Sandbox
@ShellyDeng08/youtube-connector-mcp

YouTube MCP server for Claude Code and Cursor

This repo provides a YouTube connector over MCP so your agent can search videos, inspect channels and playlists, fetch transcripts, and pull comments. It is built as a Python server with a YouTube client, tool definitions, and config for the API key and request limits.

71 stars3 forksPythonUpdated 6mo ago
Who it's for

Builders who want their agent to search and read YouTube content without leaving the editor or terminal.

What it delivers

You can pull YouTube data into agent-led research and content analysis without switching tools.

What it does

Video, channel, and playlist search

Searches YouTube content with filters for duration, date, type, and order.

Transcript retrieval

Fetches transcript text with timestamps for videos that have captions available.

Video and channel details

Returns metadata, statistics, thumbnails, subscriber counts, and upload playlists.

Comments and playlist access

Pulls video comments with pagination and lists videos in a playlist or all playlists for a channel.

How to get it

  1. 1Run
    git clone https://github.com/ShellyDeng08/youtube-connector-mcp.git
    cd youtube-connector-mcp
    python3 -m venv .venv
    source .venv/bin/activate
    pip install -e .
  2. 2Run
    youtube-connector-mcp --help
    claude mcp list  # Check if server is registered
  3. 3Set as Environment Variable (Recommended)
    # Linux/Mac - Add to ~/.bashrc, ~/.zshrc, or ~/.profile
    export YOUTUBE_API_KEY="your_api_key_here"
    source ~/.zshrc
  4. 4Run
    # Windows PowerShell - Add to $PROFILE
    $env:YOUTUBE_API_KEY="your_api_key_here"
    # Or set permanently
    [System.Environment]::SetEnvironmentVariable('YOUTUBE_API_KEY', 'your_api_key_here', 'User')
  5. 5Run
    # Windows CMD
    setx YOUTUBE_API_KEY "your_api_key_here"
  6. 6Run
    git clone https://github.com/ShellyDeng08/youtube-connector-mcp.git
    cd youtube-connector-mcp
    poetry install --with dev

README

YouTube MCP Server (youtube-connector-mcp)

MCP Badge

youtube-connector-mcp is a YouTube MCP Server / Connector that allows AI assistants like Claude, Cursor, Cline, Windsurf, Continue.dev, Grapes AI to interact with the YouTube Data API v3 via the Model Context Protocol (MCP).

🌐 Project Website: https://youtube-connector-mcp-website.vercel.app/
📦 PyPI Package: https://pypi.org/project/youtube-connector-mcp/


Quick Start

# 1. Get your YouTube API Key from Google Cloud Console
#    https://console.cloud.google.com/apis/credentials

# 2. Set your API key as environment variable
export YOUTUBE_API_KEY="your_api_key_here"

# 3. Install pipx first (if not installed)
brew install pipx  # macOS
# or: apt install pipx  # Ubuntu/Debian

# 4. Install the package
pipx install youtube-connector-mcp

# 5. Add the MCP server
claude mcp add youtube-connector-mcp youtube-connector-mcp -s user -e YOUTUBE_API_KEY="${YOUTUBE_API_KEY}"

# 6. Restart Claude Code and start using!

Prerequisites

RequirementHow to Get
Python 3.10+Download Python or brew install python3
YouTube API KeyGet it free from Google Cloud Console
MCP ClientAny MCP-compatible AI: Claude, Cursor, Cline, Windsurf, Continue.dev, etc.

Installation

Install from PyPI (Recommended)

# Install pipx first (if not installed)
brew install pipx  # macOS
# or: apt install pipx  # Ubuntu/Debian

# Then install the package
pipx install youtube-connector-mcp

# Or with pip in a virtual environment
python3 -m venv .venv
source .venv/bin/activate
pip install youtube-connector-mcp

Install from Source

git clone https://github.com/ShellyDeng08/youtube-connector-mcp.git
cd youtube-connector-mcp
python3 -m venv .venv
source .venv/bin/activate
pip install -e .

Verify Installation

youtube-connector-mcp --help
claude mcp list  # Check if server is registered

Configuration

Popular AI Coding Tools

ToolPlatformConfig
Claude CodeCLIclaude mcp add
Claude DesktopmacOS/Windowsclaude_desktop_config.json
Cursor IDEDesktopSettings UI
ClineVS Code Extension.cline/config.yaml

Note: Any MCP-compatible AI tool works! Just use the standard JSON config format below.


Option 1: Using claude mcp add (Claude Code - Easiest)

# Install for current project only
claude mcp add youtube-connector-mcp youtube-connector-mcp -s local -e YOUTUBE_API_KEY="${YOUTUBE_API_KEY}"

# Install for all your projects (recommended)
claude mcp add youtube-connector-mcp youtube-connector-mcp -s user -e YOUTUBE_API_KEY="${YOUTUBE_API_KEY}"

# Install to project's .mcp.json
claude mcp add youtube-connector-mcp youtube-connector-mcp -s project -e YOUTUBE_API_KEY="${YOUTUBE_API_KEY}"

Don't have an API key? See Creating a YouTube API Key below - it's free and takes just a few minutes.


