Installation
Install CHAOS-AI from source and run your first command.
Requirements
- Python 3.14 or higher
- Git
- PostgreSQL (for the orchestration engine) — or use the Docker setup
- An LLM provider: Claude Code, GitHub Copilot CLI, Gemini CLI, Aider, OpenRouter, or Ollama
Install from Source
CHAOS is currently distributed as a source install. Clone the repository and install in editable mode:
git clone https://github.com/luvisaisa/claude-workflow
cd claude-workflow
pip install -e .
Verify the installation:
cw --version
Install via Docker (Recommended for Production)
The Docker setup includes the full stack: FastAPI backend, PostgreSQL, and the Next.js web dashboard.
git clone https://github.com/luvisaisa/claude-workflow
cd claude-workflow
docker compose up -d
The web dashboard is available at http://localhost:3000.
The MCP server (HTTP transport) runs at http://localhost:8000.
Configure Your LLM Provider
Set your provider API keys as environment variables. Create a .env file in the project root:
# Anthropic (Claude Code)
ANTHROPIC_API_KEY=sk-ant-...
# GitHub Copilot CLI
# Authenticated via `gh auth login` — no separate key needed
# Google (Gemini CLI)
GOOGLE_API_KEY=AI...
# OpenRouter
OPENROUTER_API_KEY=sk-or-...
# Ollama (local — no key needed)
# Ensure Ollama is running: ollama serve
Verify Setup
Run a quick health check:
cw --version
cw mcp --list-tools # lists all 134 MCP tools
Next Steps
- Quick Start — Your first 5 minutes with CHAOS
- Configuration — Full configuration reference
- MCP — Connect your editor via MCP