Commit Graph

2015 Commits

Author SHA1 Message Date
Aditya Singh
facfa410a7 fix(tool-display): satisfy format/lint and address review feedback
- extract web_search/web_fetch detail resolvers into common module\n- fix node -c classification so file path remains positional\n- remove dead git subcommands set\n- keep exec summary refinements (heredoc/node check/git -C/preamble strip)\n- make tests cover node -c syntax-check path\n- run format:check, tsgo, lint, and focused e2e tests
2026-02-16 23:50:08 +01:00
Aditya Singh
24f213e7ed feat(tool-display): add intent-first details and exec summaries
- add human-readable read/write/edit/attach details with path alias support\n- add explicit web_search/web_fetch phrasing (quoted query, mode/limit)\n- make detail text title-first by returning detail-only in formatters\n- add deterministic exec summarizer (wrappers, pipelines, heredoc, git/node/python heuristics, preamble stripping)\n- extend e2e coverage for file/web/exec cases
2026-02-16 23:50:08 +01:00
Shaun Mason
feed570984 fix: syncs all credential types to agent auth.json
Previously, the synchronization of credentials to the agent's  file was limited to  OAuth profiles. This prevented other providers and credential types from being correctly registered for agent use.

This update expands the synchronization to include ,  (mappedto ), and  credentials for all configured providers.

It ensures the agent's  accurately reflects available credentials, enabling proper authentication and model discovery.

The synchronization now:
- Converts all supported credential types.
- Skips profiles with empty keys.
- Preserves unrelated entries in the target .
- Only writes to disk when actual changes are detected.
2026-02-16 23:49:54 +01:00
Daniel Sauer
12ce358da5 fix(failover): recognize 'abort' stop reason as timeout for model fallback
When streaming providers (GLM, OpenRouter, etc.) return 'stop reason: abort'
due to stream interruption, OpenClaw's failover mechanism did not recognize
this as a timeout condition. This prevented fallback models from being
triggered, leaving users with failed requests instead of graceful failover.

Changes:
- Add abort patterns to ERROR_PATTERNS.timeout in pi-embedded-helpers/errors.ts
- Extend TIMEOUT_HINT_RE regex to include abort patterns in failover-error.ts

Fixes #18453

Co-authored-by: James <james@openclaw.ai>
2026-02-16 23:49:51 +01:00
Peter Steinberger
61859377a5 refactor(test): dedupe pi-tools loop detection test setup 2026-02-16 22:39:42 +00:00
Dakshay Mehta
8947d2dea5 Agents: format process poll backoff files 2026-02-16 23:32:12 +01:00
Dakshay Mehta
23f5cc80a4 Agents: wire command poll backoff into process poll 2026-02-16 23:32:12 +01:00
Vignesh Natarajan
5a26d1c622 Agent: guard reminder promises behind cron scheduling 2026-02-16 14:07:16 -08:00
Shadow
c593709d25 Discord: add per-button component allowlist 2026-02-16 15:15:00 -06:00
Shadow
05a83b9e97 Discord: add reusable component option 2026-02-16 14:22:49 -06:00
Sk Akram
e5eb5b3e43 feat: add stuck loop detection and exponential backoff infrastructure for agent polling (#17118)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: eebabf679b
Co-authored-by: akramcodez <179671552+akramcodez@users.noreply.github.com>
Co-authored-by: gumadeiras <5599352+gumadeiras@users.noreply.github.com>
Reviewed-by: @gumadeiras
2026-02-16 15:16:35 -05:00
Shadow
72e228e14b Heartbeat: allow suppressing tool warnings (#18497)
* Heartbeat: allow suppressing tool warnings

* Changelog: note heartbeat tool-warning suppression
2026-02-16 13:29:24 -06:00
nabbilkhan
250896cf6e fix: correct contradictory test name (Greptile review)
The test verifies that cooldownUntil IS cleared when it equals exactly
`now` (>= comparison), but the test name said "does not clear". Fixed
the name to match the actual assertion behavior.
2026-02-16 12:53:45 -06:00
nabbilkhan
03cadc4b7a fix(auth): auto-expire stale auth profile cooldowns and reset error count
When an auth profile hits a rate limit, `errorCount` is incremented and
`cooldownUntil` is set with exponential backoff. After the cooldown
expires, the time-based check correctly returns false — but `errorCount`
persists. The next transient failure immediately escalates to a much
longer cooldown because the backoff formula uses the stale count:

  60s × 5^(errorCount-1), max 1h

This creates a positive feedback loop where profiles appear permanently
stuck after rate limits, requiring manual JSON editing to recover.

Add `clearExpiredCooldowns()` which sweeps all profiles on every call to
`resolveAuthProfileOrder()` and clears expired `cooldownUntil` /
`disabledUntil` values along with resetting `errorCount` and
`failureCounts` — giving the profile a fair retry window (circuit-breaker
half-open → closed transition).

