Smart MCP Server 🚀

Intelligent SQL Database Interface with Natural Language Support

A production-ready FastMCP 2 server that connects to any SQL database (MySQL, PostgreSQL, SQLite, SQL Server, Oracle, etc.) and converts natural language questions into SQL queries using AI.

✨ Features

• 🔌 Universal Database Support: MySQL, PostgreSQL, SQLite, SQL Server, Oracle, MariaDB

• 🧠 Natural Language to SQL: Ask questions in plain English, get SQL results

• 🔍 Auto-Schema Detection: Automatically scans and understands your database structure

• 🛡️ Safety First: Read-only by default, blocks dangerous operations

• ⚡ FastMCP 2 Powered: Modern, efficient MCP server implementation

• 🎯 Zero Configuration: Works with any schema without prior setup

• 🔧 Extensible: Clean, documented code for easy customization

Related MCP server: TalkDB

📚 Documentation

All documentation is organized in the docs/ folder with priority numbers for easy reading order.

Quick Links:

• docs/00_READ_ME_FIRST.md - 📖 Start here!

• docs/01_QUICKSTART.md - 🚀 Get running in 5 minutes

• docs/02_DATABASE_DOCUMENTATION_GUIDE.md - Document your DB for better queries

• docs/03_PRODUCTION_SECURITY.md - 🔒 Security guide (CRITICAL)

→ See docs/00_READ_ME_FIRST.md for complete reading order

🚀 Quick Start

1. Crate Environment

python3 -m venv .venv

On Linux/Mac:

source .venv/bin/activate

On Windows:

.venv\Scripts\activate

2. Install Dependencies

pip install -r requirements.txt

3. Edit config.yaml with LLM connector and Database Connector

you can also setup the key from .env file

4. Run the automatic database Schema generator

python generate_schema.py

5. Run the Server

python main.py

you might need to edit the main.py in the main function to change the MCP starting protocol

🎯 Example Queries

Try these natural language questions:

"Show me all customers"
"What are the top 5 most expensive products?"
"Show me orders with customer names"
"What is the total revenue by category?"
"List customers who placed orders in the last 30 days"

🛠️ Tech Stack

• FastMCP 2 - MCP server framework

• SQLAlchemy 2.x - Universal database adapter

• OpenAI API - Natural language processing

• Python 3.8+ - Core language

📦 Project Structure

SmartMCP/
├── main.py                  # Server entry point
├── config.yaml              # Configuration
├── requirements.txt         # Dependencies
├── example_database.py      # Sample data generator
├── test_setup.py           # Setup verification
├── db/adapter.py           # Database connector
├── nlp/query_parser.py     # NL to SQL parser
├── mcp/tools.py            # MCP tool definitions
└── utils/schema.py         # Schema utilities

🎓 Learn More

• QUICKSTART.md - Step-by-step setup guide

• ARCHITECTURE.md - System design deep dive

• PROJECT_SUMMARY.md - Complete feature list

• README.md - Full documentation with examples

🔒 Safety Features

• ✅ Read-only by default

• ✅ Keyword blacklist (DROP, DELETE, etc.)

• ✅ SQL injection prevention

• ✅ Row limits (max 1000 per query)

• ✅ Environment variable for secrets

📝 License

MIT License - Free to use, modify, and distribute!

Built with ❤️ and AI :P for the MCP community