* fix(issue): prevent crash from orphan issueProviderId (#7135)
The Jira image-headers effect in task-detail-panel subscribed to
selectIssueProviderById without an error handler, so a task with an
issueProviderId pointing at a deleted provider (e.g. after sync
convergence where taskIdsToUnlink didn't cover all local tasks)
propagated the selector throw to Zone.js as a crash dialog. Wrap the
inner selector observable in catchError that logs and falls back to
of(null); the downstream jiraCfg?.isEnabled guard handles the fallback.
Also drop IssueLog.log(issueProviderKey, issueProvider) from the
throwing variant of the selector: providers may carry credentials
(host, token, apiKey) and IssueLog history is exportable.
* fix(focus-mode): sync tray countdown with in-app timer during breaks
Tray title was rebuilt from a cached currentFocusSessionTime that only
refreshed when CURRENT_TASK_UPDATED fired. addTimeSpent is gated on an
active current task, so during focus-mode breaks or task-less focus
sessions the cache froze while the in-app timer kept ticking.
Add the tick action to taskChangeElectron$ so the cache refreshes every
second whenever the focus timer is running.
Fixes#7278
* fix(ci): restore GitHub Actions SHA pins undone by 0e9218bd68
Commit 0e9218bd68 silently reverted PR #7212 (github-actions-minor group
bump) along with its stated sync/client-id work. This restores the 15
workflow files to their pre-revert state.
Actions restored to newer pinned SHAs:
- actions/upload-artifact v7.0.0 -> v7.0.1
- step-security/harden-runner v2.16.1 -> v2.17.0
- softprops/action-gh-release v2.6.1 -> v3.0.0
- signpath/github-action-submit-signing-request v2.0 -> v2.1
- anthropics/claude-code-action v1.0.89 -> v1.0.93
- docker/build-push-action v7.0.0 -> v7.1.0
- easingthemes/ssh-deploy v5.1.1 -> v6.0.3
* fix: restore i18n, UI, and docs work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted the following work alongside its stated
sync/client-id changes. Files where later master commits (fec7b25f23, etc.)
already re-applied the reverted work are intentionally left untouched.
Restored:
- #7232 docs/long-term-plans/location-based-reminders.md (513 lines)
- #7199 Romanian i18n phase 3 (ro.json + ro-md.json, ~1168 lines)
- #7049 Polish translation improvements
- #7143 planner component styling (4 scss files)
- #7211 add-task-bar preserve time estimate when typing title
- #7208 task.reducer roll-up estimates for added subtasks
- #6767 focus-mode pomodoro reset button
- #7205 plugin-dev github-issue-provider TOKEN description
- #7231 mobile-bottom-nav FAB fix
- a4fe03272 iOS keyboard accessory bar (global-theme + dialog-fullscreen-markdown)
- 309670db3 ShortSyntaxEffects undefined guard
- 667a7986f Dropbox PKCE auth comment/behavior
* fix(electron): restore electron + e2e work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted the following electron/e2e work.
Files already re-fixed by later master commits are left as-is:
- e2e/tests/sync/supersync-archive-conflict.spec.ts (de33234976 + 191d129ff3)
Restored:
- e8a3e156eb fix(electron): Linux autostart IDB backing-store recovery
(re-adds electron/clear-stale-idb-locks.ts + start-app.ts wiring)
- 5ce78a5b63 fix(electron): macOS Cmd+Q / Dock > Quit hang
(setIsQuiting + before-quit delegate to close-handler)
- 46e0fa2d01 fix(sync): FILE_SYNC_LIST_FILES IPC contract
(electronAPI.d.ts + local-file-sync + preload + ipc-events)
- ea1ef16307 fix(android): session-only SAF permissions on OEM devices
- 8865dc0a50 test(e2e): supersync parallel-worker stampede guard
(SUPERSYNC_SERVER_HEALTHY env-var fallback + goto retry loop)
- af7c7687e2 test(e2e): block WS-triggered downloads in non-WS specs
- 265b44db5d test(e2e): premature waitForURL on daily-summary
(this is literally the fix the bad commit's message claimed to add)
Conflict resolutions:
- e2e/utils/supersync-helpers.ts: kept the refined getDoneTaskElement
checks from d64014d086 (later than c558bcab5e) while restoring the
goto retry loop from 8865dc0a50.
