Sandbox
@cottongeeks/ytt-mcp

MCP server for YouTube transcript fetches

ytt-mcp exposes YouTube transcript fetches through the Model Context Protocol. That lets an agent call the server instead of making you copy text by hand. The README shows setup for Claude Desktop and Raycast, and the server can also be run locally for inspection with FastMCP.

74 stars4 forksPythonUpdated 11mo ago
Who it's for

Builders who want their agent to pull transcript text from YouTube URLs.

What it delivers

You can ask your agent to read and summarize a YouTube video without copying the transcript yourself.

What it does

YouTube transcript fetches

Provides an MCP server that returns transcripts for YouTube videos.

Claude Desktop setup

Shows the `mcpServers` config block needed to add `ytt-mcp` to Claude Desktop.

Raycast support

Explains how to install the server through Raycast’s MCP extension and use it as `@youtube-transcript`.

Local dev inspector

Includes a FastMCP dev command for testing the server locally with `uv run fastmcp dev ytt_mcp.py`.

How to get it

  1. 1The most convenient way to install and run is to use uv and then invoke the package…
    uv run fastmcp dev ytt_mcp.py

README

ytt-mcp: YouTube Transcript MCP Server

MCP Server to fetch transcripts for YouTube videos.

Installing and Running

The most convenient way to install and run is to use uv and then invoke the package using uvx

Using MCP Inspector (for development/debugging/testing only)

uv run fastmcp dev ytt_mcp.py

This will generate a localhost URL that can be used to examine and test out the server.

image

Claude Desktop

Go to SettingsDeveloper, and then click on Edit Config. This will open the claude-desktop-config.json file in your default editor. Make the following addition

{
  "mcpServers": {
    …<rest of the config>…
    "ytt-mcp": {
      "command": "uvx",
      "args": ["ytt-mcp"]
    }
  }
}

Relaunch Claude config and try out the server as shown in the screenshot below

image

Raycast

If you are using Raycast, you can install the MCP server by invoking the Install Server command from the MCP extension.

image

After that you can refer to the MCP server as @youtube-transcript and interact with it. You can also use it in a Raycast AI Command with a prompt. For example, here is a prompt I use to extract and summarize a YouTube URL in the clipboard

@youtube-transcript fetch the Youtube transcript of the video: {clipboard | raw}

Comprehensively summarize the transcript with the following format:
"""
### Key Takeaways

- <EXACTLY three bullet points with the key takeaways, keep the bullet points as short as possible>
"""

### Theme Wise Breakdown
<divide the transcript into thematic sections and summarize each section comprehensively. reuse any existing section delimiters the article already has. If not add your own. keep the content of the breakdown in the same order as it appears in the webpage text.>

Some rules to follow precisely:
- ALWAYS capture the perspective and POV of the author
- NEVER come up with additional information

See video demo below

https://github.com/user-attachments/assets/e6530768-3483-4cb9-988a-7ec7a999d505

Files in the repo

Repository payload8 top-level entries
  • .python-version
  • CLAUDE.md
  • LICENSE
  • pyproject.toml
  • README.md
  • test_ytt_mcp.py
  • uv.lock
  • ytt_mcp.py

Discussion (0)

Ask about usage, or say what you built with it

Sign in to join the discussion.

No comments yet. Be the first to say what this is good for.

More connectors

Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface

86k

High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

43k

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code

14k
okf-memory/
okf-agent-memory

Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.

547
2akouwu/
reverify

Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.

1.1k

x64dbg-MCP Server is a native MCP (Model Context Protocol) plugin for x64dbg that exposes the debugger's full functionality over HTTP. Connect any MCP-compatible AI assistant and control x64dbg programmatically: set breakpoints, step through code, read memory, dump registers, and more. Built with Zig — zero dependencies, single-binary output, cros

1.9k