MCP Database Server

🚀 Universal Database Connector for AI Assistants

Connect to ANY database (SQL Server, PostgreSQL, MySQL, MongoDB, Redis, SQLite) through your AI assistant with just a simple configuration!

A universal MCP (Model Context Protocol) server that provides seamless connections to multiple database types. This server enables AI assistants like Claude to interact with your databases through a unified interface.

📦 Quick Install

# Install with uv (recommended)
uv add mcp-universal-database-server

# Or install with pip
pip install mcp-universal-database-server

🔧 Simple Setup

Option 1: Basic Setup

Add this to your ~/.cursor/mcp.json or .vscode/settings.json:

{
  "mcpServers": {
    "mcp-database-server": {
      "command": "mcp-universal-database-server",
      "args": ["serve"]
    }
  }
}

Option 2: DSN Setup (Recommended)

Use DSN strings for instant database connections:

{
  "mcpServers": {
    "mcp-database-server": {
      "command": "mcp-universal-database-server",
      "args": ["serve"],
      "env": {
        "DSN": "postgresql://user:pass@localhost:5432/mydb?sslmode=disable",
        "SQLSERVER_DSN": "sqlserver://sa:pass@localhost:1433/MyDB?sslmode=disable"
      }
    }
  }
}

That's it! Your AI assistant can now connect to any database. Just ask: "Connect to my database"

🚀 Features

• Universal Database Support: Connect to PostgreSQL, MySQL, SQLite, MongoDB, Redis, SQL Server, and more
• DSN Support: Easy configuration with Data Source Name strings - just paste your connection string!
• MCP Integration: Seamless integration with VSCode, Cursor, and other MCP-compatible tools
• Auto-Discovery: Automatically loads database connections from environment variables
• Secure Connections: Support for SSL/TLS and connection pooling
• Rich Operations: Full CRUD operations, schema management, and query execution
• Type Safety: Built with Python type hints and Pydantic models
• Easy Installation: Install via uv package manager

📦 Supported Databases

SQL Databases

• PostgreSQL - Advanced open-source relational database
• MySQL - Popular open-source relational database
• SQLite - Lightweight file-based SQL database
• Microsoft SQL Server (planned)
• Oracle Database (planned)

NoSQL Databases

• MongoDB - Document-oriented database
• Redis - In-memory key-value store
• Apache Cassandra (planned)

Cloud Databases

• Google Firestore (planned)
• Amazon DynamoDB (planned)
• Azure Cosmos DB (planned)

🛠️ Installation

Using uv (Recommended)

# Install from PyPI (once published)
uv add mcp-database-server

# Or install from source
uv add git+https://github.com/yourusername/mcp-database-server.git

Using pip

pip install mcp-database-server

🔧 Configuration

1. Generate Configuration

mcp-database-server generate-config --output mcp_config.json

2. Configure VSCode/Cursor

Add to your MCP settings file (.vscode/settings.json or Cursor equivalent):

{
  "mcp.servers": {
    "mcp-database-server": {
      "command": "mcp-database-server",
      "args": ["serve"]
    }
  }
}

3. Add Database Connections

Use the MCP tools in your AI assistant to add connections:

Add a PostgreSQL connection:
- Name: my_postgres
- Type: postgresql
- Host: localhost
- Port: 5432
- Username: myuser
- Password: mypassword
- Database: mydatabase

🚀 Usage

Starting the Server

# Start MCP server (typically called by your editor)
mcp-database-server serve

# Test a connection
mcp-database-server test-connection \
  --type postgresql \
  --host localhost \
  --port 5432 \
  my_postgres

# List supported database types
mcp-database-server list-supported

Available MCP Tools

Once configured, you can use these tools through your AI assistant:

Connection Management

• add_connection - Add a new database connection
• remove_connection - Remove a database connection
• list_connections - List all configured connections
• test_connection - Test a specific connection

Schema Operations

• get_schema_info - Get database schema information
• get_table_info - Get detailed table information
• list_databases - List available databases
• list_tables - List tables in a database/schema

Data Operations

• execute_query - Execute SQL queries or database commands
• insert_data - Insert data into tables/collections
• update_data - Update data in tables (SQL databases)
• delete_data - Delete data from tables/collections

Table Management (SQL)

• create_table - Create new tables
• drop_table - Drop tables

Collection Management (NoSQL)

• create_collection - Create new collections
• find_documents - Find documents in collections
• update_documents - Update documents in collections

📝 Examples

Example 1: PostgreSQL Connection

# Through MCP tools in your AI assistant:
# 1. Add connection
add_connection(
    name="my_postgres",
    type="postgresql",
    host="localhost",
    port=5432,
    username="myuser",
    password="mypassword",
    database="mydb"
)

# 2. Execute query
execute_query(
    connection_name="my_postgres",
    query="SELECT * FROM users WHERE active = $1",
    parameters={"active": true}
)

Example 2: MongoDB Connection

# 1. Add MongoDB connection
add_connection(
    name="my_mongo",
    type="mongodb",
    host="localhost",
    port=27017,
    username="myuser",
    password="mypassword",
    database="mydb"
)

# 2. Find documents
find_documents(
    connection_name="my_mongo",
    collection_name="users",
    filter_query={"status": "active"},
    limit=10
)

Example 3: SQLite Connection

# 1. Add SQLite connection
add_connection(
    name="my_sqlite",
    type="sqlite",
    database="/path/to/database.db"
)

# 2. Get schema info
get_schema_info(connection_name="my_sqlite")

🔒 Security

• Connection Security: All connections support SSL/TLS encryption
• Credential Management: Passwords are handled securely and not logged
• Connection Pooling: Efficient connection management with automatic cleanup
• Input Validation: All inputs are validated using Pydantic models

🧪 Development

Setup Development Environment

# Clone repository
git clone https://github.com/yourusername/mcp-database-server.git
cd mcp-database-server

# Install with uv
uv sync --dev

# Run tests
uv run pytest

# Format code
uv run black src/
uv run ruff check src/

Project Structure

mcp-database-server/
├── src/mcp_database_server/
│   ├── __init__.py
│   ├── server.py              # Main MCP server
│   ├── connection_manager.py  # Connection management
│   ├── database_factory.py    # Database factory
│   ├── types.py              # Type definitions
│   ├── cli.py                # CLI interface
│   └── databases/            # Database drivers
│       ├── base.py           # Base classes
│       ├── postgresql.py     # PostgreSQL driver
│       ├── mysql.py          # MySQL driver
│       ├── sqlite.py         # SQLite driver
│       ├── mongodb.py        # MongoDB driver
│       └── redis.py          # Redis driver
├── tests/                    # Test files
├── pyproject.toml           # Project configuration
└── README.md

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

Adding New Database Support

1. Create a new driver in src/mcp_database_server/databases/
2. Inherit from SQLDatabase or NoSQLDatabase base class
3. Implement all required methods
4. Register the driver in database_factory.py
5. Add appropriate dependencies to pyproject.toml
6. Write tests for the new driver

📞 Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: GitHub Wiki

🙏 Acknowledgments

• Model Context Protocol (MCP) for the protocol specification
• All the amazing database driver maintainers
• The Python async ecosystem

---

Made with ❤️ for the AI and database communities