- electron/start-app.ts: unioned imports (setIsQuiting +
clearStaleLevelDbLocks from theirs, fs from ours).
* fix(sync): restore sync-core work undone by 0e9218bd68
Commit 0e9218bd68 silently reverted parts of several sync fixes. Most
sync-core reverts have already been re-addressed differently on master
by later commits (1f5184f6e7, 05cd875dd6, 09f5ced2c9, 7df43358ab,
d9158d6adb, 32dbc95ed9, 8c3b08e016, f89fe1ebc3) — those files are
intentionally left untouched to avoid reverting master's newer work.
This PR restores only the pieces that are genuinely still missing:
- e8a3e156eb fix(electron): IDB backing-store autoreload (in-app piece)
operation-log-hydrator.service.ts + .spec.ts (the electron/clear-
stale-idb-locks.ts piece was restored in the prior commit)
Plus three low-risk documentation/cleanup restorations:
- operation-sync.util.ts — add "Nextcloud" to isFileBasedProvider JSDoc
- dropbox.ts — restore improved _getRedirectUri JSDoc (667a7986fb)
- dialog-get-and-enter-auth-code.component.ts — restore isNativePlatform
comment explaining why manual code entry flow is used (667a7986fb)
- file-adapter.interface.ts — remove stray "// NEW" comment (46e0fa2d01)
Intentionally NOT restored (master's newer work covers or supersedes):
- sync-trigger.service.ts / sync.effects.ts (05cd875dd6)
- sync-wrapper.service.ts (1f5184f6e7)
- sync-errors.ts (1f5184f6e7 re-added LegacySyncFormatDetectedError)
- file-based-sync-adapter.service.ts + spec (1f5184f6e7 + d9158d6adb)
- file-based-sync.types.ts (1f5184f6e7)
- operation-log.const.ts (32dbc95ed9 bumped IDB_OPEN_RETRIES to 5)
- dialog-sync-initial-cfg.component.ts (f89fe1ebc3)
* test(e2e): update Show/hide task panel button label
The TOGGLE_DETAIL_PANEL translation was renamed from "Show/Hide
additional info" to "Show/hide task panel" in dd789d2, but e2e
selectors still referenced the old string, causing every test that
opens the task detail panel to time out.
https://claude.ai/code/session_011mX4Pr24S42svmA5j7fRux
* 18.2.1
---------
Co-authored-by: Claude <noreply@anthropic.com>
Both FocusModeForegroundService and TrackingForegroundService declare
android:foregroundServiceType="specialUse" in the manifest. The 2-arg
startForeground() overload already falls back to the manifest type on
Android 14+, but passing the type explicitly via ServiceCompat is
Google's documented best practice and removes OEM-dependent fallback
ambiguity.
Extract the Build-gated call into a Service.startForegroundSpecialUse
extension in ForegroundServiceExt.kt so a future third FGS picks up
the API-version gating automatically. Callers shrink to one-liners.
WebView.getCurrentWebViewPackage() is documented to return null on
devices without an installed provider, but in practice it can throw
AndroidRuntimeException, MissingWebViewPackageException, or NPE on
devices with a corrupt or updating provider. An unhandled throw there
short-circuits WebViewCompatibilityChecker.evaluate() before
FullscreenActivity's recovery path (added in 5a5c3bca89) has a chance
to run, re-crashing the app on launch.
Wrap the call in a narrow catch for RuntimeException so the documented
throwables fall through to the PackageManager lookup while Error
subclasses still propagate. Log at debug level so broken-provider
telemetry survives.
The recent WebView, FGS, and JS-dialog guards each caught Exception,
which hid every thrown type behind a log line. Play Console only
surfaces uncaught throws, so a future regression inside any of these
paths would be silently swallowed.
Narrow to the known recoverable types:
- FullscreenActivity.initWebView: AndroidRuntimeException (incl.
MissingWebViewPackageException) + IllegalStateException.
- FullscreenActivity.onJsAlert/onJsConfirm: WindowManager.BadTokenException
+ IllegalStateException; JsResult.cancel() still fires via the
handled flag so the WebView JS thread never deadlocks.
