* fix(linear): show a saved Linear credential that cannot be decrypted The new-workspace Linear search treated a decrypt failure as an empty result, so a connected workspace still said to start typing. Record the existing decrypt message and offer Settings. * fix(linear): open Linear settings without an Active Server read The settings helper read activeRuntimeEnvironmentId, which the focus-read check counts as the Active Server setting. Resolve the source row's host id instead, which is the same host the helper already used. * fix(linear): carry a Linear read's credential error to the composer The composer showed "Start typing to search Linear issues." when the saved Linear credential could not be decrypted. The previous attempt wrote the read failure into the host-wide Linear status, which describes the Active Server and every workspace on it, while composer reads run on the repo's host and selected workspace. That caused search loops, blocked healthy hosts and workspaces, and a latch that stopped retrying. Now Linear search and list reads reject with the plain decrypt message instead of resolving empty, the composer's URL lookup asks issue reads to do the same, and the composer keeps that error per request: it clears when the next read starts and shows in the Linear and Smart tabs (and the attach-a-link search) with a Settings link that opens the Linear card on the read's host. Nothing writes read failures into the shared status any more. The Tasks page keeps showing no rows plus the error, as before. * fix(linear): keep attach actions and a steady error on Linear read failures - Type the composer test's repo fixture so CI's typecheck passes. - Attach a link: a Linear read error no longer hides the status row. The Jira-URL mode switch and the Linear-URL Retry stay; only the duplicate sentence is dropped when the empty list already shows it. - Composer: the status refresh after a first credential failure re-runs the same read. Keep the error on screen through that re-read instead of flashing a loading skeleton. * fix(linear): keep the attach status row in place; share the integrations Settings path - Attach a link: keep the status row mounted (empty) when the result list already shows the Linear read error, so the Reference type button doesn't jump across the row. - Composer: Jira and Linear Settings buttons now share one path that picks the source's host with getTaskSourceRuntimeSettings, so the Linear one no longer re-implements it. No behavior change. * fix(linear): make the Linear Settings action a list row; keep Jira URLs on the Linear tab quiet - Composer: the credential error and its Settings action are one list row (red sentence plus "Settings"). It fits the capped results popover and is reached with the arrow keys like every other row; Enter on it opens Settings. It is never highlighted by default, so a plain Enter still moves to the agent picker. - Attach a link: on the Linear tab, a pasted Jira URL shows only the existing "Select Smart or Jira" status, as before; the Linear search of the URL text no longer adds a red Linear error beside it. * fix(linear): show the Linear read error as one plain line with no Settings action The Settings action inside the results list needed its own highlight and Enter rules and still could not fit the capped popover. The list now shows one compact red line, in the Linear and Smart tabs and in attach-a-link: "Orca can't read your saved Linear login. Reconnect Linear in Settings to search issues." It names the fix in the same words as the neighbouring not-connected hint. The highlight resolver, the input's Enter handling, the composer's Settings handlers and the prop plumbing are back to main. * fix(linear): show an unreadable Linear login as a Reconnect Linear empty state on the Linear tab only The composer's Smart tab no longer mentions Linear's login; the Linear tab, Edit workspace links and the Tasks page show a calm title and sentence where results would be, matching the not-connected state. * fix(linear): drive the Tasks Reconnect card from the selected scope's read; plain words in workspace notices Host-wide status flags any broken workspace, so a healthy selected workspace with no results no longer shows Reconnect Linear. Projects and Views notices say the unreadable login in plain words. * fix(linear): keep the Tasks Reconnect card up while the same list re-reads The list effect cleared its error on every re-run, so a refresh flashed the raw status bar until the read failed again. Clear it on success instead. * refactor(linear): drop the unreachable typed-row padding from the Linear tab's Reconnect state The Linear tab never has a typed-text row, so the state always uses the roomy padding. * fix(linear): derive the Tasks Reconnect card from the plain list's read error in the board Avoids widening the selection model's existing type assertion; project and view tabs report errors inside their results. * fix(linear): say the unreadable Linear login in one short line
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
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
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 · Install guide
Community & Support
-
Discord: Join the community on Discord.
-
Twitter / X: Follow @orca_build for updates and announcements.
-
WeChat: Group 11 is full. Scan to join the Orca community WeChat group 12.
-
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.










