveReasoningOutputMode, so the map was unreachable.
Fixes#69220
* docs: clarify Gemini reasoning output modes
* fix(google): keep Antigravity reasoning tagged
* fix(google): default direct reasoning checks to native
* fix(google): import reasoning context from plugin entry
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Fixes#88355.
When a resumed Codex context-engine thread overflows and OpenClaw retries on a fresh native thread, clear the stale thread-bootstrap projection metadata from the fresh binding. This prevents later turns from treating that fresh thread as already projected when it only received the bare retry prompt.
Verification:
- Autoreview clean: no accepted/actionable findings reported.
- CI run 26717883204 green on head 5438f8ad34.
* fix(google): add gemini-3.1-flash-lite to provider catalog
Adds the missing gemini-3.1-flash-lite model definition to the
GOOGLE_GEMINI_TEXT_MODELS array. This resolves the ProviderFailoverError
when configuring google-vertex/gemini-3.1-flash-lite.
Fixes#89390
* test(google): cover Gemini flash lite catalog row
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Refactor the agent harness surface after PR #88821 by moving compaction dispatch into its own module, splitting the harness type into explicit capability interfaces, and renaming the private agent-core class declaration to `CoreAgentHarness` while preserving the exported `AgentHarness` contract.
Verification:
- `node scripts/run-vitest.mjs src/agents/harness/selection.test.ts src/agents/command/cli-compaction.test.ts src/agents/embedded-agent-runner/compact.hooks.test.ts packages/agent-core/src/agent-loop.test.ts packages/agent-core/src/harness/messages.test.ts`
- `pnpm build`
- autoreview clean
- `pnpm check:changed` passed on Testbox `tbx_01kt407hq8sv1csm287pdj3fmp`
- PR CI merge state `CLEAN`
Stabilize repeated `openclaw doctor --fix` state repairs for legacy plugin state and installed plugin index migrations.
- Import legacy-only plugin-state sidecar rows before deciding whether live conflicts require keeping the sidecar.
- Drop expired sidecar rows only when the sidecar can be archived, avoiding repeated false migration changes.
- Let richer current install records cover legacy records only when durable legacy fields are actually preserved, without erasing npm selector intent or malformed legacy metadata.
Proof:
- `node scripts/run-vitest.mjs src/commands/doctor-state-migrations.test.ts`
- `git diff --check origin/main...HEAD`
- `.agents/skills/autoreview/scripts/autoreview --mode branch --base origin/main`
- PR CI clean for head `5f3a7e0749372a40cabd7a090cae155997481b71`
Co-authored-by: Dallin Romney <dallinromney@gmail.com>
QQBot credential backups now resolve under the active OpenClaw state directory instead of the old home-global QQBot data path. This keeps isolated gateway profiles from restoring each other's QQBot appId/clientSecret backups while preserving per-state-root recovery.
Proof: focused QQBot path/storage-laziness Vitest suite passed on Node 24.15.0, focused oxlint passed, source-runtime two-root backup proof passed, exact-head CI run 26814565282 passed, and ClawSweeper re-review run 26815054980 marked proof sufficient.
Closes#84313.
Co-authored-by: Coy Geek <65363919+coygeek@users.noreply.github.com>
Surface active official external plugin version drift in gateway status diagnostics so users can see when a host/package update left npm or ClawHub plugins behind the running local gateway. The advisory uses the daemon service install records, compares against the running gateway version, gives detailed fix commands in deep status, and avoids local-state drift checks for remote gateway mode or explicit status probe URLs.
Co-authored-by: Hussein Nourelddine <hussein@gptc.com.kw>
Fixes status/update detection for npm-installed OpenClaw packages that ship npm-shrinkwrap while preserving pnpm and Bun install ownership.
Fixes#87732.
Supersedes #88283.
Proof: focused infra Vitest shard, autoreview clean, Crabbox install matrix, and PR CI all green.
Reset the session command lane when stuck-session recovery aborts and drains a ghost embedded run but queued lane work remains. This preserves pending user messages by using the existing lane recovery pump instead of leaving them stranded after recovery reports success.
Adds focused regression coverage for the abort=true, drained=true, queuedCount=1 path.
Fixes#89208.
Supersedes #89293.
Thanks @LiLan0125.
Co-authored-by: 李兰 0668001394 <li.lan3@xydigit.com>
Fix Discord progress-mode reasoning streams so delta chunks accumulate before display formatting, preserving raw Thinking/Reasoning-prefixed content and balanced truncation.\n\nFixes #83983.\n\nThanks @giodl73-repo for the fix and live Discord proof.
Reject whitespace-only cron delivery target strings before cron input normalization can trim and drop them, so bad delivery targets return INVALID_REQUEST instead of behaving as omitted fields.
Keep explicit null update clears for delivery, failure destination, and completion destination fields.
Co-authored-by: gaozixiang1 <gaozixiang1@xiaomi.com>
Co-authored-by: Lanzhi <lizhan3@xiaomi.com>
Honor explicit provider/model request timeoutSeconds when the agent run timeout is the no-timeout sentinel, and keep explicit run timeout overrides from being capped by agent defaults.
Verification:
- pnpm test src/agents/embedded-agent-runner/run/llm-idle-timeout.test.ts -- --reporter=verbose
- .agents/skills/autoreview/scripts/autoreview --mode branch --base origin/main
- CI run 26812803642 passed on the rebased PR head
- Real behavior proof run 26812917801 passed after maintainer proof override
Co-authored-by: zhongqiongbo1 <zhongqiongbo1@xiaomi.com>
Co-authored-by: Lanzhi <lizhan3@xiaomi.com>
Summary:
- The branch adds a memory-core `startup_retry` reconciliation mode and regression tests for quiet startup retries, retry-window exhaustion, and live-config retry semantics.
- PR surface: Source +9, Tests +114. Total +123 across 2 files.
- Reproducibility: yes. from source: current main routes the first startup retry through runtime reconciliatio ... st expects the warn-level `cron service unavailable` log. I did not execute tests in this read-only review.
Automerge notes:
- Ran the ClawSweeper repair loop before final review.
- Included post-review commit in the final squash: fix(memory-core): keep startup cron retries quiet
Validation:
- ClawSweeper review passed for head 7220f940d0.
- Required merge gates passed before the squash merge.
Prepared head SHA: 7220f940d0
Review: https://github.com/openclaw/openclaw/pull/89075#issuecomment-4592446250
Co-authored-by: bennewell35 <newelljben@gmail.com>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Add `ownsNativeCompaction` capability to CliBackendPlugin so backends
that manage their own transcript compaction (e.g. Claude Code) can
declare it once and OpenClaw defers instead of fighting or failing.
Today only Codex declares compaction ownership (via the embedded runner
path + agentHarnessId). Claude-cli never reaches that path because it
runs as a CLI subprocess with no harness id set, so the safeguard
summarizer fires and hard-fails the turn.
This PR:
- Adds `ownsNativeCompaction?: boolean` to the backend plugin type
- Propagates it through all 4 backend resolution paths
- In `runCliTurnCompactionLifecycle`, when a backend declares ownership
but has no harness endpoint, returns a no-op instead of falling
through to the safeguard
- Sets the flag on claude-cli (first adopter)
Codex's existing native-harness path is unchanged: when
`isNativeHarnessCompactionSession` matches, the harness compaction
endpoint is still called as before.
Generalizes the partial fix in #87785 (codex-scoped) to a capability
any backend can opt into.
* fix(qqbot): allow RFC2544 benchmark range for token fetch (#88984)
QQ Bot `bots.qq.com` token-fetch path was failing for users whose DNS resolver maps the hostname into the RFC 2544 benchmark range `198.18.0.0/15` (commonly seen with fake-IP proxy stacks: sing-box, Clash, Surge, WSL2 DNS). The default SSRF guard treats that range as private and blocks the request, surfacing as "Network error getting access_token: Blocked: resolves to private/internal/special-use IP address".
Pass a host-scoped `SsrFPolicy` (`allowRfc2544BenchmarkRange: true`) to the single hard-coded `TOKEN_URL` request, mirroring the existing `QQBOT_MEDIA_SSRF_POLICY` pattern used by the media path. Because `TOKEN_URL` is a const and not user-controlled, the relaxation cannot widen attack surface to other hosts.
Adds a regression test asserting `policy: { allowRfc2544BenchmarkRange: true }` is forwarded into `fetchWithSsrFGuard`, and updates the existing equality assertion accordingly.
Fixes#88984
* fix(qqbot): scope token ssrf policy
Repairs a batch of narrow model/provider edge cases:
- honor OpenAI and Anthropic base URL environment overrides when provider config does not set an explicit base URL
- preserve OpenRouter Anthropic cache retention while stripping unsupported transport options
- allow apply_patch for non-OpenAI providers when the tool config otherwise permits it
- prune stale same-provider model selections from configure/model picker state
- expose GitHub Copilot bundled thinking policy metadata to offline/provider-policy lookups
- repair additive SQLite shared-state upgrades for existing databases
- keep same-size rotated log readers from reusing stale content in CI tooling
Proof:
- GitHub PR checks green on exact head 46514909b0
- Crabbox delegated Blacksmith Testbox tbx_01kt3em5r9vd7g0bnykrff6jdk exited 0
- Focused local Vitest/oxlint/format proof recorded in PR body and land-ready comment
Fixes#80347.
Fixes#88357.
Fixes#45269.
Supersedes #74427, #74432, #79370, #79894, #80366, and #88359.
Summary:
- Adds defensive failed-count reads in auto-reply/ACP accounting and Feishu fallback paths, plus a focused regression test, while keeping `ReplyDispatcher.getFailedCounts` required.
- PR surface: Source +24, Tests +35. Total +59 across 5 files.
- Reproducibility: yes. from source inspection. Current main calls `dispatcher.getFailedCounts().final` and si ... issing that method follows a clear TypeError path; the source PR also supplied terminal before/after proof.
Automerge notes:
- PR branch already contained follow-up commit before automerge: fix(auto-reply): guard missing dispatcher getFailedCounts without wea…
Validation:
- ClawSweeper review passed for head 0bdfb4adeb.
- Required merge gates passed before the squash merge.
Prepared head SHA: 0bdfb4adeb
Review: https://github.com/openclaw/openclaw/pull/89318#issuecomment-4598624344
Co-authored-by: Alix-007 <li.long15@xydigit.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com>
Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com>
Approved-by: takhoffman
Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
Explicit non-Codex plugin harness runtimes now bypass stale OpenClaw provider auth cooldowns before harness startup, while Codex/OpenClaw and missing-harness gates remain fail-closed. Fixes#85105.
Fix live model inference edge cases across provider streaming, model switching, outbound delivery, and gateway tool resolution.
Includes live/provider issue fixes and leaves #89100 explicitly partial for the remaining FM-2 group routing case.
Treat targetless current-chat message-tool media telemetry as delivered for generated-media completion dedupe while preserving fallback delivery for mismatched provider/account/thread evidence.
Real behavior proof was added from the live iMessage generated-image run: inbound id 5805, exactly one outgoing media reply id 5806, and no follow-up generated-image fallback.
Co-authored-by: omarshahine <10343873+omarshahine@users.noreply.github.com>
Reviewed-by: @lobster
Keep iMessage native typing indicators alive through long tool-running gaps by bridging tool-start activity into the existing typing controller, while preserving typingMode and sendPolicy suppression semantics.
Real behavior proof was added from the live iMessage generated-image run: inbound id 5805, outgoing media reply id 5806, and requester-observed typing during the 84s tool path.
Co-authored-by: omarshahine <10343873+omarshahine@users.noreply.github.com>
Reviewed-by: @lobster
Route Slack plugin approval delivery through the shared native approval route gates while preserving Slack Block Kit buttons and plugin resolver semantics.
Verification: Slack/native approval unit tests, Slack QA Lab, and live clawd native plugin approval via Slack desktop.
* fix(agents): actionable copy for exhausted auth-profile failover
The pi-embedded runner threw a generic "No available auth profile for
<provider> (all in cooldown or unavailable)" message whenever every
configured profile was in cooldown, even though the failover machinery
had already resolved a concrete reason (auth, billing, rate_limit,
session_expired, etc.). The user-facing copy never used that reason and
never told the user how to recover.
Route the resolved reason through a single presenter
(`formatAuthProfileFailureMessage`) that composes a reason-specific
sentence with `buildProviderAuthRecoveryHint`, so FailoverError.message
ships with the right `openclaw models auth login --provider <id>` hint
when the cause is authentication/session/billing, and falls back to the
underlying provider error text otherwise. Helper moved out of
`src/commands/` into `src/agents/` because `src/agents/` cannot depend
on `src/commands/`.
* fix(agents): soften auth-profile failure copy for non-technical users
* refactor(agents): drop guidance re-export shim and de-brittle failure-copy tests
- Delete `src/commands/provider-auth-guidance.ts` and point doctor-auth, auth-choice.model-check, and models/list.status-command directly at `src/agents/provider-auth-recovery-hint.ts`. The cold-imports test moves with it.
- Rewrite `failure-copy.test.ts` to assert behavior (recovery-hint dispatch, provider mention, cause-suffix dedup) instead of pinning exact long copy strings, so wording tweaks no longer require a test update in two places.