- TrackingForegroundService / FocusModeForegroundService
ensureForegroundNotification: IllegalArgumentException (Samsung One UI
/ Android 14 blank-title case) + SecurityException. Anything else,
including ForegroundServiceStartNotAllowedException, now propagates.
Observed on Play Console: AndroidRuntimeException thrown from
WebViewFactory.getProvider during FullscreenActivity.initWebView on
devices where the system WebView package itself is broken (seen with a
WebView Beta build throwing "Already registered a list of actions in
this process" from WebViewChromiumFactoryProvider.<init>). The activity
had no recovery - the constructor threw and the app died at launch.
Mirror the existing CapacitorMainActivity pattern: catch around the
WebHelper().instanceView(this) call, and on failure present the existing
WebViewBlockActivity using compatibility.copy(status = BLOCK) so the
block screen keeps the detected provider/version, then finish().
Also harden lifecycle callbacks that dereference the lateinit webView /
javaScriptInterface fields so the new failure path can't trigger an
UninitializedPropertyAccessException via OEM-specific lifecycle quirks:
::isInitialized guards in onSaveInstanceState, onRestoreInstanceState,
onBackPressed, and callJSInterfaceFunctionIfExists (the last matching
the existing guard in CapacitorMainActivity).
Every Context.startForegroundService() call creates a per-call FGS token
that must be answered with startForeground() within the system timeout,
even if the service is already a foreground service. Earlier fixes closed
specific branches one at a time and left gaps — most notably ACTION_START
with null taskId while already tracking, which returned without calling
startForeground().
Satisfy the contract once at the top of onStartCommand in both
TrackingForegroundService and FocusModeForegroundService, so every branch
below (known and future) already has the token answered. Add
stopForegroundAndSelf() on the previously silent no-op branches
(ACTION_UPDATE / ACTION_STOP when idle) so the guard does not leak the
service as foreground with a placeholder notification.
The placeholder notification carries a content title — Samsung One UI can
render title-less notifications as blank and a few Android 14 builds throw
IllegalArgumentException from startForeground() without one, which would
be swallowed by the guard's try/catch and re-trigger the very timeout we
are preventing. Matches the crash's heavy Samsung/Lenovo-tablet bias.
File paths in this file depend on the checkout location (e.g., main vs.
git worktree), causing it to churn in unrelated commits. Already matched
by the pre-existing android/.idea gitignore rule; just needed removing
from the index.
Prevents WindowManager$BadTokenException in FullscreenActivity.onJsConfirm
and onJsAlert when JS fires alert()/confirm() after the activity starts
finishing. Also ensures JsResult is always released so the WebView JS
thread never deadlocks on lifecycle-driven dismissal.
Android clears all AlarmManager alarms (and may cancel WorkManager jobs)
whenever the app's APK is replaced — i.e. on every Play Store auto-update
or manual reinstall. BootReceiver already knows how to re-register
everything from ReminderAlarmStore and re-enqueue SyncReminderWorker, but
its intent-filter only listened for BOOT_COMPLETED, so after an update
reminders stayed silently unregistered until the user next opened the app.
Add MY_PACKAGE_REPLACED to the manifest and broaden the receiver's action
guard so alarm restoration runs on both events. No new permissions or
helpers required — MY_PACKAGE_REPLACED is a protected broadcast delivered
only by the system to the owning app.
Refs #6832
Co-authored-by: Claude <noreply@anthropic.com>
- Separate JsonParseError and SyncDataCorruptedError handlers in sync wrapper
- Handle corrupted remote data gracefully instead of throwing
- Add getOrGenerateClientId() as unified entry point, eliminating dual injection
of ClientIdService + CLIENT_ID_PROVIDER in snapshot-upload, file-based-encryption,
and sync-hydration services
- Use crypto.getRandomValues() instead of Math.random() for client ID generation
- Warn user when stored client ID is invalid and must be regenerated
- Fix flaky e2e supersync tests (premature waitForURL match on daily-summary URL)
Some OEM ROMs throw SecurityException when takePersistableUriPermission is
called, even with FLAG_GRANT_PERSISTABLE_URI_PERMISSION requested. Instead of
rejecting the call, fall through and return the URI — it remains valid for the
session. Also fix checkUriPermission to use context.checkUriPermission() which
detects both persistent grants and active session grants, so isReady() no longer
clears a working (session-only) SAF URI on those devices.
