yutu — Claude Code MCP server: features and setup
yutu lets an AI agent handle YouTube tasks such as uploading and optimizing videos, managing comments and playlists, and updating channel branding.
claude mcp add --env YUTU_CREDENTIAL=<YUTU_CREDENTIAL> --env YUTU_CACHE_TOKEN=<YUTU_CACHE_TOKEN> --env YUTU_LOG_LEVEL=<YUTU_LOG_LEVEL> --transport stdio yutu -- npx -y @eat-pray-ai/yutuWhat yutu does
yutu combines a CLI, MCP server, and AI agent for YouTube workflows. It can help upload and optimize videos, manage comments and playlists, and update channel branding. A separate skill is available for AI agents.
You’ll need a Google Cloud Platform project with the YouTube Data API v3 enabled, an OAuth Client ID for a Desktop app, and authorization via yutu auth --credential client_secret.json. After access is granted, the token is saved as youtube.token.json. Without OAuth credentials, the server can’t access YouTube. YouTube Analytics API and YouTube Reporting API can also be enabled for analytics and reports; both are optional.
What it helps with
- upload a video to YouTube
- optimize a video's settings
- review comments on a video
- manage channel playlists
- update channel branding
Example prompts for the agent
- Upload this video to YouTube and optimize its description.
- Review comments on the latest video and draft replies.
- Add these videos to a channel playlist.
How to install yutu
Platform CLI
claude mcp add --env YUTU_CREDENTIAL=<YUTU_CREDENTIAL> --env YUTU_CACHE_TOKEN=<YUTU_CACHE_TOKEN> --env YUTU_LOG_LEVEL=<YUTU_LOG_LEVEL> --transport stdio yutu -- npx -y @eat-pray-ai/yutuReplace <…> with your own values
Platform CLI
codex mcp add yutu --env YUTU_CREDENTIAL=<YUTU_CREDENTIAL> --env YUTU_CACHE_TOKEN=<YUTU_CACHE_TOKEN> --env YUTU_LOG_LEVEL=<YUTU_LOG_LEVEL> -- npx -y @eat-pray-ai/yutuReplace <…> with your own values
Config file
{
"mcpServers": {
"yutu": {
"command": "npx",
"args": [
"-y",
"@eat-pray-ai/yutu"
],
"env": {
"YUTU_CREDENTIAL": "<YUTU_CREDENTIAL>",
"YUTU_CACHE_TOKEN": "<YUTU_CACHE_TOKEN>",
"YUTU_LOG_LEVEL": "<YUTU_LOG_LEVEL>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"yutu": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@eat-pray-ai/yutu"
],
"env": {
"YUTU_CREDENTIAL": "<YUTU_CREDENTIAL>",
"YUTU_CACHE_TOKEN": "<YUTU_CACHE_TOKEN>",
"YUTU_LOG_LEVEL": "<YUTU_LOG_LEVEL>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e YUTU_CREDENTIAL=<YUTU_CREDENTIAL> -e YUTU_CACHE_TOKEN=<YUTU_CACHE_TOKEN> -e YUTU_LOG_LEVEL=<YUTU_LOG_LEVEL> yutu npx -y @eat-pray-ai/yutuReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about yutu
How do I install yutu in Claude Code?
Add the MCP server with claude mcp add yutu -e YUTU_CREDENTIAL=<YUTU_CREDENTIAL> -e YUTU_CACHE_TOKEN=<YUTU_CACHE_TOKEN> -- npx -y @eat-pray-ai/yutu. First configure YouTube OAuth and provide the credential and token paths.
What does yutu do?
yutu connects an AI agent to YouTube for uploading and optimizing videos, managing comments and playlists, and updating channel branding.
What do I need to run yutu?
Set up a Google Cloud Platform project with YouTube Data API v3 enabled and an OAuth Client ID for a Desktop app. After authorization, the token is saved as youtube.token.json.
