Gas Town is a sophisticated multi-agent workspace manager designed to orchestrate and coordinate numerous Claude Code agents working on diverse tasks within a shared development environment. It tackles the common challenge where individual AI agents lose context on restart by persisting work state using git-backed hooks, enabling reliable and resumable multi-agent workflows that scale comfortably beyond just a handful of concurrent agents. The system introduces concepts like The Mayor (the primary coordinator agent), Rigs (project containers), Crew Members (developer workspaces), and Polecats (ephemeral worker agents), organizing work into structured convoys and beads that can be tracked through command-line interfaces. By integrating with tools like tmux and supporting git worktrees, Gastown enables interactive sessions where agents and humans collaborate fluidly, with tracked progress, handoffs, and historical records.

Features

  • Git-backed persistent work state for agents
  • Multi-agent orchestration with context tracking
  • Structured roles (Mayor, Polecats, Rigs, Crew)
  • Convoy and bead work grouping
  • Tmux and CLI integration
  • Scales to dozens of AI agents

Project Samples

Project Activity

See All Activity >

Categories

System

License

MIT License

Follow Gas Town

Gas Town Web Site

Other Useful Business Software
Build Agents and Models on One Platform Icon
Build Agents and Models on One Platform

Everything you need to build production-ready agents and models. Access 200+ Google and third-party AI models and tools.

Gemini Enterprise Agent Platform is Google Cloud's comprehensive platform for developers to build, scale, govern, and optimize agents and models. Choose from Google's most advanced models and third-party models like Anthropic's Claude Model Family.
Try It Free
Rate This Project
Login To Rate This Project

User Reviews

Be the first to post a review of Gas Town!

Additional Project Details

Operating Systems

Linux, Mac

Programming Language

Go

Related Categories

Go System Software

Registered

2026-01-28