When a full-state op (SYNC_IMPORT/BACKUP_IMPORT) arrived from another
client, mergeRemoteOpClocks reset the local clock to a "minimal" form
but only preserved the current client's counter from the incoming op's
clock. If the current client had issued ops (e.g. GLOBAL_CONFIG) not
reflected in the incoming full-state op's clock, its counter was dropped,
causing subsequent ops to reuse the same counter value. Downstream clients
then saw these ops as EQUAL (duplicate) and skipped them silently.
Fix: take max(mergedClock[clientId], currentClock[clientId]) when
rebuilding the clock after a full-state op reset.
Also add __SP_E2E_BLOCK_WS_DOWNLOAD flag to WsTriggeredDownloadService
to allow E2E tests to block automatic WS-triggered downloads during
concurrent conflict scenarios.
Fix archive conflict test by blocking WS downloads on Client A during
the concurrent edit phase so it doesn't auto-receive B's rename via
WebSocket before archiving (restoring the intended conflict scenario).
Fix LWW singleton test to assert convergence rather than specific winner.
Fix renameTask helper to avoid Playwright/Angular re-render races.
Fix shepherd.js import paths that broke the Angular dev server build.
* feat(android): improve task title from share intent using EXTRA_SUBJECT
Browsers and many apps send the page title via Intent.EXTRA_SUBJECT, which
was previously ignored. Now the share handler uses subject > title > content
to build a meaningful task title instead of the generic "Check: Shared Content".
For links without metadata, a readable URL (domain + path) is used instead.
https://claude.ai/code/session_01LzFepWFjLKQSWKZn4iXEpS
* chore: update package-lock.json after npm install
https://claude.ai/code/session_01LzFepWFjLKQSWKZn4iXEpS
* refactor(android): simplify _buildTaskTitle and fix double spaces in URL
Deduplicate the subject/title preference logic and collapse multiple
spaces in _readableUrl output caused by leading slashes in pathnames.
https://claude.ai/code/session_01LzFepWFjLKQSWKZn4iXEpS
---------
Co-authored-by: Claude <noreply@anthropic.com>
Guard against null bridge/webView after onCreate by showing WebViewBlockActivity
with actionable guidance instead of crashing. Satisfy Android's foreground service
contract in TrackingForegroundService and FocusModeForegroundService by posting a
minimal notification before stopping when started without a valid action. Handle
null taskId on ACTION_START without killing an active tracking session.
Google rejects custom scheme redirect URIs that don't match the
platform's app identifier. Use the Android applicationId
(com.superproductivity.superproductivity) on Android and the iOS
bundle ID (com.super-productivity.app) on iOS, with single-slash
URI format per Google's documentation.
- Add iosClientId to OAuthFlowConfig and plugin API types
- Add iOS OAuth client ID to Google Calendar plugin
- Select client ID per platform (Android/iOS/Desktop) in bridge service
- Add Android package name intent filter in AndroidManifest
- Accept both URI schemes in OAuth callback handler
- Fix redirect handler to use IS_NATIVE_PLATFORM consistently
Lock cdkDragLockAxis to Y on small screens (<600px) for task-list and
planner components, matching the breakpoint where planner switches to
single-column and bottom nav appears. Nav menu tree always locks Y.
Uses reactive LayoutService.isXs() so orientation changes work correctly.
Throttle tree-dnd onDragMoved with requestAnimationFrame to reduce
layout thrashing from getBoundingClientRect/elementFromPoint calls.
Add safeStopSelf() to TrackingForegroundService and FocusModeForegroundService
that calls startForeground() before stopSelf() in fallback branches (missing
task ID, stop when not running, null/unknown action). Android requires
startForeground() within 10s of startForegroundService().
Replace ambiguous NullPointerException in CapacitorMainActivity.onCreate with
a descriptive IllegalStateException when Capacitor bridge fails to initialize.
Add safe-call operators in onSaveInstanceState/onRestoreInstanceState.
Add OAuth support for plugins with PKCE, token persistence in local-only
IndexedDB, and Electron/web redirect handling. Extend two-way sync with
dueDay, dueWithTime, and timeEstimate field mappings including mutually
exclusive field clearing. Support remote issue deletion on task delete
and remote deletion detection during polling.
