MD-DOCX Converter
Bidirectional Markdown ↔ Word (.docx) converter. Read Word documents directly into Claude as Markdown, or save Claude's output as a properly formatted .docx with heading styles (Title, Heading 1–9), bold, italic, tables, lists, code blocks, and images. No cloud upload — runs entirely on your machine using python.
README
MD-DOCX Converter
A Python tool for bidirectional conversion between Markdown (.md) and Microsoft Word (.docx). Designed to make it easy to move content between Word documents and AI tools like Claude, ChatGPT, and GitHub Copilot.
What it does
- Converts
.md→.docxwith correct heading hierarchy (Title, Heading 1–9) - Converts
.docx→.mdas clean GitHub Flavored Markdown (GFM) - Runs from a simple desktop shortcut — no command line knowledge needed
- Handles headings, bold/italic/strikethrough, lists, task lists, tables, blockquotes, code blocks, images, and hyperlinks
See MarkdownSyntax.md for the full element mapping and notes on what is preserved, approximated, or dropped.
Requirements
- Windows 10/11
- Python 3.11+
- The following Python packages (installed via pip):
pip install markdown-it-py python-docx
Setup
1. Clone the repository
git clone https://github.com/cjwpenner/md-docx-converter.git
cd md-docx-converter
2. Install dependencies
pip install markdown-it-py python-docx
3. Create the desktop shortcut
pip install pywin32
python create_shortcut.py
This creates an MD-DOCX Converter shortcut on your Windows desktop. pywin32 is only needed to create the shortcut — it is not required to run the converter itself.
4. Run the converter
Double-click MD-DOCX Converter on your desktop. A console window opens and prompts:
MD ↔ DOCX Converter
--------------------
Enter file path:
Paste or type the full path to your .md or .docx file and press Enter. The converted file is saved in the same directory with the extension swapped.
You can also run directly from the command line:
python md_docx_converter/converter.py
Conversion notes
Heading hierarchy
The heading level mapping is context-dependent:
- MD → DOCX: If there is exactly one
#in the document, it becomes a Word Title. All other headings shift down by one level. If there are multiple#headings, they all become Heading 1 with no Title. - DOCX → MD: If the document has a Title style, it becomes
#. All headings shift up accordingly. If there is no Title, Heading 1 becomes#.
Lossy elements
Word formatting that has no Markdown equivalent is approximated as bold:
| Word formatting | Markdown output |
|---|---|
| Underline | **bold** |
| Highlight | **bold** |
| Small caps | **bold** |
| Font colour | Stripped (text kept) |
Images
- DOCX → MD: Embedded images are extracted to a
{filename}_images/folder next to the output.mdfile. - MD → DOCX: Images referenced by relative path are re-embedded. Missing images become
[image not found: path].
MCP server (Claude / AI integration)
This tool is also available as an MCP server, letting Claude and other AI assistants read and write Word documents directly.
Install
pip install mcp-md-docx
Configure Claude Desktop
Add to %APPDATA%\Claude\claude_desktop_config.json:
{
"mcpServers": {
"md-docx": {
"command": "python",
"args": ["-m", "mcp_md_docx"]
}
}
}
Configure Claude Code
claude mcp add md-docx mcp-md-docx
Tools exposed
| Tool | What it does |
|---|---|
read_docx |
Read a .docx file — returns full Markdown text to the AI |
write_docx |
Create a .docx from Markdown text the AI has written |
convert_md_file_to_docx |
Convert a .md file on disk to .docx |
convert_docx_file_to_md |
Convert a .docx file on disk to .md |
Once configured, you can say things like:
- "Read
report.docxand summarise it" - "Turn this into a Word document and save it to my Desktop"
- "Convert all the bullet points in
notes.docxinto a table"
Project structure
md_docx_converter/
├── converter.py # CLI entry point
├── md_to_docx.py # Markdown → Word conversion
├── docx_to_md.py # Word → Markdown conversion
├── heading_mapper.py # Heading hierarchy pre-scan logic
├── image_handler.py # Image extraction and embedding
└── launch.pyw # Desktop shortcut launcher
mcp_md_docx/
├── server.py # MCP server (four tools)
└── __main__.py # Entry point for python -m mcp_md_docx
create_shortcut.py # One-time shortcut setup script
pyproject.toml # PyPI packaging config
License
This project is licensed under the GNU General Public License v3.0 (GPLv3). You are free to use, modify, and distribute this software, provided that any derivative works are also distributed under the same licence.
See LICENSE for the full licence text.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.