home / mcp / supabase mcp server

Supabase MCP Server

Queries the feature_suggestions table in your Supabase database for interactive AI tooling.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "adiletd-feature-request-collection-mcp": {
      "command": "npx",
      "args": [
        "tsx",
        "mcp-server.ts"
      ],
      "env": {
        "SUPABASE_URL": "your_supabase_url",
        "SUPABASE_ANON_KEY": "your_supabase_anon_key"
      }
    }
  }
}

You can query feature data from your Supabase project through an MCP server that connects to the feature_suggestions table. This makes it easy to fetch, filter, and present suggestions to AI tools and workflows without manual database access.

How to use

Set up the MCP server locally and start it, then connect your MCP client to the running server. You can issue a request to fetch feature suggestions and receive results that you can pass to your prompts or tools. Use the server to retrieve data from the feature_suggestions table with optional limits to control the amount of data returned.

How to install

Prerequisites include Node.js v16 or higher and npm. You also need a Supabase project along with its URL and anon key to authorize access.

1. Create a project directory and navigate into it.

2. Create a .env file with your Supabase credentials.

SUPABASE_URL=your_supabase_url
SUPABASE_ANON_KEY=your_supabase_anon_key

3. Install dependencies.

npm install

4. Start the MCP server using the recommended runtime command.

npx tsx mcp-server.ts

Additional setup notes

If you prefer a development script, you can also run the server with the npm script provided.

npm run dev

Available tools

query_feature_suggestions

Fetches rows from the feature_suggestions table with an optional limit to control how many results are returned