Getting Started
Get Authorail running locally and connect your first AI agent in under five minutes.
Prerequisites
Before you begin, make sure you have the following installed:
- Python 3.12+ — python.org/downloads
- uv — fast Python package manager (docs.astral.sh/uv)
- Claude Code — Anthropic’s CLI agent (docs.anthropic.com)
Clone and Install
Clone the repository and install dependencies with uv:
git clone https://github.com/authorail/authorail.git
cd authorail
uv sync
This creates a virtual environment and installs all required packages.
Configure Environment
Copy the example environment file and fill in your values:
cp .env.example .env
Open .env and configure the required variables:
| Variable | Description |
|---|---|
DATABASE_URL | SQLite or PostgreSQL connection string, e.g. sqlite:///./dev.db |
ADMIN_API_KEY | Secret key for admin API access — generate with python -c "import secrets; print(secrets.token_urlsafe(32))" |
Run Migrations
Apply the database schema:
uv run alembic upgrade head
This creates all tables needed for policies, audit events, quotas, and more.
Connect to Claude Code
Add Authorail as an MCP server in your Claude Code configuration. Create or update ~/.claude/claude_code_config.json:
{
"mcpServers": {
"authorail": {
"command": "uv",
"args": [
"--directory", "/path/to/authorail",
"run", "python", "-m", "authorail.server"
],
"env": {
"DATABASE_URL": "sqlite:///./dev.db",
"ADMIN_API_KEY": "your-admin-api-key"
}
}
}
}
Replace /path/to/authorail with the absolute path to your cloned repository.
First Workflow
Once connected, try a basic spend request through Claude Code:
- Create a policy — define spending limits for your agent via the admin API.
- Submit a request — ask Claude Code to make a purchase that falls under your policy.
- Check the audit trail — review the decision log to see how Authorail enforced your rules.
# Example: create a policy via the admin API
curl -X POST http://localhost:8000/api/policies \
-H "Authorization: Bearer $ADMIN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"name": "dev-sandbox",
"max_amount": 10.00,
"currency": "USD",
"action": "allow"
}'
Next Steps
- Deploy Guide — take Authorail to production
- Pricing — compare plans and features
- Blog — read about AI agent governance patterns