Neil
fa3710df5a
Backlog chunks 12-17: sidebar, hooks and four renderer/lib chunks, six auditors at 84 files each. 38 case declarations removed across 27 files (56 executed cases, one deletion was a 13-entry `it.each`), 1 test file deleted, 710 lines gone. No production code touched. The theme this wave is a test whose own scaffolding makes the decision it claims to check: - `useAutoAckViewedAgent.test.ts` — its `runAutoAckScan` helper reassembles the hook's scan loop, calling `resolveAutoAckTabTargets`, `createTerminalAttentionSurface`, `resolveViewedUnreadSubjectKey`, `shouldClearWorkspaceAttention` and `applyAgentAttentionAcknowledgement` in production's order. The test, not the hook, decides the outcome. Owner drives the real hook through `renderHook`/`rerender`. - `WorktreeList.lineage-agent-expansion-coupling.test.tsx` — a CONTROL case that was green both before AND after the fix it brackets. Expansion now lives in a module-level cache (`worktree-card-agents-expansion-state.ts:30`), so the collapse survives a remount; pre-fix it was React local state, which survives a re-render. The two arms it claims to contrast never reached different code. - `agent-paste-draft.test.ts` — a self-comparison over lazily-chunked draft arrays. - `remote-workspace-session-merge-local-survival.test.ts` — its own comment concedes the tab survives, and the assertion loops `expect(['agent','closed']).toContain(tab.id)` over a fixture containing only those two ids, so it cannot fail except on a fabricated id. Also removed: duplicate invocations where production reaches the asserted branch before the varied input is read (`resolve-zoom-target.ts:43-45` returns `'ui'` for browser tabs before any focus signal); a telemetry fallback already covered by the `it.each` row for the one call site that names no source; and `lazy-with-retry.right-sidebar-syntax-error.test.ts` whole, whose two cases drive one branch with `')'` vs `']'` while the owner cites the same crash report (e08749bb) and asserts the same pair. Two corrections to my own guidance came out of this wave, both from auditors measuring what I told them: 1. The widening-cast "detection signal" I introduced last wave is NOT greppable. An auditor checked all 200+ `as never`/`as unknown as` hits in its chunk and found zero function-signature widening; repo-wide, `as unknown as` matches 192 renderer/lib test files and the narrow `as (...args: unknown` form matches 26 that are nearly all the legitimate idiom for forwarding to a real implementation while partially mocking a module. Whether the cast target is the SUBJECT or a COLLABORATOR is semantic, not syntactic. Downgraded to a reading aid. 2. That is the ninth syntactic proxy I have proposed for a relationship that exists only between a test and its production counterpart, and the ninth to fail measurement. No such proxy exists; reading the case against production is the method. Two new defect classes recorded, both invisible in the test's own text: a FIXTURE that reimplements the production rule (`runtime-session-mirror-unverifiable-host.test.ts:47` builds inputs with the same `verification === 'verified' && !retired` expression production uses at `runtime-status-snapshot.ts:28`, so the test keeps passing against a stale premise if the rule changes — already diverging, since a sibling projection adds a third condition); and the stale CONTROL above. With anti-vacuity, that is three ways a test can silently stop testing what it claims while staying green. Coverage is partial and stated as such: 54 to 84 of 84 per chunk, every auditor listing its unread paths. `renderer/lib` files are larger than earlier areas, which is where the gap comes from. Verified: 1,076 test files / 9,765 cases pass across the touched areas; `check-reliability-gates.mjs` 140 gates; the deleted file is absent from the gate manifest, `cloud/package.json` and `mobile/tests-typecheck-baseline.txt`.
Orca
中文 · 日本語 · 한국어 · Español · Français · Português
The AI Orchestrator for 100x builders.
Run Codex, ClaudeCode, OpenCode or Pi side-by-side — each in its own worktree, tracked in one place.
Download Orca
Features
Mobile CompanionMonitor and steer your agents from your phone — get notified when an agent finishes and send follow-ups from anywhere. |
 |
Parallel WorktreesFan one prompt across five agents, each in its own isolated git worktree — compare the results and merge the winner. |
 |
Terminal SplitsGhostty-class terminals with WebGL rendering, infinite splits, and scrollback that survives restarts. |
 |
Design ModeClick any UI element in a real Chromium window to send its HTML, CSS, and a cropped screenshot straight into your agent's prompt. |
 |
GitHub & Linear, NativeBrowse PRs, issues, and project boards in-app — open a worktree from any task and review without a context switch. |
 |
SSH WorktreesRun agents on a beefy remote box with full file editing, git, and terminals — auto-reconnect and port forwarding included. |
 |
Annotate AI DiffsDrop comments on any diff line and ship them back to the agent — review, edit, and commit without leaving Orca. |
 |
Drag Files to AgentsVS Code's editor with autosave everywhere — drag files or images straight into an agent prompt. |
 |
Orca CLIAgents drive Orca too — script every workflow with |
 |
Also in the box:
- Quick open — Search across worktrees, files, agents, commands, and repo context without leaving your flow.
- Account switcher & usage tracking — See Claude and Codex usage and rate-limit resets, and hot-swap accounts without re-logging in.
- Rich repo previews — Preview Markdown, images, PDFs, and repo docs in the workspace.
- Computer Use — Let agents operate desktop apps and visible UI when a workflow needs real interaction.
- Notifications and unread state — Know when an agent finishes or needs attention, then mark threads unread to come back later.
- And many, many more — we ship daily, so this list is perpetually behind. The changelog is the real feature list.
Supported Agents
Works with any CLI agent — if it runs in a terminal, it runs in Orca.
Claude Code
Codex
Grok
Cursor
GitHub Copilot
Muse
DeepSeek Harness
ZCode
OpenCode
MiMo Code
Amp
OpenClaude
Antigravity
Pi
oh-my-pi
Hermes Agent
Devin
Goose
Auggie
Autohand Code
Charm
Cline
CodeBuddy
Codebuff
Freebuff
Command Code
Continue
Droid
Kilocode
Kimi
Kiro
Mistral Vibe
Qwen Code
Rovo Dev
+ any CLI agent
Install
Desktop — macOS, Windows, Linux
- Download from onOrca.dev
- Or grab a build directly: macOS Apple Silicon · macOS Intel · Windows (.exe) · Linux AppImage · All builds
- Running
orca serveon a headless Linux server? See the headless Linux server guide.
Or via a package manager:
# macOS (Homebrew)
brew install --cask stablyai/orca/orca
# Arch Linux (AUR) — or stably-orca-git to build from source
yay -S stably-orca-bin
Mobile Companion — iOS, Android
Pair with your desktop app to monitor and steer your agents from your phone.
- iOS: Download on the App Store
- Android: Download APK 0.0.50 · Install guide
Community & Support
-
Discord: Join the community on Discord.
-
Twitter / X: Follow @orca_build for updates and announcements.
-
WeChat: Scan to join the Orca community WeChat group 11.
-
Feedback & Ideas: We ship fast. Missing something? Request a new feature.
-
Privacy: See the privacy & telemetry docs for what anonymous usage data Orca collects and how to opt out.
-
Show Support: Star this repo to follow along with our daily ships.
Developing
Want to contribute or run locally? See our CONTRIBUTING.md guide.
The relay that pairs the mobile app with a desktop host is also in this repository under
cloud/, with a separate pnpm workspace and setup guide.
Signed Builds
Windows code signing sponored/provided by SignPath.io, certificate by SignPath Foundation.
License
Orca is free and open source under the MIT License.