* fix(sidebar): keep agent rows ordered by their session start once history wraps
Sidebar agent rows are ordered by the session's first reported state, read as
`stateHistory[0].startedAt`. `stateHistory` is a rolling window capped at
AGENT_STATE_HISTORY_MAX that trims oldest-first, so past ~20 transitions that
row is a recent turn: the row's identity clock silently becomes an activity
clock and busy agents re-sort themselves on every turn.
Latch the session origin in `firstStateStartedAt` when the entry is built and
read it through `agentEntrySessionStartedAt`, keeping the history lookup as a
fallback for entries that predate the field.
* test(sidebar): cover row order through activity and retention
* fix(sidebar): preserve session row order through activity and retention
* test: complete paired terminal status fixtures
* fix: carry accepted session origins across paired rows
* test: cover paired origin work and mobile action parity
* test(mobile): expect sessionLive in the phone chat identity after main's start-a-chat change
* refactor(agents): extract the title name-anchoring vocabulary
Pure move, no behaviour change. The name tokens, display labels and the
bare-name/whole-title resolvers come out of agent-title-evidence.ts into
agent-title-name-anchoring.ts so more than one predicate can resolve a name to
an agent through the same implementation rather than by convention.
Nothing is added or reworded on the way out; the extracted functions are
unchanged and simply exported. Verified by the existing suites: the title,
identity and sidebar tests all pass untouched.
The identity ratchet moves with it — titleHasAgentName's two
parser-implementation callsites now live in the new module — so this commit is
green on its own rather than leaving the ratchet red until the next one.
* fix(agents): stop a foreign agent name in task text from taking a pane
A Claude pane's WORKING title is a bare braille spinner plus task text, so any
agent name inside that text resolved the title to a committed foreign identity.
The guard that fixed the mirror-image bug was written for the Claude label only,
so every other name short-circuited above it and took the pane from its owner —
the tab icon in use-tab-agent.ts, and the sidebar/dashboard row in
worktree-title-derived-agent-rows.ts, which never consulted the owner at all.
Make the rule agent-neutral in both directions: a title may only take a pane
from a known owner when it PRESENTS that agent, never when the name merely
occurs in free-form task text.
titlePresentsAgent sits beside collectAgentTitleEvidence and resolves names
through the vocabulary they now share. It differs from the parser in one
respect: the parser asks who a title names given no owner, so it must refuse
anything a task sentence could forge; this predicate is told which agent to
check, so it also accepts that one name in an identity POSITION — the whole
undecorated remainder, or the head of a `Name:` frame. The colon is what
separates `Codex: fix cursor offsets` from `Claude Code compare Opencode`.
resolveTabAgentFromSignals exists twice, already drifted in owner handling.
useTabAgent calls the copy in use-tab-agent.ts, which is what paints the tab
icon; every existing test imports the copy in tab-agent-from-signals.ts. Both
are fixed, and the new suite runs its whole contract through both so a future
divergence fails instead of hiding.
* fix(agents): let the title label map admit a missing label
Narrowing the map's value type to TuiAgent surfaced an inaccuracy that was
always there: noUncheckedIndexedAccess is off, so indexing Record<string, _>
types every lookup as a hit, and `label` is arbitrary text parsed off a terminal
title. The compiler concluded the `?? 'unknown'` fallback was unreachable and
the unknown-label comparison impossible.
The runtime disagrees — the map carries a fixed set of ~10 labels and misses
everything else. Deleting the branch would return undefined where the function
returns 'unknown' today, so the type is what is wrong. Partial<Record<...>>
makes absence expressible and leaves behaviour unchanged.
Pinned by a test, since nothing else forced the map to be honest about this.
* fix(agents): let an agent's own sigil still present it against an owner
titlePresentsAgent read only anchoredNames, so it missed the evidence class the
parser ranks highest: a vendor marker, the sigil an agent emits itself and task
text cannot forge. Gemini's four glyphs and Cursor's native literal are markers,
not anchored names, so every Gemini working, idle and permission title — and
Cursor's literal — failed the gate and the pane stayed on its previous owner.