Add a combined chat.startup gateway method for Control UI startup hydration so first chat load can receive history and agents in one RPC, while falling back to chat.history for older/unadvertised gateways. Verified with focused UI/gateway tests, tsgo/oxlint/diff checks, clean autoreview, and Testbox changed gate tbx_01kt1dt6fqdtdbprsk48z8fn71.
Speed up Control UI first global chat sends by letting safe literal-global startup refresh use the fresh hello default before agents.list finishes, while keeping stale carried/cached agent ids out of that fast path. Adds chat history/send and gateway chat.send timing markers for the next latency pass.
Add chat-send first visible assistant output telemetry in the Control UI, plus Gateway diagnostics correlation attributes for chat.send dispatch spans. Verified with focused UI/Gateway tests, tsgo, oxlint, autoreview, PR checks, and Testbox-through-Crabbox check:changed.
Reduce Control UI draft-update work by guarding chat composer controls while keeping locale, session, model, settings, and busy-state invalidation. Verification: focused UI tests, format/lint/typecheck, autoreview clean, and changed gate tbx_01kt12rgjs8c077p2s0wmcsbyf.
Reduce Control UI draft-update work by guarding transcript group rendering while preserving assistant attachment availability invalidation. Verification: focused UI tests, format/lint/typecheck, autoreview clean, and changed gate tbx_01kt11qyc20ejbsbt8kd79bamx.
Revert release-time extension lane isolation for Telegram and memory, and make Telegram timer-flush tests wait for async side effects after manually firing timers.
Verification:
- pnpm test:serial extensions/telegram/src/bot.create-telegram-bot.channel-post-media.test.ts extensions/telegram/src/bot.create-telegram-bot.media-group-skip-warning.test.ts extensions/telegram/src/bot.media.stickers-and-fragments.e2e.test.ts extensions/telegram/src/bot.media.downloads-media-file-path-no-file-download.e2e.test.ts test/vitest-scoped-config.test.ts
- pnpm exec oxfmt --check on touched files
- git diff --check on touched files
Reduce Control UI typing work by avoiding slash-menu rerenders for ordinary non-command drafts. Verification: focused UI tests, format/lint/typecheck, autoreview clean, and changed gate tbx_01kt1086xrbxfzm85vynsf25hq.
Debounce draft-only Control UI chat composer persistence while snapshotting pending drafts so session changes and teardown still flush the correct state. Verified with focused UI lifecycle/composer tests, format, oxlint, tsgo core/UI test, clean autoreview, and PR checks.
* fix(diagnostics): clear embedded-run activity when recovery declares lane idle
Stuck-session recovery transitions a lane to idle via the recovery
coordinator, but only mutated the session-state store. When an aborted
embedded run was removed without markDiagnosticEmbeddedRunEnded, the
activity store kept hasActiveEmbeddedRun set, so the liveness sweep
reported idle/embedded_run and isIdleQueuedRecoverableSessionStall
re-triggered recovery indefinitely.
Reconcile the activity store from the authoritative idle declaration by
clearing the session's embedded-run owners. The existing generation
guard already excludes any newer run that re-armed activity, so a live
requeued run is preserved.
* fix(diagnostics): reconcile tool/model activity on authoritative idle cleanup
clearDiagnosticEmbeddedRunActivityForSession (renamed from
clearDiagnosticEmbeddedRunsForSession) now clears the aborted run's tool and
model markers alongside the embedded-run owners, matching the default
markDiagnosticEmbeddedRunEnded teardown. Clearing only the owner set left the
lane as idle + orphaned tool/model activity, which
isIdleQueuedRecoverableSessionStall still treats as recoverable while work is
queued, so the liveness sweep kept re-triggering recovery instead of converging.
Adds regression cases with stale tool and model markers plus queued work.
* test(phone-control): align service mocks with keyed store API
* fix(diagnostics): preserve rearmed recovery activity
* fix(diagnostics): clear recovered owner markers
* fix(diagnostics): clear recovered embedded work keys
* fix(diagnostics): ignore stale same-key recovery owners
* fix(diagnostics): preserve same-session recovery rearm
* fix(diagnostics): ignore stale queued activity starts
* fix(diagnostics): record recovery cutoffs for empty activity
* fix(diagnostics): preserve fresh recovery markers
* fix(diagnostics): prune stale activity before fresh recovery block
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(agents): clear legacy auto fallback pins
* fix(agents): repair legacy auto-fallback test mock and tighten review feedback
Add hasLegacyAutoFallbackWithoutOrigin to the live-model-switch agent-scope mock so the agents-core lane runs, simplify the redundant hasSessionModelOverride guard, use a single source of truth for the legacy-pin staleness check with a comment on the load-bearing modelKey guard, and add preservation/edge-case/guard regression coverage. Rename the misleading primary-probe agent test.
* style(agents): format rebased fallback fix
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Make first Control UI chat sends visibly queue during pending model saves, preserve early streaming deltas that arrive before chat.send ACK, and keep model-wait queued prompts scoped/retryable across session switches.
@@ -16,6 +16,10 @@ Use this with `$release-openclaw-maintainer` and `$openclaw-testing` when a rele
- Watch one parent run plus compact child summaries. Avoid broad `gh run view` polling loops; REST quota is easy to burn.
- Fetch logs only for failed or currently-blocking jobs. If quota is low, stop polling and wait for reset.
- Treat live-provider flakes separately from code failures: prove key validity, provider HTTP status, retry evidence, and exact failing lane before editing code.
- Full Release Validation parent monitors fail fast: once a required child job
fails, the parent cancels the remaining child matrix and prints the failed
job summary. Inspect that first red job instead of waiting for unrelated
matrix tails.
## Preflight
@@ -73,6 +77,9 @@ gh workflow run full-release-validation.yml \
```
Use `release_profile=stable` unless the operator explicitly asks for the broad advisory provider/media matrix. Use narrow `rerun_group` after focused fixes.
Publish with `openclaw-release-publish.yml` using `release_profile=from-validation`
unless a maintainer intentionally wants to cross-check a specific profile; the
publish workflow reads the effective profile from the full-validation manifest.
case "${PNPM_CONFIG_MODULES_DIR:?}" in "$volatile_root"/*) ;; *) echo "::error::PNPM_CONFIG_MODULES_DIR must stay under $volatile_root"; exit 1 ;; esac
case "${PNPM_CONFIG_VIRTUAL_STORE_DIR:?}" in "$volatile_root"/*) ;; *) echo "::error::PNPM_CONFIG_VIRTUAL_STORE_DIR must stay under $volatile_root"; exit 1 ;; esac
- Agents and CLI-backed runtimes recover more cleanly from interrupted tool calls, stale session bindings, compaction handoffs, and media delivery retries. (#88129, #88136, #88141, #88162, #88182)
- Channels and mobile delivery are steadier across Telegram, WhatsApp, iMessage, Slack, Discord, Microsoft Teams, Google Chat, Google Meet, and iOS realtime Talk. (#88096, #88105, #88183, #88231)
- Provider and plugin requests now bound more timers, retries, OAuth/device-code lifetimes, media downloads, local service probes, and generated-content polling paths before they can hang a run.
- Skills, session metadata, gateway runtime state, plugin metadata, and store writes do less repeated work on hot paths while keeping config and dispatch behavior stable.
- Skills and plugin loading now handle stale disabled snapshots and loader failures more clearly, so channel turns avoid disabled SecretRefs and operators get better recovery guidance. (#79072, #79173) Thanks @zeus1959.
- Skill Workshop now has a fuller Control UI flow with proposal lists, today actions, revision handoff, searchable file previews, review states, locale coverage, and reusable session routing.
- Chat and Control UI startup paths keep sends alive through history loading, stream deltas incrementally, skip markdown work while streaming, and expose calmer composer controls. (#88772, #88825)
- Provider coverage and model metadata now include MiniMax M3, account OAuth endpoints, Google/Vertex catalog fixes, OpenRouter SQLite model caching, Copilot Claude 1M capabilities, Foundry reasoning alignment, and OpenAI response replay guards. (#88480, #88512, #88851, #88860)
- iMessage monitor state, inbound queues, and plugin install ledgers moved toward SQLite-backed state so restarts and local monitors recover with less duplicate filesystem scanning. (#88794, #88797)
- Release, CI, Docker, E2E, and diagnostics lanes now cap more logs, response bodies, readiness probes, artifact checks, and status polling so failures report bounded proof instead of stalling.
- Plugin and skill installs now use an operator install policy instead of the old dangerous-code scanner path, with clearer doctor, CLI, ClawHub, and troubleshooting surfaces for package, archive, source, upload, and marketplace installs. (#89516) Thanks @joshavant.
- Chat, Control UI, Skill Workshop, Workboard, Android companion shell, and WebChat flows now preserve visible streaming text, reconcile completed sends, expose ACK timing, add Workboard keyboard movement, harden dialog accessibility, lazy-load usage views, keep current chat toggles working, and improve Android companion-first shell navigation. (#89801, #89777, #89802) Thanks @vincentkoc.
- Security, policy, and config recovery now reject corrupt shell snapshots, unsupported policy keys, unsafe exec approval precheck environments, malformed script limits, and suspicious gateway startup configs while adding data-handling conformance checks. (#89701, #87074, #81488, #87056, #89480) Thanks @RomneyDa, @giodl73-repo, and @mmaps.
- Gateway, agent, Codex, provider, model, and memory paths now recover session write-lock release failures, abandoned Codex app-server startups, stream-to-parent ACP spawns, custom-provider runtime fanout, bundled provider aliases, prompt-cache boundaries, Gemini stop sequences, Kimi cache markers, and watcher pressure warnings. (#89811, #89244) Thanks @RomneyDa and @takhoffman.
- Release, CI, Docker, Crabbox/Testbox, package, and E2E validation lanes now bound more network calls, malformed numeric limits, process groups, cleanup leaks, package hydration paths, Windows installer publishing, release asset verification, and log drains so failures produce bounded proof instead of hanging.
### Changes
- Docs: add a dedicated Skill Workshop guide covering governed skill creation, reviewable proposals, CLI, Gateway, agent tool behavior, approval policy, support files, and recovery. Thanks @shakkernerd.
- Plugins/security: replace dangerous-code scanner enforcement with operator install policy, install-policy context, doctor checks, install/update CLI wiring, ClawHub metadata paths, and package/archive/source/upload lifecycle coverage. (#89516) Thanks @joshavant.
- Telegram/channels: show commentary and reasoning in progress drafts, share progress draft compositors across channel plugins, and keep Telegram polling stop/reset boundaries cheaper and more reliable.
- Release metadata: align the root package, publishable plugin manifests, generated shrinkwraps, appcast, iOS, Android, macOS, Matrix plugin changelog, and docs/generated baselines with the 2026.6.2 beta train.
- Release/packaging: promote Windows node installer publishing, require verified Windows release asset links, and document GitHub release-note edits.
- Docs: refresh Windows Hub setup guidance and document Gateway, CLI, and plugin SDK helper contracts.
### Fixes
- Channels/outbound: keep channel sends durable when transcript mirroring fails, stop schema-padded poll modifiers from blocking normal sends, preserve WebChat `sessions_send` handoffs, preserve Discord channel-label suppression while hiding internal agent failure traces, match Discord libopus error shapes, and sanitize Discord tool progress scaffolding. (#89626, #89812, #89601) Thanks @Petru2224, @codezz, and @takhoffman.
- Telegram/Feishu: require admin rights for Telegram target writeback, keep Telegram DM exec approval allowlists working with `ask:off`, prevent Telegram preview duplication across streaming modes, isolate verbose status after streamed finals, cancel clean restart stop timers, slow polling restart storms, and wire Feishu setup runtime setters. (#88973, #89035, #89813, #89814) Thanks @pgondhi987, @zhangguiping-xydt, and @takhoffman.
- Feishu: preserve full streaming card content by sending the merged text on each update instead of only the latest delta, so card readers see complete output when intermediate frames are missed. (#90181) Thanks @mushuiyu886.
- Chat/UI/Gateway: preserve visible chat stream text, clear stale stream buffers before terminal commits, reconcile completed sends, scroll pending sends into view, harden Workboard dialog accessibility, stabilize WebChat prompt-cache affinity, overlap chat catalog startup, render chat history incrementally, lazy-load usage dashboard, and report gateway health auth diagnostics. (#89337) Thanks @RomneyDa.
- Memory/build/update: warn after startup watcher pressure checks, externalize optional Baileys image backends, restore and pin Canvas A2UI compatibility assets, keep plugin repair fetch failures nonblocking, restore Skill Workshop view switching, and keep the current chat toggle active after awaited session switches. (#89244) Thanks @RomneyDa.
- Plugins/auth: keep Hermes migration reports pointed at SQLite auth-profile stores and keep plugin auth-profile reuse tests on the current store path.
- Plugins/CLI: avoid importing the runtime plugin loader only to clear in-process caches after short-lived plugin install, enable, disable, update, and uninstall commands refresh registry metadata.
- Release/CI/E2E: restore package changelog extraction after the post-2026.6.1 version bump, keep hydrated pnpm modules under `node_modules` for ARM/Linux package lifecycle scripts, keep OpenAI live-cache prerequisites advisory while Anthropic prerequisites stay blocking, retry Windows Parallels background log appends on transient file-lock errors, bound candidate GitHub and cross-OS Discord fetches, harden ARM smoke/browser checks, show Docker build heartbeats, reset Crabbox pnpm hydrate state, and isolate Testbox/Docker/release journey artifacts.
