Document Operations MCP Server

Language / 语言: English | 中文

Document Operations MCP Server - A universal MCP server for document processing, conversion, and automation. Handle PDF, DOCX, HTML, Markdown, and more through a unified API and toolset.

Demo

Video

https://github.com/user-attachments/assets/43dfeeec-8097-413e-8519-a7de98e31136

In this demo, we showcase how to:

• Configure doc-ops-mcp in MCP clients
• Convert DOCX documents to PDF format
• Add default watermarks to converted PDF files

Table of Contents

1. Quick Start
2. System Architecture
3. Optional Integration
4. Features
5. Open Source Licenses
6. Future Roadmap
7. Docker Deployment
8. Development Guide
9. Troubleshooting
10. Contributing

1. Quick Start

First, add the Document Operations MCP server to your MCP client.

Standard config works in most MCP clients:

{
  "mcpServers": {
    "doc-ops-mcp": {
      "command": "npx",
      "args": ["-y", "doc-ops-mcp"],
      "env": {
        "OUTPUT_DIR": "/path/to/your/output/directory",
        "CACHE_DIR": "/path/to/your/cache/directory",
      }
    }
  }
}

<details> <summary>Claude Desktop</summary>

Follow the MCP install guide, use the standard config above.

</details> <details> <summary>VS Code</summary>

Follow the MCP install guide, use the standard config above.

</details> <details> <summary>Cursor</summary>

Go to Cursor Settings -> MCP -> Add new MCP Server. Name to your liking, use command type with the command npx -y doc-ops-mcp.

</details> <details> <summary>Other MCP Clients</summary>

For other MCP clients, use the standard config above and refer to your client's documentation for MCP server installation.

</details>

Configuration

The Document Operations MCP server supports configuration through environment variables. These can be provided in the MCP client configuration as part of the "env" object:

{
  "mcpServers": {
    "doc-ops-mcp": {
      "command": "npx",
      "args": ["-y", "doc-ops-mcp"],
      "env": {
        "OUTPUT_DIR": "/path/to/your/output/directory",
        "CACHE_DIR": "/path/to/your/cache/directory",
        "WATERMARK_IMAGE": "/path/to/watermark.png",
        "QR_CODE_IMAGE": "/path/to/qrcode.png"
      }
    }
  }
}

Supported Document Operations

Rewriting Features:

• Content Replacement: Support batch text replacement and regular expression replacement
• Format Adjustment: Modify document structure, heading levels, and style formatting
• Smart Rewriting: Content optimization while preserving original document format

Usage Examples

Format Conversion:

Convert /Users/docs/report.docx to PDF
Convert /Users/docs/article.md to HTML
Convert /Users/docs/presentation.html to DOCX
Convert /Users/docs/readme.md to PDF (with theme styling)

Document Rewriting:

Rewrite company names in /Users/docs/contract.md
Batch replace terminology in /Users/docs/manual.docx
Adjust heading levels in /Users/docs/article.html
Update dates and version numbers in /Users/docs/policy.md

PDF Enhancement:

Add watermark to /Users/docs/document.pdf
Add QR code to /Users/docs/report.pdf
Add company logo watermark to /Users/docs/invoice.pdf

Environment Variables

The server supports environment variables for controlling output paths and PDF enhancement features:

Core Directories

• OUTPUT_DIR: Controls where all generated files are saved (default: ~/Documents)
• CACHE_DIR: Directory for temporary and cache files (default: ~/.cache/doc-ops-mcp)

PDF Enhancement Features

• WATERMARK_IMAGE: Default watermark image path for PDF files
  • Automatically added to all PDF conversions
  • Supported formats: PNG, JPG
  • If not set, default text watermark "doc-ops-mcp" will be used
• QR_CODE_IMAGE: Default QR code image path for PDF files
  • Added to PDFs only when explicitly requested (addQrCode=true)
  • Supported formats: PNG, JPG
  • If not set, QR code functionality will be unavailable

Output Path Rules:

1. If outputPath is not provided → files saved to OUTPUT_DIR with auto-generated names
2. If outputPath is relative → resolved relative to OUTPUT_DIR
3. If outputPath is absolute → used as-is, ignoring OUTPUT_DIR

See OUTPUT_PATH_CONTROL.md for detailed documentation.