Run Claude, let it finish, run gemini in the same pane: tab and sidebar both
stayed Claude. The bug this change exists to fix, mirrored.
Admitting the whole vendor class would have reintroduced #8940 the other way,
because Claude's '. ', '* ' and idle sigil are generic — OpenCode emits them
too — so they prove activity, not identity. hasGenericClaudeStatusPrefix already
modelled that distinction inside terminal-title-agent-type.ts; it moves to
agent-title-core.ts so both callers share the one definition.
Not isGenericClaudeStatusClaim, which is that model plus a clause for a
different question: it treats a 'claude' token as upgrading a generic prefix
into a real claim, which is right when asking whether a title may MINT Claude
identity and wrong when asking whether it PRESENTS Claude. Applied here it would
have let ". port the claude prompt" reclaim an OpenCode pane.
The predicate is now one memoized set per title rather than a memoized first
line and an uncached scan, so the common path — a title presenting nobody — no
longer re-runs the name matchers over every wrapper segment on each render.
* fix(agents): gate on one identity-frame grammar, ratcheted against the corpus
Two review rounds found the same class of defect: the gate was a second model of
"does this title present an agent", poorer than the model its callers use. Round
1 missed an evidence class, round 2 missed four whole ROUTES in a different
function. A third patch would have bought the fifth route.
getAgentLabel mints identity through routes the evidence parser cannot express,
because titleHasAgentName asks whether a name occurs ANYWHERE and never where.
One Codex branch therefore serves both 'codex working' and the mid-sentence
mention this change exists to reject. Route identity cannot separate them; the
gate genuinely needs position information the resolver discards.
So the models cannot be collapsed. What they can be is one GRAMMAR plus an
enforced agreement:
agentForIdentityFrame generalizes what isClaudeIdentityFrameSegment has always
described — a name, an optional status word, an optional '- action required' —
which was agent-neutral in shape and applied only to Claude. That was the same
one-way-guard mistake as the original bug, one level down. Generalized, it
covers every minting route: bare names, 'Codex: ...', em-dash frames, name plus
status, Windows launcher forms, '|'-headed titles and Pi's native format. The
positional and vendor-marker rules fold into it rather than stacking beside it.
terminal-title-pane-claim-corpus.test.ts is the enforcement. It sweeps the whole
classification corpus, and any title getAgentLabel still mints an identity from
must remain able to claim a pane. Shared code cannot enforce this because the
two models legitimately differ; the corpus can, and it names the offending title
when a future route is added.
Corpus: 14 lost claims before this commit, 0 after. One intended gain,
'claude.bat working', listed and explained in the ratchet: the old gate accepted
'aider.ps1 ready' for aider and refused the identical form for Claude purely
because the guard was Claude-only.
* fix(agents): stop reading a wrapper prefix as an identity frame
The head grammar treated '|' as a separator an agent puts after its own name.
Wrappers PREFIX the pane's title and the innermost title comes LAST
(terminal-title-wrapper-segments.ts), so a name before a '|' is a tmux window or
an ssh host, never the pane. `claude | zsh` on a Codex-owned pane therefore
handed it to Claude through all three consumers — this change's own bug,
inverted. Wrapper-joined titles already reach the real pane title through
getEvidenceTitleSegments' suffix splitting, which is what actually resolves
`ssh host | opencode ready`; the separator was not carrying that case.
'>' goes too. Pi's native `π > session` is matched before the head grammar runs,
so the only titles it affected were `name > text` breadcrumbs, which are closer
to task text than to an identity frame. That narrows `⠋ gemini > next steps` and
`⠋ codex > plan` relative to the previous gate; both are pinned as intended
rejections, since refusing leaves the pane with its known owner, which is the
safe direction on ambiguous evidence.
