## Ziel
Jede Mission-Control-Session (Ad-hoc UND Task-Session) soll ihre Arbeit an eine frische Session
uebergeben koennen: "gib das ab an eine neue Session". Die alte Kachel wird geschlossen, an
derselben Position oeffnet eine frische Session im selben cwd, die mit einem Brief startet, den
die abgebende Session selbst geschrieben hat.
Motivation: eine lange Session mit vollem/verrottetem Kontex
ClaudeDo-Task: 8b846deb-ad10-4243-9272-9aea2a01ec93
feat(worker): Handler-Session vor dem Usage-Limit sauber uebergeben (#275); Konflikte mit start_list_handler aufgeloest, WorkerLog-Aufruf im Fire-and-forget-Pfad wieder abgesichert
Eine PowerUse-Handler-Session soll am 5h-Limit nicht mehr mitten im Merge
sterben, sondern vorher selbst sauber uebergeben:
- HandlerUsageGuard (neu) warnt die laufende Session genau einmal pro Runde,
sobald FiveHour.Utilization den Schwellwert PowerUseHandoffPct erreicht.
Die Warnung wird per HandlerUsageWarning-Broadcast an Mission Control
geschickt und dort in das Pane der Handler-TaskId getippt.
- handoff_list_handler broadcastet den Handoff dann nicht sofort, sondern parkt
ihn bis ResetsAt + 60 s; das Ergebnis traegt deferredUntil. Ohne ResetsAt
wird wie bisher sofort gesendet.
- Der geparkte Handoff liegt im HandlerRoundState (handler-rounds.json) und
wird vom Guard-Tick ausgeloest, also auch nach einem Worker-Neustart wieder
eingeplant bzw. sofort gesendet, wenn er ueberfaellig ist.
- Review-Befund aus #272: HandlerRestartCoordinator wartet nicht mehr selbst
bis ResetsAt, sondern parkt den Neustart im selben Slot. AutoRestartUsed wird
erst gesetzt, wenn der Broadcast wirklich rausging - ein Worker-Absturz
waehrend der Wartezeit laesst die Runde nicht mehr dauerhaft tot zurueck.
Zurueckgehaltener Handoff und Auto-Neustart teilen sich einen Slot, also
oeffnet nie beides dieselbe Runde. UsageLimitProofPct entfaellt zugunsten von
PowerUseHandoffPct.
- PtyTerminalSession.SendLineAsync schreibt in die laufende Session. Gegen eine
echte claude-Session verifiziert: Text und Enter muessen zwei getrennte
Writes sein, die Enter-Bytes kommen von Terminal.GenerateKeyInput.
Agents can write into the user's daily notes, but only within limits:
max 10 agent notes a day, max 500 characters, add only — no edit, no
delete. A new from_agent column separates their lines from the user's
so the cap counts the right ones and the UI can mark them.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A child merge that came back blocked/verify_failed/untracked_collision during a
parent/children unit merge used to vanish: DrainAsync only logged it server-side,
PlanningMergeAborted carried no reason, and ApproveReview/review_task always
reported StatusMerged for a task with children regardless of the real outcome,
so a failed unit merge left the parent stuck with no visible error.
- PlanningMergeOrchestrator.StartAsync/ContinueAsync/DrainAsync now return a
PlanningMergeResult(Status, Reason) instead of void, and PlanningMergeAborted
carries that reason to the UI.
- WorkerHub.ApproveReview and ExternalMcpService.ReviewTask's approve branch
propagate the real status/reason for a parent with children instead of
hardcoding "merged" (or masking a non-conflict failure as "conflict").
- StartAsync now requires the parent to already be WaitingForReview for
improvement parents too, not just planning ones, so a stale caller can no
longer trigger a partial child merge.
- HasActiveMerge now also covers the window between the last child merging and
FinalizeParentDoneAsync completing, closing a gap where a concurrent Cancel
could race the parent's own approve-to-Done transition.
- IslandsShellViewModel.OnPlanningMergeAborted flashes the reason via
FlashFooterError instead of only clearing the external-merge banner.
Merge/verify phases now broadcast over a generic (opKey, phase, current,
total) Hub event instead of a merge-specific one, so future producers
(worktree cleanup, startup recovery, planning integration) can reuse it.
IWorkerClient.MergeProgressEvent stays as a thin forwarder for existing
consumers (elapsed seconds riding in the generic "current" slot).
The merge commit message was hand-rolled per caller ("Merge task: <title>",
"Merge <branch>", "Merge subtask") and ignored the task's commit type. Every
caller now passes a blank message and TaskMergeService fills in
CommitMessageBuilder.BuildMerge -> {commitType}(list-slug): merge <title> plus the
ClaudeDo-Task trailer; the merge modal prefills it from GetMergeTargets.
A merge whose list has a verify command holds the MergeTask call for minutes (5m46s
on this repo), during which the modal only disabled its button - no spinner, no
message, so a landed merge looked like a dead app. TaskMergeService now broadcasts
MergeProgress(taskId, phase, elapsedSeconds) for the merging and verifying phases
(re-reported every 30s) plus a WorkerLog line when verify starts; the modal shows a
spinner and the localized phase.