2. System Architecture

Document Operations MCP Server adopts a pure JavaScript architecture design, providing complete document processing capabilities:

┌─────────────────────────────────────────────────────────────┐
│                    MCP Client Layer                         │
│           (Claude Desktop, Cursor, VS Code, etc.)           │
└─────────────────────┬───────────────────────────────────────┘
                      │ JSON-RPC 2.0
┌─────────────────────┴───────────────────────────────────────┐
│                 Doc-Ops-MCP Server                         │
│  ┌─────────────────┐  ┌─────────────────┐  ┌─────────────┐ │
│  │   Tool Router   │  │  Request        │  │  Response   │ │
│  │   & Handler     │  │  Validator      │  │  Formatter  │ │
│  └────────┬────────┘  └────────┬────────┘  └──────┬──────┘ │
│           │                    │                  │        │
│  ┌────────┴────────────────────┴──────────────────┴─────┐ │
│  │                Document Processing Engine             │ │
│  │  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐   │ │
│  │  │  Document   │  │   Format    │  │   Style     │   │ │
│  │  │   Reader    │  │  Converter  │  │  Processor  │   │ │
│  │  └─────────────┘  └─────────────┘  └─────────────┘   │ │
│  │                                                        │ │
│  │  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐   │ │
│  │  │    PDF      │  │  Watermark/ │  │ Conversion  │   │ │
│  │  │ Enhancement │  │   QR Code   │  │  Planner    │   │ │
│  │  └─────────────┘  └─────────────┘  └─────────────┘   │ │
└────┴───────────────────────────────────────────────────────┴─┘
                            │
┌───────────────────────────┴─────────────────────────────────┐
│                    Core Dependencies Layer                  │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐          │
│  │   pdf-lib   │  │word-extractor│  │   marked    │          │
│  │ (PDF Tools) │  │(DOCX Reader)│  │ (Markdown)  │          │
│  └─────────────┘  └─────────────┘  └─────────────┘          │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐          │
│  │   cheerio   │  │    jszip    │  │    docx     │          │
│  │(HTML Parser)│  │(ZIP Handler)│  │(DOCX Gen.)  │          │
│  └─────────────┘  └─────────────┘  └─────────────┘          │
│  ┌─────────────┐  ┌─────────────┐                           │
│  │   xml2js    │  │Custom OOXML │                           │
│  │(XML Parser) │  │   Parser    │                           │
│  └─────────────┘  └─────────────┘                           │
└─────────────────────────────────────────────────────────────┘

Architecture Overview

Core Features:

• Pure JavaScript implementation with no external system dependencies
• Complete document reading, conversion, and style processing capabilities
• Built-in PDF watermark and QR code addition functionality
• Intelligent conversion planning and path optimization

Conversion Flow:

• Direct Conversion: Supports direct conversion between most formats
• Multi-step Conversion: Complex conversions achieved through intermediate formats
• Style Preservation: Uses OOXML parser to ensure complete style integrity

3. Optional Integration

This server can work with playwright-mcp for enhanced PDF conversion capabilities. Please refer to the official playwright-mcp documentation for detailed configuration.

🔧 PDF Conversion Workflow

This server supports complete PDF conversion functionality:

1. Document Parsing: Use OOXML parser to ensure complete style preservation
2. Format Conversion: Convert documents to high-quality HTML format
3. PDF Generation: Built-in converter or optionally work with playwright-mcp
4. Enhancement Processing: Automatically add watermarks and QR codes (if configured)

How It Works

This server uses intelligent conversion architecture:

1. Smart Planning: plan_conversion analyzes conversion requirements and selects optimal paths
2. Format Conversion: Use specialized converters to handle various document formats
3. Style Preservation: Ensure style integrity through OOXML parser
4. Enhancement Processing: Automatically add watermarks, QR codes and other enhancements
5. Optional Integration: Support working with playwright-mcp for enhanced capabilities

4. Features

MCP Tools

Core Document Tools

read_document

Read various document formats including PDF, DOCX, DOC, HTML, MD, and more.

Parameters:

• filePath (string, required) - Document path to read
• extractMetadata (boolean, optional) - Extract document metadata, defaults to false
• preserveFormatting (boolean, optional) - Preserve formatting (HTML output), defaults to false

write_document

Write content to document files in specified formats.

Parameters:

• content (string, required) - Content to write
• outputPath (string, optional) - Output file path (auto-generated if not provided)
• encoding (string, optional) - File encoding, defaults to utf-8

convert_document

Convert documents between formats with enhanced style preservation.

Parameters:

• inputPath (string, required) - Input file path
• outputPath (string, optional) - Output file path (auto-generated if not provided)
• preserveFormatting (boolean, optional) - Preserve formatting, defaults to true
• useInternalPlaywright (boolean, optional) - Use built-in Playwright for PDF conversion, defaults to false

convert_docx_to_pdf

Convert DOCX to PDF with automatic watermark addition (if configured).

Parameters:

• docxPath (string, required) - DOCX file path
• outputPath (string, optional) - Output PDF path (auto-generated if not provided)
• addQrCode (boolean, optional) - Whether to add QR code, defaults to false
• preserveFormatting (boolean, optional) - Preserve original formatting, defaults to true
• chineseFont (string, optional) - Chinese font, defaults to Microsoft YaHei

convert_markdown_to_pdf

Convert Markdown to PDF with automatic watermark addition (if configured).