The corpus gain assertion is renamed to say it is corpus-scoped. Over every
string literal in the repo the change gains more than the one title this 67-title
corpus can see, mostly the same Claude symmetry, and the test should not read as
a global guarantee. Near-misses the grammar deliberately accepts are pinned too,
so the boundary is visible and moving it requires editing a list.
* fix(agents): admit identity frames before trusting terminal titles
* fix(agents): separate title identity from readiness policy
* Preserve title frame boundaries and normalization identity
* Preserve activity token whitespace at title wrapper boundaries
* test: expose typed mailbox idle evidence source
* test: scope title projection work contract to renderer
* test: distinguish task activity from agent identity
---------
Co-authored-by: Merge Sim <sim@local>
* Keep a chat's accepted pick in what it shows and sends
A pick held while a new chat is created is committed into the chat's
active option record once accepted, through the same step a normal pick
takes, and older options reads are dropped. The host also publishes the
chat's saved effort, Fast and speed as optional status fields so another
device's pick moves the option stamp and re-reads options.
* Keep chat picks ordered and capture displayed send selection
* fix: keep chat choices in one controller with accepted input capture
* fix: pass captured Codex model input to the turn request
* Keep structured chat option pickers usable during turns
* Preserve logical selection ownership across runtime transitions
* Consolidate selection lifecycle contract coverage
* Consolidate mobile structured send imports
* Preserve legacy action outcomes and opening message identity
* Consolidate repeated selection test setup and assertions
* Fix option reset admission and preserve uncertain write notices
* fix(chat): complete option contract fixture types
* test(chat): place affinity integration in cross-layer suite
* perf(mobile): skip line stats when loading Source Control status
Mobile Changes rows never render per-file line counts, so the default
git.status line-stats pass (up to two `git diff --numstat` subprocesses
plus untracked file reads) was wasted host work on every load. Hosts
that predate the param strip the unknown key and behave as before.
Co-authored-by: zhangxaochen <3416677+zhangxaochen@users.noreply.github.com>
* test(mobile): re-record Changes-load goldens for the no-line-stats status read
Co-authored-by: zhangxaochen <3416677+zhangxaochen@users.noreply.github.com>
---------
Co-authored-by: zhangxaochen <3416677+zhangxaochen@users.noreply.github.com>
* Fix Session History resume for symlinked workspaces
* Bound history alias derivation and cover resume consumers
* Reuse equal alias values across history publications
* fix(history): verify adopted source directory before commit
* fix: preserve UTF-8 paths in directory admission probe
* fix: bypass transcript preparation for non-adopted starts
* Snapshot agent message relations and verify task display prefixes
* Show native chat agent messages in navigable cards
* Keep agent message cards derived from canonical input and sidebar state
* test: correct agent card contracts and cover raw queue editing
* test: repair agent card runtime and operation contracts
* fix: preserve queued image previews and typed test measurements
* fix: use checked worker fixture and owned sender button typography
* fix: complete agent card fixtures and native presentation
* test: preserve queued edit exports and current authorship
* test: use namespace type imports in queued edit fixture
* Preserve sender progress and consolidate card contracts
* Simplify sender and notice contract instrumentation
* Remove duplicate sender and notice measurement mirrors
* Use a template literal for the dispatch boundary fixture
* Correct sender and notice verification phase accounting
* Consolidate agent message tests around user contracts
* Restore folded agent message preview on focus exit
* Consolidate agent message regression coverage
* Consolidate agent message fixture data
* Fix display text test lint diagnostics
* Use conversation view type for mail test records
* Reuse active copy control in card workflow test
* feat(agent-session-journal): save each chat's open root turn beside its epoch
A restart could only tell whether a chat was cut mid-turn by replaying its
whole journal. journal_sessions now keeps the open root turn
(open_turn_id, open_turn_since), set by the row that opens a turn and
cleared only by a finish the provider reported, in that row's own
transaction. A turn a quit, crash or Stop cut short stays open, which is
what the next launch reads as Interrupted.
