Readium MCP 📚✨

Overview

Readium MCP is a powerful Model Context Protocol (MCP) server that bridges the gap between Large Language Models and documentation repositories. It exposes the robust documentation analysis functionality of Readium for LLMs like Claude and other MCP-compatible clients.

🔍 What can it do?

• Analyze local directories, public/private Git repositories, and URLs
• Process multiple file formats (code, Markdown, text)
• Convert web pages to Markdown for analysis
• Deliver structured results with summary, file tree, and content
• Configure exclusions for directories and file types
• Set maximum file sizes and target specific subdirectories

🚀 Installation

The easiest way to install Readium MCP is via pip:

pip install readium-mcp

After installation, you can verify the installation path using:

whereis readium-mcp

This command will show you the full path to the executable, which you may need for configuration.

🔌 Configuration

Claude Desktop

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "readium": {
      "command": "readium-mcp"
    }
  }
}

Note: If the command is not found, you may need to use the full path:

{
  "mcpServers": {
    "readium": {
      "command": "/usr/local/bin/readium-mcp"  # Replace with the path from whereis
    }
  }
}

VSCode

Add to your VSCode settings.json:

{
  "mcp": {
    "servers": {
      "readium": {
        "command": "readium-mcp"  # Use full path if needed
      }
    }
  }
}

📊 Example Usage

When integrated with Claude or another LLM, you can ask it to:

• "Analyze the documentation in this GitHub repository"
• "Summarize the structure of this codebase"
• "Help me understand how this project is organized"

The LLM will use Readium MCP to scan the content and provide organized insights about the documentation and code structure.

🛠️ Advanced Configuration

Readium MCP offers extensive configuration options:

• branch: Specify which branch to analyze in Git repositories
• target_dir: Focus on a specific subdirectory
• exclude_dirs: Skip certain directories (e.g., .git, node_modules)
• exclude_ext: Ignore specific file extensions
• include_ext: Only process certain file types
• max_file_size: Limit the size of files to analyze
• use_markitdown: Convert websites to Markdown

🌟 Perfect For

• Developers seeking AI assistance with unfamiliar codebases
• Technical writers managing complex documentation
• Researchers analyzing open-source projects
• Anyone who wants to quickly grasp the structure of a large code repository