Add Google Calendar plugin (disabled from bundled builds pending legal
review) as a development reference for the plugin OAuth and two-way
sync APIs.
Additional fixes:
- Restore accidentally deleted focus-mode translation keys
- Remove full Task[] from deleteTasks action to prevent op-log bloat
- Add dueDay/deadlineDay format validation guards (#6908)
- Fix Android reminder alarm cancel intent matching
- Validate dueDay format to prevent false overdue from corrupted data
- Fix PKCE test expectation after utility consolidation
cancelReminder() created a PendingIntent without setting the action,
while scheduleReminder() set ACTION_SHOW_REMINDER. Android's
Intent.filterEquals() treats null != "action" so alarmManager.cancel()
silently failed to find the scheduled alarm.
* fix(android): prevent stale check from suppressing valid reminder notifications
The isTaskStale check in ReminderAlarmReceiver was incorrectly suppressing
all locally-scheduled reminders. After 6fd0696 enhanced extractRemindersToSchedule
to detect schedule ops from actionPayload, the stale check would find the
task's own schedule op and mark it as stale — even though the alarm IS the
current schedule.
Fix by passing triggerAtMs through the alarm intent and comparing it with
schedule ops' remindAt. A task is only considered rescheduled (stale) when
the schedule op's remindAt differs from the firing alarm's trigger time.
https://claude.ai/code/session_013216HfgtetW9KKzFr5ieLQ
* fix(android): handle dual-reminder edge case and seq cursor reset
Two additional fixes found during review:
1. Dual-reminder edge case: A task can have both a standard reminder
(remindAt) and a deadline reminder (deadlineRemindAt) with different
times. The stale check now matches on isDueDate to prevent one
reminder type from suppressing the other.
2. Seq cursor stuck after server reset: If the server is wiped (e.g.,
clean sync import), latestSeq drops below the stored lastSeq. The
worker now resets the cursor when the server's seq is lower, instead
of being permanently stuck.
https://claude.ai/code/session_013216HfgtetW9KKzFr5ieLQ
---------
Co-authored-by: Claude <noreply@anthropic.com>
The op-log system returns empty entityChanges for most actions — only
TIME_TRACKING and addTask (CRT) ops populate it. The background sync
provider was relying on entityChanges to detect remindAt changes, so
schedule/unschedule/deadline actions were invisible to the worker.
Fix by matching on actionType strings and parsing actionPayload directly:
- unscheduleTask, removeDeadline → cancel reminders
- scheduleTaskWithTime, reScheduleTaskWithTime → parse actionPayload.remindAt
- setDeadline → parse actionPayload.deadlineRemindAt
Let later ops win when dismiss and reschedule are in the same batch
(previously cancellation always won, suppressing rescheduled reminders).
Also extract task title from actionPayload as fallback for UPD ops.
The SuperSync server returns operations in full format (full field names
and NgRx action strings), but the parser was using compact format codes.
This caused all background sync parsing to silently fail, preventing
reminder cancellation and scheduling.
Add server-side stale check to ReminderAlarmReceiver before showing
notifications. When an alarm fires, the receiver does a quick server
fetch (5s OkHttp timeout, fail-open) to suppress notifications for
tasks that were deleted, done, dismissed, or rescheduled on another
device.
Also extend the background sync worker to schedule new/updated
reminders discovered in sync operations, not just cancel stale ones.
Key design decisions:
- Fail-open: any error or timeout still shows the notification
- Receiver only checks its own task; Worker owns all other reminder
management and the seq cursor
- Last-writer-wins dedup via Map keyed by (taskId, isDueDate)
- Cancellation wins over scheduling in the same batch
- Independent try/catch per cancel/schedule to prevent cascading failures
- SupervisorJob on receiver's CoroutineScope
* docs: add plan for Android background sync via WorkManager
Outlines the implementation plan for using WorkManager to periodically
sync with SuperSync server in the background, cancelling stale reminders
for tasks that were completed/deleted on other devices.
