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: flexberry-markitdown-mcp

📋 Quick Reference

New to mcp-pandoc? Check out 📖 CHEATSHEET.md for

• ⚡ Copy-paste examples for all formats

• 🔄 Bidirectional conversion matrix

• 🎯 Common workflows and pro tips

• 🌟 Reference document styling guide

Perfect for quick lookups and getting started fast!

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, odt, pptx formats)

     • reference_doc (string): Path to a reference document to use for styling (supported for docx, odt and pptx output; the file must match the output format)

     • defaults_file (string): Path to a Pandoc defaults file (YAML) containing conversion options

     • filters (array): List of Pandoc filter paths to apply during conversion

   • Supported formats, by direction:

     Format	Read	Write
     markdown	✅	✅
     html	✅	✅
     docx	✅	✅
     odt	✅	✅
     rst	✅	✅
     latex	✅	✅
     epub	✅	✅
     ipynb	✅	✅
     txt	✅	✅
     pdf	❌	✅
     pptx	❌	✅

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

🔧 Advanced Features

Defaults Files (YAML Configuration)

Use defaults files to create reusable conversion templates with consistent formatting:

# academic-paper.yaml
from: markdown
to: pdf
number-sections: true
toc: true
metadata:
  title: "Academic Paper"
  author: "Research Team"

Example usage: "Convert paper.md to PDF using defaults academic-paper.yaml and save as paper.pdf"

Pandoc Filters

Apply custom filters for enhanced processing:

Example usage: "Convert docs.md to HTML with filters ['/path/to/mermaid-filter.py'] and save as docs.html"

💡 For comprehensive examples and workflows, see CHEATSHEET.md

📊 Supported Formats & Conversions

Bidirectional Conversion Matrix

Rows are source formats, columns are targets. PDF and PPTX have no rows because pandoc cannot read them.

A Note on Write-Only Formats

PDF and PPTX can be produced but not read.

Pandoc has never shipped a PDF reader. Pandoc gained a PowerPoint reader in 3.8.3 (released 2025-12-01), but this project declares no minimum pandoc version, and Ubuntu 24.04 still ships pandoc 3.1.3, so pptx input is not offered. Tracked in #54 and #49.

PPTX output needs no particular pandoc version. The PowerPoint writer has existed since pandoc 2.0.5.

Format Categories

Requirements by Format

• PDF (.pdf) - requires TeX Live installation

• DOCX (.docx), ODT (.odt), PPTX (.pptx) - support custom styling via reference documents, and the reference must be the same format as the output

• All others - no additional requirements

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"]
    }
  }
}

💡 Quick Start: See CHEATSHEET.md for copy-paste examples and common workflows.

⚠️ 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"

# Converting to DOCX with a reference document template
"Convert input.md to DOCX using template.docx as reference and save as output.docx"

# Converting to ODT with a reference document template
"Convert input.md to ODT using template.odt as reference and save as output.odt"

# Converting to a branded PowerPoint deck
"Convert slides.md to PPTX using template.pptx as reference and save as deck.pptx"

# Step-by-step reference document workflow
"First create a reference document: pandoc -o custom-reference.docx --print-default-data-file reference.docx" or if you already have one, use that
"Then convert with custom styling: Convert this text to DOCX using /path/to/custom-reference.docx as reference and save as /path/to/styled-output.docx"

❌ 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

4. Reference Document Issues

   • Error: "Reference document not found"

   • Solution: Ensure the reference document path exists and is accessible

   • Error: "reference_doc must be a '.odt' file when output_format is 'odt'"

   • Solution: The reference document must be the same format as the output. Pandoc does not check this itself: a mismatched reference is silently ignored for DOCX output, and produces an ODT file that cannot be opened

   • Note: Reference documents work with DOCX, ODT and PPTX output formats

   • How to create: pandoc -o reference.docx --print-default-data-file reference.docx, and likewise with reference.odt or reference.pptx

Quickstart

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"
    ]
  }
}

• 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

Testing

To run the comprehensive test suite and validate all supported bidirectional conversions, use the following command:

uv run pytest tests/test_conversions.py

This ensures backward compatibility and verifies the tool's core functionality.

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.