Key design decisions:
- `cooldownUntil` and `disabledUntil` handled independently (a profile
  can have both; only the expired one is cleared)
- `errorCount` reset only when ALL unusable windows have expired
- `lastFailureAt` preserved for the existing failureWindowMs decay logic
- In-memory mutation; disk persistence happens lazily on the next store
  write, matching the existing save pattern

Fixes #3604
Related: #13623, #15851, #11972, #8434
2026-02-16 12:53:45 -06:00
Vignesh
b0a01fe482 Agents/Tools: preflight exec script files for shell var injection (#18457)
* fix(agents): don't force store=true for codex responses

* test: stabilize respawn + subagent usage assertions

* Agents/Tools: preflight exec to detect shell variable injection in scripts

* Changelog: fix merge marker formatting
2026-02-16 10:34:29 -08:00
Peter Steinberger
b0035a1e49 refactor(test): table-drive web tool defaults checks 2026-02-16 18:31:27 +00:00
Peter Steinberger
9ff473fa05 refactor(test): share sandbox config test helpers 2026-02-16 18:25:04 +00:00
Peter Steinberger
74c49c943d refactor(test): share web fetch e2e setup helpers 2026-02-16 18:09:45 +00:00
Ayaan Zaidi
16327f21da feat(telegram): support inline button styles (#18241)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: 239cb3552e
Co-authored-by: obviyus <22031114+obviyus@users.noreply.github.com>
Co-authored-by: obviyus <22031114+obviyus@users.noreply.github.com>
Reviewed-by: @obviyus
2026-02-16 22:48:47 +05:30
Gustavo Madeira Santana
8a67016646 Agents: raise bootstrap total cap and warn on /context truncation (#18229)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: f6620526df
Co-authored-by: gumadeiras <5599352+gumadeiras@users.noreply.github.com>
Co-authored-by: gumadeiras <5599352+gumadeiras@users.noreply.github.com>
Reviewed-by: @gumadeiras
2026-02-16 12:04:53 -05:00
Mariano
44ef045614 fix(canvas): port remaining iOS branch stability fixes (#18228)
* fix(canvas): prevent snapshot disconnects on proxied gateways

(cherry picked from commit 2a3c9f746a65f3301c0cfe58ebe6596fed06230f)

* fix(canvas): accept url alias for present and navigate

(cherry picked from commit 674ee86a0b776cbb738add1920a4031246125312)

---------

Co-authored-by: Nimrod Gutman <nimrod.g@singular.net>
2026-02-16 16:42:28 +00:00
Ayaan Zaidi
c8a536e30a fix(agents): scope message tool schema by channel (#18215)
Co-authored-by: Shadow <shadow@openclaw.ai>
2026-02-16 10:34:18 -06:00
Peter Steinberger
df6d0ee92b refactor(core): dedupe tool policy and IPv4 matcher logic 2026-02-16 16:14:54 +00:00
Peter Steinberger
1d37389490 test: annotate harness mocks to avoid TS2742 in CI 2026-02-16 15:19:11 +00:00
Peter Steinberger
a1ca9291f3 test(agents): fix reasoning replay input assertion helper 2026-02-16 14:59:31 +00:00
Peter Steinberger
f717a13039 refactor(agent): dedupe harness and command workflows 2026-02-16 14:59:30 +00:00
Hongwei Ma
dddb1bc942 fix(telegram): fix streaming with extended thinking models overwriting previous messages/ also happens to Execution error (#17973)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: 34b52eead8
Co-authored-by: Marvae <11957602+Marvae@users.noreply.github.com>
Co-authored-by: obviyus <22031114+obviyus@users.noreply.github.com>
Reviewed-by: @obviyus
2026-02-16 18:54:34 +05:30
sebslight
553d17f8af refactor(agents): use silent token constant in prompts 2026-02-16 08:20:24 -05:00
sebslight
3a277e394e test(agents): add cooldown expiry helper regressions 2026-02-16 08:10:52 -05:00
sebslight
d224776ffb refactor(agents): extract cooldown probe decision helper 2026-02-16 08:10:52 -05:00
zerone0x
c2a0cf0c28 fix(tts): update tool description to prevent duplicate audio delivery (#18046)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: 70c096abaa
Co-authored-by: zerone0x <39543393+zerone0x@users.noreply.github.com>
Co-authored-by: sebslight <19554889+sebslight@users.noreply.github.com>
Reviewed-by: @sebslight
2026-02-16 08:09:02 -05:00
Ítalo Souza
39bb1b3322 fix: auto-recover primary model after rate-limit cooldown expires (#17478) (#18045)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: f7a7865727
Co-authored-by: PlayerGhost <28265945+PlayerGhost@users.noreply.github.com>
Co-authored-by: sebslight <19554889+sebslight@users.noreply.github.com>
Reviewed-by: @sebslight
2026-02-16 08:03:35 -05:00
Advait Paliwal
bc67af6ad8 cron: separate webhook POST delivery from announce (#17901)
* cron: split webhook delivery from announce mode

