mcpField
M
MCP Document Reader

MCP Document Reader

Local
github.com/xt765Productivity23Updated 2w ago

About MCP Document Reader

MCP Document Reader is an MCP server that reads DOCX, PDF, TXT and Excel files so AI assistants can work with local documents. It suits users of Trae IDE or Claude Desktop and can also be imported as a Python library.

Key features

  1. 01Supports XLSX, XLS, DOCX, PDF and TXT files
  2. 02Unified read_document tool interface
  3. 03Selects the reader by file extension
  4. 04Usable as a Python library

Use cases

  1. 01Letting an AI assistant read local Word or PDF reports
  2. 02Analyzing Excel spreadsheet data in a chat
  3. 03Passing document text to an AI for summarization

MCP support

The server exposes a read_document tool over MCP that takes a file path and returns the document content.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"xt765-mcp-documents-reader": {
"command": "uvx",
"args": [
"mcp-documents-reader"
]
}
}
}
Run commanduvx mcp-documents-reader

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypimcp-documents-reader1.3.1

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

MCP Document Reader

MCP (Model Context Protocol) Document Reader - A powerful MCP tool for reading documents in multiple formats, enabling AI agents to truly "read" your documents.

🌐 Language: English | 中文

CSDN GitHub Gitee

License Python PyPI Version PyPI Downloads MCP Registry MCP Marketplace

Features

  • Multi-format Support: Supports 4 mainstream document formats: Excel (XLSX/XLS), DOCX, PDF, and TXT
  • MCP Protocol: Compliant with MCP standards, can be used as a tool for AI assistants like Trae IDE
  • Easy Integration: Simple configuration for immediate use
  • Reliable Performance: Successfully tested and running in Trae IDE
  • File System Support: Reads documents directly from the file system

📚 Documentation

User Guide · API Reference · Contributing · Changelog · License


Architecture

graph TB
    A[AI Assistant / User] -->|Call read_document| B[MCP Document Reader]
    B -->|Detect file type| C{File Type?}
    C -->|.docx| D[DOCX Reader]
    C -->|.pdf| E[PDF Reader]
    C -->|.xlsx/.xls| F[Excel Reader]
    C -->|.txt| G[Text Reader]
    D -->|Extract text| H[Return Content]
    E -->|Extract text| H
    F -->|Extract text| H
    G -->|Extract text| H
    H -->|Text content| A
    
    style A fill:#e1f5ff
    style B fill:#fff4e1
    style C fill:#f0f0f0
    style D fill:#e8f5e9
    style E fill:#e8f5e9
    style F fill:#e8f5e9
    style G fill:#e8f5e9
    style H fill:#fff9c4

Supported Formats

FormatExtensionsMIME TypeFeatures
Excel.xlsx, .xlsapplication/vnd.openxmlformats-officedocument.spreadsheetml.sheetSheet and cell data extraction
DOCX.docxapplication/vnd.openxmlformats-officedocument.wordprocessingml.documentText and structure extraction
PDF.pdfapplication/pdfText extraction
Text.txttext/plainPlain text reading

Installation

Using pip (Recommended)
pip install mcp-documents-reader
From Source
git clone https://github.com/xt765/mcp_documents_reader.git
cd mcp_documents_reader
pip install -e .

MCP Tools

This server provides the following tool:

read_document

Read any supported document type with a unified interface.

Arguments:

  • filename (string, required): Document file path, supports absolute or relative paths.

Configuration

Using in Trae IDE / Claude Desktop

Add the following to your MCP configuration file:

Option 1: Using PyPI (Recommended)

{
  "mcpServers": {
    "mcp-document-reader": {
      "command": "uvx",
      "args": [
        "mcp-documents-reader"
      ]
    }
  }
}

Option 2: Using GitHub repository

{
  "mcpServers": {
    "mcp-document-reader": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/xt765/mcp_documents_reader",
        "mcp_documents_reader"
      ]
    }
  }
}

Option 3: Using Gitee repository (Faster access in China)

{
  "mcpServers": {
    "mcp-document-reader": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://gitee.com/xt765/mcp_documents_reader",
        "mcp_documents_reader"
      ]
    }
  }
}

Usage

As an MCP Tool

After configuration, AI assistants can directly call the following tool:

# Read a DOCX file
read_document(filename="example.docx")

# Read a PDF file
read_document(filename="example.pdf")

# Read an Excel file
read_document(filename="example.xlsx")

# Read a text file
read_document(filename="example.txt")
As a Python Library
from mcp_documents_reader import DocumentReaderFactory

# Using factory (recommended)
reader = DocumentReaderFactory.get_reader("document.pdf")
content = reader.read("/path/to/document.pdf")

# Check if format is supported
if DocumentReaderFactory.is_supported("file.xlsx"):
    reader = DocumentReaderFactory.get_reader("file.xlsx")
    content = reader.read("/path/to/file.xlsx")

Tool Interface Details

read_document

Read any supported document type.

Parameters:

ParameterTypeRequiredDescription
filenamestring✅Document file path, supports absolute or relative paths

Dependencies

Core Dependencies
  • mcp >= 1.26.0 - MCP protocol implementation
  • python-docx >= 1.2.0 - DOCX file reading
  • pypdf >= 6.8.0 - PDF file reading (replaces PyPDF2)
  • openpyxl >= 3.1.5 - Excel file reading
Development Dependencies
  • pytest >= 8.0.0 - Testing framework
  • pytest-asyncio >= 0.24.0 - Async testing support
  • pytest-cov >= 6.0.0 - Coverage reporting
  • basedpyright >= 0.28.0 - Type checking
  • ruff >= 0.8.0 - Linting and formatting

License

MIT License

Contributing

Issues and Pull Requests are welcome!

Related Projects

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.