YouTube Transcript MCP Server

Researchstdiov0.2.0
代码仓库

MCP server that fetches transcripts, captions, and metadata from YouTube videos so AI agents can summarize, search, and analyze video content without watching it. Supports multiple languages, timestamped segments, and channel/playlist lookups for research and content workflows.

连接

将此配置添加到 .claude/mcp.json

{
  "mcpServers": {
    "youtube-transcript": {
      "command": "npx",
      "args": [
        "-y",
        "@kimtaeyoon83/mcp-server-youtube-transcript"
      ]
    }
  }
}

工具 (2)

get_transcript

Fetch the transcript of a YouTube video by URL or video ID.

{
  "type": "object",
  "required": [
    "video"
  ],
  "properties": {
    "video": {
      "type": "string",
      "description": "YouTube video URL or ID"
    },
    "language": {
      "type": "string",
      "description": "Preferred transcript language code (e.g., en, es)"
    }
  }
}

get_video_metadata

Retrieve title, channel, duration, and description for a YouTube video.

{
  "type": "object",
  "required": [
    "video"
  ],
  "properties": {
    "video": {
      "type": "string",
      "description": "YouTube video URL or ID"
    }
  }
}

资源

此服务器未提供任何资源。

提示词

此服务器未提供任何提示词。

服务器信息

作者
Kim Do Gyun (@kimtaeyoon83)
代码仓库
https://github.com/kimtaeyoon83/mcp-server-youtube-transcript
许可证
MIT
版本
0.2.0
星标
870
最后更新
2026年6月19日