AGI-MCP-Agent provides a Master Control Program (MCP) framework designed to orchestrate intelligent agents, their cognitive workflows, and multi-agent collaboration. The server exposes an API built with FastAPI and managed through an MCP interface, enabling you to deploy, monitor, and coordinate autonomous agents that can plan, reason, interact with tools, and persist memories. The included environment interface and coordination primitives make it possible to connect external APIs, data sources, and tools, while maintaining security through sandboxed execution and robust orchestration. Use the MCP to spawn agent lifecycles, schedule tasks, and observe performance metrics across agent fleets.

Once the server is running, you can interact with the API to create agents, assign tasks, and configure tools. The agent framework supports cognitive processing (planning, reasoning, decision-making), memory management (short- and long-term), tool integrations, perception modules, action generation, and self-monitoring. The environment interface standardizes how external systems are queried and how results are formatted, while multi-agent coordination handles communication, roles, and conflict resolution. This setup enables you to experiment with autonomous problem solving, tool chaining, and collaborative tasks across multiple agents within a single MCP instance.

Prerequisites:

• Python 3.9 or newer
• Poetry (optional but recommended) for dependency management
• PostgreSQL 12+ or SQLite for development
• Docker and Docker Compose (optional, for containerized deployment)
• OpenAI API key (for LLM-based agents)

Install and run (local development):

1. Clone the repository: git clone https://github.com/ot2net/agi-mcp-agent.git cd agi-mcp-agent

2. Install dependencies (with Poetry): make install-dev

   or manually: poetry install

3. Set up environment variables: cp example.env .env

   Edit .env with your configuration, including API keys and database URL

4. Initialize the database (if using a DB): make db-init

5. Run the server (development mode): make run-dev

   or directly: uvicorn agi_mcp_agent.api.server:app --host 0.0.0.0 --port 8000 --reload

Containerized option (Docker Compose):

1. Ensure Docker and Docker Compose are installed

2. Copy environment file and configure it: cp example.env .env

3. Start services with Docker Compose: docker-compose up -d

4. Access the API:

   • API: http://localhost:8000
   • Documentation: http://localhost:8000/docs

Tips and common considerations:

• The MCP server is Python-based and relies on uvicorn to serve the FastAPI application. Ensure the port (default 8000) is available on your host.
• Store sensitive keys (OpenAI, database passwords) in a .env file and do not commit it to version control.
• If you’re using Docker, you can customize memory/CPU limits in your docker-compose.yml for stability in resource-constrained environments.
• Use the Makefile commands to simplify development tasks (lint, test, format, migrations, etc.).
• If you encounter connection issues to the database, verify DATABASE_URL and migrations status (db-migrate, db-upgrade).
• When integrating new tools or APIs, ensure proper sandboxing and rate-limiting to prevent cascading failures across agents.
• Review the API docs at /docs to understand available endpoints for agent creation, task assignment, and environment configuration.