Docsmith MCP

Python-powered document processing MCP — Process Excel, Word, PDF documents with ease using Python.

Features

• Excel: Read/write .xlsx files with sheet support and pagination
• Word: Read/write .docx files with paragraph and table support
• PDF: Read .pdf files with text extraction and pagination
• Text Files: Read/write .txt, .csv, .md, .json, .yaml, .yml with pagination support
• Flexible Reading Modes: Raw full read or paginated for large files
• Powered by Pyodide: Runs in secure WebAssembly sandbox via code-runner-mcp

Installation

pnpm install
pnpm build

Configuration

Add to your MCP client configuration:

{
  "mcpServers": {
    "docsmith": {
      "command": "node",
      "args": ["/path/to/docsmith-mcp/dist/index.js"],
      "env": {
        "DOC_RAW_FULL_READ": "false",
        "DOC_PAGE_SIZE": "100",
        "DOC_MAX_FILE_SIZE": "50"
      }
    }
  }
}

Environment Variables

Variable	Description	Default
DOC_RAW_FULL_READ	Enable full raw read mode	false
DOC_PAGE_SIZE	Default items per page	100
DOC_MAX_FILE_SIZE	Max file size in MB	50

Tools

read_document

Read document content with automatic format detection.

{
  "file_path": "/path/to/document.xlsx",
  "mode": "paginated",
  "page": 1,
  "page_size": 50,
  "sheet_name": "Sheet1"
}

write_document

Write document content.

{
  "file_path": "/path/to/output.xlsx",
  "format": "excel",
  "data": [["Header1", "Header2"], ["Value1", "Value2"]]
}

get_document_info

Get document metadata.

{
  "file_path": "/path/to/document.pdf"
}

Architecture

docsmith-mcp/
├── python/              # Python handler scripts
│   ├── excel_handler.py
│   ├── word_handler.py
│   └── pdf_handler.py
├── src/
│   ├── index.ts        # MCP server
│   └── code-runner.ts  # code-runner-mcp client
└── dist/               # Built output

Python scripts are executed via code-runner-mcp in a Pyodide WebAssembly environment.

License

MIT