chicks2014/mcp-sql-server

Introduction to MCP with SQL Agent

Build an AI SQL agent using MCP to interact with a SQLite database through natural language.

📺 Watch the Tutorial

Key Features

• Create an AI SQL agent using MCP.
• Interact with a SQLite database using natural language.
• Use Claude 3 Sonnet to generate SQL queries.
• Execute SQL queries safely.
• Learn how to use MCP to build AI tools.

Tech Stack

• Anthropic
• Loguru
• MCP
• Python-dotenv
• Rich
• SQLite

Getting Started

Prerequisites

• Python 3.12+
• uv
• An Anthropic API key (set in .env file)

Installation

1. Clone the repository.

2. Navigate to the project directory: cd mcp-sql-server

3. Install dependencies using uv: uv sync

4. Create a .env file with your Anthropic API key:

   ANTHROPIC_API_KEY=sk-ant-api03-YOUR_API_KEY
   

Usage

1. Run the MCP client: uv run mcp_client.py
2. Enter your SQL queries in natural language.

Note that you don't need to explicitly run the server, as the client automatically runs it.

Project Structure

mcp-sql-server/
├── README.md         # Project documentation
├── mcp_client.py     # MCP client script
├── mcp_server.py     # MCP server script
├── database.db       # SQLite database
├── pyproject.toml    # Project dependencies
├── .env              # Environment variables
├── .gitignore        # Git ignore file
├── .python-version   # Python version
├── layout.zellij     # Zellij layout for development
├── code.zellij       # Zellij layout for code editing
└── demo.zellij       # Zellij layout for demo

Implementation Details

• The mcp_server.py file defines an MCP server with a query_data tool to execute SQL queries.
• The mcp_client.py file uses the Anthropic Claude 3 Sonnet model to generate SQL queries from natural language input.
• The chinook.db file is a SQLite database used for the demo.
• Zellij layouts are provided for development and demo purposes.

Contributing

1. Fork the repository.
2. Create your feature branch.
3. Submit a pull request.

License

MIT License - See file for details.