Subagent cards keyed their auto-expand on `isActive || Boolean(resultText)`.
Silent (Auggie-style) subagents return a result_text but stream no child tool
calls, so `resultText` stays populated forever and the card rendered
permanently expanded, unlike subagents that emit child tool calls (which
collapse to their header + metadata row on completion). A render-phase reset
also wiped any manual collapse the moment result_text arrived.
Key auto-expand on the running state alone and drop the reset, so completed
subagents collapse consistently and the result text is one click away. Add
component coverage for the silent-subagent collapse, expand-on-click, and
manual-collapse-persists paths.
Co-authored-by: Kandev Agent <agent@kandev.dev>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* fix(scripts): launch dev-isolated backend via __backend subcommand
The kandev binary became a bundle launcher with the restart supervisor
(ADR 0019) and exits without KANDEV_BUNDLE_DIR when invoked bare; the
isolated dev instance must run the backend entrypoint directly.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* feat(mcp): session spawning, sibling messaging, and renameable tabs
- spawn_session_kandev spawns an additional agent session on an existing
task (default: the caller's own) via LaunchSession(IntentStart), with
optional agent_profile_id (defaults to the spawner's profile) and a
session name; the initial prompt carries spawner attribution and reply
instructions.
- message_task_kandev accepts an optional session_id to target a specific
session, enabling same-task sibling messaging. Explicit targets are
pinned through dispatch so the idle-path primary-session re-resolution
cannot reroute a sibling message back to its sender (regression-tested).
- task_sessions gains a name column (migration ordered after the
agent_execution_id-drop rebuild), a session.rename WS action, and inline
tab rename UI; custom names take precedence over derived agent/model
labels across the tab, agents dropdown, mobile list, chat header, and
dockview panel titles.
- Inter-session messages stamp sender_session_name into metadata so the
sender badge shows which session sent the message.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(mcp): address review findings on session spawn/rename/messaging
- include name in task-session DTOs (list + full, no omitempty) so
persisted tab names survive list hydration and clears propagate
- sanitize sender task/session ids embedded in <kandev-system> blocks
- reject cross-workspace spawn targets; require sender_task_id
- truncate session names by runes, not bytes
- distinguish not-found from internal errors in message target lookup
- guard TabRenameInput blur-commit after Enter/Escape resolution
- cap optimistic rename length client-side and toast on rename failure
- add tests: spawn launch-failure and cross-workspace paths, renameSession
API client
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix(mcp): round-two review fixes for spawn/rename lint and hardening
- extract spawn validation/authorization helpers (cyclop/nestif limits)
- make stripSystemTag replace until stable to block nested tag evasion
- load session before rename so a durable rename can never skip its
broadcast; surface lookup failures to the caller instead
- drop redundant sessionLauncher nil check (guarded registration)
- add maxLength to the rename input matching the server cap
- add tests: nested sanitizer evasion, spawn rename-failure tolerance
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: Kandev Agent <agent@kandev.dev>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
* feat: allow multiline custom answers in clarification overlay
Swap the single-line custom-answer input for an auto-growing textarea so users
can type multi-line replies to ask_user_question. Enter submits, Shift+Enter
inserts a newline; the box grows to ~6 lines then scrolls. Adds unit and e2e
coverage.
* fix: address review feedback on multiline clarification input
- Always preventDefault on desktop plain Enter so an empty draft can't leak a
stray newline before the trim guard (Claude/CodeRabbit).
- On coarse-pointer devices let Enter insert a newline and submit via the
existing Submit button; hide the keyboard hints on touch (mobile parity).
- Preserve newlines in the resolved clarification history message via
whitespace-pre-wrap + items-start.
- Strengthen e2e to guard against fused lines; add Ctrl+Enter and touch tests.
* fix: touch submit button and IME guard for clarification input
- Guard against submitting on the Enter that confirms an IME candidate
(e.nativeEvent.isComposing) on CJK keyboards.
- Add an inline Send button on coarse-pointer devices so single-question custom
answers (which have no overlay Submit button) have a touch-reachable send
path. Fixes the regression from switching touch Enter to newline.
* refactor: extract clarification custom-input controls
- Pull the Enter/⇧↵ hints and touch Send button into a CustomInputControls
subcomponent to shrink ClarificationCustomInput.
- Compute the trimmed draft once and reuse it across the keydown handler and the
Send button (disabled state, click, styling) so the conditions can't drift.
- Add a test asserting the touch Send button is disabled while isSubmitting.
* fix: ignore auto-repeat Enter in clarification custom input
A held Enter key emits auto-repeat keydown events; guard the handler with
e.repeat so it can't fire onSubmit more than once. Adds a regression test.
* fix: suppress newline on repeat Enter and add mobile clarification e2e
- Repeat Enter on desktop now calls preventDefault before bailing, so a held key
can't leak newlines into the current or next question's textarea after the
carousel advances; it still never re-fires onSubmit/finalize.
- Add a mobile-chrome (Pixel 5) e2e covering the touch path: Enter inserts a
newline and the inline Send button submits the multiline answer.
* test: consolidate clarification e2e seed helper
Extract the shared seed → navigate → waitForLoad flow into
e2e/helpers/clarification.ts and delegate from the clarification, resize, and
mobile specs instead of duplicating it in each.
Clément Fiere