worker-start --on <server> --worktree new-child now tells the server that
orchestration picked the parent, so the server records the child's lineage as
orchestration, as a local child is, instead of cli. Older coordinators omit the
field and keep cli. The guide now shows new-child first and new-top-level as
the no-parent choice.
* Fix the speed button's unknown, Flex and composer Escape cases
Follow-up to #40052 from its review:
- An unknown speed names no tier instead of claiming Standard.
- Escape in the speed list inside the chat options popover closes only the
list, not the whole popover.
- Only Codex's Fast tier gets the on/off chip, bolt and pressed fill; Flex
and other tiers are picked from the list. The Fast test is shared with
the main process.
- The list trigger takes the chip's pressed fill from the shared variant,
and the list has an accessible name.
* Show every agent's speed as the same Speed button in every picker
Claude's and Cursor's Fast mode now use the ⚡ Fast button under a Speed
label in the chat options popover and the select-mode pickers too, instead
of a Fast mode switch, matching Codex: one button for two modes, a list for
more.
* wip: error pill
* wip: error pill tests and strings
* wip: start failures with audio are model-load failures
* wip: tests
* fix(dictation): fresh pill per recording, keep the recording when a retry fails
* fix(dictation): retry a recording cut at the cloud cap; leave Esc alone on the error pill
* refactor(dictation): share the cloud cap and the pill's close button; fetch the catalog once
* feat(dictation): a retry types only the words after those that already landed
* refactor: separate execution route and child exit owners
* refactor: move browser data into defining feature owners
* refactor: give runtime RPC entry its shutdown implementation
* refactor: use defining process types and colocate host telemetry test
* docs(process-host): point piped child contract to Node owner
* docs(process-host): clarify both piped contract owners
* fix(remote-runtime): detect a runtime that stays connected but stops answering
The shared control socket only checked ws pings, which a relay hop or a runtime
whose request path is stuck still answers, so a host could read as connected for
hours with no updates (#39785). Ready sockets now send a cheap status probe after
an idle interval; a missed deadline retires the socket so reconnect backoff runs,
and the mark stays in diagnostics until the runtime answers again. Asleep and
manually disconnected hosts are never probed.
orcad also starts a worker-thread watchdog that logs a blocked event loop and,
when run under a service manager, exits after five minutes so it is restarted
(#34410).
* fix(remote-runtime): clear the unresponsive mark on any post-ready frame and ignore late deadlines
Review follow-ups: a reconnected socket that streams frames proves the runtime
answers, so it clears the mark and lets backoff reset; the probe deadline now
sits above the ws liveness window so a dropped link is not blamed on the runtime;
a deadline that fires late after sleep is not counted; integration tests no longer
depend on wall-clock timing.
* fix(remote-runtime): gate the heartbeat on the shared retry hold and isolate the wedged-runtime test
* test(orcad): wait for stdio close before reading the watchdog child's stderr
* feat(dictation): fixed-size dictation pill with cancel, timer and finish
* fix(dictation): abort a cloud upload cancelled while finishing; time recorded audio
* refactor(dictation): reuse Esc/IME helpers, keep the pane on the target, isolate meter renders
* fix(dictation): pause the pill timer while hidden; compile under React Compiler
* feat(dictation): remove filler words from text the phone recognizes itself
Phones that dictate with their own recognizer (Automatic or This phone)
never send text through the computer, so its filler pass never saw it.
The page's on-device speech adapter now runs the same shared text
pipeline on the phone's live and final text, in the language of the
locale the recognizer started in, and honours the computer's Remove
Filler Words switch, which the host now reports in its dictation setup.
An older host that doesn't report it leaves the switch on, its default.
* test(mobile): expect both commas to go around a mid-sentence filler on the phone too
"I think we should, uh, ship it" became "I think we should, ship it":
the filler's own comma went with it, but the one before it stayed. When a
pause follows the filler too, the comma before it now goes as well.
* feat(cli): create workspaces on environment recipes
`orca worktree create --recipe <id>` creates the workspace on a new
environment from one of the repo's environment recipes, and
`orca vm recipe list` shows which recipes it accepts.
