YouTube MCP Server
Enables Claude to access and manage YouTube channel data through the YouTube Data API v3 and YouTube Analytics API. Provides tools for reading analytics, fetching video metadata, searching uploads, and updating video SEO directly from Claude.
README
YouTube MCP Server
Connect Claude to your YouTube channel. Read analytics, fetch full video metadata (including unlisted/private/draft), search your uploads, and update titles/descriptions/tags — all from Claude Desktop, Claude Code, or any MCP client.
What Is This?
A Model Context Protocol (MCP) server that connects Claude to the YouTube Data API v3 and YouTube Analytics API via OAuth2.
Instead of copying data from YouTube Studio into AI tools, just ask Claude:
- "Pull the full title, description and tags for this video"
- "What are my top performing videos this month?"
- "Update the title and tags on my latest upload"
- "Where is my traffic coming from?"
Claude reads your real private channel data and can update video SEO directly.
Claude → YouTube MCP Server → YouTube APIs → Your Channel Data
(you) (this repo) (OAuth2) (stays local)
Tools Available (10 total)
Video Metadata (Read + Write)
| Tool | What It Does |
|---|---|
get_video_details |
Full metadata for any video by ID or URL — title, full description, all tags, category, privacy status (public/unlisted/private/draft), stats, duration, thumbnail URL |
search_my_videos |
Search your own channel's videos by keyword. Returns metadata + stats for matching videos |
update_video_seo |
Update title, description, and/or tags on any video directly. Only changes fields you provide |
Channel Analytics
| Tool | What It Does |
|---|---|
get_channel_overview |
Subscribers, total views, video count, channel description, creation date |
get_all_videos |
List all videos with stats (views, likes, comments, tags, privacy status). Sort by date or views |
get_analytics_over_time |
Day-by-day views, watch time, subscribers gained/lost for any date range |
get_top_videos_analytics |
Top performing videos ranked by views with retention %, watch time, subs gained |
get_audience_demographics |
Audience breakdown: top countries, device types, age groups, gender |
get_traffic_sources |
Where viewers come from: YouTube Search, Suggested, Browse, External, Direct |
analyze_and_suggest_topics |
Pulls channel + top video data for AI-powered topic analysis |
Install
Option A: npx (Zero Install — just run it)
npx youtube-studio-mcp
Option B: Global Install
npm install -g youtube-studio-mcp
Connect to Claude
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (Mac) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"youtube-analytics": {
"command": "node",
"args": ["/full/path/to/youtube-studio-mcp/server.js"]
}
}
}
Or if installed via npm:
{
"mcpServers": {
"youtube-analytics": {
"command": "npx",
"args": ["-y", "youtube-studio-mcp"]
}
}
}
Claude Code (Terminal)
claude mcp add youtube-analytics node /full/path/to/youtube-studio-mcp/server.js
VS Code
Add to .vscode/settings.json:
{
"mcp.servers": {
"youtube-analytics": {
"command": "npx",
"args": ["-y", "youtube-studio-mcp"]
}
}
}
Restart Claude. Done!
Ready-to-Use Prompts
SEO Audit (any video)
Get the full details for this video: [paste URL or ID]
Check what keywords it's ranking for and suggest optimized title + tags
Channel Performance Report
Pull my channel overview, top 20 videos by watch time, 90-day analytics,
traffic sources and audience demographics. Give me a full performance report.
Video Topic Research
Get my top 20 videos by watch time. What patterns do you see —
topics, lengths, title styles? Suggest 10 new video ideas.
Update Video SEO
Search my videos for "elementor menu". Pull the full details.
Write an optimized title, description, and tags — then update it.
Audience Deep Dive
Show my full audience demographics — age, gender, countries, devices.
What content style and posting schedule fits my actual audience?
Underperformer Diagnosis
Get all my videos. Compare bottom 10 vs top 10 by views.
Why did the lower ones underperform? What would you change?
How the Skills + MCP Work Together
You: "Optimize this video: youtube.com/watch?v=abc123"
Claude:
1. youtube-seo-optimizer skill activates
2. Calls get_video_details → pulls current title, description, tags
3. Checks YouTube SERP for current rankings
4. Lists PROTECTED keywords (won't remove)
5. Writes 3 new title options + full description + 20 tags
6. Can call update_video_seo to apply changes directly
The skills tell Claude what to do. The MCP tools give Claude access to your data. Together they create a complete YouTube AI workflow.
Install Skills
Claude Desktop / Claude Code:
# Copy a skill to your Claude skills directory
cp skills/youtube-seo-optimizer.md ~/.claude/skills/
Or install all 8:
cp skills/youtube-*.md ~/.claude/skills/
Files
youtube-studio-mcp/
├── server.js ← MCP server (10 tools)
├── auth.js ← Run once to link YouTube account
├── package.json ← Dependencies + npm config
├── skills/ ← 8 AI skills for YouTube creators
├── LICENSE ← MIT
├── credentials.json ← YOU add this (never commit!)
└── tokens.json ← Auto-created after auth (never commit!)
License
MIT — free to use, modify, share.
Built By
Hakan Özdemir & Claude Code
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。