- Release/CI/E2E: keep Crabbox hydrate pnpm stores on the persistent cache volume while still resetting volatile modules, reducing cold installs and runner memory churn.
- Release/CI/E2E: fail secret-provider proof startup immediately when the gateway exits by signal instead of waiting for the readiness timeout.
- Release/CI/E2E: report plugin gateway gauntlet command-log write failures as failed rows instead of crashing the harness from child-process callbacks.
- Release/CI/E2E: abort stalled Kitchen Sink RPC readiness probes as soon as the gateway exits so proof failures return promptly.
- Release/CI/E2E: keep Parallels JSON-mode progress on stderr so macOS, Linux, Windows, and aggregate update smoke summaries stay parseable on stdout.
- Release/CI/E2E: fail Crabbox sparse-sync runs clearly when their temporary full checkout disappears while the child process is running, instead of pretending the child's deleted cwd can be repaired.
- Release/CI/E2E: fail PTY-backed E2E commands when transcript logs cannot be written instead of letting missing proof capture crash around a live child process.
- Release/CI/E2E: fail mock OpenAI request-log write errors with clear HTTP responses instead of leaving provider proof clients waiting on a broken socket.
- Release/CI/E2E: fail Parallels host-command log write errors through the command result path instead of leaving streaming smoke phases unresolved.
## 2026.6.1
### Highlights
- Agents and CLI-backed runtimes recover more cleanly from interrupted tool calls, stale session bindings, compaction handoffs, auth-profile failover, reasoning-tag cleanup, and media delivery retries. (#85798, #87484, #88129, #88136, #88141, #88162, #88182, #88924, #89220) Thanks @RomneyDa, @neeravmakwana, and @omarshahine.
- Channels and mobile delivery are steadier across Telegram, WhatsApp, iMessage, Slack, Discord, Microsoft Teams, Google Chat, Google Meet, QQBot, and iOS realtime Talk. (#88096, #88105, #88183, #88749, #88866, #88948, #88984, #89015, #88231) Thanks @omarshahine, @Jensenwgd, and @sliverp.
- Provider and plugin requests now bound more timers, retries, OAuth/device-code lifetimes, media downloads, local service probes, generated-content polling, provider-catalog failures, reasoning output, and model catalog paths before they can hang a run. (#88480, #88512, #88767, #88781, #88851, #88860, #89343, #89379, #89400) Thanks @vincentkoc, @charles-openclaw, @zz327455573, @849261680, and @XuZehan-iCenter.
- Skills, Skill Workshop, and plugin loading now handle proposal review, stale disabled snapshots, support-file approvals, locale/routing fixes, and loader failures more clearly, so channel turns avoid disabled SecretRefs and operators get better recovery guidance. (#79072, #79173, #88734) Thanks @zeus1959 and @shakkernerd.
- Chat and Control UI startup paths keep sends alive through history loading, stream deltas incrementally, skip markdown work while streaming, keep drafts local while typing, clear the composer after sends, trace first-output latency, cache transcript renders, prioritize first connect, and expose calmer composer controls and notification settings. (#74715, #88772, #88825, #88952, #88960, #88998, #89030, #89106) Thanks @VladyslavLevchuk, @vincentkoc, and @sallyom.
- iMessage monitor state, inbound queues, Discord thread bindings, plugin install ledgers, session metadata, gateway runtime state, plugin metadata, memory watchers, and store writes moved toward SQLite-backed or cached state so restarts and hot paths do less repeated work. (#88794, #88797, #88866, #89075, #89185, #89188, #85351) Thanks @RomneyDa and @NianJiuZst.
- Release, CI, Docker, E2E, plugin install, update, doctor, diagnostics, and security lanes now cap more logs, response bodies, readiness probes, artifact checks, status polling, child workflow waits, docker package cleanup, quiet test stalls, downgrade repair, and health probes so failures report bounded proof instead of stalling. (#84988, #87914, #87952, #88966, #89169, #89701, #89731) Thanks @LibraHo, @Niriakot, @MukundaKatta, and @RomneyDa.
### Changes
- Docs: add a dedicated Skill Workshop guide covering governed skill creation, reviewable proposals, CLI, Gateway, agent tool behavior, approval policy, support files, and recovery; refresh ClawHub cards; and add ClawHub CLI, iMessage SSH-wrapper TCC, Android helper, diff-language, and host-local media-send guidance. (#79658, #88734, #88758, #88865, #89297) Thanks @simplyclever914, @shakkernerd, @vyctorbrzezowski, @TurboTheTurtle, @RomneyDa, and @Wang-Yeah623.
- Skills: let the `skill_workshop` agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd.
- Skills: let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd.
- Skills: let pending proposals be revised in place with versioned, dated proposal frontmatter before approval. Thanks @shakkernerd.
- Plugins: externalize the GitHub Copilot agent runtime as the official `@openclaw/copilot` plugin with npm and ClawHub publish metadata.
- iOS: add hosted push relay defaults, realtime Talk playback, and a guarded WebSocket ping path for more reliable mobile sessions. (#88096, #88105, #88231)
- Workboard: add orchestration primitives and agent coordination tools for multi-agent planning and run tracking. (#87469)
- Workboard: wire task-backed board runs and show task comments in the edit modal.
- Code mode: add internal namespaces for scoped agent/global sessions and exact namespace tool dispatch. (#88043)
- Code mode: add MCP API files and docs for code-mode integrations.
- Gateway: support Tailscale Serve service names for local service routing.
- Control UI: add a Dreaming-tab agent selector and propagate the selected agent through Dreaming status, diary, and diary actions. (#78748) Thanks @stevenepalmer.
- Control UI: add calmer chat composer controls for active chat entry. (#88772)
- Control UI: add calmer chat composer controls, local draft typing state, and first-output latency instrumentation for active chat entry. (#88772, #88998) Thanks @vincentkoc.
- Plugins: add a SecretRef provider integration manifest contract and extract shared LLM core packages for provider/plugin reuse. (#82326, #88117)
- Plugin SDK: add typed presentation command actions and the bounded `resolve_exec_env` hook for plugin-provided exec environment contributions. (#88721)
- Plugins: persist the plugin install index in SQLite so installed package lookup survives reloads with less filesystem scanning. (#88794)
- Providers: add MiniMax M3 model support. (#88860)
- Tools/media: allow validated host-local text document media sends while keeping unsafe plain-text media sends blocked. (#79658) Thanks @simplyclever914.
- Doctor: add disk space health checks and stabilize post-upgrade JSON probes.
- Channels: store inbound queues in SQLite and migrate iMessage monitor state to SQLite-backed tracking. (#88797)
- Skills: add the core skills index and centralize skills runtime loading, status, filtering, and prompt formatting.
### Fixes
- Release/CI/E2E: fail early when Crabbox sparse-sync full checkouts do not have enough local disk, with guidance for moving the sync root.
- Build: render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.
- Plugins: stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.
- Control UI: serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.
- Scripts/UI: forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.
- Gateway: return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.
- Release/CI/E2E: keep temporary full-sync checkouts alive while slow Crabbox leases boot, so sparse worktree runs do not lose their sync source before file-list generation.
- Release/CI/E2E: normalize inherited Linux `C.UTF-8` locale settings before raw AWS macOS Crabbox bootstrap commands, avoiding macOS locale warnings during package-manager hydration.
- Release/CI/E2E: keep gateway watch regression checks from copying large static plugin assets inside the measured idle window.
- Update: keep core updates nonblocking when missing external plugin repair downloads or soft plugin repair warnings would otherwise stall, pin post-core plugin compatibility to the downgraded core version, and still block installed active plugin payload smoke failures. (#84431, #87914, #87952) Thanks @TurboTheTurtle, @Niriakot, and @MukundaKatta.
- Agents/providers: keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as `null` or arrays.
- Release/CI/E2E: reset incremental log readers when watched log files rotate without shrinking, so same-size replacements do not hide new readiness or RPC lines.
- Talk: preserve explicit `null` payloads on controller-created turn and output-audio lifecycle events.
- Agents/TUI: keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.
- Agents/media: keep async image, music, and video generation starts from ending the Codex turn, so mixed requests can continue with summaries or other work while media renders in the background.
- Agents/media: keep async image, music, and video generation starts from ending the Codex turn, avoid duplicate generated-media fallbacks, and let mixed requests continue with summaries or other work while media renders in the background. (#89220) Thanks @omarshahine.
- Agents/Codex: keep public OpenAI API-key profiles from being treated as native Codex app-server auth while preserving persisted Codex OAuth sessions.
- Agents/Codex: stream Codex app-server final-answer partials to live reply previews, preserve ACP metadata in SQLite, prefer real tool results over synthetic repair output, prevent aborted app-server turn handles from lingering, migrate legacy OpenAI Codex `lastGood` auth state, and preserve workspace/session metadata through ACP runtime refactors. (#88405, #88724, #88730) Thanks @vincentkoc.
- Control UI: keep collapsed tool cards labeled with the tool name and action instead of generic output text. Thanks @shakkernerd.
- Agents/Codex: surface Skill Workshop guidance in Codex app-server prompts when `skill_workshop` is available. Thanks @shakkernerd.
- Agents/auth: write auth profiles atomically, add force re-login recovery, preserve workspaces during state-only uninstall, and compact before oversized turns so recovery paths avoid partial state.
- Skill Workshop: restore and localize the Control UI board/today view switcher so review workflows keep their intended layout toggle across locales. Thanks @shakkernerd.
- Agents/auth: write auth profiles atomically, dispatch auth failures by type, add force re-login and exhausted-failover recovery, clear legacy auto fallback pins, preserve workspaces during state-only uninstall, and compact before oversized turns so recovery paths avoid partial state. (#85798, #87484, #89181) Thanks @RomneyDa and @neeravmakwana.
- Skills: skip disabled skill env overrides from stale persisted snapshots so disabled skill `apiKey` SecretRefs cannot abort embedded or channel turns. (#79072, #79173) Thanks @zeus1959.
- Skill Workshop: render the Control UI tab from filtered navigation state and keep filtered fallback routing stable.
- CLI: avoid live catalog validation during `openclaw agents add`, so adding a secondary agent no longer depends on provider catalog availability. (#76284, #88314) Thanks @zhangguiping-xydt.
- CLI: keep `plugins list --json` on the snapshot-only path so plugin sweeps avoid loading the full runtime status graph.
- CLI/desktop: bridge WSL clipboard operations through the shell and recognize manual-update launchd jobs. (#88764)
- CLI: harden CLI and plugin edge cases, and keep `plugins list --json` on the snapshot-only path so plugin sweeps avoid loading the full runtime status graph. (#88896)
- CLI/desktop: bridge WSL clipboard operations through the shell, recognize manual-update launchd jobs, and keep machine-readable startup output parseable during progress setup. (#88764, #88689) Thanks @alexzhu0.
- Plugins: make PixVerse external-plugin ClawHub metadata explicit and keep it out of bundled dist builds.
- Plugins: clarify plugin loader failure guidance so missing or incompatible plugin packages point operators at the right repair path.
- Plugins: preserve npm plugin roots after blocked installs, isolate cached tool runtime siblings, and isolate web-provider factory failures so one bad plugin does not poison sibling runtime paths. (#77237, #88807)
- Plugins: clarify plugin loader failure guidance and treat soft plugin repair warnings as nonfatal so missing or incompatible plugin packages point operators at the right repair path without blocking unrelated work. (#84431) Thanks @TurboTheTurtle.
- Plugins: preserve npm plugin roots after blocked installs, skip plugin-local `openclaw` peer symlinks during rollback snapshots, relink those peers after restore, isolate cached tool runtime siblings, isolate provider catalog projections and web-provider factory failures, and keep private LLM-core declarations bundled so one bad plugin does not poison sibling runtime paths. (#77237, #88767, #88807, #89336) Thanks @vincentkoc and @RomneyDa.
- Providers/media: cap local service, model, usage, queue, generated media, TTS, music, workflow polling, and provider OAuth request timers across hosted and local providers.
- Release/CI/E2E: bound release candidate reads, beta smoke REST calls, changelog restore, kitchen-sink and bundled plugin readiness probes, secret-provider probes, Vitest routing, and mainline test flakes. (#88127, #88137, #88155, #88160)
- Release/CI/E2E: bound release candidate reads, beta smoke REST calls, plugin npm verification commands, changelog restore, cross-OS process groups, kitchen-sink and bundled plugin readiness probes, secret-provider probes, Telegram credential timeouts, Control UI i18n and CLI startup metadata generation, Vitest routing, dependency guard admin approvals, child workflow failure detection, quiet Node test shard stalls, dist cache restores, Docker base-image/package cleanup, and mainline test flakes. (#84988, #88127, #88137, #88155, #88160, #88966, #89169) Thanks @LibraHo and @RomneyDa.
- Release/CI/E2E: keep Kitchen Sink live plugin MCP probes resolving source-checkout workspace packages and align the live gauntlet with current Kitchen Sink diagnostics.
- Release/CI/E2E: run the secret-provider integration proof through the repo pnpm runner so native macOS and Windows validation use the hydrated package-manager shim.
- Release/CI/E2E: run the Telegram desktop proof gateway through the repo pnpm runner so native macOS proof uses the hydrated package-manager shim.
- Docs/CI: run Mintlify anchor checks through the repo pnpm runner so docs link validation works when pnpm is only available through the hydrated package-manager shim.
