mcp-pandoc: A Document Conversion MCP Server

Officially included in the Model Context Protocol servers open-source project. 🎉

Overview

A Model Context Protocol server for document format conversion using pandoc. This server provides tools to transform content between different document formats while preserving formatting and structure.

Please note that mcp-pandoc is currently in early development. PDF support is under development, and the functionality and available tools are subject to change and expansion as we continue to improve the server.

Credit: This project uses the Pandoc Python package for document conversion, forming the foundation for this project.

Related MCP server: Document Conversion Assistant

Demo

🎥 Watch on YouTube

More to come...

Tools

1. convert-contents

   • Transforms content between supported formats

   • Inputs:

     • contents (string): Source content to convert (required if input_file not provided)

     • input_file (string): Complete path to input file (required if contents not provided)

     • input_format (string): Source format of the content (defaults to markdown)

     • output_format (string): Target format (defaults to markdown)

     • output_file (string): Complete path for output file (required for pdf, docx, rst, latex, epub formats)

   • Supported input/output formats:

     • markdown

     • html

     • pdf

     • docx

     • rst

     • latex

     • epub

     • txt

   • Note: For advanced formats (pdf, docx, rst, latex, epub), an output_file path is required

Supported Formats

Currently supported formats:

Basic formats (direct conversion):

• Plain text (.txt)

• Markdown (.md)

• HTML (.html)

Advanced formats (requires complete file paths):

• PDF (.pdf) - requires TeX Live installation

• DOCX (.docx)

• RST (.rst)

• LaTeX (.tex)

• EPUB (.epub)

Note: For advanced formats:

1. Complete file paths with filename and extension are required

2. PDF conversion requires TeX Live installation (see Critical Requirements section -> For macOS: brew install texlive)

3. When no output path is specified:

   • Basic formats: Displays converted content in the chat

   • Advanced formats: May save in system temp directory (/tmp/ on Unix systems)

Usage & configuration

NOTE: Ensure to complete installing required packages mentioned below under "Critical Requirements".

To use the published one

{
  "mcpServers": {
    "mcp-pandoc": {
      "command": "uvx",
      "args": ["mcp-pandoc"]
    }
  }
}

⚠️ Important Notes

Critical Requirements

1. Pandoc Installation

• Required: Install pandoc - the core document conversion engine

• Installation:

  # macOS
  brew install pandoc
  
  # Ubuntu/Debian
  sudo apt-get install pandoc
  
  # Windows
  # Download installer from: https://pandoc.org/installing.html

• Verify: pandoc --version

2. UV package installation

• Required: Install uv package (includes uvx command)

• Installation:

  # macOS
  brew install uv
  
  # Windows/Linux
  pip install uv

• Verify: uvx --version

3. PDF Conversion Prerequisites: Only needed if you need to convert & save pdf

• TeX Live must be installed before attempting PDF conversion

• Installation commands:

  # Ubuntu/Debian
  sudo apt-get install texlive-xetex
  
  # macOS
  brew install texlive
  
  # Windows
  # Install MiKTeX or TeX Live from:
  # https://miktex.org/ or https://tug.org/texlive/

4. File Path Requirements

• When saving or converting files, you MUST provide complete file paths including filename and extension

• The tool does not automatically generate filenames or extensions

Examples

✅ Correct Usage:

# Converting content to PDF
"Convert this text to PDF and save as /path/to/document.pdf"

# Converting between file formats
"Convert /path/to/input.md to PDF and save as /path/to/output.pdf"

❌ Incorrect Usage:

# Missing filename and extension
"Save this as PDF in /documents/"

# Missing complete path
"Convert this to PDF"

# Missing extension
"Save as /documents/story"

Common Issues and Solutions

1. PDF Conversion Fails

   • Error: "xelatex not found"

   • Solution: Install TeX Live first (see installation commands above)

2. File Conversion Fails

   • Error: "Invalid file path"

   • Solution: Provide complete path including filename and extension

   • Example: /path/to/document.pdf instead of just /path/to/

3. Format Conversion Fails

   • Error: "Unsupported format"

   • Solution: Use only supported formats:

     • Basic: txt, html, markdown

     • Advanced: pdf, docx, rst, latex, epub

Quickstart

Install

Option 1: Installing manually via claude_desktop_config.json config file

• On MacOS: open ~/Library/Application\ Support/Claude/claude_desktop_config.json

• On Windows: %APPDATA%/Claude/claude_desktop_config.json

a) Only for local development & contribution to this repo

ℹ️ Replace with your locally cloned project path

"mcpServers": {
  "mcp-pandoc": {
    "command": "uv",
    "args": [
      "--directory",
      "<DIRECTORY>/mcp-pandoc",
      "run",
      "mcp-pandoc"
    ]
  }
}

b) Published Servers Configuration - Consumers should use this config

"mcpServers": {
  "mcp-pandoc": {
    "command": "uvx",
    "args": [
      "mcp-pandoc"
    ]
  }
}

Option 2: To install Published Servers Configuration automatically via Smithery

Run the following bash command to install published mcp-pandoc pypi for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install mcp-pandoc --client claude

• If you face any issue, use the "Published Servers Configuration" above directly instead of this cli.

Note: To use locally configured mcp-pandoc, follow "Development/Unpublished Servers Configuration" step above.

Development

Building and Publishing

To prepare the package for distribution:

1. Sync dependencies and update lockfile:

uv sync

2. Build package distributions:

uv build

This will create source and wheel distributions in the dist/ directory.

3. Publish to PyPI:

uv publish

Note: You'll need to set PyPI credentials via environment variables or command flags:

• Token: --token or UV_PUBLISH_TOKEN

• Or username/password: --username/UV_PUBLISH_USERNAME and --password/UV_PUBLISH_PASSWORD

Debugging

Since MCP servers run over stdio, debugging can be challenging. For the best debugging experience, we strongly recommend using the MCP Inspector.

You can launch the MCP Inspector via npm with this command:

npx @modelcontextprotocol/inspector uv --directory /Users/vivekvells/Desktop/code/ai/mcp-pandoc run mcp-pandoc

Upon launching, the Inspector will display a URL that you can access in your browser to begin debugging.

Contributing

We welcome contributions to enhance mcp-pandoc! Here's how you can get involved:

1. Report Issues: Found a bug or have a feature request? Open an issue on our GitHub Issues page.

2. Submit Pull Requests: Improve the codebase or add features by creating a pull request.