The desktop app on the provisioning computer runs the same main-side
provisioning the composer uses (write-ahead record, per-step timeouts,
cancellation), registers the project on the environment's Orca server,
then creates (orca-worktree) or adopts (provisioned-root) the
workspace there. Named progress steps ride on long-poll keepalive
frames, so the CLI prints them to stderr; older CLIs ignore them. A
failure after the VM exists destroys it through the shared cleanup and
returns recipe_workspace_create_failed with the failed step, the
recipe failure code and whether the destroy finished. Ctrl+C aborts
the request and destroys a started environment.
Paired devices cannot start a recipe create. Hosts advertise
worktree.recipe-create.v1 only when the desktop wires the recipe
service, so orcad and older hosts are refused before anything runs.
Based on #39228 by jutaz, a port of #28439.
Co-authored-by: Justas Brazauskas <brazauskasjustas@gmail.com>
* fix(cli): report the VM outcome when connecting fails; record no cross-host parent
* fix(vm-recipes): destroy the VM when recording its environment fails; reset exitCode in CLI tests
* perf(types): keep recipe orchestration out of the runtime's type graph
---------
Co-authored-by: Justas Brazauskas <brazauskasjustas@gmail.com>
* feat(dictation): add an Auto shortcut mode where holding records until release and a tap toggles
* fix(dictation): apply the terminal shortcut policy and the shortcut recorder to held dictation presses
* test(dictation): drop casts from the shortcut mode tests
* fix(dictation): time Auto holds by the key, not a Cmd release macOS reports alone
* fix(dictation): start an Auto hold pressed while the last dictation finishes
* refactor(dictation): simplify Auto hold timing and reuse the shared IME check
* wip
* rm execessive tooltip
* Clarify deletion options and improve resource popover layout
- Show the scope of persistent deletion choices directly in the menu
- Keep resource details accessible in short windows
- Style disabled dropdown actions consistently
* Clarify detached commits and deletion guidance
* Preserve translated tooltip guidance
* refactor: use concrete browser and input owners
* refactor: bind remote runtime callers to concrete transport owners
* Use checked subscription mock argument types
* Type browser lease mock with its subscription contract
* refactor: bind sleeping VM status probe test to socket owner
* feat(setup): import setup scripts from more workspace config formats
* fix(setup): isolate imported Copilot steps and harden config parsing
* chore(setup): drop type assertions flagged by the quality gate
* fix(setup): keep the orcad bundle ASCII and skip unrewritable dev container variables
* fix(native-chat): clear /compact from the composer once it is sent
Conversation commands kept their draft until the host replied, while
messages release it as soon as they are on their way. Thread the
composer's onPending through the command dispatch so a sent /compact
or /clear leaves the input immediately; a refusal still hands it back.
Fixes#39122
* fix(native-chat): keep a command typed while a refused one was answering
A refused conversation command is handed back through the independent
draft merge, which replaces a lone command in the composer. With the
draft now released on send, a /clear typed while /compact was answering
was overwritten by the refused /compact. Return a refused command only
to an empty composer; anything typed since stays as typed, and the
refusal notice still says the command was not done.
* fix(native-chat): count images when a refused command checks for an empty composer
---------
Co-authored-by: mahedi-emon <113725370+mahedi-emon@users.noreply.github.com>
* feat(dictation): remove filler words from dictated text, per language
Spoken fillers (um, uh, äh, えーと, 음, 嗯 and their equivalents) are
dropped from dictated text before it is inserted, on desktop, the web
client and phones that dictate through the computer. The pass runs once
in the speech host's event sink, so every route shares it, and lives in
a new shared text pipeline that later passes join.
Only standalone sounds go, never words with meaning, and the comma after
a filler goes with it. The language comes from a single-language model
or the dictation language setting; an unknown language uses only
fillers that are no word in another language. Voice settings gain a
Remove Filler Words switch, on by default.
