This tutorial walks you through the usage of MindsDB MCP API with Cursor.

​
Enable MindsDB as an MCP Server

  1. Install MindsDB locally via Docker or Docker Desktop.

  2. When starting MindsDB, enable its MCP API as explained here.

  3. If this is your first time using MindsDB, make sure to connect your data source and/or upload files, so that you can ask questions over your data as shown in the following steps.

To get you started quickly, use our sample dataset that stores the sales manager data.

CREATE DATABASE sales_manager_data
WITH ENGINE = "postgres",
PARAMETERS = {
    "user": "demo_user",
    "password": "demo_password",
    "host": "samples.mindsdb.com",
    "port": "5432",
    "database": "sales_manager_data"
};
  1. Open Cursor, go to the Cursor Settings, open the MCP tab, and click on Add new global MCP server. Alternatively, go to the Cursor settings -> Features -> MCP Servers.

  1. Add the below content to the mcp.json file.
{
    "mcpServers": {
      "mindsdb": {
        "url": "http://127.0.0.1:47337/sse"
      }
    }
}
  1. Ensure that MindsDB is listed as an MCP server.

​
Chat with Your Data

  1. Open the Cursor chat window and select the Agent mode from the dropdown.

  1. Ask questions over your data. Note that you need to approve each call of the MCP server’s tools by clicking on Run tool.

  1. The agent provides an answer with helpful suggestions of follow-up information that can be extracted from the available data.

Was this page helpful?

Suggest editsRaise issue