When the new-workspace draft flow targets pi (e.g. user picks a GH issue
in the smart input + selects pi as the agent), the URL was supposed to
land in pi's input box as an editable draft. It didn't, because pi has
no `--prefill` flag and our bracketed-paste-after-ready path waits for
1.5s of stream silence after `\x1b[?2004h` — pi's startup banner
(Skills/Prompts/Extensions/conflicts) prints continuously for 5+ seconds,
which keeps the quiet-timer resetting until the 8s budget expires and
the paste is dropped. Claude works because of `--prefill`; codex works
because its startup is short enough to hit the quiet window.
Fix mirrors Claude's `--prefill` semantics for pi:
- Orca's pi overlay now installs a tiny `orca-prefill.ts` extension
alongside the existing titlebar spinner. On `session_start` it reads
`ORCA_PI_PREFILL` and calls `pi.ui.setEditorText(text)`, then deletes
the env var so subsequent /new sessions don't re-prefill.
- `TuiAgentConfig` gains `draftPromptEnvVar`; pi sets it to
`ORCA_PI_PREFILL`.
- `buildAgentDraftLaunchPlan` returns `{ launchCommand: 'pi', env: {...} }`
for env-var agents; the env is plumbed through `startup.env` into
`pty:spawn` (already supported by pty-connection.ts).
The launched shell command stays a clean `pi` — no `FOO='...' pi` prefix
typed into the terminal — and the readiness race is sidestepped entirely.
Co-authored-by: Orca <help@stably.ai>
* fix(gh-project): diagnose env-shadowed gh tokens in auth errors
`gh auth refresh -s project` silently no-ops when GITHUB_TOKEN/GH_TOKEN
is exported in the user's shell — gh prefers env tokens and refuses to
modify them, exiting 0. Users follow the canned remediation, see no
error, retry, and stay stuck.
Add a one-shot `gh auth status` probe (gh:diagnoseAuth IPC) that:
- Detects env-shadowed credentials and rewrites the fix to `unset
GITHUB_TOKEN` plus a grep to find where it's exported.
- Detects missing gh install, plain missing-scope on a keyring login,
and SAML SSO authorization.
- Surfaces a tailored multi-button error UI in ProjectViewWrapper and
ProjectPicker instead of one canned 'Copy command'.
Co-authored-by: Orca <help@stably.ai>
* fix(gh-project): address review feedback
- Cross-platform shell guidance: PowerShell commands on Windows
(Get-ChildItem Env:, Remove-Item Env:, [Environment]::SetEnvironmentVariable)
via navigator.userAgent platform check.
- Use `window.api.shell.openUrl` for the docs button instead of
`window.open`, matching SidebarToolbar's external-URL pattern.
- Tighten gh auth status parser: accept single-label hostnames and
optional trailing colon; recover host from the inline 'Logged in to
<host>' line so a missed section header never silently drops accounts.
- Add tests for multi-host output and host-recovery fallback.
- Drop dead command/copy locals in ProjectViewWrapper.ErrorState by
short-circuiting the auth-error case before they're computed.
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Orca <help@stably.ai>
Persist transient Tasks page position (GitHub mode, active preset/query,
Linear preset/query) in PersistedUIState so reopening Tasks restores the
user's working context instead of falling back to defaults. Source, repo
selection, team selection, and active project keep using their existing
settings paths.
Co-authored-by: Orca <help@stably.ai>
* Fix new workspace composer focus restore
* Unify new workspace source selection
* WIP: selected source pill in smart workspace name field
Co-authored-by: Orca <help@stably.ai>
* fix(new-workspace): truncate source pill so it doesn't expand the dialog
Co-authored-by: Orca <help@stably.ai>
* feat(new-workspace): add open-in-browser button to source pill, fix vertical alignment
Co-authored-by: Orca <help@stably.ai>
* refactor(new-workspace): drop redundant kind suffix, distinct PR/issue icons, tooltips on pill actions
Co-authored-by: Orca <help@stably.ai>
* fix(new-workspace): type linked URL into agent input without auto-submit
Co-authored-by: Orca <help@stably.ai>
* fix(new-workspace): use bracketed-paste for draft URL injection so it actually appears in the agent input
Co-authored-by: Orca <help@stably.ai>
* feat(agents): per-agent draft injection strategy (codex slow paste, pi/opencode type-chars)
Co-authored-by: Orca <help@stably.ai>
* fix(agents): smarter TUI-ready heuristic + bracketed paste for codex/pi/opencode
Replaces per-agent strategy guesswork with a measured readiness check:
title-idle / non-shell-foreground stable for 1.5s / 2.5s minimum floor.
Verified against codex, pi, opencode, claude in a node-pty + xterm-headless
test rig — bracketed paste lands in the input buffer for all four.
