# Agent Orchestrator > Open-source desktop application and local CLI (`ao`) to run 10+ parallel AI coding agents in isolated Git worktrees without file conflicts or API proxying. Agent Orchestrator is an open-source desktop application that lets developers run multiple AI coding agents in parallel, each in its own isolated Git worktree. It works with any CLI-based agent including Claude Code, OpenCode, and OpenAI Codex. Agents can work on different branches or features simultaneously without conflicts. Agent Orchestrator is free, does not proxy API calls, and supports macOS, Windows, and Linux. ## When to use Agent Orchestrator Reach for Agent Orchestrator when you need to: - Run several coding agents (Claude Code, Codex, OpenCode, or any CLI agent) at the same time on one repository without them stepping on each other, each agent gets an isolated Git worktree and its own branch. - Orchestrate agent work through the desktop app and local `ao` CLI: create workspaces, launch agents with a prompt, open terminals, and track tasks. - Schedule recurring agent runs (automations) that execute a prompt on a cron-like schedule in a fresh or existing workspace. - Automatically route CI failures and review feedback to the agent session that owns the branch, so the right agent can handle failures without manual babysitting. - Review diffs, manage ports, and monitor many concurrent agent sessions from one dashboard. Agent Orchestrator is not a coding agent itself; it is the local workspace and orchestration layer the agents run in. If you are an AI agent inside an AO-managed session, use the installed `ao` CLI. To learn the product, use the Documentation section below. ## Developer resources - [Agent instructions](https://useao.dev/agents.md): when and how AI agents should use Agent Orchestrator - [Blog llms.txt](https://useao.dev/blog/llms.txt): scoped index of blog posts - [GitHub](https://github.com/Untrivial-ai/agent-orchestrator): source code and releases - [Documentation map](https://github.com/Untrivial-ai/agent-orchestrator/blob/main/docs/documentation-map.md): which repository artifacts are machine-readable contracts (OpenAPI spec, AGENTS.md, skills) versus human prose, and how CI keeps them in sync - [Full LLM context](https://useao.dev/llms-full.txt): combined AO overview, developer resources, comparisons, and FAQ ## Documentation - **[Documentation overview](https://useao.dev/docs/index.html.md)**: Learn what Agent Orchestrator does, when to use it, and where to start. ### Getting Started - [Installation](https://useao.dev/docs/installation/index.html.md): Download the AO desktop app and connect your coding tools. - [Quickstart](https://useao.dev/docs/quickstart/index.html.md): Add a project, start one worker, and follow it from task to pull request. - [Platforms](https://useao.dev/docs/platforms/index.html.md): Install and use AO on macOS, Windows, Linux, and mobile. ### Using AO - [Dashboard](https://useao.dev/docs/dashboard/index.html.md): Use the AO desktop app to supervise projects, sessions, pull requests, reviews, and previews. - [Examples](https://useao.dev/docs/examples/index.html.md): Practical workflows for desktop sessions, Chat and TUI, pull requests, reviews, previews, and mobile. - **[Guides](https://useao.dev/docs/guides/index.html.md)**: Current AO workflows for agents, parallel sessions, CI, reviews, projects, Chat, Terminal UI, and mobile. - [Per-role agents](https://useao.dev/docs/guides/per-role-agents/index.html.md): Configure separate harnesses and models for worker and orchestrator sessions. - [Parallel work](https://useao.dev/docs/guides/parallel-issues/index.html.md): Run isolated worker sessions across issues without mixing branches, terminals, or conversations. - [CI recovery](https://useao.dev/docs/guides/ci-recovery/index.html.md): How AO observes failing GitHub checks and sends focused context to the owning session. - [Review loop](https://useao.dev/docs/guides/review-loop/index.html.md): Route GitHub feedback to workers and run dedicated AO reviewer agents. - [Lifecycle automation](https://useao.dev/docs/guides/reactions/index.html.md): Practical guidance for AO's built-in CI, review, conflict, and notification behavior. - [Multi-project](https://useao.dev/docs/guides/multi-project/index.html.md): Manage repositories, multi-repository workspaces, and Scratch sessions from one desktop. ### Built-in capabilities - **[Built-in capabilities](https://useao.dev/docs/plugins/index.html.md)**: Agent, runtime, workspace, tracker, SCM, notification, and terminal support compiled into AO. - **[Agents](https://useao.dev/docs/plugins/agents/index.html.md)**: The terminal harnesses built into AO and the smaller set that also supports structured Chat. - [Claude Code](https://useao.dev/docs/plugins/agents/claude-code/index.html.md): Built-in Terminal UI and structured Chat support for the user's Claude Code installation. - [Codex](https://useao.dev/docs/plugins/agents/codex/index.html.md): Built-in Terminal UI and native app-server Chat support for the user's Codex installation. - [Cursor](https://useao.dev/docs/plugins/agents/cursor/index.html.md): Built-in Terminal UI support