docx-mcp-server
Enables AI agents to read, edit, and create Microsoft Word documents (.docx) with support for rich text, tables, and images, deployable locally or via SSE.
README
Docx MCP Server 📄
A Model Context Protocol (MCP) server that enables AI agents to read, edit, and create Microsoft Word documents (.docx). It supports rich text, tables, and images, and provides flexible deployment options (Standalone Executable, Python/UVX, SSE/Remote).
✨ Features
- Document Management: Create new documents or load existing ones.
- Content Reading: Read document structure, paragraphs, tables, and images.
- Rich Editing:
- Add stylized paragraphs (Bold, Italic handled via styles).
- Add Headings (Levels 1-9).
- Insert Images (from local path or Base64).
- Create Tables with custom data.
- Image Extraction: Extract images from docx files to local disk.
- Multi-Mode Support:
- StdIO: Standard integration for local MCP clients (e.g., Claude Desktop).
- SSE: Server-Sent Events support for remote or URL-based connections.
🚀 Quick Start
Option 1: Standalone Explorer (No Python Required)
Download the latest release and run the executable directly.
- Get
docx-mcp-server.exefrom the Releases page. - Run in terminal or configure in your MCP client.
Option 2: Using UVX (Requires PyPI Publication)
⚠️ Note: This package is not currently published to PyPI by the author. To use
uvx, you must fork and publish it yourself, or install locally.
If published to PyPI, you could run:
uvx docx-mcp-server
Option 3: Development / Local Source
# Clone the repository
git clone https://github.com/OkamiFeng/docx-mcp-server.git
cd docx-mcp-server
# Run with uv
uv run docx-mcp-server
🛠️ Configuration
Claude Desktop
Using UVX (If published to PyPI)
{
"mcpServers": {
"docx": {
"command": "uvx",
"args": ["docx-mcp-server"]
}
}
}
Windows (using .exe)
{
"mcpServers": {
"docx": {
"command": "C:/Path/To/docx-mcp-server.exe",
"args": []
}
}
}
Windows (using Python/UV)
{
"mcpServers": {
"docx": {
"command": "uv",
"args": ["run", "--directory", "C:/Projects/docx-mcp-server", "docx-mcp-server"]
}
}
}
Remote / SSE Configuration
To run the server in HTTP mode (accessible via URL):
# Using Exe
docx-mcp-server.exe --transport sse --port 8000
# Using Python
uv run server.py --transport sse --port 8000
SSE Endpoint: http://localhost:8000/sse
📚 Tools Available
| Tool Name | Description |
|---|---|
create_new_document |
Create a fresh empty document. |
load_document |
Load an existing .docx file from a path. |
get_document_structure |
Get a simplified JSON-like view of paragraphs and content counts. |
add_paragraph |
Add text paragraph with optional style. |
add_heading |
Add a heading (Level 1-9). |
add_table |
Insert a table with specified rows, cols, and data. |
add_image |
Insert an image from a file path or Base64 string. |
extract_images |
Extract all images from the loaded document to a folder. |
save_document |
Save changes (overwrite or new path). |
🏗️ Development
- Clone:
git clone https://github.com/OkamiFeng/docx-mcp-server.git - Install:
uv syncorpip install -r requirements.txt - Build Exe:
pyinstaller --name docx-mcp-server --onefile server.py --icon="icon.png" --collect-all docx
📄 License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。