This MCP server provides a Metabase integration that exposes 70+ tools for managing dashboards, cards, databases, users, permissions, collections, and advanced query features. It enables natural language driven interactions with Metabase, including listing dashboards, creating and updating cards, executing queries, exporting results, managing users and permissions, and configuring dashboards with parameters and subscriptions. The server is designed to work with Metabase v0.46.x and newer, including Metabase Cloud and self-hosted deployments. To start using it, ensure the MCP server is running and authenticate with your Metabase instance using either an API key or username/password as configured in the environment variables. Once running, you can call tools like list_dashboards, create_dashboard, list_cards, execute_query, export_card_results, create_public_link, and many more to perform operations directly via the MCP interface.

Prerequisites:

• Node.js (14.x or newer) and npm installed on your machine
• Access to a Metabase instance (self-hosted or Cloud) with appropriate permissions
• The MCP server code from the mcp-metabase repository

Step-by-step installation:

1. Clone the repository git clone https://github.com/easecloudio/mcp-metabase-server.git cd mcp-metabase-server

2. Install dependencies npm install

3. Configure environment variables Create a .env file or export variables in your shell. Example: export METABASE_URL=https://your-metabase-instance.com export METABASE_API_KEY=your_metabase_api_key

   alternatively (fallback)

   export METABASE_USERNAME=your_username

   export METABASE_PASSWORD=your_password

4. Build the TypeScript project (if applicable) npm run build

   or if using ts-node in development, you may skip build

5. Run the MCP Metabase server npm run start

   or run the Node server directly if you have a built dist/server.js

6. Verify the server is running and accessible via your MCP client. You should see logs indicating Metabase integration is initialized and available tools are exposed.

Tips and considerations:

• Use API Key authentication (METABASE_API_KEY) when possible for security; the server will fall back to username/password if the API key is not set.
• Ensure METABASE_URL points to the correct Metabase instance (cloud, self-hosted, or docker deployment).
• If you update Metabase versions, review compatibility notes in the README (recommended v0.48.x or later).
• For production, rotate API keys regularly and restrict access to the MCP server endpoints.
• Validate that the environment has network access from the MCP server to the Metabase instance (firewalls, VPC settings, etc.).
• Use the 70+ tools to automate common BI tasks: dashboards and cards management, data source schemas, user permissions, and advanced exports.