gastown

Coordinate multiple Claude Code agents across rigs for automated project work.

1.0k|105|Updated Jan 2, 2026
One-click install
npx skills add https://github.com/numman-ali/n-skills --skill gastown
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gastown
Source: https://github.com/numman-ali/n-skills/tree/main/skills/tools/gastown/skills/gastown
Command: npx skills add https://github.com/numman-ali/n-skills --skill gastown

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Gas Town coordinates multiple Claude Code agents to automate cross-rig project work. It provides a unified orchestration layer that makes dispatching, monitoring, and merging tasks seamless across rigs, reducing context switching and errors.

Core Features & Use Cases

  • Cross-rig orchestration: Dispatch work to witnesses, refineries, polecats, or crew to automate end-to-end pipelines.
  • Beads-based memory and messaging: Persist work state and enable async handoffs via mail and hooks.
  • Merge pipeline control: Manage the complete flow from sling to merge to main with visibility and recovery options.
  • Troubleshooting and diagnostics: Built-in doctor checks and recovery procedures to keep the engine running.
  • Real-world use: Hook a GitHub project, track issues as beads, sling to workers, and observe per-rig workflows at scale.

Quick Start

Follow these steps to get Gas Town running and hook up your first project:

  • Install the Gas Town CLI (gt) and Beads (bd), verify versions.
  • Initialize a workspace at ~/gt with gt install ~/gt.
  • Add a rig with gt rig add myproject https://github.com/owner/repo.git.
  • Sling a bead to a polecat and observe with gt polecat list and gt status.
  • Start the refinery with gt refinery start and monitor with gt refinery status.
  • If issues arise, run gt doctor and bd doctor for health checks.

Frequently Asked Questions about gastown

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I coordinate work across multiple agents in a workflow?

Gas Town coordinates multiple Claude Code agents across rigs using a unified orchestration layer. Dispatch tasks to witnesses, refineries, polecats, or crew, monitor execution with beads-based memory, and merge results seamlessly to reduce context switching and errors.

Can I automate end-to-end project pipelines with agent orchestration?

Yes. Gas Town automates cross-rig project work by dispatching tasks, persisting state via beads-based messaging, and controlling the complete merge pipeline from sling to main with built-in visibility and recovery options.

How do I set up multi-agent orchestration for GitHub projects?

Initialize a Gas Town workspace with gt install, add a rig pointing to your GitHub repository, sling beads as tracked issues to workers, start the refinery to monitor workflows, and use gt doctor for diagnostics if issues arise.

What's the difference between orchestrating agents versus running them independently?

Orchestration with Gas Town provides unified dispatching, async handoffs via mail and hooks, centralized merge pipeline control, and built-in troubleshooting. Running agents independently loses state persistence, visibility across rigs, and recovery procedures.

Do I need a specific rig setup to use agent orchestration?

Gas Town requires the CLI (gt) and Beads (bd) installed and verified. It enforces a discoverable unit model with SKILL.md at the root containing YAML frontmatter and optional resource directories for scripts, references, and assets.

What happens if an orchestrated workflow fails mid-execution?

Gas Town includes built-in doctor checks and recovery procedures to diagnose and restart failed tasks. Run gt doctor and bd doctor for health checks, then resume work without losing prior bead state or merge progress.