Connect UCL to Cursor and access data from Notion

In this guide, you’ll learn how to connect a UCL deployment to Cursor and use its AI assistant to retrieve summary from Notion for a specific meeting; all without writing any integration code.

Connect your Notion account to UCL

• You need to authenticate your Notion account, after which the Notion App will appear as Connected.

• From Select Tools, you can enable all or any specific Actions.

Connect Cursor to UCL

• Download and install the latest version of Cursor on your desktop.

• Open Cursor and navigate to Settings in the top-right corner.

• Next, in the MCP section, click Add Custom MCP; this opens a file named mcp.json.

• From your UCL dashboard, you can navigate to the Embed section from the top navigation bar and select Cursor from the list of clients available to access the step-by-step embedding guide for Cursor with UCL.

You can select and follow the guide for any client of your choice from the embed page to integrate it with UCL.

• In the Open Setting and Add MCP Server step, you will see the Cursor command code which you can copy for your workspace and paste it to Cursor.

Checking the Multi Tenant option will enable you to embed connectors and actions for multiple tenants [users].

{
  "mcpServers": {
    "fastn": {
      "transport": "streamable_http",
      "url": "https://mcp.ucl.dev/shttp/space_id={}&tenant_id=test-tenant&auth_token=customAuthToken"
    }
  }
}

• Paste this configuration into the mcp.json file in Cursor and save it.

• You should now see the UCL Server listed as connected in Cursor. Under the tools label in Cursor, you can view all the enabled actions available with your connectors in UCL.

Use Cursor to Fetch Meeting Summary from Notion

• In Cursor, click Toggle AI Pane from the top-right corner to open the assistant.

Make sure you’ve selected an agent; it's not auto-set.

• In the assistant, it will ask for your Notion database ID after you have given a command like:

  Fetch summary from Notion for the meeting titled "Meeting Name"

• The assistant will use the "Get Tasks" or a relevant action to retrieve notes.

• The meeting summary should appear directly in your AI plane, ready for review or further processing.