This server gives Claude two document conversion tools: doc_to_markdown pulls text, formatting, tables, and images from DOCX or PDF files into Markdown, while markdown_to_docx goes the other direction with opinionated academic styling (Times New Roman, SimSun, three-line tables). It's built on Pandoc for the Markdown to DOCX direction, so you'll need that installed. Reach for this when you want Claude to extract content from uploaded documents for analysis or rewriting, or when you need to generate formatted Word documents from Markdown that Claude produces. Also ships as a standalone CLI if you want to use it outside the MCP context.
Convert between document formats — DOCX/PDF to Markdown, Markdown to DOCX.
An MCP server for AI agents, with CLI support.
| Tool | Direction | Description |
|---|---|---|
doc_to_markdown | DOCX/PDF → MD | Extract text, formatting, tables, images |
markdown_to_docx | MD → DOCX | Academic styling: Times New Roman + SimSun, three-line tables |
markdown_to_docx)pip install shuck-convert
Or from source:
git clone https://github.com/Shan-Zhu/shuck-convert.git
cd shuck-convert
pip install -e .
Add to your MCP client config:
{
"mcpServers": {
"shuck-convert": {
"command": "shuck-convert",
"args": [],
"transportType": "stdio"
}
}
}
Or with uvx:
{
"mcpServers": {
"shuck-convert": {
"command": "uvx",
"args": ["shuck-convert"],
"transportType": "stdio"
}
}
}
# DOCX/PDF to Markdown
shuck-convert doc2md report.docx
shuck-convert doc2md paper.pdf
# Markdown to DOCX
shuck-convert md2docx paper.md
pip install -e .
fastmcp dev src/shuck_convert/server.py
MIT
csoai-org/pdf-document-mcp
xt765/mcp-document-converter
io.github.xjtlumedia/markdown-formatter
io.github.ai-aviate/better-notion
suekou/mcp-notion-server
meterlong/mcp-doc