Youtube_Summarizer_MCP by cryptohero92 - MCP Server

YouTube Summarizer MCP Server

A Model Context Protocol (MCP) server that enables Claude and other language models to summarize YouTube videos by fetching transcripts and generating concise summaries using OpenAI GPT-4o or Google Gemini.

Features

YouTube Video Summarization: Automatically fetches video transcripts and generates intelligent summaries
Multiple AI Backends: Supports both OpenAI GPT-4o and Google Gemini for summarization
MCP Integration: Works seamlessly with Claude Desktop and other MCP-compatible clients
Robust Error Handling: Comprehensive error management for network issues and API failures
Easy Configuration: Simple environment variable setup

📹 Introduction Video

New to this project? Watch the introduction video for a complete walkthrough:

📺 Click the thumbnail above to watch the introduction video!

Note: If you don't see a thumbnail above, you may need to add a thumbnail image to the repository. For now, you can download and watch the video directly.

What you'll learn in this video:

✅ How to set up the YouTube Summarizer MCP Server
⚙️ Step-by-step installation and configuration process
🎬 Live demonstration of the summarizer in action with Claude Desktop
💡 Tips and best practices for optimal usage

Prerequisites

Python 3.12 or higher
uv package manager
API keys for:
- Supadata API (for YouTube transcript fetching)
- OpenAI API or Google Gemini API (for text summarization)

Installation

Clone or create the project directory:

mkdir youtube_summarizer
cd youtube_summarizer

Initialize the project with uv:
```
uv init
```

Install dependencies:

uv add google-generativeai mcp[cli] openai requests yt-dlp python-dotenv

Configuration

Create a .env file in the project root:

SUPADATA_API_KEY=your_supadata_api_key_here
GEMINI_API_KEY=your_gemini_api_key_here
OPENAI_API_KEY=your_openai_api_key_here

API Key Setup:
- Supadata API: Sign up at Supadata.ai to get YouTube transcript access
- OpenAI API: Get your API key from OpenAI Platform
- Gemini API: Get your API key from Google AI Studio
Note: You only need either OpenAI OR Gemini API key, not both. The server will use OpenAI if available, otherwise fallback to Gemini.

Claude Desktop Integration

To use this MCP server with Claude Desktop, add the following configuration to your claude_desktop_config.json:

{
  "mcpServers": {
    "youtube-summarizer": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/youtube_summarizer",
        "run",
        "mcp_server.py"
      ],
      "env": {
        "SUPADATA_API_KEY": "your_supadata_api_key_here",
        "GEMINI_API_KEY": "your_gemini_api_key_here"
      }
    }
  }
}

Important:

Replace the directory path with your actual project location
Replace the API key values with your actual keys
The env section in the config will override any .env file values

Usage

With Claude Desktop

Once configured, you can ask Claude to summarize YouTube videos:

Please summarize this YouTube video: https://www.youtube.com/watch?v=CiArUs_2jm4

Claude will automatically use the MCP server to fetch the transcript and generate a summary.

Testing the Server Directly

You can test the MCP server using the MCP inspector:

# Install MCP inspector (if not already installed)
npm install -g @anthropic/mcp-inspector

# Run the inspector
mcp-inspector

# In another terminal, run your server
uv --directory /path/to/youtube_summarizer run mcp_server.py

Project Structure

youtube_summarizer/
├── .env                                # Environment variables (create this)
├── .gitignore                         # Git ignore file
├── .python-version                    # Python version specification
├── main.py                            # Entry point (placeholder)
├── mcp_server.py                      # Main MCP server implementation
├── pyproject.toml                     # Project dependencies and metadata
├── uv.lock                           # Dependency lock file
├── youtube_summarizer_introduction.mp4 # Introduction video walkthrough
└── README.md                         # This file

API Reference

Available Tools

`summarize_youtube_video`

Summarizes a YouTube video by fetching its transcript and generating a concise summary.

Parameters:

youtube_url (string): The full URL of the YouTube video to summarize

Returns:

A formatted string containing the video title and summary

Example:

result = await summarize_youtube_video("https://www.youtube.com/watch?v=dQw4w9WgXcQ")

How It Works

Video Title Extraction: Uses yt-dlp to extract the video title
Transcript Fetching: Calls the Supadata API to get the video transcript
AI Summarization: Uses either OpenAI GPT-4o or Google Gemini to generate a summary
Response Formatting: Combines the title and summary into a readable format

Troubleshooting

Common Issues

"No transcript content found":
- The video might not have captions/transcripts available
- Check if the video is publicly accessible
"Supadata API key is not set":
- Verify your .env file contains SUPADATA_API_KEY
- Check Claude Desktop config environment variables
"No API key found for OpenAI or Gemini":
- Set at least one of OPENAI_API_KEY or GEMINI_API_KEY
- Verify the keys are correctly formatted
MCP Server Connection Issues:
- Ensure the directory path in Claude Desktop config is correct
- Check that uv is installed and accessible from the command line
- Verify all dependencies are installed with uv sync

Debug Mode

To run the server with debug output:

uv run mcp_server.py

This will show console output including API calls and error messages.

Dependencies

mcp[cli]: Model Context Protocol server framework
openai: OpenAI API client
google-generativeai: Google Gemini API client
requests: HTTP library for API calls
yt-dlp: YouTube video information extraction
python-dotenv: Environment variable loading

Development

This project was built following the MCP Python Server Development Guide.

Contributing

Fork the repository
Create a feature branch
Make your changes
Test thoroughly
Submit a pull request

Support

If you encounter issues:

Check the troubleshooting section above
Verify your API keys are valid and have sufficient credits
Ensure all dependencies are properly installed
Check the console output for detailed error messages

☕ Support This Project

If you find this YouTube Summarizer MCP Server helpful and would like to support my future work, please consider buying me a coffee! Your support helps me continue improving this tool and creating more useful MCP servers for the community.

☕ Buy me a coffee

Every contribution, no matter how small, is greatly appreciated and motivates me to keep building awesome tools! 🚀

cryptohero92/Youtube_Summarizer_MCP