Search iThub

Search tools, categories and collections. Type at least two characters.

Explore all tools
↑ ↓ to explore resultsEnter to open · Esc to close

Archon

Self-hosted workflow engine that runs AI coding agents from YAML

Visit website Save

Archon is an open source, self-hosted workflow engine that runs AI coding agents such as Claude Code and Codex from YAML workflows in isolated git worktrees.

Pricing
Open source
Deployment
Self-hosted
Platforms
Windows, macOS, Linux
23,844stars3,526forks
5views
Updated

What is Archon?

Archon is an open source workflow engine for AI coding agents, created by Cole Medin and released under the MIT license. Its site calls it the command layer for your AI coding agents. Instead of supervising every agent session, you describe how work should get done as YAML workflows, and Archon runs them with AI agents, scripts and your decisions at the points that need them. Archon is not a coding agent itself: it runs Claude Code, Codex and Pi through their own CLIs and SDKs, with the accounts you already have.

A workflow is a graph of nodes. A node can be an agent prompt, a bash command, a script, a loop, a wait for a person or a child workflow, so one workflow can plan, implement until the tests pass, ask for approval and open a pull request. Repository runs get their own git worktree by default, so several runs can work on one repository at once. The bundled sdlc pack takes an issue to a merged pull request through intake, build and merge queues.

Archon is self-hosted: one install on your machine or server, with SQLite by default and PostgreSQL as an option. It installs with a shell script on macOS and Linux, a PowerShell script on Windows, Homebrew or Docker. Runs can be started and followed from the CLI, a web console, Slack, Telegram, GitHub or Discord, and a REST API with an OpenAPI spec is available. Archon sends anonymous telemetry events by default, which an environment variable turns off.

Key capabilities

  • YAML workflows Define multi-step workflows as graphs with loops, approval gates and conditions, saved in the repository's .archon/workflows/ folder.
  • Parallel isolated runs Each repository run gets its own git worktree, so several agents can work on one repository without collisions.
  • Bundled SDLC pack Intake, build and merge queues take an issue through planning, implementation, review and CI to a merged pull request.
  • Attention queue When a run reaches a decision it cannot make, it pauses as one attention item with a recommendation and the choices.
  • Governance Approval gates, a full event log and transcript, and the cost of every run are stored in your own database.
  • Many entry points Start and follow runs from the CLI, the web console, Slack, Telegram, GitHub or Discord.
Advertisements

Where Archon fits

  • Agent orchestration Runs Claude Code, Codex and Pi behind one command layer, and you can switch providers per workflow or per node.
  • Plugin system Workflow packs, providers, forges and chat platforms plug in through a published contract and install with archon plugin add owner/repo.
  • Processes beyond code Workflows are not tied to code, so content production, research, operations and support processes fit the same engine.
  • Triggers Start queues by hand or bind them to a schedule or to issue events.
  • Deployment options Native installers for macOS, Linux and Windows, Homebrew, Docker images and source installs, with SQLite or PostgreSQL.
  • REST API A REST API on a Hono server publishes an OpenAPI 3.0 spec that tools like Swagger UI can read to generate typed clients.

Install

curl -fsSL https://archon.diy/install | bash

Reviews of Archon

No reviews yet

Be the first to share your experience with Archon.

Loading review form…

Alternatives to Archon

Compare Archon with these tools →