Multi-Database MCP Server (PostgreSQL & MySQL)

A powerful Model Context Protocol (MCP) Server designed for AI-assisted development (e.g., Cursor, Claude Desktop, Antigravity). It enables your AI assistant to safely interact directly with PostgreSQL and MySQL/MariaDB databases.

Key Features

• Dual Database Support: Supports both PostgreSQL and MySQL out of the box.
• Ready to Use: Simple pip installation with zero complex setup.
• Flexible Configuration: Configure via Client Config or Environment Variables (.env).
• Secure & Controlled: Built-in Read-Only mode, Max Row limits, and strict permission controls.
• Dual Execution Modes: Supports standard STDIO (Default for Cursor) and HTTP (for Antigravity/Docker).

Installation

Ensure you have Python 3.10 or higher installed.

pip install da2-mcp-multi-db

(Note: The package name is da2-mcp-multi-db)

Quick Start

Once installed, you can launch the servers directly using the following commands:

1. PostgreSQL

# Start PostgreSQL Server
da2-mcp-postgresql

2. MySQL / MariaDB

# Start MySQL Server
da2-mcp-mysql

Configuration Guide

This program relies on environment variables (like DATABASE_URL) to connect to your database. We provide two ways to configure this, please choose the one that fits your workflow:

Method 1: MCP Client Configuration (Recommended)

If you are using Cursor or Claude Desktop, it is recommended to define variables directly in your Client's configuration file for a cleaner setup.

Cursor Example (cursor_mcp_config.json):

{
  "mcpServers": {
    "my-postgres": {
      "command": "da2-mcp-postgresql", 
      "args": [],
      "env": {
        "DATABASE_URL": "postgresql://user:password@localhost:5432/mydb",
        "READ_ONLY": "false",
        "MAX_ROWS": "1000"
      }
    },
    "my-mysql": {
      "command": "da2-mcp-mysql",
      "args": [],
      "env": {
        "MYSQL_DATABASE_URL": "mysql://user:password@localhost:3306/mydb"
      }
    }
  }
}

Method 2: Using .env File (For CLI)

If you are running commands manually in a terminal, the program will automatically look for a .env file in the current working directory.

1. Create a file named .env in the directory where you run the command.
2. Add the following content:

# PostgreSQL Connection String
DATABASE_URL=postgresql://user:password@localhost:5432/mydb

# MySQL Connection String
MYSQL_DATABASE_URL=mysql://user:password@localhost:3306/mydb

# --- OR Use Individual Variables (Lowest Priority) ---

# PostgreSQL Config
PG_HOST=localhost
PG_PORT=5432
PG_USER=postgres
PG_PASSWORD=password
PG_DATABASE=postgres

# MySQL Config
MYSQL_HOST=localhost
MYSQL_PORT=3306
MYSQL_USER=root
MYSQL_PASSWORD=password
MYSQL_DATABASE=mysql

# Security Settings (Optional)
LOG_LEVEL=INFO
READ_ONLY=false
MAX_ROWS=1000
QUERY_TIMEOUT=30
ALLOW_CREATE_DB=false
ALLOW_DROP_DB=false

Available Tools

Once connected, your AI assistant will have access to the following capabilities:

HTTP Mode

If you need to run in HTTP Server mode (e.g., for Docker containers or remote deployment):

# Start on Port 8000
da2-mcp-postgresql --http --port 8000

You can debug this using the MCP Inspector: npx @modelcontextprotocol/inspector http://localhost:8000/mcp

License

MIT License

Copyright (c) 2026 Danny, DA2 Studio (https://da2.35g.tw)