* cron: validate webhook delivery target

* cron: remove legacy webhook fallback config

* fix: finalize cron webhook delivery prep (#17901) (thanks @advaitpaliwal)

---------

Co-authored-by: Tyler Yust <TYTYYUST@YAHOO.COM>
2026-02-16 02:36:00 -08:00
Peter Steinberger
15f8c57797 test: speed up subagent announce e2e and drop duplicate defer case 2026-02-16 09:10:11 +00:00
the sun gif man
68ea063958 🤖 fix: preserve openai reasoning replay ids (#17792)
What:
- disable tool-call id sanitization for OpenAI/OpenAI Codex transcript policy
- gate id sanitization in image sanitizer to full mode only
- keep orphan reasoning downgrade scoped to OpenAI model-switch replay path
- update transcript policy, session-history, sanitizer, and reasoning replay tests
- document OpenAI model-switch orphan-reasoning cleanup behavior in transcript hygiene reference

Why:
- OpenAI Responses replay depends on canonical call_id|fc_id pairings for reasoning followers
- strict id rewriting in OpenAI path breaks follower matching and triggers rs_* orphan 400s
- limiting scope avoids behavior expansion while fixing the identified regression

Tests:
- pnpm vitest run src/agents/transcript-policy.test.ts src/agents/pi-embedded-runner.sanitize-session-history.test.ts src/agents/openai-responses.reasoning-replay.test.ts
- pnpm vitest run --config vitest.e2e.config.ts src/agents/transcript-policy.e2e.test.ts src/agents/pi-embedded-runner.sanitize-session-history.e2e.test.ts src/agents/pi-embedded-helpers.sanitize-session-messages-images.removes-empty-assistant-text-blocks-but-preserves.e2e.test.ts src/agents/pi-embedded-helpers.sanitizeuserfacingtext.e2e.test.ts
- pnpm lint
- pnpm format:check
- pnpm check:docs
- pnpm test (fails in current macOS bash 3.2 env at test/git-hooks-pre-commit.integration.test.ts: mapfile not found)
2026-02-15 22:45:01 -08:00
Peter Steinberger
0895bb6de6 test: move skills-install fallback suite to e2e lane 2026-02-16 06:11:01 +00:00
Peter Steinberger
5212d1c79e test: make sandbox symlink-escape assertion platform-aware 2026-02-16 06:26:08 +01:00
Peter Steinberger
e8a50e41a5 test: reuse fixtures in skills install fallback suite 2026-02-16 04:03:24 +00:00
Peter Steinberger
25dc4293bf test: speed up isolated-agent and pty test suites 2026-02-16 03:58:43 +00:00
Peter Steinberger
ae6fe67550 test: align e2e coverage with supervisor session flow 2026-02-16 03:41:58 +00:00
Peter Steinberger
b5a63e18f9 test(sandbox): add array-order hash and recreate regression tests 2026-02-16 04:36:24 +01:00
Peter Steinberger
d1fca442b4 refactor(sandbox): centralize sha256 helpers 2026-02-16 04:33:47 +01:00
Sebastian
3c467baa2d test(skills): add status-to-install apt fallback coverage 2026-02-15 22:32:51 -05:00
Sebastian
c8e110e2e3 refactor(skills): extract installer strategy helpers 2026-02-15 22:32:51 -05:00
Peter Steinberger
41ded303b4 fix(sandbox): preserve array order in config hashing 2026-02-16 04:32:03 +01:00
Vignesh Natarajan
cbf58d2e1c fix(memory): harden context window cache collisions 2026-02-15 19:31:52 -08:00
Peter Steinberger
559c8d9930 fix: replace deprecated SHA-1 in sandbox config hash 2026-02-16 04:30:59 +01:00
Peter Steinberger
aef1d55300 fix(cron): normalize skill-filter snapshots and split isolated run helpers 2026-02-16 04:27:12 +01:00
McRolly NWANGWU
d19b746928 feat(skills): add cross-platform install fallback for non-brew environments (#17687)
Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: 3ed4850838
Co-authored-by: mcrolly <60803337+mcrolly@users.noreply.github.com>
Co-authored-by: sebslight <19554889+sebslight@users.noreply.github.com>
Reviewed-by: @sebslight
2026-02-15 22:25:26 -05:00
Shadow
a61c2dc4bd Discord: add component v2 UI tool support (#17419) 2026-02-15 21:19:25 -06:00