Journal schema version 5 adds the two nullable columns with a forward
migration; a new epoch starts with no turn open.
* fix(native-chat): restore chat tabs at startup without replaying any history
Startup opened every chat in the saved tab index and replayed its whole
journal into the main process to learn its status. With hundreds of tabs
and gigabytes of history, Orca froze in GC on every launch.
Startup now lists each restored chat from its record and publishes its
row as Interrupted (a turn was open when its host stopped, or the last
quit's resume offer lists it) or Done, through the status stream and the
agent-status sink, reading only each chat's journal_sessions row and its
newest row's time. A chat's history is read when something opens it; that
open settles what the gone generation left running and publishes the
journal's own row over the restored one.
No other startup path replays chats: restart offers list from their
markers without opening chats, and orchestration mail opens the chats it
must read one at a time. Opened chats are capped at 24 resident folds;
opening one past the cap closes the least recently active folds that are
only a cache, with the idle sweep's own close test.
* Fix lazy chat restoration and cache ownership
* Restore saved chat work across runtime restarts
* Wait for committed structural progress in migration fixture
* Correct restored ownership and bounded journal recovery
* Preserve bounded work selection during legacy recovery
* Complete deferred restart status publication
* Fix restart recovery test fixtures
* Use legal legacy recovery fixtures
* Correct population delete observation
* Bound legacy journal replay in a readonly worker
* Repair legacy worker proof oracles
* Reduce legacy scratch key serialization cost
* Decode population meter namespaces with scratch UTF-16 encoding
* Consolidate structured session imports for CI
* Remove redundant casts and document checked type exceptions
* Correct compiler contracts and retain journal fixture invariants
* Fix checked test fixture contracts after compiler validation
* Name population fixture identity type in exported signature
* Preserve historical quit presentation after rewind
* Fix structured session retry and remote worker delivery
* Preserve settlement and restored status contracts
* Enable legacy chat readiness on standalone hosts
* Repair structured session static checks
* Verify retained submission identity across rewinds
* Show saved chat status without replaying history at startup
* Keep chats closed when startup status finishes loading
* Keep startup owner recovery independent of history loading
* Preserve intentional quit verdicts in saved chat status
* Use a checked host fixture in startup restoration tests
* Group saved startup roster publication and verify scoped status work
* Use checked array access in status cut counters
* Complete checked saved status fixture typing
* Declare portable roster counter contract
* Use current saved chat startup APIs in contracts
* Preserve complete chat tab authority across startup
* Repair checked startup restoration test consumers
* Keep authority fixture module mocks in test entries
* Consolidate authority test fixture imports
* Check restore fixture contracts and recorded queue recovery
---------
Co-authored-by: m4air <m4air@Mac.localdomain>
Co-authored-by: Jinjing <6427696+AmethystLiang@users.noreply.github.com>
* fix(codex): launch Codex without its paste-burst guess outside Windows
Codex treats a fast run of keys followed by Enter as a paste and turns the
Enter into a newline. Keystrokes relayed from a remote client can reach the
host bunched together, so the prompt stayed in the composer as a draft.
Orca's terminals bracket every real paste, so the guess can only misfire
there. Orca now launches and resumes Codex with disable_paste_burst=true,
except on Windows, where Codex reads console input records and relies on
the burst for pastes.
Fixes#34496
* test: expect Codex launches to disable the paste burst off Windows
Refs #34496
* fix(codex): keep the paste-burst flag out of commands that end their options
Refs #34496
* fix(codex): decide the paste-burst flag from a paired server's own OS on resume
Resume quoting on a paired server keeps the client's platform, so a macOS
client resuming Codex on a Windows server would have disabled the burst that
Windows Codex needs. The resume target now carries the server's reported OS
and leaves the flag off when it is Windows or unreported.
Refs #34496
* fix(codex): add the paste-burst flag only when the command runs Codex itself
A wrapper override may not forward -c, which would break every launch.