Parameters:

• markdownPath (string, required) - Markdown file path
• outputPath (string, optional) - Output PDF path (auto-generated if not provided)
• theme (string, optional) - Theme style, defaults to "github"
• includeTableOfContents (boolean, optional) - Include table of contents, defaults to false
• addQrCode (boolean, optional) - Whether to add QR code, defaults to false

convert_markdown_to_html

Convert Markdown to HTML.

Parameters:

• markdownPath (string, required) - Markdown file path
• outputPath (string, optional) - Output HTML path (auto-generated if not provided)
• theme (string, optional) - Theme style, defaults to "github"
• includeTableOfContents (boolean, optional) - Include table of contents, defaults to false

convert_markdown_to_docx

Convert Markdown to DOCX.

Parameters:

• markdownPath (string, required) - Markdown file path
• outputPath (string, optional) - Output DOCX path (auto-generated if not provided)

convert_html_to_markdown

Convert HTML to Markdown.

Parameters:

• htmlPath (string, required) - HTML file path
• outputPath (string, optional) - Output Markdown path (auto-generated if not provided)

plan_conversion

🎯 Smart Conversion Planner - Analyze conversion requirements and generate optimal conversion plans.

Parameters:

• sourceFormat (string, required) - Source file format (pdf, docx, html, markdown, md, txt, doc)
• targetFormat (string, required) - Target file format (pdf, docx, html, markdown, md, txt, doc)
• sourceFile (string, optional) - Source file path (for generating specific conversion parameters)
• preserveStyles (boolean, optional) - Whether to preserve style formatting, defaults to true
• includeImages (boolean, optional) - Whether to include images, defaults to true
• theme (string, optional) - Conversion theme, defaults to github
• quality (string, optional) - Conversion quality requirement (fast, balanced, high), defaults to balanced

process_pdf_post_conversion

Parameters:

• playwrightPdfPath (string, required) - Generated PDF file path
• targetPath (string, optional) - Target PDF file path (auto-generated if not provided)
• addWatermark (boolean, optional) - Whether to add watermark, defaults to false
• addQrCode (boolean, optional) - Whether to add QR code, defaults to false
• watermarkImage (string, optional) - Watermark image path
• qrCodePath (string, optional) - QR code image path

PDF Enhancement Tools

add_watermark

🎨 PDF Watermark Addition Tool - Add image or text watermarks to PDF documents.

Parameters:

• pdfPath (string, required) - PDF file path
• watermarkImage (string, optional) - Watermark image path (PNG/JPG)
• watermarkText (string, optional) - Watermark text content
• watermarkImageScale (number, optional) - Image scale ratio, defaults to 0.25
• watermarkImageOpacity (number, optional) - Image opacity, defaults to 0.6
• watermarkImagePosition (string, optional) - Image position, defaults to fullscreen

add_qrcode

📱 PDF QR Code Addition Tool - Add QR codes to PDF documents.

Parameters:

• pdfPath (string, required) - PDF file path
• qrCodePath (string, optional) - QR code image path
• qrScale (number, optional) - QR code scale ratio, defaults to 0.15
• qrOpacity (number, optional) - QR code opacity, defaults to 1.0
• qrPosition (string, optional) - QR code position, defaults to bottom-center
• addText (boolean, optional) - Whether to add explanatory text, defaults to true

System Requirements

System Requirements

• Node.js ≥ 18.0.0
• Zero external system dependencies - All processing via npm packages
• Optional Integration: playwright-mcp for enhanced PDF conversion

Core Technology Stack

• pdf-lib - PDF operations and enhancement
• word-extractor - DOCX document text extraction
• marked - Markdown parsing and rendering
• cheerio - HTML parsing and manipulation
• docx - DOCX document generation
• jszip - ZIP file processing
• xml2js - XML parsing and conversion
• Custom OOXML Parser - Advanced DOCX style preservation

Installation

# Global installation
npm install -g doc-ops-mcp

# Or using pnpm
pnpm add -g doc-ops-mcp

# Or using bun
bun add -g doc-ops-mcp

Architecture Components

• MCP Server Core: Handles JSON-RPC 2.0 communication and tool registration
• Smart Router: Routes requests to optimal processing modules
• Conversion Engine: Contains specialized converters for different document types
• Style Processor: Ensures style preservation during format conversion
• Security Module: Provides path validation and content security handling

5. Open Source Licenses

Project License

• This Project: MIT License
• Compatibility: Available for commercial and non-commercial use

Third-Party Dependencies

License Compatibility

• ✅ Commercial Use: All dependencies support commercial use
• ✅ Distribution: Free to distribute and modify
• ✅ Patent Protection: Apache-2.0 provides patent protection
• ⚠️ Notice: Original license notices must be retained