* fix(dictation): take the filler language from what the engine is told, and drop the comma before a final filler
* fix(dictation): keep 嗯 as a reply, clean all-caps transcripts, and read the sentence start in constant time
* refactor(dictation): one list of sentence-end marks in filler removal, and drop a duplicate off-switch test
* fix(dictation): write non-ASCII filler punctuation as escapes so the orcad bundle stays ASCII
ORCAD_CHILD_ENTRY_POINTS builds session-search-worker-entry.js, but
ORCAD_ARTIFACTS never listed it, so templates, the materializer, and the
remote installer dropped it and SSH-hosted session search failed with
'Session search worker entry not found'. Add the artifact and a test that
every built child entry is listed.
A suspended recipe VM kept getting reconnect attempts: shared-control
socket retries, health probes and SSH logins. Providers that resume a VM
on a new SSH login or connection woke it again, so suspend saved nothing
(#36363).
While a VM's record is suspended (or its resume failed) and no wake is
under way, Orca now makes no background dial to it:
- SSH: connect and the reconnect ladder refuse that VM's own SSH host,
which covers tunnels, managed-server checks, census and browser bridge.
- Shared control: no retry after close, and no socket open, including a
retry timer armed before the sleep.
- Status probe: answers asleep without dialing.
A wake (opening the workspace, Resume) is pending before it dials, so the
wake's own connect passes. Afterwards the server's control socket and
status are reopened.
A VM recorded as up whose SSH host Orca gave up reaching (the reconnect
ladder used every attempt) may have been stopped or reclaimed by its
provider. Before, opening the workspace did nothing for it (#34089). Now
it runs the recipe's resume and reconnects. If that fails, the record is
left as it was, so normal reconnects go on.
Fixes#36363
Refs #34089
The renderer's crew follow reconnect scheduler imported the shared-control
socket state module, which reaches the Node-only sealed-frame reader and
its node:zlib codec. Vite dev throws on that import in the web client,
the desktop renderer's crew panel chunk and the component fixtures.
Move the reconnect timer function next to the scheduler, its only user.
* Remove low-value ACP, hook and renderer test suites
* Remove low-value shared automation and agent test cases
* Remove low-value packaging and mobile probe test cases
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
Cloning always went into <parent>/<repo name>, so a same-named repository
from another owner (or any non-empty folder of that name) could not be
cloned into the same parent. Before cloning, Orca now checks the target
on the clone host. If it is taken, the clone forms show a Folder name
field with a free suggestion (<repo>-<owner>, then <repo>-2...) and clone
into that instead. RPC repo.clone and `orca project setup-clone` accept
the folder name too.
* fix(native-chat): show newly installed skills in the Codex $ menu before the first message
Resting Codex chats served the background probe's skills/list from a
5-minute cache, so a skill installed in that window showed "No matching
skills", even in a new chat. Each menu open now compares a cheap stamp
of Codex's skill folders and skill/plugin config with the one taken
before the last scan, and rescans only when it changed.
Probe requests always force a reload, and probe pools are keyed by the
catalog's launch generation instead of the whole env, whose 10 s refresh
started a second Codex on every rescan.
Fixes STA-14359
* fix(native-chat): keep the structured session hook within its line limit
* feat(cli): list and reload installed browser extensions
* docs(browser): document extension command contracts
* docs(browser): describe extension test fixture factory
* refactor(cli): validate extension reload once and trim list output
The RPC params schema owns the ID-or-path and ID format checks. Reload
drops --profile and the constant scope field, since it always reloads
every profile where the extension is enabled. List no longer returns
icon data URLs or permission lists.
* test(browser): keep five extension reload tests at the command owner
* fix(cli): route orca extension through the packaged Linux launcher
* fix(browser): refresh Settings after an extension reload from the CLI
* test(browser): type the reload fixture list without a cast
---------
Co-authored-by: Kelvin Amoaba <97001695+AmoabaKelvin@users.noreply.github.com>
The host stops the old agent (about 0.6 s for Claude, up to 3 s with slow
MCP servers) before it writes the divider, so the old transcript lingered.
