PDFMonkey MCP Server

The official Model Context Protocol (MCP) server for PDFMonkey. It lets AI assistants such as Claude, Cursor or any other MCP client list your templates, generate documents and check your quota through the PDFMonkey API.

• Runs locally over stdio, with your API key
• Listed in the MCP Registry as io.github.pdfmonkey/pdfmonkey-mcp
• Documentation: pdfmonkey.io/docs/integrations/mcp

The MCP server works with existing templates: it can read them but not create or edit them. To let an agent build and preview templates, use the PDFMonkey CLI.

Get your API key

Sign up at dashboard.pdfmonkey.io/register, then copy your API key from your account page.

Setup

The server needs Node.js 18 or later and reads your key from the PDFMONKEY_API_KEY environment variable.

Claude Code

claude mcp add pdfmonkey -e PDFMONKEY_API_KEY=your-api-key -- npx -y pdfmonkey-mcp-server

Claude Desktop, Cursor and other clients

Add this to your client’s MCP configuration (claude_desktop_config.json for Claude Desktop, .cursor/mcp.json for Cursor):

{
  "mcpServers": {
    "pdfmonkey": {
      "command": "npx",
      "args": ["-y", "pdfmonkey-mcp-server"],
      "env": {
        "PDFMONKEY_API_KEY": "your-api-key"
      }
    }
  }
}

Claude Desktop’s configuration file lives in ~/Library/Application Support/Claude/ on macOS and %APPDATA%\Claude\ on Windows.

Docker

No image is published yet, so build it locally:

docker build -t pdfmonkey-mcp-server .

{
  "mcpServers": {
    "pdfmonkey": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "-e", "PDFMONKEY_API_KEY", "pdfmonkey-mcp-server"],
      "env": {
        "PDFMONKEY_API_KEY": "your-api-key"
      }
    }
  }
}

See DOCKER-MCP-TOOLKIT.md to use it with the Docker MCP Toolkit.

Tools

The server also exposes four prompts: list-templates-formatted, show-documents-status, generate-invoice and workspace-overview.

Good to know:

• Download URLs expire after one hour. Download the file right away, or use a share link (Pro+ and Premium plans).
• If a generation takes longer than 3 minutes, call

... View full README on GitHub