- Agents: keep configured fallback model metadata typed so provider params, context-token caps, and media input limits do not break changed-gate typechecks.
- Agents: accept hidden `sessions_send` body aliases before validation while keeping the model-facing `message` schema canonical. (#88229) Thanks @zhangguiping-xydt.
- Chat/UI: preserve startup chat sends during history loading, unblock the initial Control UI chat send, stream chat deltas incrementally, skip markdown parsing while streaming, honor Chromium executable overrides, and detect system Chromium for E2E.
- Channels: preserve long Feishu streaming replies, send visible fallbacks when accepted Feishu turns produce no final reply, tolerate iMessage self-chat timestamp skew, decode Nostr `npub` allowlists correctly, and suppress raw provider errors during channel delivery. (#87896)
- Config/status/doctor: skip unresolved shell references in state-dir dotenv files, resolve gateway auth secrets during deep status audits, respect explicit PI runtime policy, report runtime tool-schema errors, and keep post-upgrade JSON stable. (#88288)
- Gateway/session state: list commands from the Gateway plugin registry, harden MCP loopback toolschemas, hide phantom agent-store rows from `sessions.list`, make task persistence failures explicit, and carry session UUIDs on interactive dispatch events.
- Chat/UI: preserve startup chat sends during history loading, unblock the initial Control UI chat send, stream chat deltas incrementally, skip markdown parsing while streaming, keep drafts local while typing, guard composer rerenders, cache chat transcript renders, record pending-send paint timing, show the Communication Notifications tab, honor Chromium executable overrides, and detect system Chromium for E2E. (#74715, #88952, #88960, #88998) Thanks @VladyslavLevchuk and @vincentkoc.
- Channels: stop schema-padded poll modifiers from turning normal `send` actions into invalid poll sends. (#89601) Thanks @codezz and @takhoffman.
- Channels: preserve long Feishu streaming replies, recover failed progress draft starts, send visible fallbacks when accepted Feishu turns produce no final reply, preserve external `sessions_send` routes, persist Discord thread bindings in SQLite, tolerate iMessage self-chat timestamp skew, preserve colon-prefixed slash commands in mention parsing, decode Nostr `npub` allowlists correctly, and suppress raw provider errors during channel delivery. (#87896, #88749, #88803, #88866) Thanks @MonkeyLeeT.
- Config/status/doctor: skip unresolved shell references in state-dir dotenv files, resolve gateway auth secrets during deep status audits, surface disabled Codex plugin routes in doctor lint, respect explicit PI runtime policy, report runtime tool-schema and gateway health credential errors, clear recovered embedded-run activity, migrate voice-call call logs through doctor, and keep post-upgrade JSON stable. (#88731, #88761, #88820, #88288, #89731) Thanks @brokemac79, @openperf, and @RomneyDa.
- Gateway/session state: list commands from the Gateway plugin registry, harden MCP loopback tool schemas, hide phantom agent-store rows from `sessions.list`, make task persistence failures explicit, support Tailscale Serve service names, guard Browser/Chrome pending attach aborts, and carry session UUIDs on interactive dispatch events. (#88305) Thanks @rohitjavvadi.
- Gateway/plugins: narrow plugin lookup memoization to the stable plugin/runtime inputs, avoiding repeated lookup work without mixing disabled or filtered plugin state.
- OpenAI/TTS: handle speed directives for OpenAI TTS voices. (#74089)
- CI/Crabbox: keep default runner capacity on the Azure credit-backed on-demand D4 lane with the Azure SSH port and a Git-independent full check job, so broad validation avoids low-priority spot quota stalls, hydrate port mismatches, non-Git hydrated workspaces, and stale AWS region hints.
- CI/Crabbox: route Crabbox wrapper and Testbox workflow edits to their regression tests so changed-test gates do not silently run zero specs.
- Gateway/sessions: allow shared-secret bearer callers to read and stream session history without an explicit scope header. (#81815) Thanks @medns.
- Agents/embedded runner: classify HTML auth provider responses as `auth_html` and return a re-authentication hint instead of the CDN-blocked copy that `upstream_html` returns. Cloudflare Access login pages, nginx basic-auth challenges, and gateway login walls all produce HTML auth bodies that were previously misdiagnosed as transient CDN blocks. (#79900) Thanks @martingarramon.
- TUI/streaming watchdog: dismiss the `This response is taking longer than expected` notice as soon as a chat event for the same run arrives, so the message no longer sits next to the recovered response when the run was only briefly silent. Refs #67052, #69081 (closed), prior attempt #69026. Thanks @jpruit20 and @romneyda.
- Agents/auth profiles: replace the bare `No available auth profile for <provider> (all in cooldown or unavailable)` TUI error with plain-language copy that explains what happened in user terms (sign-in expired, provider asking us to slow down, billing issue on the account, etc.) and suggests the matching `openclaw models auth login --provider <provider>` recovery command for sign-in and billing causes, while falling back to the underlying provider error for cases without a clear recovery path. Thanks @romneyda.
- Agents/Pi: tolerate OpenClaw-owned transcript writes while embedded prompts are released for model I/O, keeping long-running Feishu, Slack, Telegram, and cron turns from failing with false session-takeover errors. Fixes #84059. (#84250) Thanks @tianxiaochannel-oss88.
@@ -30,7 +30,8 @@ Supported channels include: WhatsApp, Telegram, Slack, Discord, Google Chat, Sig
New install? Start here: [Getting started](https://docs.openclaw.ai/start/getting-started)
Preferred setup: run `openclaw onboard` in your terminal.
OpenClaw Onboard guides you step by step through setting up the gateway, workspace, channels, and skills. It is the recommended CLI setup path and works on **macOS, Linux, and Windows (via WSL2; strongly recommended)**.
OpenClaw Onboard guides you step by step through setting up the gateway, workspace, channels, and skills. It is the recommended CLI setup path and works on **macOS, Linux, and Windows**.
Windows desktop users can start with the native [Windows Hub](https://docs.openclaw.ai/platforms/windows) companion app for setup, tray status, chat, node mode, and local MCP mode.
Works with npm, pnpm, or bun.
## Sponsors
@@ -164,7 +165,7 @@ Run `openclaw doctor` to surface risky/misconfigured DM policies.
- **[Voice Wake](https://docs.openclaw.ai/nodes/voicewake) + [Talk Mode](https://docs.openclaw.ai/nodes/talk)** — wake words on macOS/iOS and continuous voice on Android (ElevenLabs + system TTS fallback).
- **[Live Canvas](https://docs.openclaw.ai/platforms/mac/canvas)** — agent-driven visual workspace with [A2UI](https://docs.openclaw.ai/platforms/mac/canvas#canvas-a2ui).
<li>Agents and CLI-backed runtimes recover more cleanly from interrupted tool calls, stale session bindings, compaction handoffs, and media delivery retries. (#88129, #88136, #88141, #88162, #88182)</li>
<li>Channels and mobile delivery are steadier across Telegram, WhatsApp, iMessage, Slack, Discord, Microsoft Teams, Google Chat, Google Meet, and iOS realtime Talk. (#88096, #88105, #88183, #88231)</li>
<li>Provider and plugin requests now bound more timers, retries, OAuth/device-code lifetimes, media downloads, local service probes, and generated-content polling paths before they can hang a run.</li>
<li>Skills, session metadata, gateway runtime state, plugin metadata, memory watchers, and store writes do less repeated work on hot paths while keeping config, dispatch, and Linux file-watch behavior stable. (#89185, #89188, #85351) Thanks @RomneyDa and @NianJiuZst.</li>
<li>Skills and plugin loading now handle stale disabled snapshots and loader failures more clearly, so channel turns avoid disabled SecretRefs and operators get better recovery guidance. (#79072, #79173) Thanks @zeus1959.</li>
<li>Skill Workshop now has a fuller Control UI flow with proposal lists, today actions, revision handoff, searchable file previews, review states, locale coverage, and reusable session routing.</li>
<li>Chat and Control UI startup paths keep sends alive through history loading, stream deltas incrementally, skip markdown work while streaming, keep drafts local while typing, clear the composer after sends, trace first-output latency, prioritize first connect, and expose calmer composer controls. (#88772, #88825, #88998, #89030, #89106) Thanks @vincentkoc and @sallyom.</li>
<li>Provider coverage and model metadata now include MiniMax M3, account OAuth endpoints, Google/Vertex catalog fixes, OpenRouter SQLite model caching, Copilot Claude 1M capabilities, Foundry reasoning alignment, and OpenAI response replay guards. (#88480, #88512, #88851, #88860)</li>
<li>iMessage monitor state, inbound queues, and plugin install ledgers moved toward SQLite-backed state so restarts and local monitors recover with less duplicate filesystem scanning. (#88794, #88797)</li>
<li>Release, CI, Docker, E2E, plugin install, and diagnostics lanes now cap more logs, response bodies, readiness probes, artifact checks, status polling, child workflow waits, docker package cleanup, quiet test stalls, and rollback snapshots so failures report bounded proof instead of stalling. (#88966) Thanks @RomneyDa.</li>
</ul>
<h3>Changes</h3>
<ul>
<li>Docs: add a dedicated Skill Workshop guide covering governed skill creation, reviewable proposals, CLI, Gateway, agent tool behavior, approval policy, support files, and recovery, and refresh the ClawHub showcase cards. (#88734) Thanks @shakkernerd and @vyctorbrzezowski.</li>
<li>Skills: let the <code>skill_workshop</code> agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd.</li>
<li>Skills: let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd.</li>
<li>Skills: let pending proposals be revised in place with versioned, dated proposal frontmatter before approval. Thanks @shakkernerd.</li>
<li>Skills: add Skill Workshop with pending proposals, CLI/Gateway review actions, rollback metadata, and the <code>skill_workshop</code> agent tool. Thanks @shakkernerd.</li>
<li>Skill Workshop: add the Control UI navigation, styled dashboard, proposal today view, revision dialog, file preview modal, searchable preview files, reusable session handoff, and localized strings.</li>
<li>Plugins: externalize Tokenjuice as the official <code>@openclaw/tokenjuice</code> plugin with npm and ClawHub publish metadata.</li>
<li>Plugins: externalize the GitHub Copilot agent runtime as the official <code>@openclaw/copilot</code> plugin with npm and ClawHub publish metadata.</li>
<li>iOS: add hosted push relay defaults, realtime Talk playback, and a guarded WebSocket ping path for more reliable mobile sessions. (#88096, #88105, #88231)</li>
<li>iOS: support native iPad display layouts.</li>
<li>Workboard: add orchestration primitives and agent coordination tools for multi-agent planning and run tracking. (#87469)</li>
<li>Workboard: wire task-backed board runs and show task comments in the edit modal.</li>
<li>Code mode: add internal namespaces for scoped agent/global sessions and exact namespace tool dispatch. (#88043)</li>
<li>Code mode: add MCP API files and docs for code-mode integrations.</li>
<li>Control UI: add a Dreaming-tab agent selector and propagate the selected agent through Dreaming status, diary, and diary actions. (#78748) Thanks @stevenepalmer.</li>
<li>Control UI: add calmer chat composer controls, local draft typing state, and first-output latency instrumentation for active chat entry. (#88772, #88998) Thanks @vincentkoc.</li>
<li>Plugins: add a SecretRef provider integration manifest contract and extract shared LLM core packages for provider/plugin reuse. (#82326, #88117)</li>
<li>Plugins: persist the plugin install index in SQLite so installed package lookup survives reloads with less filesystem scanning. (#88794)</li>
<li>Providers: add MiniMax M3 model support. (#88860)</li>
<li>Doctor: add disk space health checks and stabilize post-upgrade JSON probes.</li>
<li>Channels: store inbound queues in SQLite and migrate iMessage monitor state to SQLite-backed tracking. (#88797)</li>
<li>Skills: add the core skills index and centralize skills runtime loading, status, filtering, and prompt formatting.</li>
</ul>
<h3>Fixes</h3>
<ul>
<li>Release/CI/E2E: fail early when Crabbox sparse-sync full checkouts do not have enough local disk, with guidance for moving the sync root.</li>
<li>Build: render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.</li>
<li>Plugins: stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.</li>
<li>Control UI: serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.</li>
<li>Scripts/UI: forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.</li>
<li>Gateway: return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.</li>
<li>Release/CI/E2E: keep temporary full-sync checkouts alive while slow Crabbox leases boot, so sparse worktree runs do not lose their sync source before file-list generation.</li>
<li>Release/CI/E2E: normalize inherited Linux <code>C.UTF-8</code> locale settings before raw AWS macOS Crabbox bootstrap commands, avoiding macOS locale warnings during package-manager hydration.</li>
<li>Release/CI/E2E: keep gateway watch regression checks from copying large static plugin assets inside the measured idle window.</li>
<li>Update: keep core updates nonblocking when a missing external plugin repair download stalls, while still blocking installed active plugin payload smoke failures.</li>
<li>Agents/providers: keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as <code>null</code> or arrays.</li>
<li>Release/CI/E2E: reset incremental log readers when watched log files rotate without shrinking, so same-size replacements do not hide new readiness or RPC lines.</li>
