* fix(agents): apply stale-run liveness to aborted subagent orphan recovery
Skip stale unended subagent runs during orphan recovery and registry
restore, even when they carry abortedLastRun. Previously, restart-aborted
runs were exempt from the stale-unended age check, allowing hours-old
aborted child sessions to be resurrected after long downtime.
Fixes#90766
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* fix(agents): finalize stale aborted runs instead of only skipping them
Previously the stale-run check in recoverOrphanedSubagentSessions only
incremented the skipped counter. Stale active runs were left unended
because scheduleOrphanRecovery only retries failedRuns, not skipped runs.
Now stale runs are finalized via finalizeInterruptedSubagentRun so they
don't remain orphaned in the registry.
Ref: #90766 review feedback
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* fix(agents): await stale-run finalization in orphan recovery
Await finalizeInterruptedSubagentRun for stale aborted runs and report
failedRuns when finalization does not update the registry, so the
scheduler retry path can recover from finalization failures.
Co-authored-by: Cursor <cursoragent@cursor.com>
* test(agents): faithful restart-path proof for stale orphan recovery
Drive the real recoverOrphanedSubagentSessions against the real subagent
registry, the real isStaleUnendedSubagentRun policy, and a real on-disk
session store, mocking only the outbound gateway transport and transcript
reader. Proves finalizeInterruptedSubagentRun actually ends the stale
aborted run in the registry (endedAt set, outcome error) instead of
resuming it, while a fresh aborted run still resumes.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* chore: amend author email
* fix(agents): scope orphan finalization to run generation
* fix(agents): preserve stale restart recovery ownership
* test(agents): isolate restart recovery scheduling
* fix(agents): make stale restart finalization durable
* fix(agents): keep stale retries generation-scoped
* test(agents): await restart recovery scheduling
* fix(agents): verify interrupted finalization
* fix(agents): defer interrupted finalization during restart
* fix(agents): preserve lifecycle type narrowing
* style(agents): use nonmutating recovery ordering
* chore: keep release note in PR body
---------
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Pick-cat <266665499+Pick-cat@users.noreply.github.com>
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(anthropic): delegate adaptive CLI effort
Strip static --effort args for adaptive runs so Claude Code resolves effort from its environment, settings, and model default. Preserve configured effort for off or absent thinking and replace it only for concrete OpenClaw levels.
Fixes#103245
Co-authored-by: Dan Korotin <korotin.daniil@gmail.com>
* fix(anthropic): make effort dispatch exhaustive
---------
Co-authored-by: Dan Korotin <korotin.daniil@gmail.com>
* fix(exec): auto-approve recognized read-only boolean flags on default safe bins
Default safe bins (cut, head, tail, tr, uniq, wc) auto-approve stdin-only
text-filter invocations, but the short-option validator only had an accept
path for value-consuming flags: a cluster of pure boolean short flags fell
through to a terminal reject, so common read-only forms like 'wc -l',
'tr -d', 'uniq -c' and 'sort -n' were force-routed to manual approval even
though the bins are stdin-only and the dangerous flags are already denied.
Add an allowedBooleanFlags allowlist to the safe-bin profile model and
populate it for the default bins with their read-only boolean flags. The
short-cluster validator now accepts recognized boolean flags and the long
validator reuses its existing boolean-flag accept branch (previously
unreachable). Unrecognized short flags (e.g. 'tr -S') stay fail-closed, and
denied flags are still rejected first.
* fix(exec): keep safe-bin allowedBooleanFlags off the config-facing fixture type
The boolean-flag allowlist for default safe bins leaked onto SafeBinProfileFixture, the type used for tools.exec.safeBinProfiles, while the strict zod schema and the config normalizer never accepted or preserved the key. Move allowedBooleanFlags to an internal BuiltinSafeBinProfileFixture used only for the curated built-in profiles; custom config profiles keep the allowedValueFlags/deniedFlags model.
Adds tests proving built-in profiles still honor the boolean allowlist and that a custom profile cannot widen it.
* fix(exec): keep tail follow approval-gated
* test(exec): stabilize safe-bin trust fixtures
* test(exec): isolate safe-bin argv fixtures
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(doctor): gate cross-state-dir legacy imports behind explicit doctor opt-in
A gateway or CLI command started with OPENCLAW_STATE_DIR pointing at a
non-default directory used to import legacy files FROM the default
~/.openclaw state dir (exec-approvals.json, plugin-binding-approvals.json)
and archive the originals with a .migrated suffix. Any isolated, test, or
staging run on a host with production state silently captured and archived
the production files.