Refs #34496
* fix(codex): leave the paste-burst flag off compound commands and unreported SSH hosts
Refs #34496
* test: leave pass-through resume input untouched in the banner spec
Refs #34496
* test(mobile): expect AI-vault Codex resumes to disable the paste burst
Refs #34496
An overdue chunk reply now asks the connection for one decisive liveness
probe instead of waiting for the idle watchdog, so the resend lands inside
the host's grace. Native clients probe directly; the OTA page asks the
shell with a new notify that older shells refuse with a diagnostic only.
* fix(mobile): resend dictation audio a dropped connection left unanswered
The host feeds chunks in seq order, each seq once, holding audio that
arrives ahead of a resend for up to the reconnect grace, and says so in
its start reply. To such a host the phone resends chunks rejected as
delivery-unknown after each reconnect, so speech sent on a dying
connection is no longer lost.
* fix(mobile): cap held dictation chunks by count as well as bytes
* fix(mobile): skip every gap needed to keep held dictation audio within the cap
* feat(mobile): stop tap-mode dictation after a minute of silence
A forgotten mic kept recording and held the screen awake until stopped. Hold mode is left alone: a finger on the button isn't a forgotten mic.
* fix(mobile): time dictation silence from the last speech window the meter measured
* refactor(mobile): simplify the dictation silence timer
- Commit-message generation keeps the host's optional errorCode, errorParams and
errorDetail and translates every known text-generation code, matching desktop's
rules: unknown codes, older hosts or missing values show the host's English
- Background tasks, subagent groups, the model label and agent-session write
notices use localized mobile adapters instead of the shared English sayers
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* Unify dedicated agent launch choices and argument badges
* Enable action model and effort menus with argument migration
* Bound argument choice edit range processing
* Preserve independent action effort through launch and migration
* Preserve literal config arguments and explicit launch permissions
* Respect repository action argument groups and share launch controls
* Remove duplicate action launch command coverage
* Unify Source Control action choices and repository overrides
* Use precise types in action choice contracts
* Consolidate action choice test imports
* Correct action choice types and integration test boundaries
* Complete prepared terminal test result contract
* Align action integration expectations with canonical choices
* test(mobile): type the file preview retry test helpers so the test typecheck passes (#35431)
(cherry picked from commit 0774fc9373)
* Commit edited action arguments before planning and launch
Reconcile deliberate raw edits once into the submitted action recipe so replacement model and effort flags reach Save and Start together. Preserve captured sources and untouched load conflicts.
* Name portable Source Control dialog fixture results
* fix(mobile): retire file save continuations when preview closes
* fix(mobile): retire file saves only on unmount, warn when Discard interrupts a save
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): retire closed-tab composer state
* Require confirmed tab retirement before composer cleanup
* fix(mobile): keep drafts of tabs the host has not listed yet
Optimistically created tabs can precede their host listing, so an accepted inventory without them must not retire their composer state.
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): release retired route storage mirrors
* fix(mobile): keep route hydration through unrelated writes and route changes
Stamp writes per owned key so an unrelated write no longer voids a cold
route's hydration into host-profile-unreadable, and retain the next
route's keys before releasing the previous route's so shared host-wide
keys stay mirrored.
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* perf(mobile): skip model projection for retired discovery owners
* test(mobile): drop spy-only model discovery projection test
Existing hook suite already covers stale-result exclusion; the new file only asserted a namespace spy call.
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* feat(mobile): show a voice meter and timer while dictating, with start and stop haptics
The phone meters the PCM it already streams with desktop's analyzer,
now shared from src/shared, so both shells get a level without a native
event crossing the page bridge. Recording start and stop tap the phone,
iOS now allows haptics during input, and "Dictation inserted" only
shows for live terminal input, where the text can be hidden.
* refactor(mobile): keep the dictation hooks under the line cap after #35502
Moves the start tap and stop tick into useMobileDictationHaptics and shares one current-finish guard in finishDictation.
