System Requirements
- Python 3.12+
- uv package manager (Install uv)
- Node.js (optional, for Node.js-based MCP servers)
Step-by-Step Installation
1. Clone the Repository
2. Install Dependencies
- FastAPI and Uvicorn (web framework)
- OpenAI SDK (LLM integration)
- MCP (Model Context Protocol) libraries
- Optional: Redis, SQLAlchemy, JWT libraries
3. Configure Environment
Create a.env file from the template:
.env with your configuration. See Environment Variables for details.
Minimum required configuration:
4. Configure MCP Servers (Optional)
Editmcp_config.json to add MCP servers:
5. Run the Server
Development mode (with auto-reload):http://localhost:8000/docs for interactive API documentation.
Docker Installation
Using Docker Compose
Using Dockerfile
Verification
Health Check
Check the health endpoint:Test Agent Endpoint
Optional: Enable Persistence
If you want server-side session persistence:-
Set up Redis:
-
Set up PostgreSQL:
-
Enable persistence:
Optional: Enable User Authentication
If you want JWT-based user authentication:-
Set JWT secret:
-
Configure Google OAuth (optional):
-
Configure email service:
-
Enable user auth:
Troubleshooting
Port Already in Use
If port 8000 is already in use:OpenAI API Key Not Found
Make sure your.env file contains:
MCP Server Connection Issues
Check yourmcp_config.json syntax and ensure MCP servers are accessible. See MCP Servers Configuration for troubleshooting.
Database Connection Issues
If using persistence, verify:- Redis is running:
redis-cli ping(should returnPONG) - PostgreSQL is accessible:
psql -h localhost -U user -d agent_chassis
Next Steps
- Read the Quick Start Guide for your first API call
- Explore the API Reference for all endpoints
- Check out Configuration for all options