<li>Talk: preserve explicit <code>null</code> payloads on controller-created turn and output-audio lifecycle events.</li>
<li>Agents/TUI: keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.</li>
<li>Agents/media: keep async image, music, and video generation starts from ending the Codex turn, so mixed requests can continue with summaries or other work while media renders in the background.</li>
<li>Agents/Codex: keep public OpenAI API-key profiles from being treated as native Codex app-server auth while preserving persisted Codex OAuth sessions.</li>
<li>Agents/Codex: stream Codex app-server final-answer partials to live reply previews, preserve ACP metadata in SQLite, prefer real tool results over synthetic repair output, prevent aborted app-server turn handles from lingering, migrate legacy OpenAI Codex <code>lastGood</code> auth state, and preserve workspace/session metadata through ACP runtime refactors. (#88405, #88724, #88730) Thanks @vincentkoc.</li>
<li>Control UI: keep collapsed tool cards labeled with the tool name and action instead of generic output text. Thanks @shakkernerd.</li>
<li>Agents/Codex: surface Skill Workshop guidance in Codex app-server prompts when <code>skill_workshop</code> is available. Thanks @shakkernerd.</li>
<li>Skill Workshop: restore and localize the Control UI board/today view switcher so review workflows keep their intended layout toggle across locales. Thanks @shakkernerd.</li>
<li>Agents/auth: write auth profiles atomically, dispatch auth failures by type, add force re-login recovery, preserve workspaces during state-only uninstall, and compact before oversized turns so recovery paths avoid partial state. (#89181) Thanks @RomneyDa.</li>
<li>Skills: skip disabled skill env overrides from stale persisted snapshots so disabled skill <code>apiKey</code> SecretRefs cannot abort embedded or channel turns. (#79072, #79173) Thanks @zeus1959.</li>
<li>Skill Workshop: render the Control UI tab from filtered navigation state and keep filtered fallback routing stable.</li>
<li>CLI: avoid live catalog validation during <code>openclaw agents add</code>, so adding a secondary agent no longer depends on provider catalog availability. (#76284, #88314) Thanks @zhangguiping-xydt.</li>
<li>CLI: keep <code>plugins list --json</code> on the snapshot-only path so plugin sweeps avoid loading the full runtime status graph.</li>
<li>CLI/desktop: bridge WSL clipboard operations through the shell, recognize manual-update launchd jobs, and keep machine-readable startup output parseable during progress setup. (#88764, #88689) Thanks @alexzhu0.</li>
<li>Plugins: make PixVerse external-plugin ClawHub metadata explicit and keep it out of bundled dist builds.</li>
<li>Plugins: clarify plugin loader failure guidance so missing or incompatible plugin packages point operators at the right repair path.</li>
<li>Plugins: preserve npm plugin roots after blocked installs, skip plugin-local <code>openclaw</code> peer symlinks during rollback snapshots, relink those peers after restore, isolate cached tool runtime siblings, and isolate web-provider factory failures so one bad plugin does not poison sibling runtime paths. (#77237, #88807)</li>
<li>Cron: keep update delivery validation scoped, harden restart state, and retire MCP runtimes on isolated cron cleanup.</li>
<li>Memory: serialize QMD update/embed writes per store, reduce Linux watcher fan-out, retry transient FileProvider-backed reads, preserve phase signals on read errors, harden envelope metadata sanitization, reattach Linux native watchers when directories are recreated, and rewrite generated transcript paths on rollover so memory/search state survives concurrent gateway and CLI activity. (#66339, #85931, #89185, #89188, #85351) Thanks @openperf, @amittell, @RomneyDa, and @NianJiuZst.</li>
<li>Memory: keep vector-disabled FTS indexes from resolving embedding providers during sync and search.</li>
<li>Providers: bound generated media downloads from OpenAI, Runway, xAI, MiniMax, BytePlus, DashScope-compatible, FAL, OpenRouter, Google, Vydra, and Comfy providers.</li>
<li>Providers: resolve Google defaults to <code>google-generative-ai</code>, register Vertex static catalog rows, align Foundry reasoning metadata, skip DeepSeek V4 thinking params on Foundry fallback, use MiniMax account OAuth endpoints, preserve Copilot Claude 1M capabilities, suppress disabled Ollama reasoning output, forward Gemini stop sequences, strip Kimi-incompatible Anthropic cache markers, keep OpenAI stop-finished tool calls, and avoid replay ids when the Responses store is disabled. (#88480, #88512, #76612) Thanks @coder999999999, @BryanTegomoh, and @vliuyt.</li>
<li>Providers: cap GitHub Copilot OAuth request timeouts before creating abort signals.</li>
<li>Cron: retry recurring jobs after transient model rate limits before waiting for the next scheduled slot.</li>
<li>Agents/Codex: keep live session locks during cleanup, recover interrupted CLI tool transcripts, preserve Codex auth and compaction session identity, clear orphan tool state, cap app-server idle timers, and keep media completion delivery retryable. (#88129, #88136, #88141, #88162, #88182)</li>
<li>Chat/UI: show Gateway chat failures as visible assistant messages in the Control UI instead of only setting an invisible error state.</li>
<li>Channels: cap Telegram, Discord, WhatsApp, Signal, Feishu, Google Chat, Microsoft Teams, QQBot, Nostr, Zalo, Zalouser, and Nextcloud-style request/retry timers; preserve SMS approval reply routes; and retry WhatsApp QR login 408 timeouts. (#88183)</li>
<li>Security/config parsing: reject unsafe OAuth/token lifetimes, retry-after delays, inbound timestamps, response body sizes, command timeout config, sandbox observer token TTLs, and gateway WebSocket calls after close.</li>
<li>Providers/media: cap local service, model, usage, queue, generated media, TTS, music, workflow polling, and provider OAuth request timers across hosted and local providers.</li>
<li>Release/CI/E2E: bound release candidate reads, beta smoke REST calls, plugin npm verification commands, changelog restore, cross-OS process groups, kitchen-sink and bundled plugin readiness probes, secret-provider probes, Telegram credential timeouts, Control UI i18n and CLI startup metadata generation, Vitest routing, dependency guard admin approvals, child workflow failure detection, quiet Node test shard stalls, docker package cleanup, and mainline test flakes. (#88127, #88137, #88155, #88160, #88966) Thanks @RomneyDa.</li>
<li>Release/CI/E2E: keep Kitchen Sink live plugin MCP probes resolving source-checkout workspace packages and align the live gauntlet with current Kitchen Sink diagnostics.</li>
<li>Release/CI/E2E: run the secret-provider integration proof through the repo pnpm runner so native macOS and Windows validation use the hydrated package-manager shim.</li>
<li>Release/CI/E2E: run the Telegram desktop proof gateway through the repo pnpm runner so native macOS proof uses the hydrated package-manager shim.</li>
<li>Docs/CI: run Mintlify anchor checks through the repo pnpm runner so docs link validation works when pnpm is only available through the hydrated package-manager shim.</li>
<li>Agents: keep configured fallback model metadata typed so provider params, context-token caps, and media input limits do not break changed-gate typechecks.</li>
<li>Agents: accept hidden <code>sessions_send</code> body aliases before validation while keeping the model-facing <code>message</code> schema canonical. (#88229) Thanks @zhangguiping-xydt.</li>
<li>Chat/UI: preserve startup chat sends during history loading, unblock the initial Control UI chat send, stream chat deltas incrementally, skip markdown parsing while streaming, keep drafts local while typing, guard composer rerenders, honor Chromium executable overrides, and detect system Chromium for E2E. (#88998) Thanks @vincentkoc.</li>
<li>Channels: stop schema-padded poll modifiers from turning normal <code>send</code> actions into invalid poll sends. (#89601) Thanks @codezz.</li>
<li>Channels: preserve long Feishu streaming replies, send visible fallbacks when accepted Feishu turns produce no final reply, tolerate iMessage self-chat timestamp skew, preserve colon-prefixed slash commands in mention parsing, decode Nostr <code>npub</code> allowlists correctly, and suppress raw provider errors during channel delivery. (#87896)</li>
<li>Config/status/doctor: skip unresolved shell references in state-dir dotenv files, resolve gateway auth secrets during deep status audits, respect explicit PI runtime policy, report runtime tool-schema errors, and keep post-upgrade JSON stable. (#88288)</li>
<li>Gateway/session state: list commands from the Gateway plugin registry, harden MCP loopback tool schemas, hide phantom agent-store rows from <code>sessions.list</code>, make task persistence failures explicit, and carry session UUIDs on interactive dispatch events.</li>
<li>Gateway/plugins: narrow plugin lookup memoization to the stable plugin/runtime inputs, avoiding repeated lookup work without mixing disabled or filtered plugin state.</li>
<li>OpenAI/TTS: handle speed directives for OpenAI TTS voices. (#74089)</li>
<li>CI/Crabbox: keep default runner capacity on the Azure credit-backed on-demand D4 lane with the Azure SSH port and a Git-independent full check job, so broad validation avoids low-priority spot quota stalls, hydrate port mismatches, non-Git hydrated workspaces, and stale AWS region hints.</li>
<li>CI/Crabbox: route Crabbox wrapper and Testbox workflow edits to their regression tests so changed-test gates do not silently run zero specs.</li>
<li>CI/workflows: route workflow sanity helper edits to their guard tests and cover composite-action input interpolation checks.</li>
<li>CI/tooling: route CI scope, dependency, changelog, and docs helper edits to their owner tests instead of silently skipping changed-test coverage.</li>
<li>CI/tooling: route package, release, and install helper edits to their owner tests so changed-test gates cover publish and installer script changes.</li>
<li>CI/tooling: route shared script library edits through their owner tests so lock, process, safety, and scan helpers do not skip changed-test coverage.</li>
<li>CI/tooling: skip expensive import-graph scans once a changed diff already requires broad fallback, keeping local changed-test planning fast while still collecting explicit owner tests.</li>
<li>CI/tooling: route script edits through conventional owner tests when matching <code>test/scripts</code> or <code>src/scripts</code> coverage already exists.</li>
<li>CI/tooling: honor option terminators in the memory FD repro script so follow-on arguments are not reparsed.</li>
<li>Release/CI/E2E: assert plugin lifecycle runtime inspect output instead of only capturing it.</li>
<li>Release/CI/E2E: make gateway-network prove the advertised health RPC and retry early WebSocket closes without burning full open timeouts.</li>
<li>Release/CI/E2E: honor option terminators across release, Parallels smoke, plugin gauntlet, and extension-memory scripts.</li>
<li>Release/CI/E2E: fail plugin gateway gauntlet QA chunks when the requested suite summary is missing or invalid.</li>
<li>Performance: prebuild QA runtime probes with generated plugin assets but without CLI startup metadata.</li>
<li>Performance: skip declaration bundling for runtime-only CLI startup and gateway watch build profiles.</li>
<li>Faster Gateway and replies: startup avoids repeated plugin, channel, session, usage-cost, warning, scheduled-service, and filesystem scans; visible replies separate user-facing sends from slower follow-up work; Gateway runtime/session caches churn less under load.</li>
<li>Transcripts are core: transcript-backed meeting summaries, source-provider chunks, cleaned user turns, media provenance, Codex mirrors, WebChat replies, and CLI/TUI replay now use one more reliable transcript path.</li>
<li>More channels are production-ready: Telegram keeps typing/progress context and forum topics, iMessage handles attachment roots, remote media staging, and duplicate local Messages sources, WhatsApp restores group/media behavior, Discord improves voice playback and model picking, and Signal/iMessage/WhatsApp get reaction approvals.</li>
<li>Better voice and Talk: realtime Talk runs can be inspected, steered, cancelled, or followed up from Web UI and Discord voice; wake-name handling is more tolerant without letting ambient speech trigger agents.</li>
<li>Safer content boundaries: Browser snapshot reads honor SSRF policy, system-event text cannot spoof nested prompt markers, fetched file text is wrapped as external content, ClickClack inbound sender allowlists run before agent dispatch, stale device tokens are rejected, and serialized tool-call text is scrubbed from replies.</li>
<li>Providers, Codex, and local models are steadier: named auth profiles, OpenAI sampling params, Codex app-server resume/timeout/usage-limit recovery, dynamic tool-schema guards, xAI usage-limit surfacing, Ollama top-p normalization, and local approval resolution reduce provider-specific dead ends.</li>
<li>Transcripts: add core transcript capture and source-provider support for transcript-backed meeting summaries, including the renamed Transcripts docs, CLI surface, source-provider chunks, and cleaned user-turn persistence.</li>
<li>Auth: add named model login profiles and supported credential migration for Hermes, OpenCode, and Codex auth profiles, with explicit opt-out and non-interactive controls. (#85667) Thanks @fuller-stack-dev.</li>
<li>Diagnostics: trace gateway secret preparation, classify skill/tool usage, surface model stream progress, add OpenTelemetry LLM content spans, and expose alertable telemetry for blocked tools, failover, stale sessions, liveness, oversized payloads, and webhook ingress. (#83019, #80370, #86191)</li>
<li>Channels: add Signal reaction approvals, iMessage thumb approval reactions, and WhatsApp thumb approval reaction support so mobile approval flows work without textual <code>/approve</code> commands. (#85894, #85952, #85477)</li>
