skill

chaincrew

Claude Code Hooks-based autonomous AI agent system

taro-taryo/chaincrewcommunitylistedunreviewed

Set up

Skills are installed by placing their folder where your agent loads skills.

Download ZIP from GitHub Open the repository

The download comes from github.com/taro-taryo/chaincrew itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 26, 2026

ChainCrew

Multi-agent orchestration powered by Claude Code Hooks.

Blog Post (Zenn, Japanese) — Design philosophy, implementation details, and operational results

Quick Setup

Prerequisites: Python 3.10+ / Claude Code CLI

# 1. Fork and clone
git clone https://github.com/YOUR_USERNAME/chaincrew.git
cd chaincrew
./setup.sh

# 2. Start a session
claude

# 3. Monitor (in a separate terminal)
./monitor.sh

Hooks

Four hooks manage session state and autonomous task control.

Hook	Implementation	Role
SessionStart	`session_start.py`	Restore mission, pending tasks, violation history
UserPromptSubmit	`inject_context.py`	Input analysis, task decomposition, context injection
PostToolUse	`post_edit.py` / `post_task.py`	Auto-commit, subagent quality verification
Stop	`session_end.py`	Session state persistence

Details: Hooks Implementation Guide — processing flow, function list, output examples

Customization

Edit config/mission.json to adapt ChainCrew to your project.

{
  "mission": "Your mission here",
  "kpis": { "primary_metric": "Key metric", "target_value": 100 }
}

To integrate into an existing project:

pip install git+https://github.com/taro-taryo/chaincrew.git

Documentation

License

MIT