* fix(macos): honor user accent precedence in config snapshot and live-update chat window
The Control UI user accent (ui.prefs.accent) landed in #128432/#128577 with
precedence user accent -> operator ui.seamColor -> theme default, and the
gateway's talk.config payload already applies it. The macOS app had two gaps:
- ChannelsStore.applyUIConfig read raw ui.seamColor from the config.get
snapshot and clobbered the user accent set from talk.config depending on
arrival order. It now resolves ui.prefs.accent ?? ui.seamColor via a
testable helper mirroring the gateway precedence.
- The native chat window read AppStateStore.seamColorHex once at window
construction, so accent changes never live-updated. MacChatSurface now
reads the @Observable store in body, deleting the one-shot userAccent
plumbing.
Docs: configuration-reference.md documents the precedence for native-app
chrome. Regression test fails pre-fix (snapshot returned the operator seam
color instead of the user accent).
* fix(macos): refresh config from gateway config.changed events
Addresses the review finding that no macOS consumer turned the gateway's
hash-only config.changed broadcast into refreshed shared state, so a
Control UI accent change never reached an open native chat window while
the app ran. ChannelsStore now subscribes to gateway pushes and re-fetches
config.get on config.changed, reconnect snapshots, and sequence gaps.
The refresh applies non-force so an in-progress local settings draft wins
(the gateway rejects stale-hash writes anyway). The in-flight reload queue
gains a closed pending level (none/refresh/force) so a refresh arriving
during a load is coalesced instead of dropped, and a requeued refresh
cannot clobber a dirty draft the way the old boolean force-pending did.
* fix(nodes): report camera positions the hardware actually reached
`camera.ptz.control` returned a position it never verified, and
`camera.snap`/`camera.clip` could capture from a camera the caller did
not ask for. Both told the agent an action succeeded when it had not.
PTZ read its post-write status from the same UVC connection that issued
the write. Gimbal cameras echo a pending setpoint back on the writing
connection, so the check confirmed its own write. Those cameras also
service camera-terminal controls only while a video stream is active, and
no capture session was held, so writes could be discarded entirely while
reads returned phantom values.
Hold a frame-discarding capture session across every PTZ operation, close
the writing controller, and verify through a fresh connection against each
axis's advertised resolution. An axis that misses now reports through the
existing CAMERA_PTZ_PARTIAL outcome with observed versus requested values
and what to check next.
Apple camera selection accepted an explicit deviceId and silently fell
back to the default camera when nothing matched. Linux already rejected
this, and CameraPTZService already rejected it in the same app. Centralize
exact selection in OpenClawKit so macOS and iOS both fail with a
device-not-found error; the facing/default fallback stays only for
requests that supply no deviceId.
camera.ptz.status now activates the camera and its privacy indicator for
the duration of the read. That is the cost of returning real positions.
* fix(nodes): tell callers how to recover from an unknown camera ID
Device IDs change when cameras are reconnected, so a bare
device-not-found error dead-ends the caller. Both Apple errors and the
docs now point at camera.list for current IDs.
Addresses the ClawSweeper P2 finding on #128595.
Fixes#128194.
Local-mode onboarding has two rightful actors on the CLI page: the page's
own install flow, which may open an install-target prompt on unreleased
builds, and the connection-mode commit, which starts the gateway through
ConnectionModeCoordinator the moment the user picks This Mac. When the
gateway comes up on its own (dev-root builds, or an externally attached
gateway), AI setup auto-connects and finish() runs — but close() was
silently ineffective because AppKit ignores NSWindow.close() while a sheet
is attached. The result, reproduced live with os_log instrumentation: a
completed onboarding (dashboard open, onboardingSeen set) with a zombie
onboarding window showing a dead CLI page and a stale channel-choice sheet.
Two coupled fixes. OnboardingController.close() now ends any attached sheet
before closing, so completion always tears the window down. And a running
local gateway now resolves a pending install prompt directly: a new tested
static (shouldResolveInstallPromptForRunningGateway) extends the existing
gateway-status revise path to the choosingTarget phase, marking the step
installed and dismissing the moot sheet; runCLIInstall returns quietly in
that case instead of writing a cancellation over a resolved step. Genuine
declines keep the cancelled status.
* fix(macos): attach the onboarding CLI install prompt to the window
On an unreleased build, the onboarding CLI page resolved its install target
through NSAlert.runModal() — a detached app-modal panel that is absent from
the app's AX window list and freely covered by system permission dialogs.
Live repro: with two TCC prompts stacked over it, the main thread sat parked
in runModal for 30+ minutes while the page showed an active "Install
OpenClaw" spinner, the close button was disabled, and busyReason claimed an
install was in flight — before the user had chosen anything. A spinner that
means "answer a dialog you cannot see" is a silent-failure trap.
The prompt (both the confirm-stable alert and the channel chooser) is now a
sheet attached to the onboarding window via beginSheetModal, so it stays
z-ordered with the window and AX-visible. The busy state is honest: a new
.choosingTarget phase renders the install row as pending instead of running,
and installingCLI/close-disabled/busyReason only engage after a target is
actually chosen. Non-onboarding callers (checkAndPromptIfNeeded) keep
runModal by passing no window. All alert text and button order unchanged.