<li>Agents/API: forward OpenAI sampling params through the Gateway and expose estimated context-budget status for active agent runs. (#84094)</li>
<li>TUI/status: queue prompts submitted while an agent is busy and show explicit fast-mode state plus richer systemd Gateway hygiene in status output. (#86722, #87115, #86976)</li>
<li>Exec approvals: hide durable approval actions that are unavailable for the current prompt and keep approval runtime tokens local-only so stale prompts cannot offer misleading controls. (#86270, #86359)</li>
<li>Plugin SDK: add reaction approval helpers and keep diagnostic event root exports discoverable across function-name and alias-bound module graphs. (#86735, #87084)</li>
<li>Android/iOS: add the Android pair-new-gateway action and improve mobile Talk mode surfaces, including iOS realtime Talk mode and Android offline voice/gateway recovery. (#86798, #86355) Thanks @ngutman.</li>
<li>Performance: cache plugin metadata snapshots, package realpaths, stable gateway metadata, model cost indexes, channel resolution, usage-cost indexes, and session/auth hot-path facts so common Gateway and reply paths do less rediscovery. (#84649, #85843, #86517, #86678)</li>
<li>Voice: expose shared realtime turn-context tracking through the realtime voice SDK and reuse it for Discord speaker attribution and wake-name context recovery.</li>
<li>Voice: reuse shared realtime output activity tracking in Google Meet command and node audio bridges, including recent-output checks for local barge-in detection.</li>
<li>Voice: expose shared realtime output activity tracking through the realtime voice SDK and reuse it for Discord playback activity and barge-in decisions.</li>
<li>Voice: expose shared realtime consult question matching, speakable-result extraction, and alias-aware forced-consult coordination through the realtime voice SDK, then reuse it in Gateway Talk, Voice Call, and Discord voice paths.</li>
<li>Voice: share activation-name matching and consult-transcript screening through the realtime voice SDK so Discord, browser voice, and meeting surfaces can reuse one implementation.</li>
<li>Cron: default <code>cron.maxConcurrentRuns</code> to 8 so scheduled automations and their isolated agent turns can make progress in parallel without explicit configuration.</li>
<li>QA-Lab: add <code>qa coverage --match <query></code> so focused proof selection can discover matching scenarios from existing metadata before running live or remote lanes.</li>
<li>Discord/model picker: surface an alpha-bucket select (e.g. <code>A–G (12) · H–N (18) · O–Z (5)</code>) when the provider list or a provider's model list exceeds 25 items, so configs with <code>provider/*</code> wildcards stay one click from the right page instead of paginating through prev/next; falls back to numeric chunks when every item shares the same first letter.</li>
<li>Control UI: add an ephemeral Activity tab for sanitized live tool activity summaries without persisting raw telemetry. Fixes #12831. Thanks @BunsDev.</li>
<li>Build: include <code>ui:build</code> in the <code>full</code> and <code>ciArtifacts</code> profiles of <code>scripts/build-all.mjs</code> so <code>pnpm build</code> always rebuilds <code>dist/control-ui</code> after <code>tsdown</code> cleans <code>dist</code>, removing the second-command requirement and the missing-asset failure mode for source/runtime installs and CI artifact uploads. (#85206)</li>
<li>iOS: improve Talk mode with direct realtime voice sessions, compact toolbar status, and responsive voice waveform feedback. (#86355) Thanks @ngutman.</li>
<li>Media: replace the Sharp image backend with Rastermill for metadata, resizing, EXIF orientation, and PNG alpha-preserving optimization so OpenClaw no longer installs Sharp or the WhatsApp Jimp fallback for image processing. (#86437)</li>
<li>Codex: update the bundled Codex CLI to 0.134.0 and keep native compaction disabled for budget-triggered app-server turns so OpenClaw owns the recovery boundary. (#86772)</li>
</ul>
<h3>Fixes</h3>
<ul>
<li>Memory/security: reject prompt-like text submitted through the explicit <code>memory_store</code> tool before embedding or storage, matching the existing auto-capture prompt-injection filter. (#87142)</li>
<li>Gateway/security: enable the default auth rate limiter for remote non-browser and HTTP gateway auth failures when <code>gateway.auth.rateLimit</code> is unset, while preserving the loopback exemption. (#87148)</li>
<li>Security/content boundaries: validate Browser snapshot tab URLs against SSRF policy before ChromeMCP or direct CDP reads, sanitize queued system-event text so untrusted plugin/channel labels cannot spoof nested prompt markers, wrap fetched file text and metadata as external content, apply ClickClack <code>allowFrom</code> sender allowlists before agent dispatch, reject RPCs from invalidated device-token clients during rotation, require staged sandbox media refs, and scrub serialized tool-call text from replies. (#78526, #87094, #87062, #83741, #70707, #86924) Thanks @zsxsoft, @ttzero25, and @mmaps.</li>
<li>Transcripts/user turns: persist CLI, WebChat, media, follow-up, hook, and Codex-mirror user turns to the admitted session target; keep cleaned transcript text, inline image routing, provenance metadata, replay hooks, and fallback paths idempotent when runtimes fail or restart.</li>
<li>TUI/status/onboarding/UI: queue busy TUI prompts instead of dropping them, preserve the configured default model during onboarding, show failed tool results as errors, show config-open failures in Control UI, keep status JSON plugin scans healthy, preserve xAI usage-limit errors locally, and expose explicit fast-mode/systemd state. (#86722, #87000, #85786, #87108, #87001, #86614, #87115, #86976)</li>
<li>Plugin commands/SDK: preserve plugin LLM command auth, bind native plugin command dispatch to the host agent's LLM auth, keep <code>onDiagnosticEvent</code> exports discoverable through <code>Function.name</code>, stabilize diagnostic event root aliases, correlate pathless read diagnostics, suppress transient runner failures in channel command paths, and repair local approval resolution. (#85936, #87084, #86977, #87069, #86771)</li>
<li>Codex/providers: keep WebChat delivery hints out of user prompts, avoid false queued-terminal idle timeouts, share the native hook relay registry, quarantine unsupported dynamic tool schemas, preserve Claude resumed-session system prompts, normalize greedy Ollama <code>top_p</code>, preserve per-agent thinking defaults for ingress runs, and avoid native compaction takeover on budget-triggered Codex turns. (#87096, #73950, #87049, #86689, #86772)</li>
<li>Gateway/perf/release: reuse startup-warning metadata and prepared auth stores, avoid cloning live-switch and lifecycle session caches on read paths, defer warning and scheduled-service fallback imports, trim Gateway session/startup/runtime CPU churn, skip duplicate turn session touches, stop chat timeout fallback cascades, drop stale subagent announce history, bound benchmark/watch/kitchen-sink teardown waits, bound macOS/package/onboarding/plugin smoke commands, bound install finalization probes, resolve Parallels npm-update commands from guest <code>PATH</code>, and bootstrap raw AWS macOS Node/pnpm commands through <code>/usr/bin/env</code>. (#86997)</li>
<li>Reply/source delivery: keep TUI, Control UI, media, TTS, transcript, and Codex source-reply finals live without duplicate terminal events or stale replay artifacts.</li>
<li>Agents/replay: repair legacy tool results before replay, preserve <code>sessions_spawn</code> transcript payloads, restore current guard checks, stage sandboxed workspace media, and keep duplicate transcripts tool display metadata from reappearing. (#82203, #86934, #87025) Thanks @martingarramon, @vincentkoc, and @joshavant.</li>
<li>Agents/sessions: handle active-fallback failures in <code>sessions_send</code> so fallback routing reports the real failure and does not leave callers with an ambiguous dropped send. (#86638)</li>
<li>Agents/hooks/subagents: enforce default hook agent allowlists, recover failed subagent lifecycle completions, and keep node task lifecycle cleanup from closing the Gateway listener. (#86101)</li>
<li>Codex: project newer OpenClaw chat history into resumed app-server threads and keep Codex turn timeouts inside the Codex runtime boundary so timeouts do not poison shared app-server clients or fall through to unrelated provider fallback. (#86677, #86476) Thanks @TurboTheTurtle and @pashpashpash.</li>
<li>Config/doctor/update: narrow profiled tool-section doctor repair, keep runtime-injected legacy web-search provider config out of user-authored config validation, and keep prerelease tags excluded from stable updater resolution. (#87030, #86818, #86559) Thanks @joshavant, @luoyanglang, and @stevenepalmer.</li>
<li>CLI/Windows: add a Windows-only stack-size respawn for stack-heavy startup paths, default CLI logs to local timestamps, and validate timeout/banner TTY state more strictly. (#87031, #85387) Thanks @giodl73-repo and @vincentkoc.</li>
<li>Locking/security: require owner identity proof before stale plugin lock removal, memoize session lock owner arguments, and avoid writing default exec approval stores unless policy state actually changed. (#86814, #86964) Thanks @Alix-007 and @vincentkoc.</li>
<li>Install/release: bound Docker package build, inventory, pack, and tarball preparation with process-group timeouts; pin shrinkwrap patch drift to the pnpm lock; harden macOS restart and dSYM packaging; and run release Docker/live timeout wrappers in the foreground so child processes cannot wedge gates.</li>
<li>Telegram/network: treat <code>ENETDOWN</code> as a transient pre-connect network failure so Telegram sends, gateway unhandled-rejection handling, and cron network retries follow the same recovery path as sibling network outages. (#86762) Thanks @TurboTheTurtle.</li>
<li>Telegram: preserve inbound text entities, overlapping DM replies, account topic cache sidecars, outbound reply context, targeted bot-command mentions, durable group retry targets, forum topic names, and native progress callbacks. (#83873, #85361, #85555, #85656, #85709, #86299, #86553) Thanks @SebTardif, @luoyanglang, and @neeravmakwana.</li>
<li>iMessage: read image attachments from local Messages attachment roots, dedupe duplicate local Messages-source accounts, seed direct DM history, fix image/group media attachment commands, advance catchup cursors after live handling, and keep slash-command acknowledgements in the source conversation. (#82642, #85475, #86569, #86705, #86706, #86770) Thanks @homer-byte, @TurboTheTurtle, @swang430, and @OmarShahine.</li>
<li>WhatsApp/QQ/Twitch/IRC/Slack: restore WhatsApp ack identity and group-drop warnings, make QQ Bot media respect <code>OPENCLAW_HOME</code>, serialize Twitch auth disconnects, store IRC channel routes canonically, and keep Slack downloaded files out of reply media. (#83833, #85309, #85777, #85794, #85906, #86318, #86697) Thanks @sliverp, @neeravmakwana, and @Kailigithub.</li>
<li>Discord/voice: improve voice playback and wake replies, bucket large model picker menus, merge media captions into one message, route metadata through configured proxies, restore numeric channel sends, suppress self-reply echoes, and tighten wake matching without breaking fuzzy wake phrases. (#80227, #86238, #86487, #86571, #86595, #86601)</li>
<li>Agents/runtime: enforce session lock max-hold reclaim, release embedded-attempt locks on all exits, treat aborted subagent runs as terminal, avoid runtime model hydration on hot paths, disclose scoped session list counts, derive overflow budgets from provider errors, and keep fallback errors scoped to the active model candidate. (#70473, #85764, #86014, #86134, #86427, #86944) Thanks @openperf, @fuller-stack-dev, @zhangguiping-xydt, and @ferminquant.</li>
<li>Config/update/doctor: retry config recovery after failed backup restore, skip shell env fallback on Windows, exclude prerelease tags from the stable git channel, support deep config edits, warn instead of aborting on unreadable cron stores, prune stale bundled plugin paths, and avoid duplicate restart prompts when the Gateway is already healthy. (#85739, #85787, #86060, #86260, #86384, #86533) Thanks @liaoyl830.</li>
<li>Install/release: support Alpine CLI installs and runtime floors, prefer trusted startup argv runtime fallback roots, reject stale CLI node runtimes, avoid npm <code>min-release-age</code> installer failures, bound npm/package/Docker install phases, restore config parent ownership in Docker, seed Docker lockfile package tarballs before prune, make release/plugin prerelease checks fail closed instead of hanging or false-greening, and use host-visible Crabbox local work roots for Docker-backed proof. (#85491)</li>
<li>Windows daemon: keep Scheduled Task gateway launches running on battery power and avoid workgroup-machine prompts for a domain user during task installation. (#59299)</li>
<li>Security: avoid printing Gateway tokens in Docker, validate plugin model-pattern regexes safely, escape transcript metadata field names, harden session allowlist glob matching, audit Claude permission overrides under YOLO, and require explicit allow for ACP auto approvals. (#85849, #85934, #86046, #86557)</li>
<li>Media/images: replace Sharp with Rastermill, keep EXIF normalization best-effort, normalize HEIC/HEIF before image descriptions, route Codex image API keys through OpenAI, preserve image compression metadata, and auto-scale live tool result caps. (#85776, #86037, #86437, #86857, #86923)</li>
<li>Memory: prevent semantic vector indexes from silently degrading when embeddings are unavailable, stop doctor OOMs on large session stores, preserve sidecar hooks/artifacts, write fallback dream diaries, use CJK-aware dreaming dedupe, and avoid per-file watcher FD fan-out. (#80613, #82928, #85060, #85704, #85967, #86701) Thanks @brokemac79, @openperf, and @yaaboo-gif.</li>
