Supabase MCP Server

A Model Context Protocol (MCP) server that provides tools for interacting with a Supabase database. This server enables AI assistants to perform database operations through a standardized interface.

NOTE: This Supabase MCP server was created as a demonstration of my AI IDE coding workflow. It is still a work in progress which I will expand on in future videos on my channel.

Features

• Read Table Rows: Query data from Supabase tables with optional filtering, pagination, and column selection
• Create Table Records: Insert new records into Supabase tables
• Update Table Records: Modify existing records in Supabase tables based on filters
• Delete Table Records: Remove records from Supabase tables based on filters

Prerequisites

• Docker or Docker Desktop
• Supabase account and project

Installation

1. Clone the repository:

   git clone https://github.com/coleam00/supabase-mcp.git
   cd supabase-mcp
   

Docker Setup

1. Build the Docker image:

   docker build -t mcp/supabase .
   

Usage

Running as an MCP Server with Docker

The Supabase MCP server can be integrated with AI assistants using the Model Context Protocol.

1. Include the below configuration in your MCP config (in Claude Desktop, Windsurf, etc.)

Be sure to build the container with the installation steps first!

{
  "mcpServers": {
    "supabase": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "-e", "SUPABASE_URL", "-e", "SUPABASE_SERVICE_KEY", "mcp/supabase"],
      "env": {
        "SUPABASE_URL": "YOUR-SUPABASE-URL",
        "SUPABASE_SERVICE_KEY": "YOUR-SUPABASE-SERVICE-ROLE-KEY"
      }
    }
  }
}

1. Replace YOUR-SUPABASE-URL and YOUR-SUPABASE-SERVICE-ROLE-KEY with your actual Supabase credentials.

2. The AI assistant can now access the Supabase database through the MCP server using the provided tools.

For more information on the Model Context Protocol, visit modelcontextprotocol.io.

Available Tools

Read Table Rows

read_table_rows(
    table_name: str,
    columns: Optional[List[str]] = None,
    filters: Optional[Dict[str, Any]] = None,
    limit: Optional[int] = None,
    offset: Optional[int] = None
)

Example:

# Read active users
read_table_rows(
    table_name="users",
    columns=["id", "name", "email"],
    filters={"is_active": True},
    limit=10,
    offset=0
)

Create Table Records

create_table_records(
    table_name: str,
    records: Union[Dict[str, Any], List[Dict[str, Any]]]
)

Example:

# Create a new user
create_table_records(
    table_name="users",
    records={
        "name": "John Doe",
        "email": "john@example.com",
        "is_active": True
    }
)

Update Table Records

update_table_records(
    table_name: str,
    updates: Dict[str, Any],
    filters: Dict[str, Any]
)

Example:

# Update user status
update_table_records(
    table_name="users",
    updates={"status": "premium"},
    filters={"is_active": True}
)

Delete Table Records

delete_table_records(
    table_name: str,
    filters: Dict[str, Any]
)

Example:

# Delete inactive users
delete_table_records(
    table_name="users",
    filters={"is_active": False}
)

Development

Project Structure

supabase-mcp/
├── supabase_mcp/
│   ├── __init__.py
│   ├── server.py              # Main MCP server implementation
│   └── tests/                 # Unit tests
├── Dockerfile                 # Docker configuration for MCP server
├── example_mcp_config.json    # Example MCP configuration
├── requirements.txt           # Python dependencies
├── .env.example               # Example environment variables
├── README.md                  # Project documentation
├── PLANNING.md                # Project planning
└── TASKS.md                   # Task tracking

Running Tests

pytest supabase_mcp/tests/

Model Context Protocol Integration

The Supabase MCP server implements the Model Context Protocol, which allows AI assistants to interact with Supabase databases in a standardized way.

How It Works

1. The MCP server exposes tools for database operations (read, create, update, delete)
2. AI assistants connect to the MCP server using the stdio transport
3. The AI assistant can invoke the tools to perform database operations
4. The MCP server handles the communication with Supabase and returns the results

MCP Configuration

The example_mcp_config.json file shows how to configure an AI assistant to use the Supabase MCP server:

{
  "mcpServers": {
    "supabase": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "-e", "SUPABASE_URL", "-e", "SUPABASE_SERVICE_KEY", "mcp/supabase"],
      "env": {
        "SUPABASE_URL": "YOUR-SUPABASE-URL",
        "SUPABASE_SERVICE_KEY": "YOUR-SUPABASE-SERVICE-ROLE-KEY"
      }
    }
  }
}

This configuration tells the AI assistant:

• To use Docker to run the MCP server
• To pass the Supabase credentials as environment variables
• To use the mcp/supabase Docker image

Using with AI Assistants

AI assistants that support the Model Context Protocol can use this server to:

1. Query data from Supabase tables
2. Insert new records into tables
3. Update existing records
4. Delete records

The assistant will have access to the tools documented in the "Available Tools" section above.

Environment Variables

License

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