Co-authored-by: Orca <help@stably.ai>
* refactor(agents): drop unused per-agent draft strategy abstraction
The TUI-ready heuristic in agent-paste-draft.ts works for every tested
agent (claude/codex/pi/opencode), so the AgentDraftInjectionStrategy
field, type-chars + bracketed-paste-slow code paths, and per-agent
overrides are dead. Keep the `agent` arg on pasteDraftWhenAgentReady
for future per-agent escape hatches without touching every call site.
Co-authored-by: Orca <help@stably.ai>
* feat(agents): skip draft URL injection for copilot + cursor-agent
Both TUIs open with a 'Do you trust this folder?' menu on first launch
that consumes keystrokes as menu input — pasting a URL there either
selects an arbitrary option or quits the session. Mark them with
skipDraftUrlInjection so the workspace still opens cleanly; the user
types/pastes the URL themselves once past the trust menu.
Co-authored-by: Orca <help@stably.ai>
* feat(agents): native --prefill for claude, trust pre-write for cursor/copilot
Replaces the empirical TUI-ready waits with two deterministic mechanisms:
1) `claude --prefill <text>` flag — Claude launches with the URL already in
its input box, no submit. Eliminates the readiness/paste race entirely
for the most common agent.
2) DECSET 2004 (`\x1b[?2004h`) detection on the PTY data stream for every
other agent. That escape is the protocol-level "input layer ready,
accepting bracketed paste" handshake — emitted by claude/codex/pi/
opencode/gemini/cursor-agent/copilot the moment the input box mounts.
We tap it via a sidecar subscription on pty-dispatcher (no interference
with the primary xterm handler) and paste as soon as it lands. The
8s budget is now an upper bound, not a target.
Cursor-agent and Copilot's "Do you trust this folder?" menus are bypassed
by writing the same trust artifacts the CLIs themselves write after the
user accepts:
- Cursor: `~/.cursor/projects/<slug>/.workspace-trusted` (slug = abs path
with leading `/` stripped, remaining `/` → `-`).
- Copilot: append cwd to `trustedFolders` in `~/.copilot/config.json`
(the same array the bundled `addTrustedFolder` writes).
Verified against the cursor-agent CLI bundle (versions/2026.04.17-787b533/
index.js: `_=".workspace-trusted"`) and the @github/copilot 1.0.32 bundle
(`isFolderTrusted` / `addTrustedFolder` both read/write `trustedFolders`).
Both check via realpath() before string-comparing, so the trust preset
canonicalizes too.
skipDraftUrlInjection is dropped — both agents now get the draft URL
paste once the trust menu is pre-resolved.
Tests: 24 passing across tui-agent-startup, agent-trust-presets,
pty-dispatcher routing.
Co-authored-by: Orca <help@stably.ai>
* fix(agents): wait for post-?2004h render burst to settle before paste
OpenCode emits DECSET 2004 at ~500ms during alt-screen setup, then runs
a 1.3s splash render with NO bytes on the PTY, then paints the actual
input box at ~1.85s. Pasting on the bare ?2004h signal lands during the
silent gap and the bytes are dropped.
The fix: take ?2004h as the necessary precondition, then wait for the
TUI's render burst to finish — defined as 1500ms of stream silence
after the most recent post-?2004h byte. This captures both the fast
TUIs (claude/pi/codex emit setup escapes in one burst then go quiet)
and the slow ones (opencode emits, sleeps for the splash, emits again,
then goes quiet).
Verified against opencode/claude/pi in a node-pty rig: paste lands on
the first try with the new strategy. The hard 8s timeout still caps
the wait when an agent fails to launch.
Co-authored-by: Orca <help@stably.ai>
* fix(agents): guard agentTrust IPC so stale preload doesn't crash launch
If the preload bundle is older than the renderer (a real situation in
electron-vite dev because preload changes only apply on full restart,
not HMR), `window.api.agentTrust` is undefined and the launch crashes
with "Cannot read properties of undefined (reading 'markTrusted')"
before the worktree even opens.
Guard the call sites in launch-work-item-direct and useComposerState
to skip the trust pre-write when the IPC isn't exposed, and wrap the
invoke in try/catch so an IPC error never blocks the launch — the user
just sees the trust menu and accepts it manually, same as before this
feature shipped.
Co-authored-by: Orca <help@stably.ai>
* feat(tasks): route 'Use' through the New Workspace dialog instead of yolo-create
The Use CTA on the Tasks page used to create+activate a worktree
synchronously, which surprised users — the worktree appeared in the
sidebar before they had a chance to confirm name / agent / setup. The
unified New Workspace dialog landed in this branch already supports
opening with a linked work item pre-filled (see openComposerForItem /
openComposerForLinearItem), so just route Use through it.
