* perf(rpc): compile Zod request schemas lazily
* chore(deps): pin zod 4.5.4 and except it from the release-age gate
4.5.4 is the first release fixing isRecursiveSchema (upstream 84e416f, #6500),
which compile() calls on every schema — on 4.5.0 it fired .default() factories
at compile time. Verified: compile-time factory calls 0 on 4.5.4, 1 on 4.5.0.
`serializes FETCH_HEAD callers before they enter admission` assumed that two
same-repo fetches join the FETCH_HEAD lock lane in call order. They do not.
`runWithGitFetchHeadLock` first `await`s `fetchLockPath`, which walks the
filesystem (`realpath`, `stat` per parent directory, `readFile` of `commondir`,
`realpath` again) before it calls `runWithGitOperationLock`, and the lane is
registered only after that walk resolves. For a non-existent `/repo` that is
five libuv threadpool round-trips per caller. Two callers issued back to back
run their chains concurrently, so lane order is threadpool completion order,
not call order.
When the `interactive` fetch won that race it entered the lane ahead of the
`background` fetch. On the first caller's release it reached admission
immediately and, being interactive, took the free network headroom slot instead
of queueing, while the background fetch stayed parked on the lock. `queued`
therefore settled at 0 and never reached the asserted 1. Measured inversion
rate for the bare lock-path walk was 54/500 on an idle machine; the test itself
failed 5/20 locally, always at the same assertion, matching the two CI failures
on unrelated PRs (#17530, #17630) at the same line.
Fix the premise rather than the symptom: stub only the key derivation, keeping
the real FIFO `runWithGitOperationLock` that the test actually exercises, so the
lane is registered synchronously with the call. Key derivation keeps its own
coverage in `src/shared/git-fetch-head-lock.test.ts`. This also stops the fetch
tests in this file from sharing one global `/.git/FETCH_HEAD` lane with each
other and from touching the real filesystem.
Verified deterministic: 40/40, then 30/30 clean runs, plus 25/25 with twelve CPU
hogs and a concurrent `src/main/git/command-runner/` run saturating the box.
* perf(wsl): warn at project-add when the tree sits on a Windows drive
Worktree placement now puts new workspaces inside the distro, but a project
whose own tree is on C:\ still pays the 9p/drvfs crossing on every git command
it runs — measured at ~20x for a clean `git status` against the same tree on
ext4. Nothing in the UI says so, so the project just feels slow.
Warn once, right after the add succeeds, naming the distro the project's git
actually runs in. The advisory is wrapped so it can never fail the add.
Two path shapes cross the boundary and both warn: a Windows drive path under a
WSL project runtime, and the UNC spelling of a distro's own drvfs mount
(\\wsl.localhost\Ubuntu\mnt\c\...), which crosses it however the runtime is set.
A tree already inside the distro, a drive path under Windows-host git, a plain
UNC share, and every POSIX/SSH path stay silent.
* chore(i18n): register the WSL filesystem boundary advisory keys in en.json
Track scroll-restore generation to invalidate stale callbacks that were
resetting the scroll position to 0 when reopening a detail page. Prevent
restoration while a detail page is open.
Update automation test to use runtime.call RPC instead of removed preload
CRUD method. Hide browser import hint in E2E profile to prevent overlay from
intercepting test setup clicks.
Validate scheduled values match current state before executing idle
timeout callbacks. Use useLayoutEffect to synchronously update refs,
preventing outdated searches when rapid keystrokes overwrite timers.
* refactor(agent-hooks): drop the unused per-agent hook status IPC surface
No renderer, CLI, or mobile caller invoked window.api.agentHooks.*Status; main
already reads install status through MANAGED_AGENT_HOOK_STATUS_READERS. The
14 handlers had also drifted (kimiStatus existed in main/preload but not in
AgentHooksApi or the web stub).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* refactor(tui-agent-config): default launchCmd and expectedProcess to detectCmd
32 of 36 entries repeated the binary name three times. Entries are now
authored in a source form where both default to detectCmd and resolved once
at module load, so TUI_AGENT_CONFIG keeps its exact shape for consumers
(verified equal to the previous table).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* refactor(mobile): derive the agent order, labels, and picker from src/shared
The mobile mirror (and its regex-over-desktop-source parity test) predates
mobile importing runtime values from src/shared, which it now does in a dozen
modules. Only the favicon-domain map stays mobile-local because desktop's lives
in the renderer catalog next to bundled ?url imports. The parity test now
imports the real registries and also checks label parity.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* test(web): align preload surface after hook IPC removal
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: Brennan Benson <79079362+brennanb2025@users.noreply.github.com>
Review follow-ups on the create-field shaping:
- Seed only `reporter`. isVisibleJiraCreateField matches every required
non-system field, so the previous filter also pre-filled required custom
user pickers (Reviewer, Requested by) that Jira never defaults.
