Auto-Company: open-source firm of 14 AI agents runs 24/7 with one markdown file as memory
JafarNajafov · x · 2026-10-29
A new open-source project, Auto-Company, claims to run a fully autonomous AI company on your laptop around the clock.
- 14 autonomous agents, each modeled on a real expert: CEO after Jeff Bezos, CTO after Werner Vogels, inversion after Charlie Munger, full-stack after DHH, marketing after Seth Godin, product design after Don Norman
- Minimal memory: a single shared markdown consensus file replaces any vector DB; each loop the team reads consensus, forms a 3-5 person squad, executes, updates memory, sleeps briefly, repeats
- 30+ shared skills (deep research, security audit, financial modeling, SEO) loadable mid-cycle
- Built on Claude Code (default) and Codex CLI, with a local dashboard on macOS and Windows/WSL
The author says most multi-agent setups fail via endless discussion loops or heavy vector-DB context plumbing; the consensus-memory mechanism avoids both.
More from coding & agent
- Routing coding agents' questions by decision type: 762 questions, only 59 needed the owner — DynoKore · 2026-10-29
- 1.2B tokens later: multi-agent coding run rebuilds a 3D game but hits diminishing returns — teortaxesTex · 2026-10-29
- Unreal Engine runs in the browser via a single Claude agent prompt — chongdashu · 2026-10-29
- Evals and Tracing 101: A Practical Guide for Engineers Building With AI — blaizedsouza · 2026-10-29
- Stop repeating yourself to every AI: a 4-step fix for syncing memory across Claude, ChatGPT and personal agents — jonathanmalkin · 2026-10-29
- hunch: GPU-free ~20ms Jev-like decisions built on EmbeddingGemma 2 — iamrobotbear · 2026-10-29