Cross-state-dir imports now run only with the new crossStateDirImports
opt-in: openclaw doctor --fix (or an interactive doctor confirm) performs
the import; the implicit CLI/gateway preflight leaves the default dir
untouched and emits a notice pointing at doctor instead.
* test(doctor): include notices in legacy-state detector mocks
* test: shorten loopback token fixtures below secret-scanner thresholds
* feat: carry gateway client capabilities through the CLI loopback backend
CLI-backed model runs now transport the originating client's declared
capabilities along the existing per-field loopback contract: RunCliAgentParams
gains clientCaps, prepare emits OPENCLAW_MCP_CLIENT_CAPS, the loopback header
template forwards x-openclaw-client-caps, the request context parses and
normalizes it (grant-authenticated callers keep ignoring spoofable headers),
and the loopback tool cache keys on a stable caps serialization so capless and
capped requests never share tool lists. Capability-gated tools such as
show_widget now work on CLI backends; the CLI session binding hash includes
caps, consistent with messageProvider. Cron and command-attempt runs stay
capless (fail closed).
Fixes#102577
Adds client-capability-gated tool availability: gateway clients declare
capabilities at connect (new inline-widgets cap), chat.send stamps them into
the run context, and every tool assembly path (embedded runner, queued
followups, Codex app-server harness, plugin-only construction plans) drops
tools whose requiredClientCaps the originating client did not declare. The
Canvas plugin ships the first such tool, show_widget: agents pass SVG or an
HTML fragment plus a title; the plugin hosts it as a bounded, retention-scoped
Canvas document and returns the existing canvas preview handle, which web chat
renders as a sandboxed iframe fitted to the widget's reported content height.
Widget frames never get allow-same-origin (per-preview sandbox ceiling,
including the sidebar path) and the Canvas host serves widget documents with a
CSP sandbox header so direct navigation runs in an opaque origin. Verified
live end-to-end on a Testbox with gpt-5.5 (screenshots on the PR). CLI-backed
model backends do not carry client caps yet and stay fail-closed (#102577).
Closes#101790
* fix(gateway): support native Windows exec approvals
Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
* chore: defer changelog entry to release
* test: use tracked approvals temp directories
---------
Co-authored-by: Vincent Koc <vincentkoc@ieee.org>
* feat(browser): direct extension→gateway relay path for remote Chrome (#53599)
Let the OpenClaw Chrome extension pair directly to a remote gateway over
wss:// with no OpenClaw node host on the browser machine — the managed-hosting
path from #53599 (extension is the only thing installed on the laptop).
- Gateway route /browser/extension registered by the browser plugin with
auth:"plugin" + no nodeCapability, so the gateway does not pre-enforce token
auth (browser WebSockets cannot send an Authorization header). The upgrade
handler self-validates the host-local relay secret from ?token=, origin-checks
chrome-extension://, resolves the extension profile, then attaches the socket
to the same ExtensionRelayBridge the loopback relay uses. All CDP synthesis,
tab-group scoping, and the in-process Playwright /cdp client are unchanged.
- `openclaw browser extension pair --gateway-url wss://host` prints a
wss://host/browser/extension#<secret> string; the path ends in /extension so
the extension's existing pairing parser accepts it with zero extension code
changes.
- relay-server: extract attachExtensionWebSocket + export requestToken /
isAllowedExtensionOrigin / EXTENSION_RELAY_MAX_PAYLOAD_BYTES so loopback and
gateway paths share one bind + one frame cap.
- runtime-lifecycle: dispose the shared gateway WebSocketServer on shutdown.
- docs: three remote topologies (same host / direct-to-gateway / via node host).
Coverage: 6 unit tests for the handler's path/503/403/404/401/attach branches.
The full extension→bridge→CDP→Chrome loop over /browser/extension was live-proven
with a real Chrome + the built extension. The real gateway upgrade→handleUpgrade
dispatch for an auth:"plugin" unprotected route is verified against core
(server-http.ts, plugins-http.ts, route-auth.ts).
* fix(browser): harden remote extension pairing
Tracks per-skill usage from the skill.used diagnostic event (trusted-only
delivery, file-scoped identity), sweeps workshop-created skills daily from
gateway maintenance (active -> stale 30d -> archived 90d, pinned bypass,
restore-only unarchive, files never touched), filters archived skills from
snapshots fail-open, reports workspace-scoped overlap candidates, and adds
openclaw skills curator CLI, additive gateway methods, and a warn-only
doctor finding. Zero new config keys; SQLite/Kysely state only.