- Skip the seed when the target project is on another site. The viewer
comes from the active site, and the host shapes against the target's
client, so a foreign accountId is rejected on Cloud and can silently
resolve to a different person by username on Server/DC.
- Render JiraUserPicker only for scalar user fields. It holds one user, so
an array-of-user field collapsed to a single member; those keep the
existing comma-separated text path, which still reaches toUserFieldValue's
array branch.
Adds docstrings across the touched Jira functions to satisfy the
docstring-coverage pre-merge check.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HdMzQN6T3jRVahCK2sNyur
- Jira rejects a bare string for reporter/user-picker fields on issue
create, so shape customFields values into {accountId}/{name} objects
for keys the caller flags via userFieldKeys.
- Seed required user fields with the authenticated viewer by default
and add a searchable user picker (jira.searchUsers) so users aren't
forced into free text for reporter/custom user fields.
* fix(agents): keep OMP identity and forward ask/approval events (STA-4130)
A live OMP pane was re-owned as Pi because the generic pi-compatible
fallback always won, ask events blocked without a question payload, and
OMP suppressed tool_approval_* unless an extension registered handlers.
Mark Pi as the title-group fallback so a specific OMP identity is not
downgraded, publish OMP ask input as the existing questions envelope, and
forward tool_approval_requested/resolved onto blocked/working.
STA-4130
Related to #14278
Co-authored-by: devatnull <59279509+devatnull@users.noreply.github.com>
* fix(agents): keep launch Pi ownership over OMP wrapper frames (STA-4130)
The pi-compatible fallback treated every generic Pi owner as inferred, so an
explicit launch-Pi pane (and a launchless Pi pane with an OMP-shaped title)
was re-owned as OMP. Launch provenance now stays authoritative; only an
inferred status-frame owner yields to a specific sibling, and same-group
titles no longer count as reuse.
STA-4130
* fix(agents): drop Pi wrapper idle titles while OMP hook is active (STA-4130)
Title-completion suppression compared pick-a-winner ownership, so a Pi ready
frame looked like a different agent than a live OMP hook and fired a spurious
task-complete notification. Reuse checks now use the title-identity group.
STA-4130
* fix(agents): restore OMP approval forwarding after merge
* fix(agents): restore title-owner API after merge
* test(agents): update identity inventory ratchet
---------
Co-authored-by: devatnull <59279509+devatnull@users.noreply.github.com>
Co-authored-by: Merge Sim <sim@local>
* fix(omp): read Pi/OMP static state-title markers and retire stale spinners
OMP 17.2.12 replaced its animated braille title frames with static markers
on WSL/ConPTY (`π : working`, `π > idle`, `π ! needs input`). Orca read all
three as idle, so a working OMP pane lost its status, and a synthetic title
spinner started by an earlier hook kept rotating after its status row was
gone.
Classify the markers from one shared table so a later upstream punctuation
change is a row, not a reparse, and stop the spinner when the hook row it
stands in for is cleared or dismissed.
Fixes#13890
* test(omp): preserve static state titles during normalization
---------
Co-authored-by: Merge Sim <sim@local>
* feat(browser): edit a page annotation's comment and intent inline in the tray
Fixing a typo in a browser annotation required deleting it and re-picking
the element in the page. Each tray row now has a hover-revealed edit
button that swaps the row to an inline editor (comment textarea seeded
with the current text, intent toggle, Save/Cancel), reusing the compose
card's submit shortcut and length budget. Edits route through a new
updateBrowserPageAnnotation store action that no-ops on a missing id and
sanitizes the merged annotation exactly like the add path.
Escape inside the editor cancels only the row edit: the annotate mode's
window-level capture listener now exempts the edit container, since
stopPropagation from a descendant cannot reach a capture listener.
Editing is deliberately limited to comment and intent: element targets,
cross-navigation persistence, and the prompt builder are untouched (the
builder reads live off the array, so edited text flows through).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_017rEaTnSaSneEMEJvebzM53
* test(browser): pin the annotate-mode Escape exemption for annotation edits
The data-slot annotation-edit clause in useGrabMode's capture keydown
listener had no test that dispatched a keydown through it, so deleting
the clause left the suite green. These tests invoke the real registered
capture handler with an Escape whose target sits inside (and outside) an
annotation-edit container; the inside case uses a button, which
isEditableKeyboardTarget does not cover, so only the exemption clause
can keep grab mode alive. Removing the clause now fails the suite.
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Neil
Jinwoo Hong
2sumtech
Jinjing
Brennan BensonandMerge Sim
m4air
Aldin SaracevicandClaude Opus 5
SahirandClaude Fable 5