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* docs: refine plan based on multi-agent architecture review
- Pin frontend hook to SyncProviderManager.currentProviderPrivateCfg$
- Add skipWhileApplyingRemoteOps() guard for selector-based effect
- Add HRX (dismiss reminder) and deadlineRemindAt to parser
- Add batch operation ds field handling
- Add ProGuard rules step for release builds
- Add error handling note for BootReceiver WorkManager enqueue
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* docs: improve plan after second review round
Key changes:
- Remove Step 5 (ReminderAlarmStore lookup) - hash-based cancellation is
sufficient, no store lookup needed
- Add BackgroundSyncProvider interface for extensibility to Dropbox/WebDAV
- Per-account lastServerSeq keyed by baseUrl hash
- Add SyncReminderScheduler helper to deduplicate scheduling logic
- Document edge cases: account switching, token expiry, force-kill,
gapDetected, notification race conditions
- Add extensibility section explaining Dropbox feasibility and architecture
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* feat(android): add background sync to cancel stale reminders via WorkManager
When a task is completed/deleted on desktop, Android reminders for that task
would still fire because the mobile app didn't know about the change. This adds
a WorkManager periodic job (every 15 min) that fetches operations from the
SuperSync server and cancels AlarmManager alarms + notifications for tasks that
are done, deleted, archived, or had their reminders dismissed.
Architecture:
- BackgroundSyncProvider interface for extensibility (Dropbox/WebDAV possible later)
- SuperSyncBackgroundProvider: lightweight HTTP client + operation parser
- SyncReminderWorker: CoroutineWorker that fetches ops and cancels reminders
- SyncReminderScheduler: helper to schedule/cancel the WorkManager job
- AndroidSyncBridgeEffects: mirrors SuperSync credentials to native SharedPreferences
- Per-account lastServerSeq prevents account-switching bugs
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* chore: update package-lock.json after npm install
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* fix(android): address review issues and add tests for background sync
Fixes from code review:
- Fix distinctUntilChanged to properly handle provider switching
(non-SuperSync emissions now treated as equal to prevent repeated clears)
- Add null-safe ops array parsing in SuperSyncBackgroundProvider
- Add write/call timeouts to OkHttpClient (30s write, 60s call)
- Remove no-op `hash and hash` line in Kotlin hash function
Tests:
- Add cross-platform parity tests for notification ID hash function
(pinned expected values that must match Kotlin implementation)
- Add AndroidSyncBridgeEffects unit tests covering:
- distinctUntilChanged comparator behavior
- credential set/clear decision logic
- observable filtering integration (dedup, null filtering)
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* fix(android): fix hash overflow and operation parser payload paths
Two critical bugs found during final review:
1. Hash function abs(Int.MIN_VALUE) bug: In Kotlin, abs(Int.MIN_VALUE)
returns Int.MIN_VALUE (still negative) due to 32-bit overflow.
Fixed by converting to Long before abs(), matching JS behavior where
Math.abs works on 64-bit floats.
2. Operation parser payload structure: The parser was looking at
p.isDone and p.task.changes directly, but the actual compact
operation format wraps everything in p.entityChanges[] and
p.actionPayload. Rewrote to use p.entityChanges as primary source
(consistent structure with entityType, entityId, changes fields)
and p.actionPayload.task.changes as secondary source.
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* docs: add long-term plan for Android background sync improvements
Covers three phases: fast app startup via cached sync seq,
push-based notification cancellation via FCM, and extending
background sync to Dropbox/WebDAV providers.
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
* fix(android): address review feedback for background reminder sync
- Remove PLAN.md from repo root (C1)
- Use EncryptedSharedPreferences for access token storage with
fallback to standard SharedPreferences on broken KeyStore (C2)
- Reset lastServerSeq when access token changes to prevent stale
seq on account switch with same server URL (I1)
- Add max iteration guard (100) to pagination loop to prevent
infinite loops from server bugs (I2)
- Use type predicate filter instead of non-null assertions (I3)
- Add exponential backoff (5min) to WorkManager schedule (S1)
- Add comments linking action type codes to frontend source (S2)
- Extract distinctUntilChanged comparator to named function with
JSDoc explaining the suppression semantics (S3)
https://claude.ai/code/session_01RkrVBB492k8RBA8zt5swwe
---------
Co-authored-by: Claude <noreply@anthropic.com>