* feat(browser): restore driver "extension" via loopback Chrome extension relay
Reintroduces browser profile driver "extension" (removed in 2026.3.22) as a
loopback relay that drives the user's signed-in Chrome through an MV3 extension
instead of the remote-debugging port. This avoids Chrome's blocking "Allow
remote debugging?" prompt, which cannot be clicked when the operator drives
OpenClaw from a phone. Automated tabs live in an "OpenClaw" tab group (the
consent boundary), mirroring the Codex/Claude-in-Chrome model.
- relay bridge synthesizes the CDP browser target surface for Playwright
connectOverCDP and forwards session-scoped commands to chrome.debugger
- relay server binds loopback only; both sides authenticate with a token
derived (HMAC-SHA256) from gateway auth, so the raw credential never reaches
Chrome; extension origin + loopback Host checks guard the upgrade
- built-in "chrome" profile; distinct relay ports per extension profile;
relay reconciles on auth rotation / cdpPort change and prunes removed profiles
- doctor + status surface the extension transport; doctor keeps repairing the
retired relay endpoint URL on legacy "extension" profiles
Refs #53599
* feat(browser): bundle the OpenClaw MV3 Chrome extension
Thin MV3 extension (chrome-extension/): a WebSocket client to the loopback
relay plus chrome.debugger forwarding and OpenClaw tab-group management. All
CDP target synthesis lives server-side in the relay bridge, so the extension
stays a dumb transport (the removed 2026.3 extension put that logic in a
1000-line untestable service worker). Popup handles pairing and per-tab share
toggle; `openclaw browser extension path|pair` load and pair it. A build copy
hook stages it into dist so the load path is stable.
Refs #53599
* docs(browser): document the Chrome extension profile
Adds docs/tools/chrome-extension for the restored extension driver (install,
pair, tab-group consent model, security posture) and wires it into the browser
docs profile section and nav.
Refs #53599
* feat(browser): make the extension relay work on remote browser nodes
Derives the relay auth token from a host-local secret in the credentials dir
(created on first use) instead of gateway auth. Each machine that runs a
browser — the gateway host and every browser node host — owns its own token, so
the extension pairs with whichever machine hosts its Chrome and no gateway
credential travels to a node. The node host already runs the shared browser
control bootstrap, so this is all that was missing for cross-machine control.
Also removes the "relay needs gateway auth before it can start" failure mode:
startup and `openclaw browser extension pair` ensure the secret exists.
Refs #53599
* fix(browser): harden relay secret creation and satisfy CI lint/typecheck
- Make the host-local relay secret creation atomic (O_CREAT|O_EXCL + adopt the
winner on EEXIST) so the gateway service and `extension pair` CLI cannot mint
divergent tokens on a fresh host (would 401 until restart); credentials dir
created mode 0700. (adversarial review finding)
- Resolve type-aware oxlint findings across the relay + extension: unknown catch
vars, addEventListener over ws.on* in the MV3 worker, void async listeners,
drop useless returns/spreads, Object.assign over map-spread, safe ws frame
decode (Buffer[]/ArrayBuffer), toSorted.
- Add extensionRelayDefaultPort/extensionRelayPorts to remaining test config
literals; type the extension relay-core module (.d.ts, excluded from dist);
regenerate docs_map.
* fix(browser): satisfy OpenGrep security policy on the relay
- Hash both operands before timingSafeEqual so token comparison has no
length short-circuit (GHSA-JJ6Q-RRRF-H66H).
- Bound the relay WebSocketServer with maxPayload (64 MiB, headroom for CDP
screenshots/bodies) against oversized frames (GHSA-VW3H-Q6XQ-JJM5).
- Rewrite the config test env helper to avoid the skill-env-host-injection
shape (GHSA-82G8-464F-2MV7); it is a test-only env swap.
Reactive corrections ("that's not what I asked", "stop doing X") now
count as durable signals: expanded extraction patterns, vocabulary
routing to existing workspace skills, per-skill grouping. Both capture
modes share one invariant: a bounded signal-fingerprint ring on the
session entry prevents replaying applied/rejected corrections, pending
autocapture-owned proposals are revised instead of skipped, /learn-style
turns suppress duplicate agent-end capture, and extraction runs before
any skill discovery. Autonomy off keeps the suggest-tier offer; autonomy
on files/revises proposals directly.