for the Cursor Agent CLI. - [Aider](https://useao.dev/docs/plugins/agents/aider/index.html.md): Built-in Terminal UI support for a local Aider installation. - [OpenCode](https://useao.dev/docs/plugins/agents/opencode/index.html.md): Built-in Terminal UI and native ACP Chat support for the user's OpenCode installation. - **[Runtimes](https://useao.dev/docs/plugins/runtimes/index.html.md)**: AO automatically uses tmux on macOS/Linux and ConPTY on Windows. - [tmux runtime](https://useao.dev/docs/plugins/runtimes/tmux/index.html.md): The built-in Terminal UI runtime on macOS and Linux. - **[Workspaces](https://useao.dev/docs/plugins/workspaces/index.html.md)**: Git worktrees for repositories and AO-managed directories for Scratch projects. - [Git worktrees](https://useao.dev/docs/plugins/workspaces/worktree/index.html.md): Isolated branches and files for sessions in Git projects. - **[Trackers](https://useao.dev/docs/plugins/trackers/index.html.md)**: The current GitHub-only, opt-in issue intake capability. - [GitHub issue intake](https://useao.dev/docs/plugins/trackers/github/index.html.md): Opt-in, assignee-scoped worker creation from GitHub issues. - **[SCM](https://useao.dev/docs/plugins/scm/index.html.md)**: GitHub PR, CI, review, and mergeability support built into AO. - [GitHub SCM](https://useao.dev/docs/plugins/scm/github/index.html.md): Pull request observation, lifecycle feedback, and explicit PR actions. - **[Notifications](https://useao.dev/docs/plugins/notifiers/index.html.md)**: Durable in-app notifications and Electron toasts built into AO. - [Notification center](https://useao.dev/docs/plugins/notifiers/dashboard/index.html.md): Persistent notification history and live updates in the desktop app. - **[Terminals](https://useao.dev/docs/plugins/terminals/index.html.md)**: The built-in desktop terminal backed by AO's WebSocket mux. - [Desktop terminal](https://useao.dev/docs/plugins/terminals/web/index.html.md): Session terminals streamed through the daemon's mux to Electron. ### Reference - [CLI reference](https://useao.dev/docs/cli/index.html.md): Commands provided by the current Go CLI. - **[Configuration](https://useao.dev/docs/configuration/index.html.md)**: Configure projects, agents, session interfaces, and local daemon behavior in current AO. - [Projects](https://useao.dev/docs/configuration/projects/index.html.md): Configure a registered project, its worker and orchestrator agents, setup, rules, reviews, and intake. - [Lifecycle automation](https://useao.dev/docs/configuration/lifecycle-automation/index.html.md): How AO currently routes CI failures, review feedback, and merge conflicts to the correct session. - [Connect Mobile](https://useao.dev/docs/configuration/remote-access/index.html.md): Pair the native AO mobile app through the authenticated, opt-in LAN listener. - [Architecture](https://useao.dev/docs/architecture/index.html.md): How the Go daemon, Electron app, storage, controllers, and adapters fit together. ### Help - [Troubleshooting](https://useao.dev/docs/troubleshooting/index.html.md): Diagnose desktop, daemon, agent, session, GitHub, preview, and mobile problems. - [FAQ](https://useao.dev/docs/faq/index.html.md): Common questions about the current Agent Orchestrator desktop product. - [Migration](https://useao.dev/docs/migration/index.html.md): Move from the frozen npm CLI and older AO architecture to the desktop app. - [Changelog](https://useao.dev/docs/changelog/index.html.md): Where to find current AO desktop and mobile release notes. ## Comparisons - [Agent Orchestrator vs AgentsMesh](https://useao.dev/compare/ao-vs-agentsmesh/) - [Agent Orchestrator vs Conductor](https://useao.dev/compare/ao-vs-conductor/) - [Agent Orchestrator vs Emdash](https://useao.dev/compare/ao-vs-emdash/) - [Agent Orchestrator vs Superset](https://useao.dev/compare/ao-vs-superset/) ## FAQ ### I already use an IDE like Cursor, is this for me? AO is designed to work with your existing tool. We natively support deep-linking to IDEs like Cursor so you can open your workspaces and files in your IDE. AO sits above individual tools, use whatever agent you like, AO keeps the workflow the same. ### Which AI coding agents are supported? AO works with any CLI-based coding agent including Claude Code, OpenCode, OpenAI Codex, Cursor, Aider, Goose, and 18 more. If it runs in a terminal, it runs in AO. 23 harnesses total, with per-project agent choice. ### How does the parallel agent system work? Each agent runs in its own isolated Git worktree, which means they can work on different branches or features simultaneously without conflicts. AO's orchestrator spawns workers, routes CI failures and review feedback to the right session, and lets you monitor the entire fleet from one board. ### Is Agent Orchestrator free to use? Yes. AO is free and open source under Apache 2.0. It runs as a local daemon on your machine, your code never leaves localhost. No account, no cloud, no credit card required. ### Can I use my own API keys? Absolutely. AO doesn't proxy any API calls. You use your own API keys directly with whatever AI providers you choose. This means you have full control over costs and usage.