MySQL MCP Server

A Model Context Protocol (MCP) server for MySQL databases. Provides safe, read-only access to MySQL databases with automatic query sanitization, Google Cloud SQL Proxy support, and remote HTTP deployment options.

Features

• Read-Only Access: All queries are sanitized to ensure only read operations are allowed
• Google Cloud SQL Proxy: Built-in support for connecting to Cloud SQL instances
• Remote HTTP Mode: Deploy as a remote service with API key authentication
• Docker Support: Production-ready Docker images for easy deployment
• Token Optimization: Results are automatically limited to minimize token usage

Quick Start

Local Usage with Claude Code

Add to your project's .mcp.json:

{
  "mcpServers": {
    "mysql": {
      "command": "npx",
      "args": ["-y", "@matpb/mysql-mcp-server"],
      "env": {
        "MYSQL_HOST": "localhost",
        "MYSQL_PORT": "3306",
        "MYSQL_USER": "your_username",
        "MYSQL_PASSWORD": "your_password",
        "MYSQL_DATABASE": "your_database"
      }
    }
  }
}

Remote HTTP Server

Deploy as a remote service and connect from anywhere:

{
  "mcpServers": {
    "mysql": {
      "type": "http",
      "url": "https://your-server.example.com/mcp",
      "headers": {
        "X-API-Key": "your-api-key"
      }
    }
  }
}

Available Tools

Configuration

Environment Variables

Copy .env.example to .env and configure:

cp .env.example .env

MySQL Connection

Query Settings

HTTP Transport (Remote Mode)

Google Cloud SQL Proxy

Docker Deployment

Project Structure

mysql-mcp-server/
├── docker/
│   ├── Dockerfile              # Standard Docker image
│   ├── Dockerfile.digitalocean # Image with baked-in credentials
│   ├── docker-compose.yml      # Docker Compose configuration
│   └── .env.example            # Docker environment template
├── src/                        # Source code
└── .env.example                # Environment template

Build and Run Locally

# Copy and configure environment
cp docker/.env.example .env

# Edit .env with your settings:
# - Set API_KEYS for authentication
# - Configure MySQL connection
# - (Optional) Configure Cloud SQL Proxy

# Build and start
cd docker
docker compose up -d

# View logs
docker compose logs -f

# Stop
docker compose down

Deploy to Cloud Platforms

Step 1: Prepare Credentials (if using Cloud SQL)

# Create a service account
gcloud iam service-accounts create mysql-mcp-server \
  --display-name="MySQL MCP Server"

# Grant Cloud SQL Client role
gcloud projects add-iam-policy-binding YOUR_PROJECT_ID \
  --member="serviceAccount:mysql-mcp-server@YOUR_PROJECT_ID.iam.gserviceaccount.com" \
  --role="roles/cloudsql.client"

# Generate key file
gcloud iam service-accounts keys create docker/service-account.json \
  --iam-account=mysql-mcp-server@YOUR_PROJECT_ID.iam.gserviceaccount.com

Step 2: Build for AMD64

Most cloud platforms run AMD64. Build with:

docker buildx build --platform linux/amd64 \
  -f docker/Dockerfile.digitalocean \
  -t your-registry.com/mysql-mcp-server:latest \
  --push .

Step 3: Configure Environment Variables

Set these in your cloud platform:

Step 4: Connect from Claude Code

{
  "mcpServers": {
    "mysql": {
      "type": "http",
      "url": "https://your-deployment-url.com/mcp",
      "headers": {
        "X-API-Key": "your-secure-api-key"
      }
    }
  }
}

Test Your Deployment

curl -X POST https://your-server.com/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "X-API-Key: your-api-key" \
  -d '{"jsonrpc":"2.0","method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"test","version":"1.0"}},"id":1}'

Google Cloud SQL Proxy

Local Development

For local development, use Application Default Credentials:

# Install gcloud CLI
brew install google-cloud-sdk  # macOS

# Authenticate
gcloud auth application-default login

Then configure your .mcp.json:

{
  "mcpServers": {
    "mysql": {
      "command": "npx",
      "args": ["-y", "@matpb/mysql-mcp-server"],
      "env": {
        "CLOUD_SQL_PROXY_ENABLED": "true",
        "CLOUD_SQL_INSTANCE": "your-project:region:instance",
        "MYSQL_USER": "your_username",
        "MYSQL_PASSWORD": "your_password",
        "MYSQL_DATABASE": "your_database"
      }
    }
  }
}

Production with Service Account

{
  "env": {
    "GOOGLE_APPLICATION_CREDENTIALS": "/path/to/service-account.json",
    "CLOUD_SQL_PROXY_ENABLED": "true",
    "CLOUD_SQL_INSTANCE": "your-project:region:instance"
  }
}

Security

Query Sanitization

The server automatically rejects:

• Data modification: INSERT, UPDATE, DELETE, DROP, CREATE, ALTER, TRUNCATE
• Transactions: BEGIN, COMMIT, ROLLBACK
• File operations: INTO OUTFILE, INTO DUMPFILE
• Lock operations: LOCK TABLES, FOR UPDATE

Allowed Operations

• SELECT queries
• SHOW commands
• DESCRIBE/DESC commands
• EXPLAIN queries
• WITH clauses (CTEs)

Best Practices

1. Use HTTPS in production (via reverse proxy or cloud platform)
2. Generate strong API keys: openssl rand -hex 32
3. Restrict CORS to specific domains
4. Rotate API keys regularly
5. Never commit .env files or service account credentials

Development

# Install dependencies
npm install

# Build
npm run build

# Run in stdio mode (local)
npm start

# Run in HTTP mode
MCP_TRANSPORT=http-stream API_KEYS=test-key npm start

# Watch mode
npm run watch

Troubleshooting

Cloud SQL Proxy Errors

"could not find default credentials"

• Run gcloud auth application-default login
• Or set GOOGLE_APPLICATION_CREDENTIALS to your service account file

"Access denied for user"

• Verify MySQL username/password
• Check that your service account has Cloud SQL Client role

HTTP Transport Errors

"Session not found"

• The MCP client must send proper session headers
• Test with the MCP Inspector: npx @modelcontextprotocol/inspector

License

MIT