* feat(mobile): localize tasks, new workspaces and source control
Moves the last ~670 strings (task lists and boards, issue and PR sources, the
new-workspace flow, and source control) to the translation API, with all ten
translations. What remains in the coverage baseline is reviewed non-copy: key
names, schema errors and log lines.
- Source control section titles recompute on a language change
* fix(mobile): align catalog terminology across screens
A whole-catalog pass per language reconciled terms that differed between batches
(host/desktop, dictation, transcripts, address form) with the glossaries and the
desktop catalogs. The page bundle verifier now reports catalog bytes separately.
* fix(mobile): keep fetched history and source error slugs language-independent
History rows are derived from the fetched page at render so relative times follow a
language switch; quality-gate follow-ups for the tasks and source control migration.
* fix(mobile): name Android notification channels in the UI language
* chore(mobile): drop catalog entries for removed source-control copy
* fix(mobile): re-localize the composer source hook on current main
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
Moves ~970 strings in session screens, the terminal, agent chat (including the
copy mobile used to render straight from shared English records), the browser pane
and the file explorer to the translation API, with all ten translations.
- Chat tool-run sentences, task lists, receipts, notices and turn status take
translated words; the shared sentence joiner accepts an optional continuation so
German nouns and Turkish casing survive mid-sentence (desktop unchanged)
- "Used the browser" vs "… 21 times" and resolved PR comment headers use explicit
keys instead of plural rules, which would drop the count for Russian 21, 31, …
- Action sheet rows carry stable ids instead of being keyed by translated labels
- Rename inputs keep the raw title rather than a translated placeholder
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* Support explicit remote branch reuse across workspaces
- Preserve selected branch identity, tracking, and busy-branch handling
- Add explicit branch-off intent with v2 runtime capability gating
* Style existing workspace action in branch reuse UI
* Preserve explicit branch reuse with versioned host support
- Gate strict reuse on a new request capability while retaining legacy behavior
- Warn when a reused remote branch cannot be configured as the push upstream
* Clarify selected branch checkout options
* Offer existing or new branch choices for worktrees
* Preserve new-branch intent when reusing a branch name
* Adapt branch composer tests and catalog after rebase
* Refresh mobile RPC recordings for branch intent capabilities
* Remove retired Russian localization entries
* Remove retired German and Turkish localization entries
* Add branch selection translations for existing workspaces
* Localize mobile worktree branch selection
* fix(mobile): refresh custom shortcuts when returning to a session
* fix(mobile): adopt accepted custom-key writes while the session is mounted
A focus or foreground reread could land between queued Settings writes and
keep an intermediate list. Subscribe to accepted writes, and fold the refresh
regressions into the existing lifecycle test fixture.
* test(mobile): type session scope custom keys without an assertion
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): retire obsolete file preview reads and grant retries
* fix(mobile): keep the post-refresh preview read as the one applied read
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): retire browser pointer queues with their pane
* test(mobile): use T[] array types in pointer retirement test
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* feat(mobile): show dictated words as you speak and insert them at the cursor
Newer desktops already send the text so far with each audio chunk reply;
the phone now shows it: grey in the chat composer or command field where
it will land, and on the live-input status row. The final transcript
replaces the selection the field had when recording started, and is
appended when the field was never focused or changed since.
* refactor(mobile): pass the dictation preview by context and split its hooks out
The preview reaches composers through a context instead of four layers of
props, and the composer's and session's dictation logic move into their
own hooks, which keeps each file under its line cap.
* fix(mobile): keep dictation selections and live words to their own terminal
* test(mobile): keep the composer dictation test typechecked and under the line cap
* fix(mobile): keep the chat field controlled while dictation previews over it
The native composer field swapped its value for rich-text children while
showing heard words, so Android could duplicate or drop typed characters and
selection offsets could drift from the draft. Draw the preview as a hidden-
from-accessibility overlay shared with the web field, and only hide the
field's own text.
