docx-comments
Provides comprehensive read/write access to Word documents, including comments, track changes, and reply threads.
README
docx-comments-mcp
An MCP server for Claude Desktop that provides comprehensive read/write access to Word documents, including comments, track changes, and reply threads — features that python-docx doesn't fully expose.
<!-- mcp-name: io.github.kosh-jelly/docx-comments -->
Features
- Read documents: Extract text, comments (with reply threads), and track changes
- Add comments: Anchor comments to specific text in the document
- Reply to comments: Create threaded replies on existing comments
- Track changes: Make edits with insertions and deletions tracked
- Resolve comments: Mark comments as done
- Accept/reject changes: Apply or undo tracked changes
Installation
# Clone the repository
git clone https://github.com/your-username/docx-comments-mcp.git
cd docx-comments-mcp
# Install with uv
uv sync
Usage with Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"docx-comments": {
"command": "uv",
"args": ["--directory", "/path/to/docx-comments-mcp", "run", "docx-comments-mcp"]
}
}
}
Available Tools
read_document
Read a Word document and extract content, comments, and track changes.
Parameters:
path(required): Path to the .docx fileinclude_text(default: true): Include full document textinclude_comments(default: true): Include comments with anchorsinclude_track_changes(default: true): Include insertions/deletions
Returns:
{
"metadata": {
"path": "/path/to/file.docx",
"author": "Original Author",
"created": "2025-01-15T10:30:00Z",
"modified": "2025-01-18T14:22:00Z",
"word_count": 4523
},
"paragraphs": [
{"index": 0, "text": "The paragraph content...", "style": "Heading 1"}
],
"comments": [
{
"id": 0,
"author": "Dr. Smith",
"date": "2025-01-16T09:15:00Z",
"text": "Consider citing Main & Hesse here",
"anchor_text": "disorganized attachment patterns",
"anchor_paragraph": 12,
"resolved": false,
"replies": [
{
"id": 1,
"parent_id": 0,
"author": "Josh",
"date": "2025-01-17T11:00:00Z",
"text": "Added citation — see revision"
}
]
}
],
"track_changes": [
{
"id": 5,
"type": "deletion",
"author": "Dr. Smith",
"date": "2025-01-16T09:20:00Z",
"text": "invariably",
"paragraph": 8
}
]
}
create_comment
Add a comment anchored to specific text in a Word document.
Parameters:
path(required): Path to the .docx fileanchor_text(required): Text to anchor the comment to (must exist and be unique)comment_text(required): The comment contentauthor(default: "Claude"): Comment author nameoutput_path(optional): Save to new file; if omitted, creates timestamped backup and overwrites
Returns:
{
"success": true,
"comment_id": 3,
"anchored_to": "the exact text that was matched",
"paragraph": 15,
"output_path": "/path/to/output.docx"
}
create_reply
Add a reply to an existing comment.
Parameters:
path(required): Path to the .docx fileparent_comment_id(required): ID of comment to reply toreply_text(required): The reply contentauthor(default: "Claude"): Reply author nameoutput_path(optional): Save to new file; if omitted, creates backup
create_track_change
Make an edit with track changes enabled (insertion, deletion, or replacement).
Parameters:
path(required): Path to the .docx filefind_text(required): Text to find and modifyreplace_with(required): Replacement text (empty string for deletion)author(default: "Claude"): Change author nameoutput_path(optional): Save to new file; if omitted, creates backup
mark_comment_resolved
Mark a comment as resolved/done.
Parameters:
path(required): Path to the .docx filecomment_id(required): ID of comment to resolveoutput_path(optional): Save to new file; if omitted, creates backup
accept_change
Accept a tracked change (apply permanently).
Parameters:
path(required): Path to the .docx filechange_id(required): ID of the track change to acceptoutput_path(optional): Save to new file; if omitted, creates backup
reject_change
Reject a tracked change (undo the change).
Parameters:
path(required): Path to the .docx filechange_id(required): ID of the track change to rejectoutput_path(optional): Save to new file; if omitted, creates backup
Safety Features
- Automatic backups: When modifying a file without specifying
output_path, a timestamped backup is created (e.g.,document.backup_20250119_143022.docx) - Atomic writes: Uses temporary files and atomic moves to prevent corruption
- Unique anchor matching: Comments require unique anchor text to prevent ambiguity
Development
# Install dev dependencies
uv sync
uv pip install pytest pytest-asyncio
# Run tests
uv run pytest -v
# Run specific test file
uv run pytest tests/test_reader.py -v
Architecture
src/docx_comments_mcp/
├── __init__.py
├── server.py # MCP server with tool definitions
├── reader.py # Read operations (document, comments, track changes)
├── writer.py # Write operations (add comments, track changes)
└── xml_helpers.py # Low-level OOXML parsing utilities
License
MIT
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.