MCP Documents Reader

You can read and extract content from Excel, Word, PDF, and text documents using the MCP Documents Reader. This tool connects to the MCP ecosystem, allowing AI assistants to access document content in a consistent, MCP-compliant way for tasks like summarization, extraction, and question answering.

How to use

To leverage this MCP server with your MCP client, configure one or more MCP endpoints in your client’s settings. You can run a local MCP service or point to a remote MCP server. Once configured, you can call the read_document tool to load the contents of a supported document. The tool accepts a filename and returns the document’s text and structure in a unified format, enabling downstream AI workflows such as answering questions, extracting data, or generating summaries.

How to install

Prerequisites you need before installing: Python 3.8 or higher and an MCP-enabled AI tool such as Trae IDE.

Option 1 Install from PyPI (recommended) then you can start using the MCP Document Reader in your MCP workflows.

pip install mcp-documents-reader

Option 2 Install from source if you prefer to build from the repository. Clone the project, install dependencies, and install the package in editable mode.

# Clone the repository
git clone https://github.com/xt765/mcp_documents_reader.git
# or use Gitee
# git clone https://gitee.com/xt765/mcp_documents_reader.git

cd mcp_documents_reader

# Install dependencies
pip install -e .

Configuration

Set up your MCP client to run the MCP Documents Reader as a server. You have multiple options, depending on how you want to deploy: using the PyPI package, or using git-based sources.

Option 1 Use PyPI package in Trae IDE or your MCP client that supports MCP servers.

Option 2 Use the GitHub repository to run the MCP Documents Reader directly from source.

Option 3 Use the Gitee repository to run the MCP Documents Reader directly from source.

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

Environment variables

The reader can read documents from a directory you specify. The default directory is ./documents. If you want to customize the path, set the DOCUMENT_DIRECTORY environment variable.

Usage

After configuring, you can call the read_document tool to load a document. Provide the filename with the path to the document you want to read. The tool supports Excel, Word, PDF, and TXT files.

Example usage (filename variations): read_document(filename="example.docx"), read_document(filename="example.pdf"), read_document(filename="example.xlsx"), read_document(filename="example.txt").

Available tools

read_document

Unified tool to read any supported document type (Excel, DOCX, PDF, TXT) by specifying the filename. Returns the document content and structure for downstream processing.