Co-authored-by: Peter Steinberger <steipete@gmail.com>
Adds the middle tier between capture-off and autonomous capture: when
autonomy is disabled, detected durable-instruction signals record a
one-shot pendingSkillSuggestion on the session entry (signal-hash
fingerprint prevents transcript-history replay), and the next
non-heartbeat turn atomically consumes it and injects one bounded
user-role line offering to save the skill. The agent offers, the user
decides; skill_workshop approval flow unchanged; no new config.
* feat(skills): diagnose skill_workshop hidden by tool policy (#87570)
Workshop can be enabled and auto-capturing while tools.profile hides the
skill_workshop tool; every inspection surface looked healthy. plugins
inspect and openclaw doctor now name the excluding policy layer (global/
agent/provider profile, allowlist, denylist) and the exact alsoAllow
grant to add, via a shared resolveSkillWorkshopToolPolicyAvailability
helper that /learn's guard now reuses instead of composing policy
itself. Diagnosis only; no policy behavior change.
* ci: retrigger
* feat(goals): keep active session goals in per-turn context (#100409)
Active goals now inject one bounded user-role context line on every
non-heartbeat turn, refreshed at queued/interrupt admission via
provenance-tracked generated lines so operator stops take effect
immediately. Adds qa/scenarios/goals continuance scenarios and
goals.* coverage IDs.
* docs: regenerate docs map for goal context section
* fix(qa): use requiredChannelDriver instead of execution.channel pin in goal scenarios
Approval wait now fits inside the Codex dynamic-tool watchdog (70s +10s
gateway grace under the 90s kill), approval cards carry proposal id,
skill name, description, file count, and body size (spoof-safe
rendering), and timeouts return a structured pending-not-failed outcome
instead of a bare error. Expired requests cannot execute late; no
auto-apply; generic plugin approvals unchanged.
* feat(commands): add /learn to draft skills from recent work (#100408)
/learn rewrites the turn into a standards-guided Skill Workshop authoring
instruction: the agent gathers named sources (or distills the current
conversation) and files ONE pending skill proposal via skill_workshop.
Approval flow unchanged; sandboxed/tool-restricted agents get a clear
unavailable reply. Extracts the harness OpenClaw-tools predicate into
shared helpers and reserves the command name against plugin shadowing.
* docs: regenerate docs map for /learn section
* fix(browser): drain download saves and use monotonic cursor for act response
- Add drainPendingDownloadSaves() to wait for in-flight saveAs before sampling
- Use monotonic downloadSeq cursor instead of bounded-list length
- Propagate downloads info in POST /act response for click/batch/evaluate
Fixes#93250
* fix(lint): add braces around single-line if returns
Fixes eslint(curly) failures in drainPendingDownloadSaves and
pickNewDownloads. PR #93307 required CI gate.
Ref: ClawSweeper P1 review finding
* fix(browser): scope act download metadata to action
* fix(browser): broadcast downloads to all active captures to prevent misattribution
When concurrent /act calls overlap on the same page, using
state.actionDownloadCaptures.at(-1) assigned downloads to the wrong
action's capture. Push managed save promises to all active captures
so the triggering action always receives its download metadata.
Also adds regression tests: broadcast-to-all-captures, sequential
capture isolation, and strengthen the dispose test to assert no
re-capture after disposal.
* fix(browser): prevent unhandled rejection when download capture action throws before drain
Move managedSave.catch() before the captures-branch check so the
rejection handler always runs, preventing an unhandled promise
rejection when the action throws before drain() is called. Simplify
the handler by removing the now-dead return + catch at the bottom.
* fix(browser): type downloads in BrowserActResponse, fix lint unused var
- Add optional downloads field to BrowserActResponse type contract so
typed callers of browserAct() can consume the new payload without casts.
- Remove unused afterDispose variable in pw-session tests (lint fix).
* fix(test): correct post-dispose download assertion
capture.promises is not cleared by dispose(), so re-draining a disposed
capture still returns pre-dispose results. This is benign — callers
should drain before disposing. Update the test to assert the actual
behavior (still shows old results, new download not captured).
* test(browser): cover /act download metadata response
* refactor(browser): report action-owned downloads safely
* fix(browser): close action download ownership races
* test(browser): type action download capture mock
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