* refactor(mobile): build the dictation route context beside its destination
Keeps the session dictation hook under the line cap after rebasing onto main.
* fix(mobile): land dictation where its preview showed after switching chat or terminal
The final words now go to the target captured at start: the chat it began
in (its own draft when another chat is showing) rather than whichever
surface is visible. The command field preview only uses the saved selection
in the terminal it was read from, so it shows the words appended in another
terminal, as they will land.
* fix(mobile): send terminal dictation to the chat that replaced its terminal
Routing only by the captured target dropped the words when a chat tab
replaced the terminal mid-dictation, since no terminal draft was left to
take them.
* fix(mobile): release skill scans from retired transport generations
* fix(mobile): read skill-scan transport generation when the request starts
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): refuse PDF opens from hidden pages
* test(mobile): assert one refused hidden PDF open instead of scaled loops
---------
Co-authored-by: OrcaWin <293788423+OrcaWin@users.noreply.github.com>
* fix(mobile): preserve long terminal and chat pastes
Agents can keep only the last ~1 KiB read of unframed input (#13332).
Frame long single-line chat messages like multi-line ones, and frame Live
input appends over 1 KiB (an OS paste) with the terminal's paste rule,
including held text flushed before a terminal click.
Co-authored-by: Carlos Baraza Haro <carlos.baraza@ieee.org>
* fix(mobile): end the live input session after a framed paste
Agents such as Codex collapse a framed paste into one token, so later
per-character erases from the field mirror desynced the line. A long
append to a terminal in bracketed paste mode now commits the whole field
in one frame and clears the field, recording it like other session ends
so a late native echo cannot send the paste twice; without that mode the
paste stays unframed and editable. Chat sends measure the sanitized bytes
when deciding to frame.
---------
Co-authored-by: Carlos Baraza Haro <carlos.baraza@ieee.org>
Co-authored-by: Kelvin Amoaba <97001695+AmoabaKelvin@users.noreply.github.com>
* fix(mobile): give agent questions one answering surface that fits the screen
A shown prompt card now replaces the chat composer, as on desktop. The
composer stays mounted underneath so its draft and attachments survive,
and comes back when the card collapses, clears, or dictation is running.
The card keeps the conversation visible above it (60% cap) and takes the
whole area while its reply is being typed. Question and Ask cards scroll
inside that bound and keep a focused reply in view as the card shrinks.
Co-authored-by: Neil <neil@stably.ai>
Co-authored-by: Wooseong Kim <innocarpe@gmail.com>
Co-authored-by: fancivez <fancivez@gmail.com>
* fix(mobile): open each ask step at its question
The ask card's body now scrolls inside a bounded card, so a new step must not inherit the previous step's scroll offset.
* fix(mobile): keep the queued tray shrinkable and approval actions reachable
The composer wrapper passes the remaining height down again, the queued-message tray unmounts while a card owns input so a focused editor cannot keep typing out of sight, and an approval card scrolls as a whole when its choices exceed the card's bound.
* fix(mobile): keep the composer for an approval this build cannot answer
Its choices are disabled and only a new turn can cancel it, as desktop's pendingPromptsAllUnanswerableHere rule keeps the composer for.
* fix(mobile): keep the approval card's choices pinned and the composer for a terminal-only ask
A long plan shrinks inside the approval card again instead of pushing its choices below a scroll. An ask only the terminal can answer leaves the composer (and Stop) in place, like an approval this build cannot draw.
* fix(mobile): keep the queued tray mounted under a prompt card
Unmounting it dropped its in-flight guards, so Resume could run while Clear all was still deleting. A hidden focused editor loses focus with display none.
* fix(mobile): keep the composer while dictation lands, and each ask step at its question
Transcribed text arriving after the mic stops stays visible in the composer. A focused reply is revealed only when its card shrinks, so a step that grows the card keeps its reset to the question.