Threads a nextPhase parameter (wait/merge/wait_final/merge_final, validated
by the new MergeHelperPhase) from handoff_list_handler through
HubBroadcaster/WorkerHub into InteractiveLaunchSpecService, which now picks
the next session's system prompt (MergeHelperWait/MergeHelperMerge) and
model (HandlerWaitAlias/HandlerMergeAlias) from it instead of hardcoding the
old two-phase Execute prompt -- this also fixes a build break left by the
prior prompt-split task, which removed PromptKind.MergeHelperExecute without
updating its only caller.
Also sets --model/--effort/--permission-mode explicitly for every list-handler
session (Triage included) via PermissionModeResolver instead of inheriting the
CLI's ambient model and hardcoding "auto", and adds Task to the merge-helper
allowlist so the Merge phase can delegate diff reviews to subagents.
review_task/continue_merge on a planning parent always leaves conflicts in
the tree, and the UI auto-opened the resolver on every PlanningMergeConflict
broadcast regardless of who started the merge -- so a running Claude session
resolving a unit-merge conflict could race a human editing the same shared
checkout in a resolver window neither of them asked for.
PlanningMergeOrchestrator.StartAsync now takes an externallyDriven flag (set
by ExternalMcpService's MCP-driven review_task path, left false for the UI's
ApproveReview) that rides along on the PlanningMergeConflict broadcast. The
UI only auto-opens the resolver when it's false; otherwise it shows a
persistent banner with a manual "Open resolver" button, cleared on
PlanningMergeAborted/PlanningCompleted. A new GetActiveExternalConflictsAsync
query (checked against GitService.IsMidMergeAsync rather than the in-memory
flag alone) lets the UI resync the banner on reconnect instead of trusting a
one-shot broadcast that isn't replayed after a restart.
The childless single-task conflict path was checked and needed no change --
it only broadcasts the generic TaskUpdated, never PlanningMergeConflict.
The merge-helper ("Let Claude handle it") system prompt now calls a new
handoff_list_handler MCP tool once every surviving task is enhanced,
instead of continuing into Phases 3-5 in the same session -- avoiding
paying for Phases 0-2's dedupe/rewrite context on every polling round of
the run/review/merge phases.
The tool broadcasts HandoffRequested; Mission Control opens a second
ConPTY tile for the SAME handler task id (no new task, HandlerBaseCommit
untouched) running a fresh handoff brief that starts at Phase 3. The
original tile stays open. Adds PromptKind.MergeHelperHandoff,
InteractiveLaunchSpecService.BuildForMergeHelperHandoffAsync, and the
GetMergeHelperHandoffLaunchSpec hub method.
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.
The embedded ConPTY terminal replaced the in-app streaming interactive session, so
delete the dead stack: StreamingClaudeSession, InteractiveSessionService,
ProcessClaudeStreamTransport, IClaudeStreamTransport, ILiveSession, LiveSessionRegistry,
IdleSessionReaper (+ WorkerConfig.InteractiveIdleTimeoutMinutes), the WorkerHub
interactive methods + HubBroadcaster events, IWorkerClient interactive members, the
TaskMonitorViewModel composer + SessionTerminalView composer markup, and the old
'Run interactively' entry. AskUser/PendingQuestionRegistry, the autonomous path,
planning, ResumeTaskInTerminal, and all ConPTY code are kept. Localization pruned.
StreamingClaudeSession raises onQueueChanged (pending snapshot) and onUserMessageSent
(on delivery, incl. the seeded first prompt); InteractiveSessionService forwards these
as InteractiveQueueChanged/InteractiveMessageSent broadcasts. Lets the UI show queued
messages above the input and move a message into the transcript only when actually
delivered to Claude. Client events + fakes updated.
InteractiveSessionService resolves a task's list working dir + seeded prompt,
spawns a StreamingClaudeSession (claude stream-json in the list dir, model+auto
as before), registers it in LiveSessionRegistry, streams output over TaskMessage,
and broadcasts InteractiveSessionStarted/Ended (an exit watcher fires Ended). The
hub's OpenInteractiveTerminalAsync now starts this in-app session; SendInteractiveMessage
and StopInteractiveSession route to it. The external Windows-Terminal interactive
launch (LaunchInteractiveAsync / InteractiveLaunchContext / OpenInteractiveAsync) is
removed; planning sessions keep their terminal launch.
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'.
Runner stack (non-worktree path): IClaudeProcess + ClaudeProcess spawning the
CLI with --output-format stream-json, prompt via stdin, parses the final
type:"result" line into RunResult. LogWriter appends ndjson to
~/.todo-app/logs/<taskId>.ndjson. TaskRunner orchestrates DB transitions
(MarkRunning -> MarkDone/Failed) and pushes TaskStarted/Message/Finished/
Updated via HubBroadcaster. Worktree-backed lists short-circuit with a
"Slice E" failure message until git support lands.
QueueService (BackgroundService) holds two in-memory slots (_queueSlot +
_overrideSlot) guarded by a lock. Uses PeriodicTimer + SemaphoreSlim wake
signal so WakeQueue() triggers an instant pickup. RunNow throws
InvalidOperationException when override busy; CancelTask cancels the linked
CTS which kills the child process tree.
WorkerHub extended with GetActive, RunNow (translated to HubException
variants), CancelTask, WakeQueue. HubBroadcaster exposes typed push methods.
Tests: 26 pass (12 new). QueueServiceTests cover override-busy,
schedule-filter, FIFO sequentiality, cancellation, plus a FakeClaudeProcess
that blocks on a TCS for deterministic slot-state assertions.
MessageParserTests cover result extraction + malformed/non-result lines.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Mika Kuns