* style(macos): wrap sheetPresentationWindow property body
Writing to a pipe whose reader has already exited raises SIGPIPE, which
kills the whole process instead of throwing. MacNodeHostWorker already
guarded its stdin pipe with F_SETNOSIGPIPE; the codex App Server client
and the MLX TTS helper transport did not, so a child exiting mid-write
could take down the app. Suppressing the signal exposed that an
undelivered App Server request write was reported as appServerUnavailable
even though the frame was provably never sent, so it now requeues once
onto a fresh child instead of failing the caller.
Test-side pipe write ends whose readers are spawned children (or a
readability handler that can close the pipe mid-test) get the same
suppression so a racing reader exit fails the assertion instead of
killing swiftpm-testing-helper with signal 13, which is what caused the
macos-swift CI lane's intermittent unrelated-test crashes (e.g. PR
#126559, run 32341197738 job 96340683947).
* fix(apple): gate gateway RPC polling on the hello method catalog
Released 2026.7.x gateways authorize before method dispatch and reject
unknown methods with INVALID_REQUEST / "missing scope: operator.admin",
which never names the method, so the shared Swift kit's error-text
matchers could not detect an unsupported gateway: queued sends wedged in
a delay-free sessions.branches.list reconcile loop, question.list burned
its retry budget per health event, and progressCard.get fired a rejected
fetch per event.
Port the Android fix pattern (#126540): generalize the progressCard-only
transport seam into tri-state gatewayAdvertisesMethod(_:), make hello
catalog parsing distinguish absent (nil) from empty, route branch
listing through a catalog-checking dispatch point with a typed
BranchListingUnadvertisedError, tighten error-text matching to the
modern "unknown method:" shape, and skip question.list/progressCard.get
when unadvertised. Regression tests encode the exact 2026.7.1-2 wire
shape and fail pre-fix (A/B verified).
* fix(apple): keep explicit unsupported branch-listing replies releasing sends
ClawSweeper P1: the tightened matcher dropped the shipped acceptance of
explicit unsupported/unimplemented GatewayResponseError replies that
name sessions.branches.list, which would wedge queued sends on a
pre-catalog gateway emitting that shape. Collapse the matcher onto the
bridged localizedDescription (errorDescription always prefixes the
method name), preserving both legacy qualifier shapes while still
rejecting bare missing-scope denials and the old false-positive-prone
INVALID_REQUEST arm; add the releasing-send regression test.
* fix(macos): surface concrete Gateway start failure reason in onboarding
GatewayProcessManager already retains the specific registration/readiness
failure (e.g. "launchd disabled", a launchd enable error, a readiness
timeout) in lastFailureReason, and Settings/menu bar UI already read it.
Onboarding discarded it: LocalGatewayActivation.failed collapses every
cause to the same generic "Retry setup" message, so a missing LaunchAgent
registration is indistinguishable from any other startup failure.
Surface the retained reason in the onboarding status text so the failure
is diagnosable without going through Settings.
* fix(macos): record command-resolution failures in lastFailureReason
GatewayProcessManager set status but not lastFailureReason when
GatewayEnvironment.resolveGatewayCommand() returns no command (missing
runtime/CLI), unlike the launchd-disabled and launchd-enable-error
branches a few lines below. Onboarding's new failure message therefore
rendered the generic text or a stale reason from an earlier attempt
for this failure class. Mirror the sibling branches and record
resolution.status.message.
Also fixes the macos-swift SwiftFormat lint failure: the comment block
directly above gatewayStartFailureMessage needed to be a doc comment
(///), matching the repo's existing convention for declaration-adjacent
comments.
* fix(macos): bind Gateway start failure reason to its activation attempt
LocalGatewayActivation.failed carried no data, so both onboarding call
sites reread the mutable GatewayProcessManager.shared.lastFailureReason
singleton after activateLocalGateway() returned. A later gateway-start
attempt can overwrite that singleton before the caller gets around to
reading it, so a stale wait could surface a newer attempt's reason (or
vice versa) attributed to the wrong onboarding attempt.
Widen LocalGatewayActivation.failed to carry reason: String?, captured
inside activateLocalGateway() the instant waitUntilReady() resolves to
false, and have both onboarding call sites map that bound value instead
of rereading the singleton. CLIInstallPrompter's two `!= .failed`
comparisons become `if case .failed = activation` pattern matches since
`.failed` is no longer a payload-free value; its existing `case .failed:`
message switch is unaffected, since bare-case patterns still match
regardless of associated data.
* fix(macos): satisfy SwiftFormat lint on CLIInstaller.swift
Converts the LocalGatewayActivation.failed declaration comment to a
doc comment and wraps activateLocalGateway's closing signature per
config/swiftformat, matching the same docComments convention already
applied elsewhere in this PR. No behavior change.