Linear: STA-14361
* Move session search SQLite into dedicated workers
* test: spawn the session search test worker with the shared worker helper
* fix: pass only drain promises to the channel's drain wait
* Repair session search failure and lifecycle boundaries
* Stop indexing tool output and reclaim existing rows
* test: say why the session search worker tests need a longer timeout
* Bound search cleanup and preserve failure and digest semantics
* Use the database row type for tool purge batches
* test: drop the same-connection mid-page write case now that each page reads one snapshot
* test: prove a mid-page delete from the writer can't change that page's hits
* Repair transcript cancellation cleanup and search test runtime admission
* fix: cancel active transcript finalization and enrichment waits
* fix: close transcript handles when cancellation wins delivery
* fix: preserve scan cancellation errors across parser waits
* test: prove cold session search bypasses an active writer batch
* fix: preserve search continuation through background tool cleanup
* fix: preserve search pages during orphan reclamation
* fix: bound typo repair posting visibility probes
* fix: preserve exact search terms when visibility is inconclusive
* test: drop orphan-reclaim cursor staleness case superseded by frozen continuation
* Bound session search work before ranking candidates
* Preserve the limit notice for empty search windows
* Bound session eligibility lookups and reuse FTS scoring statistics
* Index session scopes for population and candidate searches
* Derive search status from the live index after clear
* Preserve disabled search status during settings transitions
* Share planned expressions between spelling probes and search
* Keep identifier fragments out of spelling presence checks
* Report search initialization through its pending transition
* Bound spelling vocabulary work with the search posting budget
* Preserve search fallback while bounding distinct vocabulary seeks
* fix(session-search): drain removals before the worker queue fills; keep aborted close waits from failing teardown
* fix(session-search): hold writer commands when the worker queue is full instead of stopping the writer
* chore(ratchets): baseline the session search worker's one-shot entry check
* fix(browser): show site favicons on remote host browser tabs
Host-run browser tabs mirrored to a paired client always showed the
globe: the host's session-tab snapshot carried no favicon. Hosts now
publish the active page's icon URL as an optional faviconUrl, capped at
2 KB and dropped when it points at the host's loopback. Headless hosts
pick it up from page-favicon-updated and publish only when it changes;
desktop hosts include it in the existing graph-sync key. Clients keep
their current icon when an older host omits the field.
* fix(browser): publish a remote tab icon on arrival, not on every swap
Read the headless favicon through the agent browser bridge so the
runtime stays free of Electron imports, publish only when a page gains
an icon so animated icons ride later publishes, and drop IPv4-mapped
loopback icon hosts.
* fix(browser): clear a remote tab icon when a cross-origin load starts
A load that fails on the new origin never commits, so clearing only on
did-navigate kept the old site's icon beside the failed URL. Also give
the runtime tests' bridge fakes the new favicon getter.
* test: give the session-tab bridge fakes the favicon getter
* fix(vm-recipes): record each VM before create runs so a quit or crash can't leak it
Orca now saves a "provisioning" record for a recipe VM before running its
create command, and updates it when create finishes. If Orca quits or
crashes mid-create, or create fails without returning a result, the VM
stays listed in Settings > Cloud VMs as Failed with its instance id, so the
user can run Cleanup (destroy gets the instance id) or remove it from the
list. On startup, provisioning records with no create running are marked
as interrupted.
Records without a recipe result live in a separate file that older builds
never read, so downgrading keeps working. Records also store the repo path
so Cleanup, Copy command, suspend and resume still work after the repo is
removed or re-added under a new id (#17293).
* fix(vm-recipes): keep no runtime row when create is cancelled or rejected before its script starts
* fix(vm-recipes): harden write-ahead records from review
- Never overwrite a provisioning-records file this build cannot read; refuse new records instead.
- Remove the provisioning-records file when empty instead of rewriting it on every store write.
- Only fall back to the saved repo path while it still defines the same recipe.
- Don't let Cleanup or Attach act on a create that is still running or never returned a result.
- Show Provisioning before Cleanup disabled; stamp running records with the save time.
