A steer or follow-up accepted while a turn is closing is now claimed by a
fresh turn at the driver's clean exit instead of stranding in the inbox;
cancellation and pre-step rejection still park accepted work. Continuable
subagent follow-ups accept image parts: the wire is upload-shaped, the Host
admits and persists each batch before inbox acceptance, and delivery is
refused when the child model declines image input. The queue dock renders
durable image thumbnails instead of an [image] text marker.
Fixes#3186
A title now comes from an attached session's live projection cut or a cold
one's durable checkpoint, and from nothing else. Attachment is decided by the
session store at read time, so a session that attached after the listing is no
longer answered from a checkpoint its log has moved past — the stale-title case
`api-session.list` already handles this way.
The log fold and its per-log memo are gone. Folding one title costs a whole
log, and this call sits under every keystroke; a session no projection answers
for is labeled by its id and regains its title the first time it is opened.
`lib` leaves the default exclusions: Ruby gems and many npm packages keep
sources there, and the miss would be silent and total. A traversal whose root
is unreadable now rejects instead of publishing an empty index over entries
that are still good, which is what the stale-while-revalidate path claimed but
could not do while every readdir error was swallowed. A drill marks the menu
drilled only when its edit actually reached the draft.
Refs #3154
Refs #3180
- Map an absent persistence backend to queue-item-not-found for updateQueue.
- Re-fence the published Agent after shared cold resume settles.
- Document cold queue activation and regenerate its public API catalog.
- Supersede the attached-only claim in the Web subagent conversations note.
The Host session list already carries each session's `updatedAt`, which is
the number its own rows show; reading it there keeps the two surfaces from
disagreeing and avoids making a context capability depend on the BFF
assembly that owns the `sessionListMetadata` projection. A session the list
does not carry falls back to the candidate's creation time.
Refs #3154
Session candidates labelled from projection checkpoints instead of a full
log fold per keystroke, with the uncheckpointed remainder folded once and
memoized while its log stays cold. The file index keeps answering while an
invalidated traversal rebuilds behind the caret, and its default exclusions
now cover build outputs so deep sources stay reachable.
Rows carry only what distinguishes them: a file names its parent directory,
a session names its workspace only when that workspace is not the current
one, and a drilled listing names none because its new breadcrumb does.
Resolves#3180
Related to #3154
The provider-card seat's keyConfigured now derives from the reference the
page would use — the profile's apiKeyEnv, or the page's derived
<ROUTE>_API_KEY while the profile names none — so the add-provider draft
agrees with its own editor about an existing conventional credential (the
store joins the derived describe in the same batched call, as
ProviderRow.derivedCredential). ModelsSectionProps makes the renderSlot seat
required so a direct render that forgets it fails to compile; the one such
render in provider-form.client.spec regained a real mount and the test
boilerplate collapsed to renderSlot={() => null}. The extension-slots Agent
Note now states the keyed cell's real override rule (same priority throws,
a different priority shadows), and docs/subsystems/slots.md carries the two
new seats in its hierarchy, both languages.
The executor seam collapses run()/start() (and the planned begin()) into one
resolve() + execute(): the handle IS the live process, and the foreground
result, the background cursor, and the deadline's promotion offer are
projections over the same spawn. ShellExecSpec.onExpiry ('kill' default /
'offer' / 'none') replaces the implicit method split; one stdout budget applies
everywhere, the 'start ignores timeoutMs' wart dies, and sync spawn throws are
contained into the same settled-killed handle as async rejections (result()
rejects with the original error, identity preserved for the sandbox
classification).
The promotion offer settles exactly once and never rejects — offer at an
'offer' deadline while running, undefined when the process settles first — so
'await execution.promotion' alone distinguishes the outcomes. An offer must be
answered synchronously; an unanswered one auto-declines back to the
kill-on-timeout behavior. accept() ends the deadline and detaches the caller's
abort signal; decline() kills and classifies timedOut.
tool-bash and tool-pwsh promote on timeout by default (promoteOnTimeout: true,
inert without run_in_background or a job registry): at the offer the running
handle registers as a job with the same hooks as run_in_background, mirrors
into ctx.activities (the web task list streams it, stop control included), and
the call returns { kind: 'promoted', jobId, timeoutMs, output } carrying the
output so far — the job cursor continues exactly after it. Any promotion
failure declines and falls back to the plain timeout kill. Descriptions and
the timeoutMs schema text advertise the semantics only when they hold;
recorded-session and web snapshot fixtures re-recorded with them.
Also repairs the four CI failures from the human-kill push: ui-activity's
missing dsh-session peer declaration (static + windows observational), the
observe owner-arm/outer-catch and killJob-stub coverage gaps, tool-workflow's
owned-run arm, and the timing-dependent projection-cache interval flush test
(now poll-based for coarse Windows timers).
Peer evidence and rejected alternatives (Codex's session model, a third seam
method, a second threshold knob) in the Agent Note
2026-08-26-shell-execute-timeout-promotion.
The web task list gains a two-press stop control on running job rows, built on
one seam change: JobRegistry.kill now takes { reason?, reported? } instead of a
positional reason. reported defaults to true (the model's job_kill claims the
terminal report exactly as before); the web path passes reported: false because
a human kill has no model-visible channel, so the settlement flows through the
existing tool-jobs completion notice (wakeup/inject, budgets, truncation all
unchanged). A recorded kill reason merges into a killed settlement's detail
(producer facts first: 'signal: SIGTERM; cancelled by the user'), so the notice
and the web row both say who stopped the work.
- session-controller: session.killJob Remote beside cancel, live-only Agent
lookup, job-not-found/jobs-unavailable rejections joined to the closed RPC
error table; ClientSession.killJob and the ISession face carry it to the UI.
- ui-activity: arm/confirm stop button (3s disarm, pending disable, failure
hint), rendered only on running job rows; standalone activities keep no kill
handle and the README limitation now says so.
- The background-job-list web e2e now drives the whole human path through the
button; goldens show the merged killed detail end to end.
- Agent Note 2026-08-26-human-job-kill records the decision against Kimi's
suppress-own-notification precedent and Claude Code's silent-kill failure;
the 2026-08-08 blocker note and the unified-task-list note point at it.