The launchWorkItemDirect helper stays exported for ProjectViewWrapper,
which has its own UX where the immediate-create flow is the right call.
Co-authored-by: Orca <help@stably.ai>
* test(agents): include `agent` field in autohand startup-plan assertion
Merging main brought in the Autohand Code agent test (PR #1382), which
predated this branch's addition of `agent` to AgentStartupPlan.
Aligning the assertion fixes the lone CI test failure on this PR.
Co-authored-by: Orca <help@stably.ai>
* refactor(agents): drop unused expectedProcess arg + snapshot sidecar set
Two minor follow-ups from self-review:
1. `pasteDraftWhenAgentReady` no longer reads `expectedProcess` — readiness
is gated on DECSET 2004 alone now, not on PTY foreground process. Drop
it from the signature and from the two callers (launch-work-item-direct,
new-workspace).
2. The pty-dispatcher's sidecar fan-out iterates the live Set, which is
safe against deleting the current element but not against a watcher
that synchronously subscribes a sibling. Snapshot via Array.from
before the loop. Cheap (Set is tiny) and removes the latent footgun.
No behavior change.
Co-authored-by: Orca <help@stably.ai>
* test(e2e): match the unified smart-name input's new placeholder
The CreateFromTab refactor in this branch replaced the separate "Workspace
name" Input with a single SmartWorkspaceNameField whose default-mode
placeholder is "Type a name, #1234, branch, GitHub or Linear URL". The
worktree-create e2e test was still anchoring on the old "Workspace name"
text and could not find the input.
Update the placeholder regex to match the new copy. Free-form text typed
into smart mode is treated as a workspace name by submitQuick — same
contract the test used before.
Verified locally: targeted e2e passes in 2.2s.
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Orca <help@stably.ai>
* Add support for custom desktop notification sounds
* perf(notifications): cache custom sound + restart-on-play
Avoids re-reading the configured audio file (up to 10MB) from disk and
re-transferring it over IPC on every notification. Adds a path-only
resolver so repeated dispatches with an unchanged sound skip the heavy
load entirely.
For burst handling, follows the VS Code AccessibilitySignalService /
GNOME canberra pattern: one shared HTMLAudioElement per sound, restarted
from t=0 on each play, with an in-flight guard that drops new plays
while the sound is still ringing. This self-dedupes by the sound's own
duration without any magic time constant — distinct sounds remain free
to overlap. The Test button passes force: true so an explicit user
action always plays through.
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Neil <4138956+nwparker@users.noreply.github.com>
Co-authored-by: Orca <help@stably.ai>
* Add an Appearance setting for the app/workbench UI font family. It will be independent from terminal typography and will not change Monaco, diffs, markdown editor text, or terminal panes.
* Update IDE font copy
* feat(cli): add browser tab profile controls
* feat(cli): add tab profile automation primitives
* refactor(cli): narrow tab profile automation scope
* chore: retrigger PR checks
* review: harden tab profile automation CLI
- Wait for tab re-registration after browser.tabSetProfile so a follow-up tab list --show-profile reads the new sessionProfileId from BrowserManager instead of the stale one from the previous webview
- Wait for tab registration after browser.tabProfileClone, matching browser.tabCreate, so the cloned browserPageId is operable when the CLI returns
- Short-circuit browser.tabSetProfile when the tab is already on the requested profile so we do not tear down and remount the webview for a no-op switch
- Switch TabShow.worktree from OptionalPlainString to OptionalString to match every other tab schema; empty --worktree should fall back to the active worktree, not pass through as the empty string
- Add max-lines disable to browser.test.ts (file grew past 300 lines after adding the new tab-profile and tab-show tests)
* review: fix useIpcEvents test setup for tab profile API
CI failure: useIpcEvents.test.ts threw at module load with TypeError: window.addEventListener is not a function. The chain: the rebased useIpcEvents.ts imports destroyPersistentWebview from webview-registry, which calls window.addEventListener at module load. The test stubs window via vi.stubGlobal as a plain object without addEventListener, so the typeof window check passes but the call throws.
- webview-registry.ts: tighten the module-load guard to also check that window.addEventListener is callable, so importing this module from a non-DOM-ish test env (vitest node env with stubbed window) does not throw at module load
- useIpcEvents.test.ts: add the new onRequestTabSetProfile and replyTabSetProfile stubs to all 8 window.api.ui mocks so the new IPC subscription registered by useIpcEvents resolves
* review: restore profile CRUD lost during rebase onto 1397-merged main
The rebase brought commit 3242aa27 (refactor: narrow tab profile automation scope) onto a main that already had the lifecycle CRUD from 1397. The refactor commit removes BrowserProfileList/Create/Delete types, runtime methods, RPC registrations and schemas, plus the help/specs entries, because those were the precursor versions in commit 1 of this branch. Post-rebase those removals land on the hardened versions inherited from main, breaking 1397.