* fix(setup): refresh Codex registry with staged install
* fix(macos): verify inference before onboarding handoff
* fix(setup): use native Codex home for subscription auth
* fix(codex): honor attempt-scoped setup config
* fix(macos): align onboarding handoff with reopen
* fix(setup): await prepared model convergence
* fix(ui): avoid false auth state for empty catalog
* fix(setup): scope catalog convergence to Codex gateway
* fix(setup): publish the committed runtime catalog
* fix(models): project configured static runtime models
* fix(codex): expose app-server model catalog
* fix(models): preserve Codex auth across reloads
* fix(ci): align Codex onboarding checks
* test(ui): stabilize dock suppression environment
* fix(codex): honor discovery config in app-server model catalog
The manifest documents discovery.enabled (bundled fallback list) and
discovery.timeoutMs (default 2500ms) for model discovery; the new catalog
path used the generic 60s request timeout and ignored the enable gate.
Also drop the test-only listModels injection seam in favor of vi.mock.
* fix(setup): refuse prepared Codex auth over an explicit remote transport
configureCodexCliPreparedAuth silently rewrote an explicitly configured
websocket/unix app-server to local stdio (keeping a dangling url), moving
the credential boundary onto this host. Fail setup with actionable
guidance instead; also surface the root cause when the prepared model
catalog refresh fails after activation.
* refactor(agents): one canonical model-catalog identity key
Three near-identical key helpers existed (models-list-result,
models-list-configured-static, harness/model-catalog). Export
resolveModelCatalogIdentityKey from the route-policy owner, collapse the
duplicate dedupe loops into dedupeByKey, make donor enrichment Map-based,
and inline the one-off harness-augment wrapper.
* fix(macos): restore custodian handoff for fresh activations
Landing every finish on the plain dashboard stranded the custodian
first-run flow (memory import, channels, permissions, hatch). Fresh
activations now hand off to custodian onboarding; live-verified
pre-existing setups reopen the normal dashboard, matching the removed
already-configured shortcut. Tests pin the destination per path.
Also isolate the post-startup Codex login test from developer machines:
ambient OPENAI_API_KEY and a real Codex login made it assert-fail.
---------
Co-authored-by: Peter Steinberger <steipete@gmail.com>
* fix(apps): render legacy plan events when the Gateway lacks the progress-card store
progressCard.get ships in no released Gateway tag, so iOS/macOS on any current
release silently showed no status card after the durable-card migration. Restore
legacy stream:"plan" rendering, synthesized into the unified progress card and
gated on a tri-state capability probe so a dual-emitting Gateway never fights the
durable store. Mirrors the Android fix in #125444.
* fix(apps): invalidate progress-card capability on gateway route replacement
A replacement route may be a different Gateway, so a cached known-absent store
must not authorize the legacy plan fallback against a new Gateway that dual-emits
both sources. Clear and re-probe on .routeChanged.
* feat(apps): migrate iOS/macOS plan surface to the durable progress card
Replace the legacy stream:"plan" agent-event pipeline (runId-scoped state,
run-gated pill) with the sessionKey-scoped progress-card store: the shared
chat surface now renders progressCard.get snapshots, refetches on
progressCard.changed pokes with revision dedupe, clears on null-revision
pokes, and persists the card after the run completes. The card renders
markdown through the shared markdown view plus typed steps. Legacy Apple-side
plan handling (agent-event case, run-snapshot plan reconciliation,
OpenClawChatPlanStep parsing) is deleted; gateway emission stays for Android.
Removes the ios progressCard.changed coverage allowlist entry so the check
enforces the handler.
* chore(i18n): refresh native inventory for the progress-card rename
* fix(apps): keep the last progress card when a refresh fails
A transient progressCard.get failure no longer clears an already-rendered
durable card; only a successful null fetch or a null-revision poke clears it.
Make validated CLI cache writes idempotent and reuse the startup-scoped node worker launch across route retries, keeping the signed menu-bar app near-zero CPU while idle.\n\nCloses #124592
The computer.act v1 wire contract is gone, but the naming that survived it
still described a version split instead of the real one: screen-coordinate
execution versus window/element-scoped execution. Both are live rungs of the
same ladder.
- Extract the screen-coordinate half of the 1334-line ComputerActionService
into ComputerScreenActionExecutor (dispatch, typing, scroll, coordinate
mapping, button-hold watchdog, raw CoreGraphics primitives). Moved code is
unchanged apart from threading the queue authority check as a parameter
instead of reaching back into the queue.
- ComputerActionService keeps its name and becomes the coordinator that owns
the execution queue, the permission probe, and the shared error vocabulary.
- Rename ComputerActionServiceV2 to ComputerWindowActionExecutor, isV2Request
to isWindowScopedRequest, isComputerActV2Only to isWindowScopedOnly, and
ComputerActionError.invalidV2Request to .invalidRequest. The emitted
COMPUTER_INVALID_REQUEST: prefix is unchanged.
- cua-computer: v2-actions.ts becomes window-actions.ts, handleV2Act becomes
handleWindowAct, and the stale v1Params local in handleDesktopAct becomes
desktopParams.
- Note at the computer.act idempotency key that its v1 prefix versions the key
composition, not the wire contract.
Behavior-neutral: no logic edits, no new branches, no changed error strings.
Peter Steinberger