Build and maintain Convex-powered applications with enhanced AI capabilities through MCP integration.

What is Convex MCP?

Convex MCP enables Cursor to communicate with Convex's API directly through the Model Context Protocol. This integration lets AI assistants in Cursor access and manage Convex resources without requiring you to switch contexts.

Setting Up Convex MCP

Step 1: Access MCP Settings

Open Cursor and navigate to Settings → Features → MCP section to manage your Model Context Protocol connections.

Step 2: Add Convex Server

Click the “+ Add New MCP Server” button and fill in the configuration details for Convex.

Step 3: Configure Connection

Enter the following information:

  • Name: Convex
  • Type: Select the appropriate transport (usually SSE)
  • URL: Enter the Convex MCP endpoint
  • API Key: Add your Convex API key if required

Step 4: Verify Connection

After adding the server, it should appear in your MCP servers list. Click the refresh button to confirm the tools are available.

Using Convex with Cursor

Once you've set up the Convex MCP connection, Cursor can interact with Convex services directly. Here's how to use it:

  • Automatic detection: Cursor will recognize when your prompts relate to Convex and will suggest using the appropriate tools.
  • Direct reference: You can explicitly mention Convex in your prompts to activate the integration.
  • Task approval: When Cursor wants to use Convex MCP, it will request your approval before executing the action.

Example Use Cases

Here are some ways you can leverage the Convex MCP integration:

  • Ask Cursor to fetch information from your Convex account
  • Request Cursor to perform routine Convex operations
  • Use Cursor to analyze and work with your Convex data directly

Note:

MCP tools may not be compatible with all AI models in Cursor. For best results, use with Claude models which have the strongest MCP support.

Search tools

Modal for searching tools