TaskRunner.RunAsync created the worktree (PrepareRunDirectoryAsync)
before claiming Running via StartRunningAsync. RunNow dispatches by
task id with no atomic claim of their own, so a Queued task racing
the queue picker's atomic SQL claim could hit WorktreeManager's
branch-collision self-heal, which force-removes and recreates the
winner's live worktree mid-run.
Move the claim before any resource creation and bail out immediately
when it's rejected. OverrideSlotService.RunNow also fast-rejects a
task already Running in the DB (defense in depth). RunCancellationRegistry
now refuses (and logs) a double registration instead of silently
overwriting the first CTS, so a losing dispatch's cleanup can no
longer unregister the winner's cancellation token.
Remove the redundant TaskUpdated broadcast in TaskRunner.ContinueAsync's
queue-claim path, consolidate InteractiveLaunchSpecService's seven
MCP_TOOL_TIMEOUT literals into one constant (fixing the merge-helper handoff
spec's stale 200000ms value), and surface OpenQuickClaudeSession's two
failure cases via ErrorReported/footer instead of a silent no-op, with a
less ambiguous icon.
QueuePicker's raw-SQL Queued->Running claim bypasses TaskStateService.StartRunningAsync,
the only other place that broadcasts TaskUpdated on this transition, so the task-list
badge stayed on "Queued" until the run finished. Send TaskUpdated for alreadyClaimed
dispatches in TaskRunner.RunAsync/ContinueAsync to close that gap.
Adds a read-only get_effective_run_config(taskId) tool that reports the
model/max-turns/effort/permission-mode/agent-path/system-prompt/skills a
task will actually run with, each tagged with its source (task/list/
preset/global), plus max-turns' raw requested value and clamp status.
Extracted the model/max-turns/agent-path resolution out of
TaskRunner.ResolveConfigAsync into EffectiveRunConfigResolver so the run
path and the new reporting tool share one codepath instead of risking
drift, per docs/explore-notes/worker-task-pipeline.md's max-turns trap.
StreamAnalyzer now reads subtype/terminal_reason/errors from the CLI's result
event, and TaskRunner builds a specific ErrorMarkdown from them: max_turns names
the turn budget and points at set_task_config + requeue, api_error passes
through the provider's own message (which carries the reset time), and any
other terminal_reason is appended to the previous generic text instead of
staying invisible. Falls through to the old "exited with code N and no result"
text when there's no terminal_reason at all (a real crash).
task_runs.tokens_in only ever held the API's uncached "input_tokens" field
(off by a factor of ~400,000 on a resumed session), and tokens_out summed
only the last result event instead of the whole session. TaskRunner now
reads each run's cache-read/cache-write/input/output totals from the
session transcript via a new ITranscriptUsageReader.ReadSessionTotalsAsync,
storing the delta against prior runs on the same session so a --resume
doesn't double-count. New task_runs.cache_read_tokens/cache_write_tokens
columns; the Session tab now shows the raw total (what actually counts
against the 5h/7d usage limit) with a breakdown tooltip.
Runaway sessions were the single biggest cost driver: model_presets was
never persisted (stayed code-only), default_max_turns shipped at 100, and
ResolveMaxTurns had no upper bound, so a task/list override could run
hundreds of turns unchecked.
- TaskRunner.ResolveMaxTurns now clamps the resolved value to
AppSettings.MaxTurnsCeiling (new column, default 80) and logs a warning
with task id / requested / effective value when it clamps.
- default_max_turns default lowered from 100 to 40 (entity, EF config,
and the seeded row via the new AddMaxTurnsCeiling migration).
- AppSettingsRepository.GetAsync backfills model_presets with the
shipping defaults on first read instead of leaving the column null.
- Settings > General's per-model preset table and the task/list agent
editor now show a hint when a set max-turns value exceeds the ceiling.
Adds GetUsageSnapshot/GetModelUsage/GetTaskUsage to WorkerHub (backed by a
shared UsageSnapshotBuilder), a UsageUpdated broadcast fired after every
UsageMonitorService poll cycle, and records the resolved model on each
task_runs row so per-model/per-task usage can be reported from history.
A full model id like claude-sonnet-4-6 in list_config.model never matched any
ModelPresets row (only bare aliases did), so every run under it landed on the
hardcoded 30-turn/DefaultLevel fallback instead of the sonnet preset row -
this is what killed two autonomous tasks at the 30-turn limit.
ModelRegistry.TryNormalizeAlias (non-throwing: exact match, then substring
match against a full model id) lets ModelPresets.For resolve a full model id
to its alias's preset row. For a model that still doesn't resolve, the
hardcoded 30 is replaced by a caller-supplied fallbackMaxTurns, and TaskRunner
now passes AppSettings.DefaultMaxTurns there - so that setting has a real
effect instead of being dead, matching the direction already noted in
docs/open.md.
