
YouTube MCP
io.github.mrsknetworkPublished by GitHub account mrsknetworkv1.0.14Updated Oct 24, 2026
Interact with YouTube data via the official Data API v3
Overview
Connects an assistant to YouTube data through the official Data API v3, for searching videos, pulling transcripts and metadata, and reading comments and channel
- What you can do
- Download full spoken transcripts of videos, choosing language and manual or auto-generated captions, without credentials. Search videos, channels and playlists by keyword, date range, duration, quality and region, and list trending videos by country and category. Retrieve video metadata and stats, channel profiles, playlists and playlist items, and channel activity and shelf sections. Read top-level comments and replies, and list available caption tracks. Look up supported languages and regions, and list a channel's public subscriptions or your own via OAuth. List your channel's paying members and membership tiers, and revoke the stored OAuth token.
- When to use it
- Useful when you want an assistant to research YouTube content, summarize or quote a video from its transcript, compare channels and videos by stats, or review comments and trending topics. OAuth mode is for creators or channel owners who need their own subscriptions and membership data.
- Requirements
- Local Node.js process run via npx from the npm package @mrsknetwork/ytmcp. Guest transcript mode needs yt-dlp and no credentials. Public data tools need a Google API key with YouTube Data API v3 enabled, passed as GOOGLE_API_KEY. Private subscriptions and memberships need an OAuth 2.0 web client with GOOGLE_CLIENT_ID and GOOGLE_CLIENT_SECRET and the redirect URI Network access required.
Installation
In SourceWeft
- Open YouTube MCP in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
YouTube MCP
Connect AI assistants to YouTube: search, transcripts, metadata, and more.
What can it do?
Once connected, your AI can:
- Extract clean, word-for-word video transcripts. No credentials required.
- Search YouTube by keyword, date range, duration, quality, and region.
- Fetch trending videos for any country and category.
- Retrieve video stats, metadata, comments, captions, and channel details.
- Look up channels by
@handle(e.g.@MrBeast). - Access your own subscriptions, members, and membership tiers via OAuth.
Getting Started
The server supports three access tiers. Pick the one that fits your use case.
Guest Mode (No Setup)
Works out of the box. The get_video_transcript tool uses yt-dlp to extract transcripts without any API credentials.
API Key (Recommended)
Unlocks all public data tools. Best for search, metadata, comments, trending videos, and transcripts.
1. Get a Google API Key
- Open the Google Cloud Console.
- Enable YouTube Data API v3 for your project.
- Go to Credentials and create an API Key.
2. Add to your MCP client config or .env file
Pass GOOGLE_API_KEY via env in your MCP client config, or place GOOGLE_API_KEY=YOUR_KEY inside a .env file in the project directory.
OAuth 2.0
Required for private subscriptions, memberships, and account activity.
1. Create an OAuth 2.0 Client
- Open the Google Cloud Console.
- Enable YouTube Data API v3 for your project.
- Go to Credentials > Create Credentials > OAuth 2.0 Client ID.
- Set Application type to Web application.
- Add this exact Redirect URI:
http://localhost:31415/oauth2callback. - Go to OAuth Consent Screen > Test Users and add your Gmail address.
2. Add to your MCP client config
Credentials are passed securely via environment variables in the MCP config. They are never stored in files.
3. First-time login
On your first tool call, the AI will share a login link. Click it, authorize the app in your browser, then tell the AI you are done. Your session is saved. You will not need to log in again unless you revoke access.
4. Signing out
Ask your AI to call revoke_oauth_token to sign out and delete your stored credentials at any time.
Available Tools
Building from Source
Instructions for running the server locally.
License
Licensed under the MIT License.
Source: README.md at commit c4ca227
Tools
0Version history
1- v1.0.14LatestSep 16, 2026
Score
50% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts