Office Document Processing MCP Server

An MCP (Model Context Protocol) server for Office document processing, enabling creation and editing of Word, Excel, and PowerPoint documents within MCP Clients without leaving the AI assistant environment.

Overview

Office-Editor-MCP implements the Model Context Protocol standard to expose Office document operations as tools and resources. It serves as a bridge between AI assistants and Microsoft Office documents, allowing you to create, edit, format, and analyze various Office documents through AI assistants.

Features

Word Document Operations

Document Management

• Create new Word documents with metadata (title, author, etc.)
• Extract text content and analyze document structure
• View document properties and statistics
• List available documents in a directory
• Create document copies

Content Creation

• Add headings with different levels
• Insert paragraphs with optional styling
• Create tables with custom data
• Add images with proportional scaling
• Insert page breaks

Text Formatting

• Format specific text sections (bold, italic, underline)
• Change text color and font properties
• Apply custom styles to text elements
• Search and replace text throughout documents

Excel Operations

Workbook Management

• Create new Excel workbooks
• Open existing Excel files
• Add/delete/rename worksheets

Data Processing

• Read and write cell contents
• Insert/delete rows and columns
• Sort and filter data
• Apply formulas and functions

PowerPoint Operations

Presentation Management

• Create new PowerPoint presentations
• Add/delete/rearrange slides
• Set slide themes and backgrounds

Content Editing

• Add text and graphic elements
• Insert tables and charts
• Add animations and transitions

Advanced Features

• OCR recognition (extract text from images)
• Document comparison (compare differences between documents)
• Document translation
• Document encryption and decryption
• Table data import/export (database interaction)

Installation Guide

Prerequisites

• Python 3.10 or higher
• pip package manager
• Microsoft Office is only required for Windows COM features such as PDF export, animations, transitions, and legacy Office formats
• Tesseract OCR is required when using ocr_recognize_text

Basic Installation

# Clone the repository
git clone https://github.com/theWDY/office-editor-mcp.git
cd office-editor-mcp

# Install dependencies
pip install -r requirements.txt

The repository exposes four independent MCP servers: word_server.py, excel_server.py, powerpoint_server.py, and general_server.py. Configure only the servers you need. The checked-in .cursor/mcp.json contains a complete four-server example.

Configuration

Configuration in Cursor

Method 1: UI Configuration

1. Open Cursor
2. Go to Settings > Features > MCP
3. Click "+ Add New MCP Server"
4. Fill in the configuration information:
   • Name: Office Assistant (modify as preferred)
   • Type: Select stdio
   • Command: Enter the full path to run the server, for example:

     python /path/to/office-editor-mcp/word_server.py
     

     Note: Replace with your actual file path

Method 2: JSON Configuration File (Recommended)

1. Create a .cursor folder in the project directory (if it doesn't exist)
2. Create an mcp.json file in that folder with the following content:

{
  "mcpServers": {
    "office-assistant": {
      "command": "python",
      "args": ["/path/to/office-editor-mcp/word_server.py"],
      "env": {"OFFICE_EDIT_PATH": "/path/to/output"}
    }
  }
}

Configuration in Claude for Desktop

1. Edit the Claude configuration file:

   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json

2. Add the following configuration:

{
  "mcpServers": {
    "office-document-server": {
      "command": "python",
      "args": [
        "/path/to/office-editor-mcp/word_server.py"
      ]
    }
  }
}

1. Restart Claude to apply the configuration.

All paths handled by general_server.py must stay inside OFFICE_EDIT_PATH. Move and delete operations are disabled unless OFFICE_ALLOW_DESTRUCTIVE=true is explicitly configured.

New encrypted files use a versioned Fernet container with a random salt and Scrypt password derivation. Existing encrypted files remain readable and can be re-encrypted to upgrade them. Batch operations are limited to 100 files and 16 workers per request.

Usage Examples

After configuration, you can issue commands to your AI assistant like:

Word Document Operations

• "Create a new document called 'quarterly_report.docx' with a title page"
• "Add a heading and three paragraphs to the document"
• "Insert a 4x4 table with sales data"
• "Make the word 'important' in paragraph 2 bold and red"
• "Search and replace all instances of 'old term' with 'new term'"

Excel Operations

• "Create a new Excel workbook named 'financial_analysis.xlsx'"
• "Insert 'Quarterly Sales' as a title in cell A1"
• "Create a table with department sales data and calculate the sum"
• "Create a bar chart for the sales data"
• "Sort the data in column B in descending order"

PowerPoint Operations

• "Create a presentation named 'project_presentation.pptx'"
• "Add a new slide with the title 'Project Overview'"
• "Insert the company logo in slide 2"
• "Add a fly-in animation to the title"

API Reference

Word Document Operations

# Document Creation and Properties
create_document(filename, title=None, author=None)
get_document_info(filename)
get_document_text(filename)
get_document_outline(filename)
list_available_documents(directory=".")
copy_document(source_filename, destination_filename=None)

# Content Addition
add_heading(filename, text, level=1)
add_paragraph(filename, text, style=None)
add_table(filename, rows, cols, data=None)
add_picture(filename, image_path, width=None)
add_page_break(filename)

# Text Formatting
format_text(filename, paragraph_index, start_pos, end_pos, bold=None, 
            italic=None, underline=None, color=None, font_size=None, font_name=None)
search_and_replace(filename, find_text, replace_text)
delete_paragraph(filename, paragraph_index)
create_custom_style(filename, style_name, bold=None, italic=None, 
                    font_size=None, font_name=None, color=None, base_style=None)

Excel Operations

# Workbook Operations
create_workbook(filename)
open_workbook(filename)
save_workbook(filename, new_filename=None)
add_worksheet(filename, sheet_name=None)
list_worksheets(filename)

# Cell Operations
read_cell(filename, sheet_name, cell_reference)
write_cell(filename, sheet_name, cell_reference, value)
format_cell(filename, sheet_name, cell_reference, **format_args)

PowerPoint Operations

# Presentation Operations
create_presentation(filename)
open_presentation(filename)
save_presentation(filename, new_filename=None)
add_slide(filename, layout=None)

Troubleshooting

Common Issues

1. Missing Styles

   • Some documents may lack required styles for heading and table operations
   • The server will attempt to create missing styles or use direct formatting
   • For best results, use templates with standard Office styles

2. Permission Issues

   • Ensure the server has permission to read/write to document paths
   • Use the copy_document function to create editable copies of locked documents
   • Check file ownership and permissions if operations fail

3. Image Insertion Problems

   • Use absolute paths for image files
   • Verify image format compatibility (JPEG, PNG recommended)
   • Check image file size and permissions

Debugging

Enable detailed logging by setting the environment variable:

export MCP_DEBUG=1  # Linux/macOS
set MCP_DEBUG=1     # Windows

Implementation Progress

• ✅ Build MCP server basic framework
• ✅ Successful integration with AI assistants
• ✅ Basic Word document operations
• ✅ Basic Excel workbook operations
• ✅ Basic PowerPoint presentation operations
• ✅ Advanced features enhancement
• ✅ Performance optimization
• ✅ Cross-platform compatibility testing

Contributing

Contributions are welcome! Feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Model Context Protocol for protocol specification
• python-docx for Word document processing
• openpyxl for Excel processing
• python-pptx for PowerPoint processing

Note: This server interacts with document files on your system. Always verify that requested operations are appropriate before confirming them in AI assistants or other MCP clients.