6. Future Roadmap

Core Features

• 🔄 Enhanced Conversion Quality: Improve style preservation for complex documents
• 📊 Excel Support: Complete Excel read/write and conversion functionality
• 🎨 Template System: Support for custom document templates
• 🔍 OCR Integration: Image text recognition capabilities

System Improvements

• 🌐 Multi-language Support: Internationalization and localization
• 🔐 Security Enhancements: Document encryption and access control
• ⚡ Performance Optimization: Large file handling and memory optimization
• 🔌 Plugin System: Extensible processor architecture

Version Roadmap

• v2.0: Complete Excel support and template system
• v3.0: OCR integration and multi-language support
• v4.0: Advanced security features and plugin system

7. Docker Deployment

Quick Start

Using Pre-built Image

# Pull the latest image
docker pull docops/doc-ops-mcp:latest

# Run with default configuration
docker run -d \
  --name doc-ops-mcp \
  -p 3000:3000 \
  docops/doc-ops-mcp:latest

Building from Source

# Clone the repository
git clone https://github.com/JefferyMunoz/doc-ops-mcp.git
cd doc-ops-mcp

# Build the Docker image
docker build -t doc-ops-mcp .

# Run the container
docker run -d \
  --name doc-ops-mcp \
  -p 3000:3000 \
  -v $(pwd)/documents:/app/documents \
  doc-ops-mcp

Docker Compose Deployment

Create a docker-compose.yml file:

version: '3.8'

services:
  doc-ops-mcp:
    image: docops/doc-ops-mcp:latest
    container_name: doc-ops-mcp
    ports:
      - "3000:3000"
    volumes:
      - ./documents:/app/documents
      - ./config:/app/config
    environment:
      - NODE_ENV=production
      - PORT=3000
    restart: unless-stopped
    
  # Optional: Add Nginx for reverse proxy
  nginx:
    image: nginx:alpine
    container_name: doc-ops-nginx
    ports:
      - "80:80"
    volumes:
      - ./nginx.conf:/etc/nginx/nginx.conf:ro
    depends_on:
      - doc-ops-mcp
    restart: unless-stopped

Environment Variables

Volume Mounts

Mount local directories for persistent storage:

# Documents directory for file processing
docker run -d \
  --name doc-ops-mcp \
  -p 3000:3000 \
  -v $(pwd)/documents:/app/documents \
  -v $(pwd)/output:/app/output \
  doc-ops-mcp

Docker Configuration Examples

Production Deployment

# Production setup with Docker Swarm
docker swarm init
docker stack deploy -c docker-compose.yml doc-ops

# Scale the service
docker service scale doc-ops_mcp=3

Health Checks

The container includes built-in health checks:

# Check container health
docker ps

# View health check logs
docker inspect --format='{{.State.Health.Status}}' doc-ops-mcp

# Manual health check
docker exec doc-ops-mcp curl -f http://localhost:3000/health || exit 1

8. Development Guide

Local Development

# Clone the repository
git clone https://github.com/your-org/doc-ops-mcp.git
cd doc-ops-mcp

# Install dependencies
npm install

# Run in development mode
npm run dev

# Build the project
npm run build

# Run tests
npm test

Project Structure

src/
├── index.ts          # MCP server entry point
├── tools/            # Tool implementations
│   ├── documentConverter.ts
│   ├── pdfTools.ts
│   └── ...
├── types/            # Type definitions
└── utils/            # Utility functions

Adding New Tools

1. Create a new tool file in src/tools/
2. Implement the tool logic
3. Register the tool in src/index.ts
4. Add test cases
5. Update documentation

9. Troubleshooting

Common Issues

1. Port conflicts: Change the host port in docker-compose.yml
2. Permission issues: Ensure volume mounts have correct permissions
3. Memory issues: Increase Docker memory allocation

Debug Mode

# Run with debug logging
docker run -d \
  --name doc-ops-mcp \
  -p 3000:3000 \
  -e LOG_LEVEL=debug \
  doc-ops-mcp

# View logs
docker logs -f doc-ops-mcp

10. Contributing

How to Contribute

1. Fork the Project
2. Create a Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit Your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Intellectual Property License

By submitting a Pull Request, you agree that all contributions submitted through Pull Requests will be licensed under the MIT License. This means:

• You grant the project maintainers and users the right to use, modify, and distribute your contributions under the MIT License
• You confirm that you have the right to make these contributions
• You understand that your contributions will become part of the open source project
• You waive any claims to exclusive ownership of the contributed code

If you cannot agree to these terms, please do not submit a Pull Request.

Code Standards

• Use TypeScript
• Follow ESLint configuration
• Add appropriate tests
• Update relevant documentation

Reporting Issues

• Use GitHub Issues
• Provide detailed error information and reproduction steps
• Include system environment information

License

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