MCP-Based Agentic AI System

Production-ready, self-hosted AI infrastructure with Model Context Protocol.

Quick Start

1. Install Dependencies

pip install -r requirements.txt

2. Configure Environment

cp .env.example .env
Edit .env with your configuration

3. Initialize Database

python scripts/init_db.py

4. Start Services

Start MCP Server
python -m server.mcp_server
Start API Server (in new terminal)
python -m api.main

5. Test the System

pytest tests/

Related MCP server: MCP Agent Memory

Architecture

• MCP Server: WebSocket-based protocol server

• REST API: FastAPI application for HTTP access

• Agent System: Autonomous AI agents with memory

• Tool Registry: Extensible function execution

• State Management: Redis + PostgreSQL persistence

API Documentation

Once running, visit: http://localhost:8000/docs

Configuration

All settings managed through environment variables:

• Database: PostgreSQL connection

• Redis: Caching and sessions

• LLM: Model provider and settings

• Security: JWT tokens and CORS

Deployment

Docker

docker-compose up -d

Kubernetes

kubectl apply -f kubernetes/

Monitoring

Prometheus metrics available at :9090/metrics

Support

For issues and questions, see docs/ directory.