home / mcp / officereader mcp server

OfficeReader MCP Server

OfficeReader-MCP used by AI agent

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "asunainlove-officereader-mcp": {
      "command": "python",
      "args": [
        "-m",
        "officereader_mcp.server"
      ],
      "env": {
        "OFFICEREADER_CACHE_DIR": "<path-to-cache>"
      }
    }
  }
}

OfficeReader MCP is a server that converts Microsoft Office documents (Word, Excel, PowerPoint) into Markdown while smartly extracting and optimizing embedded images. It preserves document structure, exposes useful metadata, and caches conversions for fast reuse, making it ideal for integrating Office content into Markdown-based workflows and knowledge bases.

How to use

You interact with OfficeReader MCP through an MCP client by sending a conversion request for a supported Office document. Typical workflows include converting a single file to Markdown, extracting text and images for reporting, or retrieving metadata without a full conversion.

Practical usage patterns you can perform include converting an Excel workbook to markdown with images preserved, extracting content from a PowerPoint deck, and fetching document metadata such as author and creation date. Conversions are cached, so repeated requests are served quickly without reprocessing the original file.

How to install

Prerequisites: Python 3.10 or higher and an MCP-capable client setup (Claude Desktop or Claude Code).

Step 1: Install the package by cloning the repository and installing in editable mode.

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

# Install in editable mode
pip install -e .

Step 2: Configure your MCP client to run the server. Create or edit the MCP config with the following settings to start the OfficeReader MCP server locally.

{
  "mcpServers": {
    "officereader": {
      "command": "python",
      "args": ["-m", "officereader_mcp.server"],
      "env": {
        "OFFICEREADER_CACHE_DIR": "/path/to/cache"
      }
    }
  }
}

Additional setup for Claude clients

If you use Claude Desktop, add the MCP server configuration to your Claude Desktop config file. You’ll specify the command to run the server and the environment variable for the cache directory.

{
  "mcpServers": {
    "officereader": {
      "command": "python",
      "args": ["-m", "officereader_mcp.server"],
      "env": {
        "OFFICEREADER_CACHE_DIR": "/path/to/cache"
      }
    }
  }
}

Restart and verify

After saving the configuration, restart Claude Desktop or Claude Code to load the MCP server. You can then issue conversion requests from your Claude chat interface.

Quick start examples

Convert an Excel file to markdown with images included, starting from the path you provide.

Extract text and images from a PowerPoint presentation to review slides and embedded visuals.

Retrieve metadata for a Word document to verify authorship and creation date without performing a full conversion.

Available tools

convert_document

Convert any supported Office document to Markdown format. You provide the file path and optional image extraction and output preferences.

read_converted_markdown

Read the full content of a previously converted markdown file from the cache.

list_conversions

List all cached conversions with details for quick reference.

clear_cache

Remove all cached conversions to free disk space.

get_document_metadata

Extract metadata from a document without performing a full conversion.

get_supported_formats

Return a list of all supported file formats and extensions.