* fix(vm-recipes): keep new store code off direct sync fs calls and render-path translate
* fix(vm-recipes): detect a missing provisioning-records file without a type assertion
* fix(remote-runtime): keep a resumed VM's device pairing while its host key is unchanged
* fix(remote-runtime): probe before dropping the transport, keep saved key spelling, leave SSH wakes as they were
* Keep inherited fork actions with the original chat
* Keep fork boundary tests aligned with inherited actions
* Keep inherited fork actions with their original chat
* Keep fork refusal tests within their typecheck projects
* Allow phone forks for supported agents beyond Claude and Codex
* Gate inherited tip forks on host original-chat support
* Require current host proof for inherited fork offers
* Bind inherited fork actions to the current host runtime
* Keep unsent browser forks definitive and offer accurate retry guidance
* Offer copied tip forks only at the fork point and name their source
* Offer forks on all completed copied replies
* Explain which original response can be forked
* feat(vm-recipes): optional per-step timeouts for recipe lifecycle commands
A recipe in orca.yaml can set timeoutSeconds (one number, or a map of
create/suspend/resume/destroy). When a step runs past its limit, Orca
stops the whole process tree the same way Cancel does and reports a
recipeTimedOut failure naming the step and the limit. Recipes that do not
set timeoutSeconds keep running with no limit, exactly as before.
Saved runtime records keep a recipe snapshot that older builds parse
strictly, so the snapshot leaves timeoutSeconds out; suspend, resume and
destroy read the current limit from the repo's orca.yaml instead.
Suspend and resume exit failures now go through the failure catalog with
the same English text.
* test(vm-recipes): fire the grandchild timeout only after the grandchild exists
* fix(vm-recipes): only reuse orca.yaml timeouts for the same recipe a runtime was created from
* Project chat-mode worker liveness from the structured host
* Keep chat reach checks out of per-publication attention and Crew rebuilds
* Install the session host before judging structured workers in worker-list and worker-show
* Keep worker-list rows to the liveness fix
Rich-text pastes now keep their hyperlinks as Markdown links, and pasting a
URL over selected words links them, as other message editors do.
Fixes#39153Fixes#39150
* fix(orchestration): wake reattached Hermes terminal workers
Record the provider-proven terminal grid on the headless attach path before
restoring the screen. Without it, trusted-screen readiness rejected idle
Hermes workers after a runtime restart and stranded both send and reply mail.
Reuse the spawn geometry policy without resizing the surviving PTY. Cover
status and question-reply wakes plus busy and draft composer holds.
* fix(orchestration): retain Hermes activity across native titles
Hermes writes a status marker followed by an arbitrary session name and
model. These titles can lack a recognized agent-name token, clearing live
activity after the mailbox pointer is typed and leaving its Enter deferred.
Classify native markers only for an already identified Hermes terminal.
Reproduce the exact OSC 1/2 titles during both status and reply delivery,
retain busy/draft holds, and update the existing partial attach fixture.
* refactor(pty): keep runtime attach seeding under the line cap
The rebased attach path pushed operations.ts over the 300-line oxlint cap. Move the size commit and headless seed into their own module so the wake behavior stays and the file fits.
* fix(orchestration): retarget Hermes title imports after the agent-status move
Upstream moved agent-title-core and terminal-agent. The rebased title helper still imported the old paths, so typecheck could not resolve them.
* fix(pty): keep a resize that lands during runtime attach
Attach waits on the daemon before committing snapshot geometry. A resize in that window already published the live grid, so the later snapshot commit must not write it back.
* fix(pty): detect an attach-time resize by cache identity
A resize that returns to the pre-attach dimensions still publishes a new cache entry. Comparing only cols and rows let the snapshot commit overwrite that live grid.
* test(pty): assert attach reflow and Hermes idle status in existing cases
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* refactor(orchestration): keep the Hermes title rule with the other title providers
* fix(pty): give a resize that landed during attach a readable mirror
Reflowing the seeded snapshot marked the grid unrepainted, so the screen
rules kept refusing it after the TUI repainted for the resize.