Option 2: Standard JSON Configuration

适用于任何 MCP 兼容的 AI 工具

{
  "mcpServers": {
    "youtube-connector-mcp": {
      "command": "youtube-connector-mcp",
      "env": {
        "YOUTUBE_API_KEY": "your_api_key_here"
      }
    }
  }
}

各工具配置文件位置:

ToolConfig File Path
Claude Code~/.claude/mcp_config.json
Claude Desktop~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
Cursor IDESettings → MCP Servers (UI)
Cline (VS Code).cline/config.yaml

Option 3: Claude Desktop

  1. Open Claude Desktop
  2. Go to Settings → Developer → Edit Config
  3. Or edit the config file directly:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "youtube-connector-mcp": {
      "command": "youtube-connector-mcp",
      "env": {
        "YOUTUBE_API_KEY": "your_api_key_here"
      }
    }
  }
}

Option 4: Cursor IDE

  1. Open Cursor Settings (Cmd/Ctrl + ,)
  2. Go to MCP Servers section
  3. Add a new server using the JSON format above

Option 5: Cline (VS Code Extension)

Add to your .cline/config.yaml:

mcpServers:
  youtube-connector-mcp:
    command: youtube-connector-mcp
    env:
      YOUTUBE_API_KEY: "your_api_key_here"

API Key Setup

Set as Environment Variable (Recommended):

# Linux/Mac - Add to ~/.bashrc, ~/.zshrc, or ~/.profile
export YOUTUBE_API_KEY="your_api_key_here"
source ~/.zshrc
# Windows PowerShell - Add to $PROFILE
$env:YOUTUBE_API_KEY="your_api_key_here"
# Or set permanently
[System.Environment]::SetEnvironmentVariable('YOUTUBE_API_KEY', 'your_api_key_here', 'User')
# Windows CMD
setx YOUTUBE_API_KEY "your_api_key_here"

Or Put Directly in MCP Config:

{
  "mcpServers": {
    "youtube-connector-mcp": {
      "command": "youtube-connector-mcp",
      "env": {
        "YOUTUBE_API_KEY": "AIzaSyC-Your-Actual-API-Key-Here"
      }
    }
  }
}

Security Note: Using environment variables is safer as it keeps your key out of version control.


Creating a YouTube API Key

  1. Go to Google Cloud Console
  2. Create a new project or select an existing one
  3. Enable YouTube Data API v3
  4. Go to Credentials and create an API key
  5. (Optional) Restrict the key to YouTube Data API v3 for better security

Environment Variables

VariableRequiredDefaultDescription
YOUTUBE_API_KEYYes-YouTube Data API v3 key
YOUTUBE_RATE_LIMITNo100Max requests per second

Features

Core Capabilities

ToolDescription
youtube_searchSearch videos, channels, playlists with filters (duration, date, type, order)
youtube_get_videoGet detailed video metadata, statistics, thumbnails, and content details
youtube_get_channelGet channel info, subscriber count, upload playlists, statistics
youtube_get_transcriptRetrieve actual video transcript text with timestamps
youtube_get_commentsFetch video comments with pagination support
youtube_get_playlistGet playlist details and complete video list
youtube_list_playlistsList all playlists for a specific channel

Use Cases

  • Research: Search and analyze YouTube content programmatically
  • Content Analysis: Extract transcripts and comments for AI processing
  • Channel Monitoring: Track channel statistics and new uploads
  • Data Mining: Gather YouTube data for your projects
  • Automated Workflows: Integrate YouTube data into AI-assisted workflows

Usage Examples

CategoryExample Prompts
Search"Search for Python tutorials" / "Find recent AI videos" / "Channels about cooking with 100k+ subscribers"
Video"Get details for this video: URL" / "What's the view count?" / "Get the transcript"
Channel"How many subscribers does @MKBHD have?" / "Recent uploads from this channel" / "Channel statistics"
Playlist"List all playlists for this channel" / "Get videos in this playlist"

Troubleshooting

MCP Server Not Found

Error: No MCP servers configured

Solutions:

  1. Verify ~/.claude/mcp_config.json exists
  2. Check JSON syntax is valid
  3. Run claude mcp list to see registered servers
  4. Restart Claude Code after updating config

Python Not Found

Error: command not found: python

Solutions:

  1. Use python3 instead of python
  2. Provide full path: which python3 (Mac/Linux) or where python (Windows)

Module Not Found

Error: ModuleNotFoundError: No module named 'mcp'

Solutions:

  • If using pipx: pipx reinstall youtube-connector-mcp
  • If using pip in venv: Activate virtual environment first source .venv/bin/activate

API Quota Exceeded

Error: 403 Forbidden - quota exceeded

Solutions:

  1. Check Google Cloud Console quota
  2. Default: 10,000 units/day
  3. Consider upgrading for higher limits

Transcript Not Available

Error: "No transcript available" or "Transcripts are disabled"

Solutions:

  1. Video may not have captions enabled
  2. Auto-generated captions may take 24+ hours after upload
  3. Try a video known to have captions

Transcript Request Blocked

Error: "YouTube is blocking requests from your IP"

Solutions: See youtube-transcript-api documentation for proxy options.


Development

Setup

git clone https://github.com/ShellyDeng08/youtube-connector-mcp.git
cd youtube-connector-mcp
poetry install --with dev

Run Tests

poetry run pytest

Publishing to PyPI

# Bump version (PyPI doesn't allow re-uploading the same version)
poetry version patch  # 0.3.0 → 0.3.1
poetry version minor  # 0.3.0 → 0.4.0
poetry version major  # 0.3.0 → 1.0.0

# Build and publish
poetry build
poetry publish

License

MIT License - see LICENSE for details.


Links

SEO Keywords

YouTube MCP Server, YouTube MCP Connector, MCP YouTube API, Claude MCP YouTube, Cursor MCP YouTube, AI YouTube API, Model Context Protocol YouTube

Files in the repo

Repository payload11 top-level entries
  • .github
  • docs
  • src
  • tests
  • .env.example
  • .gitignore
  • CHANGELOG.md
  • pyproject.toml
  • README.md
  • requirements.txt
  • TECH_NOTES.md

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
tirth8205/
code-review-graph

Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.

31k
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