<li>Agents/sessions: include visibility metadata on restricted <code>sessions_list</code> results so scoped counts are clearly reported without widening access or exposing hidden-session counts. (#86944) Thanks @ferminquant.</li>
<li>Gateway/DNS: validate wide-area discovery domains before deriving zone paths or writing zone files, so invalid <code>discovery.wideArea.domain</code> and <code>dns setup --domain</code> values fail with a DNS-name diagnostic instead of falling through to unrelated configuration errors. Thanks @mmaps.</li>
<li>Agents/BTW: route fallback side-question streams through the embedded stream resolver so Anthropic-compatible MiniMax requests use the same capped transport as normal chat. (#86312) Thanks @neeravmakwana.</li>
<li>Telegram: treat <code>/command@TargetBot</code> bot-command entities as explicit mentions for the addressed bot so <code>requireMention</code> groups no longer drop targeted commands or captions. Fixes #84462. (#86553) Thanks @luoyanglang.</li>
<li>CI: bound Docker/Bash E2E tarball npm installs with <code>OPENCLAW_E2E_NPM_INSTALL_TIMEOUT</code> so package, onboarding, plugin, and upgrade lanes fail instead of hanging on a stuck npm install.</li>
<li>CI: fail Parallels npm-update smoke jobs after the guest command timeout and cleanup backstop instead of only logging a timeout line.</li>
<li>CI: bound kitchen-sink RPC HTTP probes so stalled gateway readiness or response bodies fail and retry instead of wedging the walker.</li>
<li>CI: keep <code>OPENCLAW_TESTBOX=1 pnpm check:changed</code> delegating to Blacksmith Testbox through Crabbox without forwarding local Testbox or worker env into the remote command.</li>
<li>CI: send KILL after the TERM grace period for manual checkout fetch timeouts so stuck Testbox and workflow checkout retries cannot hang behind a wedged <code>git fetch</code>.</li>
<li>CI: send KILL after the TERM grace period for Bun global install smoke command timeouts so trapped <code>openclaw</code> child processes cannot wedge the scheduled install smoke.</li>
<li>iMessage: thread current channel/account inbound attachment roots into the image tool so iMessage-saved attachments under <code>~/Library/Messages/Attachments</code> (including the wildcard <code>/Users/*/Library/Messages/Attachments</code> root) are read through the existing inbound path policy instead of being rejected as <code>path-not-allowed</code>. Literal <code>localRoots</code> stays workspace-scoped. Fixes #30170. (#86569)</li>
<li>QQ Bot: respect <code>OPENCLAW_HOME</code> for outbound media path resolution so <code><qqmedia></code> sends no longer silently fail when <code>HOME</code> and <code>OPENCLAW_HOME</code> differ (Docker / multi-user hosts). Persisted QQ Bot data (sessions, known users, refs) stays anchored on the OS home for upgrade compatibility. Fixes #83562. Thanks @sliverp.</li>
<li>Update: report the primary malformed <code>openclaw.extensions</code> payload error without adding a duplicate missing-main diagnostic. (#86596) Thanks @ferminquant.</li>
<li>Gateway: dampen repeated unauthenticated device-required probes per URL while preserving explicit-auth and paired recovery paths. (#86575) Thanks @ferminquant.</li>
<li>IRC: store inbound channel routes with the canonical <code>channel:#name</code> target and join transient channel sends before writing. (#85906) Thanks @Kailigithub.</li>
<li>Usage: surface unknown all-zero model pricing as missing cost entries instead of a confident <code>$0</code> total. (#85882) Thanks @MichaelZelbel.</li>
<li>Agents/Codex: honor yolo app-server approval policy only for the full <code>never</code> plus <code>danger-full-access</code> case. (#85909) Thanks @earlvanze.</li>
<li>Gateway/Gmail: clear Gmail watcher renewal intervals on re-entry so hot reloads do not leak lifecycle timers. (#82947) Thanks @SebTardif.</li>
<li>Logging: exit cleanly on broken stdout/stderr pipes without masking existing failure exit codes. (#80059) Thanks @pavelzak.</li>
<li>Gateway/security: escape transcript metadata field names while extracting oversized session line prefixes. (#85934) Thanks @SebTardif.</li>
<li>Plugins/security: validate manifest model pattern regexes with the safe-regex compiler so unsafe patterns are ignored before matching. (#86046) Thanks @SebTardif.</li>
<li>Discord: route gateway metadata REST lookups through the configured Discord proxy so proxied accounts do not fall back to direct <code>discord.com</code> connections before opening the WebSocket. Fixes #80227. Thanks @Clivilwalker.</li>
<li>Agents/media: hydrate current-turn image attachments from filename-derived MIME types so active vision can see generated or forwarded images whose source omitted an image content type. (#84812) Thanks @marchpure.</li>
<li>Agents/fs: point workspace-only scratch-path guidance at in-workspace temp directories while keeping host-root writes rejected by the tool guard. (#86501) Thanks @tianxiaochannel-oss88.</li>
<li>Agents/media: keep async cron media completions scoped to their run session while preserving direct delivery for stale generated-media success and failure notifications. (#86529) Thanks @ai-hpc.</li>
<li>Gateway: emit plugin <code>session_end</code>/<code>session_start</code> hooks when <code>agent.send</code> rotates or replaces a session id, keeping hook lifecycle state aligned with <code>sessions.changed</code> notifications. Fixes #83507. (#85875) Thanks @brokemac79.</li>
<li>OpenShell/SSH: reject malformed generated exec commands before sandbox/session setup so unresolved workflow placeholders fail fast instead of reaching the remote shell. Fixes #72373. Thanks @brokemac79.</li>
<li>Google: stop normalizing <code>gemini-3.1-flash-lite</code> to the retired preview endpoint and update Flash Lite alias guidance to the GA model id. Fixes #86151. (#86240) Thanks @SebTardif.</li>
<li>Installer: make Alpine apk installs cover Git, verify the Node runtime floor, try <code>nodejs-current</code>, and report Alpine version guidance when repositories only provide older Node packages.</li>
<li>Agents/status: prefer the active Claude CLI OAuth auth label over an unused Anthropic env API-key label for equivalent runtime aliases. Fixes #80184. (#86570) Thanks @brokemac79.</li>
<li>Agents/media: send direct fallback for generated media still missing after an active requester wake fails. (#85489) Thanks @fuller-stack-dev.</li>
<li>Agents: derive overflow compaction budgets from provider-reported and synthetic over-budget token counts so confirmed context overflows compact before retrying. (#70473) Thanks @fuller-stack-dev.</li>
<li>Agents/Codex: recover Codex context-window prompt errors through overflow compaction and surface reset guidance when recovery is exhausted. (#85542) Thanks @fuller-stack-dev.</li>
<li>Agents/Codex: allow Codex app-server runs to bootstrap from <code>CODEX_API_KEY</code> or <code>OPENAI_API_KEY</code> when no Codex auth profile is configured.</li>
<li>Agents/Codex: keep selected Codex runtime routing on OpenAI-Codex while preserving direct OpenAI API-key compaction fallback. (#86408) Thanks @funmerlin and @VACInc.</li>
<li>Agent transcript: include OpenClaw agent session logs when finding local transcript candidates.</li>
<li>Crabbox: bootstrap raw AWS macOS shell commands wrapped in absolute <code>time</code> paths so RSS probes can run Node and pnpm on fresh macOS runners.</li>
<li>Crabbox: bootstrap raw AWS macOS shell commands even when setup statements precede Node or pnpm usage.</li>
<li>TUI/local: skip unnecessary secret resolution, gateway model catalog loading, bootstrap, and skill scans in explicit local-model runs so startup reaches the model request faster.</li>
<li>Sessions/doctor: load large session stores without clone amplification during read-only doctor checks and reclaim stale <code>sessions.json.*.tmp</code> sidecars. Fixes #56827. Thanks @openperf.</li>
<li>Tests: clean successful plugin gateway gauntlet isolated temp roots while keeping an explicit preservation switch for failed/debug runs.</li>
<li>Plugins/perf: reuse derived plugin metadata snapshots for the lifetime of the process so reply-time skill setup no longer rescans plugin metadata on every turn.</li>
<li>Discord/OpenAI voice: keep wake-name master consults using the current speaker context after ignored ambient transcripts and shorten the default capture silence grace.</li>
<li>Doctor: skip redundant Gateway restart prompts when a recent supervisor restart leaves the Gateway healthy. Fixes #86518. (#86533) Thanks @liaoyl830.</li>
<li>Cron: restore suspended cron lanes to the configured/default concurrency instead of falling back to one after quota or circuit-breaker auto-resume.</li>
<li>Gateway: keep session-only Control UI tool-start mirrors flowing during diagnostic queue pressure instead of silently dropping non-terminal tool updates.</li>
<li>Discord: merge streamed text captions into following media block replies so captions and attachments send as one message. (#86487) Thanks @neeravmakwana.</li>
<li>Gateway: avoid sending duplicate tool-event frames to Control UI connections that are subscribed by both run and session.</li>
<li>Discord/OpenAI voice: accept longer leading wake-name mistranscripts such as "Open Club" for OpenClaw.</li>
<li>Agents/OpenAI-compatible: stop ModelStudio-compatible chat requests before sending system/tool-only payloads that have no usable user or assistant turn. (#86177) Thanks @TurboTheTurtle.</li>
<li>Gateway/plugins: reuse plugin package realpath checks while building installed plugin indexes so startup avoids repeated filesystem resolution work.</li>
<li>Kilo Gateway: send string <code>stop</code> sequences as arrays so Kilo accepts OpenAI-compatible chat completions. (#86461) Thanks @SebTardif.</li>
<li>Discord/OpenAI voice: accept leading fuzzy wake-name transcripts such as "Monty" or "Moti" for a Molty agent while keeping ambient speech gated.</li>
<li>Media understanding: convert HEIC and HEIF images to JPEG before image description providers run so iPhone photos work in direct and configured image-description flows. (#86037)</li>
<li>Agents: release embedded-attempt session locks from outer teardown so post-prompt exceptions cannot wedge later requests behind <code>SessionWriteLockTimeoutError</code>. Fixes #86014. Thanks @openperf.</li>
<li>Discord/OpenAI voice: rotate Realtime sessions at provider max duration without logging the expected session-expiry event as an error.</li>
<li>Sessions: skip metadata-only entries during QMD-slugified session lookup so one incomplete row does not block transcript hit resolution. (#86327) Thanks @abnershang.</li>
<li>Agents/media: derive bundled plugin local-media trust from plugin tool metadata instead of importing the full plugin registry on subscription paths. (#84409) Thanks @samzong.</li>
<li>Image tool: keep config-backed custom-provider API keys usable for auto-discovered vision models, including deferred image-tool execution without env keys or auth profiles. (#85733)</li>
<li>Memory/local embeddings: run local GGUF embeddings in an isolated worker sidecar and degrade to configured fallback or keyword search on worker failure so native embedding crashes do not take down the Gateway. (#85348) Thanks @osolmaz.</li>
<li>Gateway: clear the runtime config snapshot before <code>SIGUSR1</code> in-process restarts so config changes survive the next gateway loop. (#86388) Thanks @XuZehan-iCenter.</li>
<li>Models: show OAuth delegation markers as configured <code>models.json</code> auth while keeping runtime route usability checks strict. (#86378) Thanks @rohitjavvadi.</li>
<li>Cron: seed active scheduled and manual cron task rows with a progress summary so status surfaces do not look blank while jobs run. (#86313) Thanks @ferminquant.</li>
<li>Cron: preserve unsupported persisted cron payload rows during routine store writes while keeping those rows non-runnable. Fixes #84922. (#86415) Thanks @IWhatsskill.</li>
<li>Updater: exclude prerelease git tags from stable channel resolution so source updates do not check out newer alpha/rc/preview/canary tags. (#86260) Thanks @stevenepalmer.</li>
<li>Security/Audit: flag webhook <code>hooks.token</code> reuse of active Gateway password auth in <code>openclaw security audit</code> while keeping password-mode startup compatibility. (#84338) Thanks @coygeek.</li>
<li>QQBot: derive the outbound reply watchdog from configured agent and provider timeouts so slow local model replies are not cut off at five minutes. Fixes #85267. (#85271) Thanks @SymbolStar.</li>
<li>Agents/heartbeat: stop heartbeat turns after the first valid <code>heartbeat_respond</code> so repeated response loops do not burn tokens. (#86357) Thanks @udaymanish6.</li>
<li>Tasks: keep retained lost tasks out of default status health counts, explain their cleanup window during maintenance, and prune lost task records after 24 hours instead of the general 7-day terminal retention.</li>
<li>Memory-core: keep REM dreaming focused on live light-staged memories and mark staged entries as considered so old recall history no longer dominates fresh candidates. (#86302) Thanks @SebTardif.</li>
<li>Memory: abort sync instead of downgrading an existing semantic vector index to FTS-only when the configured embedding provider is temporarily unavailable. (#85704) Thanks @yaaboo-gif.</li>
<li>Telegram: propagate forum topic names through the account-scoped topic cache for native command context and topic create/edit actions. (#86299) Thanks @SebTardif.</li>
<li>Slack: keep downloaded read-only files out of reply media so Slack file reads do not echo files back to the conversation. (#86318) Thanks @neeravmakwana.</li>
<li>Cron: accept leading-plus relative durations such as <code>+5m</code> for one-shot <code>--at</code> schedules. (#86341) Thanks @mushuiyu886.</li>
<li>Agents/media: preserve async-started media tool metadata so background generation starts no longer surface generic incomplete-turn warnings while replay stays unsafe. (#85933) Thanks @fuller-stack-dev.</li>