* fix(orchestration): write the Hermes title markers as escapes for the ASCII server bundle
* test(pty): retarget Hermes attach test imports after the orchestration module moves
* fix(orchestration): restore a reattached agent's idle status from its trusted ready screen
A screen-ruled agent idle before a runtime restart may write no title after
reattach (Hermes repaints its title only on a state change) and has no restored
hook report, so it stayed without live status and mail was never offered.
When no report restores it, the agent's own strong ready rule on the trusted,
correctly sized grid now gives it a live idle status; busy, draft and unknown
screens still give nothing. Quiet-gated rules and a grid not yet readable get
one recheck.
The attach test no longer injects a fresh title after attach.
* fix(orchestration): keep rechecking a reattached ready screen until it settles
One recheck could land inside a redraw's quiet window or before the grid became
readable, leaving a quiet-gated agent without live status. Recheck within a
bounded window instead, and cover a Cline redraw after attach.
* fix(agent-state): read agy's idle composer with background task rows as ready
agy 1.4 lists running background tasks between the composer and the hint
row, so composer_ready never matched and a reattached agy with a task
running held its mail until the task ended. A sibling rule allows up to
six task rows; the hint row still decides, so mid-turn, draft and picker
screens hold.
---------
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Co-authored-by: Jinwoo-H <jinwoo0825@gmail.com>
* fix(orchestration): refuse a reply to a finished worker's report and name the follow-up that works
* fix(orchestration): say what the action under a finished report does
last-status.json was rewritten whole (~190 KB) up to every 5 s while agents ran. Status now
lives in agent-hooks/last-status.db with one row per pane and per authority commitment, and a
save writes only rows whose serialized content changed. Routine saves wait 3 minutes; authority
changes still save at once, and stop, quit and hook-disable flush. Rows past the 7-day TTL are
no longer saved. An existing last-status.json (written by this or an older build) is imported
on start and then removed.
* perf(session-history): answer Workspace and Project from their own sessions
Workspace and Project send scopeOnly. The scanner then skips listing and parsing cwd-bucket agents (Claude, Pi, CodeBuddy) outside the scope, keeps the recency pass for the other agents but returns only its in-scope rows, and checks name-matched Claude buckets file by file when their newest transcripts name no in-scope cwd. The tab move lookup and All keep the full scan.
* test(session-history): pin that scope-only drops out-of-scope recency rows
* fix(session-history): keep every row main shows when scope-only skips bucket agents
Parse the whole recency window and keep all of its in-scope rows, bound unverified bucket files apart from confirmed ones and let their parse decide, and keep a scope's rows on screen across a depth change.
* fix(session-history): keep Pi listed for scope-only and let scoped rows stand in across scopes
OMP shares Pi's folder under PI_CODING_AGENT_DIR and Pi's listing decides which files OMP parses, so scope-only keeps listing Pi. Workspace and Project rows now stand in for any scope but All, since the renderer filter narrows them.
* fix(session-history): serve scope-only from a cached full list and NFC-fold bucket names
All and Project send the same paths, so a cached All list answers Project without a rescan. Bucket names and scope prefixes are compared NFC-folded, since CodeBuddy and Pi keep the cwd's own Unicode form.
* fix(session-history): match NFD-named buckets by encoding both Unicode forms
Claude replaces each non-alphanumeric code unit with a dash, so normalizing an encoded name can't recover an NFD cwd. Encode the NFD spelling of each scope path too, and drop the post-encoding NFC fold.
* refactor(session-history): build the scope matcher only for scope-only scans and tighten comments
* fix(claude): ask before matching Claude Code's theme in every profile
New profiles turned the sync on by default and wrote custom:orca into the
global ~/.claude/settings.json, so a fresh dev build, test copy or second
profile restyled Claude Code for an Orca whose toggle was off, and that Orca
could not restore it. Every profile now starts off and gets the one-time offer.
* docs(claude): the theme sync offer now reaches every profile
* refactor(main): organize SSH transport and managed server owners
* Remove orphan deploy readiness fixture and shrink test baselines
* refactor(wsl): import lock records from their defining owner
* refactor(wsl): use the canonical deferred update contract