ClaudeDo never passed --effort, so every session inherited whatever effortLevel
the user's Claude Code config happened to carry. Settings -> General now holds one
row per model alias (haiku medium/20, sonnet high/30, opus high/40, fable high/25)
supplying the global effort and turn defaults; list- and task-level max-turns
overrides still win, and the agent editor's inherited badge follows the model.
--effort is applied to autonomous runs and to every ConPTY spec (task session,
planning start/resume, ad-hoc, list handler). The model itself is deliberately not
forced on interactive sessions. The single global 'Max turns' field is replaced by
the table, and 'fable' joins ModelRegistry.Aliases.
The migration also adds the is_manual columns used by the next commit.
A running task can call mcp__claudedo_run__AskUser(question) to block (up to 3
min) on a human answer. PendingQuestionRegistry holds the pending question +
TaskCompletionSource; the tool broadcasts TaskQuestionAsked, awaits the answer
(WorkerHub.AnswerTaskQuestion resolves it), and returns it as the tool result —
or a 'proceed on your judgment' fallback on timeout. The run stays Running
throughout (no status/schema change). ClaudeProcess raises MCP_TOOL_TIMEOUT so
the 60s HTTP-MCP cap doesn't kill the wait; the run MCP is now wired for every
task, not just standalone ones. System prompt updated to reconcile 'unattended'.
TaskRunner appends attached files (absolute paths) to the run prompt as the
read-only Reference files section. Task and list deletes now remove the
on-disk attachment dir eagerly, and a startup AttachmentOrphanRecovery sweep
drops any attachments/<taskId>/ whose task no longer exists (covers list
cascade and planning-discard paths).
The picker claims Queued->Running atomically before dispatch; the new
StartRunningAsync guard then rejected every queue-dispatched run. Add
alreadyClaimed to RunAsync/ContinueAsync (queue passes true, override
slot keeps the guard) and align the routing tests.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Befund (bestätigt): src/ClaudeDo.Worker/Runner/TaskRunner.cs:101 (RunAsync) und :211 (ContinueAsync) ignorieren das TransitionResult von _state.StartRunningAsync. Race-Szenario: Der QueuePicker claimt Queued→Running atomar; ruft der Override-Pfad (RunNow) kurz danach RunAsync für denselben Task auf, schlägt StartRunningAsync fehl (0 rows affected), der Runner startet Claude aber trotzdem → derselb
ClaudeDo-Task: 44f86be2-7f3d-462e-98b3-eb94c0174eea
1. ArgumentList (fix injection): ClaudeArgsBuilder.Build() now returns
IReadOnlyList<string>; ClaudeProcess populates ProcessStartInfo.ArgumentList
instead of Arguments, so values like system prompts are never shell-split.
DailyPrepPrompt, RefinePrompt, and WeekReportService migrated similarly.
All IClaudeProcess fakes updated.
2. ContinueAsync exception guard: wrap RunOnceAsync in try/catch matching
the RunAsync pattern so an unexpected exception never leaves the task
stuck in Running status.
3. Planning chain cascade: OnChildFinishedAsync now calls CancelAsync on
the immediate blocked successor when a child fails or is cancelled,
triggering a recursive cascade that clears the entire remaining chain
instead of leaving it wedged.
4. FailAsync guard: restrict valid source states to Running and Queued;
WaitingForReview -> Failed is now rejected, preventing an invalid
transition that could corrupt the review workflow.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Resolved sub-tasks are no longer appended to the prompt. Extracted into a
shared TaskPromptComposer so the UI's description preview can render the same
'what Claude gets' text.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- TaskRunner: extract worktree-vs-sandbox selection into
PrepareRunDirectoryAsync so RunAsync reads linearly (a small helper, not
a Strategy pattern — overkill for a two-way branch).
- App: drop the public static ServiceProvider locator; inject the provider
via constructor through AppBuilder.Configure(() => new App(services)).
Parameterless ctor + BuildAvaloniaApp() retained for the XAML designer.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Clear ReviewFeedback only after a successful re-run so a failed/cancelled
run keeps it for a manual retry.
- Clear stale StartedAt/FinishedAt when rejecting a task back to the queue.
- Only non-planning standalone tasks gate on review (guard PlanningPhase).
- Hide "send to queue" for WaitingForReview tasks so review isn't bypassed.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Standalone tasks now enter WaitingForReview on success; re-queued tasks
carrying reviewer feedback resume the prior Claude session with that
feedback as the next turn.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Set the task's log path when the run is created (not at completion) so the
session terminal can replay live output when the user navigates away and back
mid-run. Remove the now-redundant inline per-row live tail (LiveTail /
HasLiveTail / TaskMessageEvent) and scroll the terminal to end after the next
layout pass so wrapping lines aren't clipped.
Unify the near-identical HandleFailure/MarkFailed into a single MarkFailed that
always persists the failed state and never throws, and replace the inline
null-if-blank checks in ListMcpTools with the existing extension.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Drops TagEntity, TagRepository, and tag wiring across data layer, worker,
and UI. Adds RemoveTags migration to clean up schema.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Slice 2 of the worker state consolidation refactor (spec sections 2 and 8).