<li>Docker E2E: dedupe scheduler lane resources so npm/service package lanes are not over-counted and serialized unnecessarily.</li>
<li>QA/diagnostics: add a collector-backed OpenTelemetry smoke lane, make the OTLP payload leak check scenario-aware, and keep source QA builds from failing on optional dependency imports resolved through pnpm's temp module path.</li>
<li>Crabbox: bootstrap Git metadata for sparse remote changed gates so raw synced workspaces can run <code>pnpm check:changed</code> from the intended diff.</li>
<li>xAI/LM Studio: avoid buffering ordinary bracketed or <code>final</code> prose until stream completion while watching for plain-text tool-call fallbacks.</li>
<li>Doctor: warn and continue when the cron job store exists but cannot be read so later health checks still run. Fixes #86102. (#86384) Thanks @1052326311.</li>
<li>Discord: suppress a bot's previous reply body and referenced media from prompt context when a user replies to that bot message, while keeping reply metadata for routing. (#86238) Thanks @fuller-stack-dev.</li>
<li>Discord: restore bare numeric channel IDs for outbound message-tool sends while keeping explicit DM targets unambiguous. (#86571) Thanks @joshavant.</li>
<li>Docker E2E: avoid rebuilding the Control UI twice while preparing the shared OpenClaw package tarball for package-backed scenario runs.</li>
<li>Tests: avoid rebuilding the Control UI twice during the installer Docker smoke now that <code>pnpm build</code> includes <code>ui:build</code>.</li>
<li>Tests: give QA config mutation RPCs enough native Windows budget to finish gateway config writes and restart settle after hot scenario runs.</li>
<li>Tests: keep the gateway restart-inflight QA scenario focused on restart recovery on native Windows by allowing expected embedded prompt handoff errors and using the Windows-safe timeout budget.</li>
<li>QA-Lab: make the synthetic OpenAI provider honor generic <code>reply exactly:</code> directives after required kickoff reads so restart-recovery scenarios do not fall through to generic repo-summary prose.</li>
<li>Gateway: abort active <code>agent</code> RPC runs during forced restart shutdown so stale in-process turns cannot keep writing a session after the Gateway lifecycle restarts.</li>
<li>Crabbox: sync clean sparse worktrees through a temporary full checkout even when reusing an existing lease so tracked build-time files are not omitted.</li>
<li>Build: route <code>scripts/ui.js</code> through the shared pnpm runner and keep Control UI chunking helpers in sparse-included source so native Windows Corepack builds can produce <code>dist/control-ui</code>.</li>
<li>Tests: give the memory fallback QA scenario enough turn budget to exercise native Windows gateway runs instead of failing on the client timeout while the mock agent is still dispatching.</li>
<li>Tests: collect QA gateway CPU/RSS metrics on native Windows and give the channel baseline enough turn budget to report slow gateway runs instead of timing out before proof.</li>
<li>Install/update: bypass npm <code>min-release-age</code> policies with <code>--min-release-age=0</code> instead of <code>--before</code> so hosted installers keep working on npm versions that reject the combined config. (#84749) Thanks @TeodoroRodrigo.</li>
<li>Diagnostics: reclaim wedged session lanes when stale active-run bookkeeping blocks queued work despite no forward progress. Fixes #85639. Thanks @openperf.</li>
<li>WebChat: keep message-tool replies visible in the chat while still summarizing internal tool results for the model. Fixes #86347. Thanks @shakkernerd.</li>
<li>Gateway/perf: fail startup benchmark samples when the Gateway process exits before benchmark teardown, including signal deaths after readiness probes.</li>
<li>Gateway/perf: fail restart benchmark samples when the Gateway exits before benchmark teardown, including clean exits and signal deaths after successful restart probes.</li>
<li>Agents/tests: keep model catalog visibility on static selection helpers so catalog visibility checks avoid the broad model-selection barrel import.</li>
<li>Agents/commitments: serialize commitment store load-modify-save writes so concurrent heartbeat and CLI updates no longer lose dismissal, sent, or attempt state. (#81153) Thanks @ai-hpc.</li>
<li>xAI/LM Studio: promote plain-text tool-call fallbacks into structured tool calls and strip leaked internal tool syntax before user-facing delivery. (#86222) Thanks @fuller-stack-dev.</li>
<li>CLI: suppress benign self-update version-skew warnings during package post-update finalization.</li>
<li>Gateway/perf: tighten restart and startup benchmark failure handling so long profiling runs, failed probes, and fresh Linux runners no longer produce false passing or <code>n/a</code> results.</li>
<li>Checks: keep intentional Knip unused-file findings optional so full CI and sparse proof workspaces stay aligned.</li>
<li>Docker: restore writable <code>~/.config</code> in runtime images. Fixes #85968. Thanks @hkoessler and @Bartok9.</li>
<li>Plugin SDK: keep legacy root diagnostic subscriptions connected when built plugin SDK aliases resolve diagnostic helpers through a separate module graph.</li>
<li>Diagnostics: export alertable OTel and Prometheus signals for blocked tools, model failover, stale sessions, liveness warnings, oversized payloads, and webhook ingress while fixing shared OTLP endpoints with query strings.</li>
<li>Tests: normalize macOS canonical temp paths in exec allowlists, fs-safe trash assertions, installed plugin matching, Telegram topic-name stores, and built ACPX MCP server expectations so native macOS proof runners cover the intended behavior.</li>
<li>Codex/app-server: preserve message-tool-only source reply delivery mode on active runs so sub-agent completion wakeups can steer the active Codex turn instead of being rejected. (#86287) Thanks @ferminquant.</li>
<li>Tests: sample the Windows kitchen-sink RPC gateway directly and serialize RSS probes so native runs keep the memory guard active.</li>
<li>Tests: normalize bundled plugin lifecycle probe paths and state-root lookup so native Windows release sweeps accept valid packaged plugin installs.</li>
<li>Agents/Claude CLI: route live native Bash permission requests through OpenClaw exec policy so Claude turns no longer stall on <code>control_request</code>, and document that OpenClaw exec policy is authoritative. Fixes #80819. (#86330, from #81971) Thanks @guthirry and @sallyom.</li>
<li>Security audit: warn when YOLO OpenClaw exec policy overrides a restrictive raw Claude <code>--permission-mode</code> for managed live sessions. (#86557) Thanks @sallyom.</li>
<li>Config: keep benign legacy metadata write anomalies out of default doctor and config command output while preserving explicit anomaly logging for diagnostics.</li>
<li>Codex: log when implicit app-server <code>never</code> approvals are promoted for OpenClaw tool policy, including whether the trigger was a <code>before_tool_call</code> hook or trusted tool policy.</li>
<li>Codex harness: make subscription usage-limit errors without reset times explain that OpenClaw cannot determine the reset and point users to wait until Codex is available, use another Codex account, or switch to another configured model/provider. Thanks @amknight.</li>
<li>Google Vertex: support production ADC modes such as Workload Identity Federation, service-account credentials, and metadata-server ADC for the native Vertex transport. (#83971) Thanks @damianFelixPago.</li>
<li>Telegram: route normal <code>[telegram][diag]</code> polling diagnostics through <code>runtime.log</code> while keeping non-diag warnings and persistence failures on <code>runtime.error</code>, so healthy polling startup no longer looks like an error. Fixes #82957. (#82958) Thanks @galiniliev.</li>
<li>Providers/Ollama: strip inline Kimi cloud reasoning prefixes from streamed and final visible replies while keeping ordinary Kimi answers append-only. (#86286) Thanks @jason-allen-oneal.</li>
</ul>
<ul>
<li>Gateway: require Talk secret authority before setup-code handoff can include Talk secrets. (#85690) Thanks @ngutman.</li>
<li>Agents: keep fallback error reporting scoped to the active model candidate so stale prior-provider quota/auth text is not reported for later fallback attempts. (#86134) Thanks @zhangguiping-xydt.</li>
<li>iMessage: dedupe watcher startup when <code>channels.imessage.accounts</code> lists both <code>default</code> and a named account that point at the same local Messages source, so the gateway no longer spawns two <code>imsg rpc</code> processes or doubles inbound replies; the dedupe is scoped to watcher startup, leaving duplicate accounts addressable for outbound sends, status, and capability listings, and <code>openclaw doctor</code> flags the redundant account with a rebinding hint. Fixes #65141. (#86705) Thanks @swang430.</li>
</ul>
<p><a href="https://github.com/openclaw/openclaw/blob/main/CHANGELOG.md">View full changelog</a></p>
@@ -218,6 +218,7 @@ Current OpenClaw Android implication:
- Google Play build excludes SMS send/search, Call Log search, and recent-photo access unless the product is intentionally positioned and approved under the relevant policy exception.
- The repo now ships this split as Android product flavors:
-`play`: removes `READ_SMS`, `SEND_SMS`, `READ_CALL_LOG`, `READ_MEDIA_IMAGES`, `READ_MEDIA_VISUAL_USER_SELECTED`, and `READ_EXTERNAL_STORAGE`; hides SMS, Call Log, and Photos surfaces in onboarding, settings, and advertised node capabilities.
- Installed-app listing is user controlled. `device.apps` is advertised only after the user enables **Settings > Phone Capabilities > Installed Apps**. The command defaults to launcher-visible apps and does not require `QUERY_ALL_PACKAGES`.
-`thirdParty`: keeps the full permission set and the existing SMS / Call Log / Photos functionality.
Policy links:
@@ -252,9 +253,9 @@ Pre-req checklist:
4) Open the app **Screen** tab and keep it active during the run (canvas/A2UI commands require the canvas WebView attached there).
5) Grant runtime permissions for capabilities you expect to pass (camera/mic/location/notification listener/location, etc.).
6) No interactive system dialogs should be pending before test start.
7) Canvas host is enabled and reachable from the device (do not run gateway with `OPENCLAW_SKIP_CANVAS_HOST=1`; startup logs should include `canvas host mounted at .../__openclaw__/`).
7) Canvas host is enabled and reachable from the device for remote Canvas checks (do not run gateway with `OPENCLAW_SKIP_CANVAS_HOST=1`; startup logs should include `canvas host mounted at .../__openclaw__/`).
8) Local operator test client pairing is approved. If first run fails with `pairing required`, preview the latest pending request, approve the printed request ID, then rerun:
9) For A2UI checks, keep the app on **Screen** tab; the node now auto-refreshes canvas capability once on first A2UI reachability failure (TTL-safe retry).
9) For A2UI checks, keep the app on **Screen** tab; the node uses its bundled app-owned A2UI page for message application.
```bash
openclaw devices list
@@ -286,8 +287,8 @@ Common failure quick-fixes:
-`pairing required` before tests start:
- list pending requests (`openclaw devices list`), then approve with the exact ID (`openclaw devices approve <requestId>`) and rerun.
-`A2UI host not reachable` / `A2UI_HOST_NOT_CONFIGURED`:
-ensure the Canvas plugin host is running and reachable, keep the app on the **Screen** tab. The app refreshes the Canvas plugin surface URL once before failing; if it still fails, reconnect app and rerun.
-`A2UI host not reachable` / `A2UI_HOST_UNAVAILABLE`:
-keep the app foregrounded on the **Screen** tab and rerun. A2UI commands use the bundled app-owned A2UI page; the Gateway Canvas host is still needed for remote Canvas checks, but not for A2UI message application.
@@ -635,6 +773,13 @@ private fun PhoneCapabilitiesScreen(
listOf(
SettingsToggleRow("Camera","Allow camera tools when requested.",Icons.Default.CameraAlt,cameraEnabled,::setCameraAccess),
SettingsToggleRow("Precise Location","Share precise location while location is enabled.",Icons.Default.LocationOn,locationPreciseEnabled,::setPreciseLocation),
SettingsToggleRow(
"Installed Apps",
if(installedAppsSharingEnabled)"OpenClaw can list launcher-visible apps."else"App list stays on this phone.",
Icons.Default.Storage,
installedAppsSharingEnabled,
viewModel::setInstalledAppsSharingEnabled,
),
SettingsToggleRow("Keep Awake","Keep the node available during active work.",Icons.Default.Bolt,preventSleep,viewModel::setPreventSleep),
Text(text=if(pendingRunCount>0)"OpenClaw is working"else"Ready when you are",style=ClawTheme.type.title.copy(fontSize=20.sp,lineHeight=24.sp),color=ClawTheme.colors.text)
-`progress` keeps one editable status draft for tool progress, clears it at completion, and sends the final answer as a normal message
-`streaming.preview.toolProgress` controls whether tool/progress updates reuse the same edited preview message (default: `true` when preview streaming is active)
-`streaming.preview.commandText` controls command/exec detail inside those tool-progress lines: `raw` (default, preserves released behavior) or `status` (tool label only)
-`streaming.progress.commentary` (default: `false`) opts into assistant commentary/preamble text in the temporary progress draft
- legacy `channels.telegram.streamMode` and boolean `streaming` values are detected; run `openclaw doctor --fix` to migrate them to `channels.telegram.streaming.mode`
Tool-progress preview updates are the short status lines shown while tools run, for example command execution, file reads, planning updates, patch summaries, or Codex preamble/commentary text in Codex app-server mode. Telegram keeps these enabled by default to match released OpenClaw behavior from `v2026.4.22` and later.
If Codex app-server plugin inventory is unavailable during planning, migration
falls back to cached bundle advisory items instead of failing the whole
migration.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.