YouTube Studio MCP
Local MCP server for YouTube metadata, thumbnails, comments, and analytics.
YouTube Studio MCP
A local Model Context Protocol (MCP) server for managing YouTube channels from AI tools such as Codex, Claude Desktop, Cursor, and other MCP-compatible clients.
YouTube Studio MCP lets your assistant inspect channel performance, review recent uploads, update video metadata, upload thumbnails, post comments, and read analytics through your own Google OAuth credentials.
Demo
User: Show my YouTube channel overview and summarize the last 10 uploaded videos.
Assistant uses:
- youtube_auth_status
- youtube_channel_overview
- youtube_list_videos
Result:
The assistant can review recent uploads, public stats, metadata, privacy status, and improvement opportunities.
See Demo for a setup walkthrough.
Features
- Read authenticated channel profile, statistics, branding, and uploads playlist.
- List recent videos with metadata, privacy status, content details, and public stats.
- Update video title, description, tags, category, default language, and privacy status.
- Upload custom thumbnails from local image files.
- Read channel and per-video YouTube Analytics reports.
- Post and list top-level video comments.
- Runs locally over stdio with no hosted backend.
- Uses only the Python standard library at runtime.
Why this exists
Creators often want AI help with repetitive YouTube Studio work: auditing metadata, improving SEO, comparing video performance, preparing descriptions, and keeping thumbnails/titles consistent. This MCP server gives an AI assistant controlled access to those workflows while keeping credentials on your own machine.
Requirements
- Python 3.10 or newer.
- A Google account with access to the YouTube channel.
- A Google Cloud project with:
- YouTube Data API v3 enabled.
- YouTube Analytics API enabled.
- A Google OAuth Desktop app client JSON.
- An MCP-compatible client.
Quick start
Clone the repo:
git clone https://github.com/i1s-abhishek/youtube-studio-mcp.git
cd youtube-studio-mcp
Create the local secrets folder if it does not exist:
mkdir -p secrets
Download your Google OAuth Desktop client JSON and save it as:
secrets/client_secret.json
Authenticate with Google:
python3 scripts/auth.py auth
Configure your MCP client to run:
python3 scripts/server.py
See Google OAuth Setup and MCP Client Configuration for detailed steps.
MCP configuration
Example stdio config:
{
"mcpServers": {
"youtube-studio": {
"command": "python3",
"args": ["./scripts/server.py"],
"cwd": "/absolute/path/to/youtube-studio-mcp",
"env": {
"YOUTUBE_CLIENT_SECRETS": "./secrets/client_secret.json",
"YOUTUBE_TOKEN_FILE": "./secrets/token.json"
}
}
}
}
Replace /absolute/path/to/youtube-studio-mcp with the path where you cloned this repository.
Available tools
See Tools for the full tool list.
Core tools include:
youtube_channel_overviewyoutube_list_videosyoutube_get_videoyoutube_update_videoyoutube_upload_thumbnailyoutube_channel_analyticsyoutube_video_analyticsyoutube_post_commentyoutube_list_comments
Example prompts
Show my YouTube channel overview and summarize the last 10 uploads.
Review my last 28 days of analytics and suggest what I should improve next.
Update this video's title, description, tags, and language.
Upload this local thumbnail image to video VIDEO_ID.
OAuth scopes
This server requests:
https://www.googleapis.com/auth/youtubehttps://www.googleapis.com/auth/youtube.force-sslhttps://www.googleapis.com/auth/youtube.readonlyhttps://www.googleapis.com/auth/yt-analytics.readonly
These scopes are broad because the server supports both read and write YouTube Studio actions. Only run this server on machines you trust.
Safety notes
- Do not commit
secrets/client_secret.json. - Do not commit
secrets/token.json. - Review tool calls before allowing metadata updates or comments.
- Keep a backup of important titles, descriptions, and tags before bulk updates.
Verification
Before publishing, this repo was checked to confirm that only secrets/.gitkeep is tracked under secrets/; real OAuth files such as secrets/client_secret.json and secrets/token.json are ignored by git.
Repository topics
Suggested GitHub topics:
mcp, model-context-protocol, youtube, youtube-api, youtube-analytics, youtube-studio, ai-tools, creator-tools, python
Publishing
Maintainers can publish or update the GitHub repository with:
scripts/publish_github.sh i1s-abhishek youtube-studio-mcp
The script creates a public GitHub repo if needed, pushes main, and applies relevant discovery topics.
Sharing
Use Launch Copy when submitting this project to MCP directories or posting it on social platforms.
Contributing
Contributions are welcome. See CONTRIBUTING.md.
Security
Please read SECURITY.md before using this with a production channel.
License
MIT. See LICENSE.
相关服务器
Janee API Security
MCP server that sits between AI agents and APIs. Agents request access, Janee makes the call with the real credentials, agents never see the secrets.
Ecovacs Robot Control
Control and query the status of Ecovacs cleaning robots using the MCP protocol.
Pace
Pace is the first MCP connector that brings wearable health data directly into Claude — no third-party dashboards, no manual exports, no extra apps. Most health apps lock your data behind their own UI. Pace breaks that wall: connect once, and Claude can analyze your sleep, activity, workouts, nutrition and recovery in natural language — with full visualizations inline.
WattyEnergie MCP Server
Dutch energy comparison tools. Calculate costs, compare contracts, get dynamic EPEX prices.
Compliance Intelligence
Access 692+ compliance frameworks, 13,700+ controls, and 280,000+ cross-framework mappings via MCP. Query ISO 27001, NIST CSF, GDPR, SOC 2, HIPAA, PCI DSS and more.
AgentTrust
Challenge-response quality verification for AI agents and MCP servers.
ProfitSpot MCP
Cross-chain DeFi intelligence for AI agents — 86 chains, 6500+ pools, Monte Carlo simulations
agimem
Hosted MCP server for persistent, shared AI agent memory
Fabric MCP Server
The Fabric MCP server exposes the following MCP tools that allow AI assistants to interact with your Equinix Fabric resources. Each tool corresponds to a specific Fabric API endpoint.
Agent Central
Hosted MCP server for Amazon sellers using Claude, ChatGPT, and other AI clients.