Restore:
- runtime-types.ts: BrowserSessionProfile import; ProfileList/Create/Delete result types
- orca-runtime.ts: ProfileList/Create/Delete result type imports; browserProfileList/Create/Delete methods
- browser-core.ts: ProfileCreate, ProfileDelete schema imports; browser.profileList/profileCreate/profileDelete RPC registrations
- browser-schemas.ts: ProfileCreate, ProfileDelete zod schemas
- help.ts: list/create/delete subcommand lines under Browser Automation
- specs/browser-basic.ts: list/create/delete spec entries
---------
Co-authored-by: Nikolatesla-lj <Nikolatesla-lj@users.noreply.github.com>
Co-authored-by: Jinwoo-H <jinwoo0825@gmail.com>
- Scope pruneLastVisitedTimestamps per-repo so not-yet-hydrated SSH repos retain persisted Cmd+J recency instead of being wiped at startup.
- Parse lastVisitedAtByWorktreeId leniently: drop only bad entries rather than failing the whole workspace session on one corrupted timestamp.
- Route sidebar agent-tab clicks through activateAndRevealWorktree so cross-repo activation and nav history are not silently skipped.
Co-authored-by: Orca <help@stably.ai>
* fix(cmd-j): rank empty-query worktrees by focus recency
Persist a per-worktree focus-recency timestamp and use it as the primary
ordering signal for Cmd+J's empty-query Worktrees section, so SSH and
other quiet worktrees surface based on user focus rather than background
activity. See docs/cmd-j-empty-query-ordering.md.
Co-authored-by: Orca <help@stably.ai>
* fix(tests): guard lastVisitedAtByWorktreeId and mock markWorktreeVisited
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Orca <help@stably.ai>
* feat(browser): viewport-size emulation via CDP
Adds a Viewport Size submenu in the browser toolbar "…" menu.
Presets apply width/height/deviceScaleFactor/mobile/touch via
Emulation.setDeviceMetricsOverride + setTouchEmulationEnabled, and
swap the UA to a mobile iPhone CriOS string on mobile presets.
Responsive clears the override. Selection persists per-tab and
re-applies on dom-ready so it survives navigations.
Co-authored-by: Orca <help@stably.ai>
* chore(browser): rename 'Responsive' viewport option to 'Default'
Co-authored-by: Orca <help@stably.ai>
* fix(browser): harden viewport emulation — serialize, validate, client-hints
- Chain per-tab setViewportOverride calls to prevent rapid-toggle races
- Validate viewport metrics at IPC trust boundary (reject non-finite/out-of-range)
- Emit userAgentMetadata alongside mobile UA to avoid UA/CH mismatch
- Always reapply on dom-ready (incl. null) to clear stale emulation
- Persist viewportPresetId in session schema (optional+nullable for back-compat)
- Convert preset submenu to DropdownMenuRadioGroup for a11y
- Log debugger.attach failures and cover with a unit test
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Orca <help@stably.ai>
Renames the experimental pet overlay to "sidekick" with themed character
names (Claude the Mage, OpenCode the Rogue, Gremlin the Trickster).
Covers IPC channels, preload API, persisted UI state, settings flag,
on-disk userData path, components, and types.
Deletes the standalone pet-overlay design mock.
Co-authored-by: Orca <help@stably.ai>
* feat: add experimental pet overlay
Adds an opt-in 3D pet overlay pinned to the bottom-right. Gated behind
an experimental flag so three.js + GLB models stay out of the renderer
bundle for users who never enable the feature. Ships with four bundled
models plus user-uploaded custom GLBs via a pet:import IPC; a status-bar
segment provides model picker + hide toggle.
See docs/design/pet-overlay.md for the full design.
Co-authored-by: Orca <help@stably.ai>
* refactor(pet): replace 3D GLB models with 2D webp images
Co-authored-by: Orca <help@stably.ai>
* chore(pet): compress pet webp images and remove unused assets
Reduce resources/claude.webp, gremlin.webp, opencode.webp sizes; drop
the obsolete pet-overlay design doc and compress-pet-glb script now
that the GLB pipeline has been replaced by 2D webp images.
Co-authored-by: Orca <help@stably.ai>
---------
Co-authored-by: Orca <help@stably.ai>
Jinjing
Brennan BensonandOrca
Jinwoo HongandOrca
Neil
Igor Costa
SSWSer
Leynier Gutiérrez GonzálezandClaude Opus 4.7
