MCP Integration

PaperQuire includes a built-in Model Context Protocol (MCP) server, letting AI agents generate branded, professional PDFs directly. Claude, ChatGPT, Copilot, and any MCP-compatible client can call PaperQuire as a tool.

How It Works

The MCP server runs as a CLI subcommand:

paperquire mcp-server

It communicates via JSON-RPC over stdio, following the MCP specification. The server reuses PaperQuire's Electron-based render pipeline (the same Chromium engine behind the desktop app) to produce identical output.

When an AI agent needs a PDF, the flow is:

The agent generates Markdown content
It calls the PaperQuire render tool with the Markdown and any options
PaperQuire renders it through a template, produces a PDF, and returns the file path
The agent can report the result back to the user

Setup

PaperQuire must be installed on your machine. The MCP server is included in the app — no additional installation needed.

Claude Desktop

Add PaperQuire to your Claude Desktop MCP configuration (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "paperquire": {
      "command": "/Applications/PaperQuire.app/Contents/Resources/bin/paperquire",
      "args": ["mcp-server"]
    }
  }
}

On Windows, use the full path to the PaperQuire executable instead.

Restart Claude Desktop. You should see "paperquire" listed in the MCP tools panel.

Claude Code

If paperquire is on your PATH (e.g. via Homebrew), add it to your project's .mcp.json:

{
  "mcpServers": {
    "paperquire": {
      "command": "paperquire",
      "args": ["mcp-server"]
    }
  }
}

VS Code (GitHub Copilot)

In your VS Code settings.json or workspace .vscode/mcp.json:

{
  "mcp": {
    "servers": {
      "paperquire": {
        "command": "paperquire",
        "args": ["mcp-server"]
      }
    }
  }
}

Homebrew Users

If you installed via Homebrew, the paperquire command is already on your PATH:

brew install --cask paperquire/paperquire/paperquire

You can use "command": "paperquire" directly in your MCP config.

Tools

The MCP server exposes four tools:

render

Convert Markdown to PDF (or HTML). Returns the output file path and size.

Parameter	Type	Required	Description
`markdown`	string	Yes	Markdown content to render
`outputPath`	string	No	Output file path (default: `./output.pdf`)
`template`	string	No	Template ID (default: `executive-report`)
`title`	string	No	Cover page title
`subtitle`	string	No	Cover page subtitle
`author`	string	No	Document author
`date`	string	No	Document date
`company`	string	No	Company name
`cover`	boolean	No	Include cover page
`toc`	boolean	No	Include table of contents
`pageSize`	"Letter" \| "A4"	No	Paper size
`primaryColor`	string	No	Accent color (hex)
`watermark`	string	No	Diagonal watermark text
`allowHtml`	boolean	No	Allow raw HTML in Markdown
`format`	"pdf" \| "html"	No	Output format (default: `pdf`)

Example response:

{
  "path": "/Users/you/projects/output.pdf",
  "format": "pdf",
  "sizeKb": 59,
  "template": "executive-report"
}

list_templates

Returns all available templates with IDs, names, and descriptions. No parameters.

Example response:

[
  { "id": "executive-report", "name": "Executive Report", "description": "Refined serif report..." },
  { "id": "technical-design", "name": "Technical Design Document", "description": "..." },
  ...
]

show_template

Returns full details of a specific template, including design tokens, cover style, and page configuration.

Parameter	Type	Required	Description
`templateId`	string	Yes	Template ID to look up

batch_render Pro

Pro feature — Batch rendering requires a Pro or Enterprise license. Compare plans

Render multiple Markdown documents in a single call. Accepts an array of files (either inline Markdown or file paths) and shared render options.

Parameter	Type	Required	Description
`files`	array	Yes	Array of `{ markdown?, path?, outputPath? }`
`template`	string	No	Template ID for all documents
`title`	string	No	Shared cover title
`author`	string	No	Shared author
`company`	string	No	Shared company name
`cover`	boolean	No	Include cover page
`toc`	boolean	No	Include table of contents
`pageSize`	"Letter" \| "A4"	No	Paper size
`primaryColor`	string	No	Accent color (hex)

Resources

paperquire://config

Returns the merged .paperquire.yml configuration for the current working directory. This lets AI agents discover project-level defaults (template, branding, metadata) before rendering.

Example Workflows

Generate a report from a prompt

Ask your AI agent:

"Write a project status report for Q2 2026 and render it as a PDF with the executive-report template."

The agent will generate the Markdown, then call the render tool to produce the PDF.

Discover templates before rendering

"What document templates are available? Show me the details of the technical design template, then render my spec as a PDF using it."

The agent calls list_templates, then show_template to inspect the template, then render with that template ID.

Batch render documentation

"Render all the markdown files in my docs/ folder as branded PDFs."

The agent reads the directory, then calls batch_render with the file paths.

Free Tier Limits

The free tier allows 3 renders per day via MCP (same limit as the CLI and desktop app). Upgrade to Pro for unlimited renders.

Troubleshooting

Server not found

Make sure the command path in your MCP config points to the actual PaperQuire binary. Run it directly to verify:

paperquire version

Server times out on first call

The first tool call takes a few seconds because PaperQuire initializes a Chromium renderer. Subsequent calls reuse the same renderer and are much faster.

Tools not appearing

Restart your MCP client (Claude Desktop, VS Code) after updating the config. Check the client's MCP logs for connection errors.