Cursor

Add pdf-mcp to Cursor with one click, sign in with OAuth, and create your first PDF.

On this page

Connect Cursor to pdf-mcp to generate, read, and edit PDFs from your agent chat. You need Cursor installed and a pdf-mcp account.

Connect

Add to Cursor

Open the link above and allow your browser to launch Cursor.
Review the proposed pdf-mcp server and confirm installation.
In Cursor’s Plugins & MCPs settings, enable the server and start its sign-in action.
Sign in to pdf-mcp in your browser, select the workspace, and approve access.

The installation link contains only the server address. Your browser sign-in authorizes access. Cursor documents this format in its MCP install-link guide.

Manual configuration

If the link does not open Cursor, add this server to ~/.cursor/mcp.json for all projects, or .cursor/mcp.json for one project. Keep any existing servers.

JSON

{
  "mcpServers": {
    "pdf-mcp": {
      "url": "https://api.pdf-mcp.io/mcp"
    }
  }
}

Save the file, enable the server, and complete the same browser sign-in. See the Cursor MCP documentation for configuration and tool permissions.

Try it

Open an agent chat with pdf-mcp enabled and ask:

Use pdf-mcp to create a one-page PDF titled “Hello from Cursor”. Store it and return the browser view link.

Confirm the tool call if Cursor asks. PDF generation uses the connected workspace’s credits.

Troubleshooting

No tools: Check that pdf-mcp is enabled in MCP settings and that browser authorization completed.
Sign-in keeps returning: Clear the server’s saved authentication and reconnect.
API key required by your setup: Cursor supports environment variables in headers. Follow the API-key fallback; keep secrets out of shared configuration.

See MCP troubleshooting for credit and permission errors.