Adds Worker/State/ITaskStateService + TaskStateService as the single component
that mutates Status, PlanningPhase, and BlockedByTaskId. Each transition is one
atomic ExecuteUpdate with a WHERE filter on the expected source status, so
parallel claims are TOCTOU-free. Side effects (queue wake on -> Queued, hub
TaskUpdated broadcast, chain advance + parent completion on terminal child)
are owned by the service so callers no longer need to remember them.
Migrated callers (mechanical, behavior preserved):
- TaskRunner: HandleSuccess/HandleFailure/MarkFailed/RunAsync/ContinueAsync
- StaleTaskRecovery: bulk recover stale Running tasks
- TaskResetService: status flip (worktree cleanup stays in service)
- PlanningSessionManager.StartAsync: status flip via state, token write via repo
- PlanningChainCoordinator.OnChildFinishedAsync: routes the next-sibling write
through state.UnblockAsync (Slice 4 finishes the rewrite)
- ExternalMcpService.UpdateTaskStatus: Queued case via state.EnqueueAsync
Repo Mark*Async helpers (MarkRunning/MarkDone/MarkFailed/FlipAllRunningToFailed)
are now internal; ClaudeDo.Data grants InternalsVisibleTo to ClaudeDo.Worker
and ClaudeDo.Worker.Tests for the existing repo-level tests.
DI: TaskStateService is registered as Singleton in both the main app and the
external-MCP app; the queue-wake delegate captures sp -> QueueService.WakeQueue
to break the TaskStateService -> QueueService -> TaskRunner -> TaskStateService
construction cycle. PlanningChainCoordinator takes Func<ITaskStateService> for
the same reason; Slice 3 will replace both with IQueueWaker.
Tests: TaskStateServiceTests covers happy + reject for every transition, the
parallel StartRunningAsync claim race, child-terminal chain advancement, and
stale recovery. Existing service/repo tests are updated to construct the new
state-service via a TaskStateServiceBuilder helper. Pre-existing constructor
drift in QueueService/ExternalMcp/PlanningHub tests is patched to keep the
test project building (the surrounding test logic is otherwise untouched).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Introduces ~/.todo-app/prompts/{system,planning,agent}.md as the canonical
location for prompt content. The settings modal exposes "Open in editor"
shortcuts for each, and TaskRunner merges system.md (always) and agent.md
(for "agent"-tagged tasks) into the effective system prompt alongside the
existing global/list/task layers.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Coordinates Waiting -> Queued transitions between sibling subtasks: when a child finishes Done, the next Waiting sibling is promoted to Queued. WorkerHub.QueuePlanningSubtasksAsync exposes this to the UI; TaskRunner advances the chain on completion. Also tightens the planning-session prompt: planner must use MCP tools, not direct edits.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Remove premature RunCreated broadcast from WorkerHub.RunNow and the
duplicate calls in RunAsync retry block and ContinueAsync. RunOnceAsync
now owns the broadcast for every run, fired immediately after the row
insert so the UI never receives an event for a non-existent row.
Worker and App Program.cs: replace SqliteConnectionFactory+SchemaInitializer
with AddDbContextFactory<ClaudeDoDbContext> + Database.Migrate(). Repos
changed from AddSingleton to AddScoped.
All singleton services (QueueService, StaleTaskRecovery, WorktreeManager,
TaskRunner) and singleton ViewModels (MainWindowViewModel, TaskDetailViewModel,
TaskListViewModel, TaskEditorViewModel) now take IDbContextFactory<ClaudeDoDbContext>
and create short-lived contexts per operation.
Test infrastructure: DbFixture now uses EF migrations instead of SchemaInitializer;
all test classes create contexts via DbFixture.CreateContext().
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- claude process: run stdout/stderr reads without ct; rely on
kill-on-cancel closing the pipes to unblock them — previously
ReadLineAsync(ct) could hang, stalling task slots and shutdown
- task runner: terminal db writes (task_runs, MarkDone, MarkFailed,
SetLogPath) now use CancellationToken.None; RunOnceAsync catches
OCE and finalizes the run row so ContinueAsync can resume
- task repository: GetNextQueuedAgentTaskAsync is now a single
UPDATE ... RETURNING statement — closes TOCTOU window where two
loop iterations could dispatch the same queued task
- queue service: dispose CancellationTokenSource in slot-completion
ContinueWith to stop leaking wait handles
- git service: register ct.Kill(processTree), drain reads without ct,
always reap via WaitForExitAsync(None) — no more git zombies on
cancelled worktree ops
- worktree manager: branch name uses full task id (dashes stripped)
instead of 8-char prefix, eliminating collision risk
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Per-task checklist backend: subtasks table with CASCADE delete,
SubtaskEntity + SubtaskRepository (connection-per-op, async), DI
registration in App and Worker, TaskRunner composes a '## Sub-Tasks'
markdown block into the Claude prompt when subtasks exist.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
mika kuns