#7070th of 107 in Coding agents and AI editors

autonomous-company

Loop that runs Claude Code or Codex CLI as a 14-role product team

Stars
3.2k
License
MIT
Last commit
Oct 2026
Last release
Oct 2026
Language
Python
Stars, last 29 weeks+2,843
31 Mar 2026: 310 stars7 Apr 2026: 395 stars14 Apr 2026: 471 stars21 Apr 2026: 521 stars28 Apr 2026: 574 stars5 May 2026: 633 stars12 May 2026: 649 stars19 May 2026: 650 stars26 May 2026: 654 stars2 Jun 2026: 849 stars9 Jun 2026: 1,197 stars16 Jun 2026: 1,294 stars23 Jun 2026: 1,302 stars30 Jun 2026: 1,305 stars7 Jul 2026: 1,308 stars14 Jul 2026: 1,311 stars21 Jul 2026: 1,943 stars28 Jul 2026: 2,182 stars4 Aug 2026: 2,218 stars11 Aug 2026: 2,235 stars18 Aug 2026: 2,250 stars25 Aug 2026: 2,315 stars1 Sept 2026: 2,759 stars8 Sept 2026: 2,939 stars15 Sept 2026: 3,021 stars22 Sept 2026: 3,075 stars29 Sept 2026: 3,097 stars6 Oct 2026: 3,110 stars13 Oct 2026: 3,125 stars20 Oct 2026: 3,153 stars

310 on 31 Mar 20263,153 on 20 Oct 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

Runs a continuous shell loop that calls Claude Code (default) or Codex CLI each cycle with a charter, 14 expert-persona role prompts and 30+ skills. State persists in memories/consensus.md between cycles. A local dashboard and Product Center on port 8810 show cycles, usage, budgets and logs, and the repo ships 22 example apps produced by the agents.

Who it is for: Developers with Claude Code or Codex quota who want an unattended agent loop

Strengths

  • Runs as launchd or systemd --user daemon with restart, rate-limit wait and circuit breaker
  • Supports Claude Code and Codex CLI; Cursor and OpenAI-compatible adapters are optional
  • Local Product Center tracks usage and budgets, with four parallel projects by default
  • Includes 22 runnable example products with startup notes and limitations

Weaknesses

  • Consumes model quota continuously; outcomes depend on engine and budget limits
  • Team creation depends on the model; workflow is prompt guidance, not enforced
  • Supports macOS and Windows via WSL2 only; no Docker or compose files
  • Steering means editing consensus.md after stopping; mid-cycle edits can conflict

What it needs

  • no GPU
  • Needs Claude Code or Codex CLI, Python 3.10+, Git, make
  • Models: Claude Code, Codex CLI, Cursor (optional adapter), OpenAI-compatible (optional adapter)
  • port 8810

Also in Coding agents and AI editors

See all 107
Also in Coding agents and AI editors
RankProjectScore
1claw-codeRust CLI coding agent harness using Anthropic and OpenAI-compatible APIs195.2k starsMIT83 out of 100
2ArchonYAML workflow engine that runs AI coding agents through fixed dev steps23.7k starsMIT83 out of 100
3Open SWEAgent that turns issues into pull requests, reviews PRs and monitors CI11.1k starsMIT81 out of 100
#4oh-my-openagentTerminal coding agent with multi-model orchestration and git-backed memory70.2k starscustom license80 out of 100
#5gooseRust AI agent with desktop app, CLI and API, extended through MCP55.5k starsApache-2.080 out of 100