* fix(mobile): keep queued work in view under a question, and end an ask step's edit on navigation
Desktop leaves queued messages and background-task controls visible while a prompt card shows; only the composer steps aside. Moving between ask steps ends the reply edit so typing can't land in a field scrolled out of view.
---------
Co-authored-by: Kelvin Amoaba <97001695+AmoabaKelvin@users.noreply.github.com>
Co-authored-by: Wooseong Kim <innocarpe@gmail.com>
Co-authored-by: fancivez <fancivez@gmail.com>
* fix(mobile): show live input caret and preserve selection edits
Display the existing input directly so its native caret shows. Accessory
Backspace/Delete edit at the field's caret, read from the input when pressed,
so edits that keep the caret (IME composition) still delete, and the caret is
restored for subsequent typing.
Port and adapt community PR #28294 to the current command dock.
Co-authored-by: Kyino <kyino1025@gmail.com>
* fix(mobile): keep live input caret keys local and the dock height fixed
Hardware Left/Right/Home/End moved both the visible field caret and the
PTY cursor, so the mirror erased at the wrong place. With field text they
now move only the field caret, like Backspace and Delete; Right/End at the
field end still reach the PTY so shells can accept an autosuggestion. On
Android the key binding lets plain caret keys through to that decision
instead of sending them to the PTY. The dictation and upload status covers
the field instead of adding a row, so starting or stopping either no longer
resizes the terminal.
* test(mobile): annotate the DOM field cast in the accessory input test
* chore(mobile): allowlist the live input field strings for localization coverage
---------
Co-authored-by: Kyino <kyino1025@gmail.com>
Co-authored-by: Kelvin Amoaba <97001695+AmoabaKelvin@users.noreply.github.com>
* fix(mobile): preserve browser typing across delayed keyboard sends
A failed send restores an old draft over newer typing, null success is treated as refusal, and rapid presses duplicate sends. Preserve newer typing, interpret acknowledgement explicitly, and fence page/client ownership.
* test(mobile): preserve browser drafts through same-computer connection cutover
* test(mobile): record browser keyboard acknowledgement semantics
* fix(mobile): warn only when a browser keyboard send's delivery is unknown
A refused send restores the text silently as before; only a lost reply or connection cutover shows the unconfirmed toast, using the app's existing unconfirmed-delivery wording.
* fix(mobile): keep a browser keyboard send's text through a reconnect
The pane remounts per page, so a reconnect that swaps the client keeps the same draft; only unmount should stop a late reply from restoring it.
* fix(mobile): warn only for written browser keyboard sends with lost replies
A cutover error already carries the delivery-unknown mark when its request was written, so the extra cutover check only warned about text that never left the phone. Classifying outside the resolved path also keeps an unconfirmed send from clearing the error banner.
* fix(mobile): let newer browser typing send while an earlier send is in flight
Claiming the draft from a ref stops a repeat tap before re-render without blocking a second send of newly typed text.
---------
Co-authored-by: Kelvin Amoaba <97001695+AmoabaKelvin@users.noreply.github.com>
Load the Hermes plural-rules polyfill from the phone app entry instead of
mobile-i18n, and transform mobile-i18n and its catalog loader without the
Expo tsconfig in root vitest.
Code blocks get a header with the language and copy button, syntax
colours from the file reader's highlighter, a line-number column and
horizontal scrolling. Short tables fill the message width, cells get more
room, quotes sit on a panel, bullets are dots, and Mermaid diagrams use
the app's borders and type.
Co-authored-by: Nico Hu <haoyu.hu@lexmount.com>
* feat(mobile): show chat images that live on the paired desktop
* fix(mobile): read every image type desktop chat shows
* refactor(runtime): share the bounded preview read and chat-image type check
* test: keep the chat-image tests typed and off importActual
Kelvin Amoaba
Brennan Benson
Neilandzhangxaochen
OrcaWinandOrcaWin
