Youtube Transcript MCP 11460a5b

MCPAI代理

提供方 guardkit

AI代理 stdio · 本地进程 无需鉴权 社区 / 第三方 Claude DesktopClaude
查看接入方式

一个通过MCP服务器获取YouTube视频字幕并使用Claude提取结构化洞察的工具,提供6种预设分析领域和24个洞察类别。

一句话选型结论

按资料字段整理

Youtube Transcript MCP 11460a5b 属于AI代理类 MCP 服务器,stdio · 本地进程,无需鉴权,可直接接入 Claude Desktop、Claude,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类AI代理 传输方式stdio · 本地进程 鉴权无需鉴权 部署方式本地部署
GitHub 星标 0 仓库热度,无公开仓库时显示未标注
工具数 4 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

AI代理
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量4暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

视频分析Python本地部署自然语言处理商业洞察技术学习Claude集成

安装与配置

stdio · 本地进程 无需鉴权
客户端在本机拉起子进程,数据不出本机
python3 --version

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 Python,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后客户端会列出约 4 个工具,随便调一个只读工具确认能返回结果。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
给代理接外部能力

让代理在任务中途调用搜索、数据库等真实工具。

多步任务编排

把一串操作拆成步骤,由代理按顺序执行。

自动化重复流程

把固定流程交给代理跑,减少手工操作。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 资料标注 4 个工具,职责相对集中,接入后不容易误操作。
  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Claude,换客户端不用重新研究配置。
  • 资料标注无需鉴权,装好就能连。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在AI代理场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Youtube Transcript MCP 11460a5b当前 4 stdio · 本地进程 本地部署 有 2026-05-17
Filesystem 560c4363 13 未标注 本地部署 无 2026-05-17
Server Everything (everything) 11 stdio · 本地进程 本地部署 有 2026-05-17
Time 7365cb6d 2 SSE · 服务端推送 远程部署 有 2026-05-18
Everything Claude Code 未标注 stdio · 本地进程 本地部署 有 2026-05-17
Openmemory 557e1585 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

AI代理
字段来自项目仓库与公开索引,中文说明由本站标注
提供方guardkit上游标注的维护方 guardkit
分类AI代理本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式无需鉴权远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · Python本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-05-17上游最后核验时间

详细介绍

9.6 KB
来自项目仓库与上游目录的原文,未做删改

youtube-insights-mcp

An MCP server that fetches YouTube transcripts and extracts structured insights using Claude. Provides 4 MCP tools and a full CLI, with 6 focus area presets across 24 insight categories.

Features

  • Transcript extraction with intelligent language fallback
  • Structured insight analysis across 24 categories (business, investment, technical, YouTube growth, AI learning)
  • 6 focus area presets: general, entrepreneurial, investment, technical, youtube-channel, ai-learning
  • Multiple URL formats: youtube.com, youtu.be, embed links, or just a video ID
  • Dual mode: MCP server for Claude Desktop + standalone CLI
  • Claude Code skill: /yt-insights slash command for interactive insight extraction
  • Smart chunking for long transcripts with overlap for context continuity
  • Proxy support for environments with IP restrictions

Prerequisites

You need Python 3.10 or later. Check if it's installed:```bash
python3 --version


**If Python is not installed:**

- **macOS**: `brew install python` (requires Homebrew) or download from python.org
- **Windows**: Download from python.org (check "Add to PATH" during install)
- **Linux**: `sudo apt install python3 python3-pip` (Debian/Ubuntu) or `sudo dnf install python3 python3-pip` (Fedora)

## Installation

We recommend installing in a virtual environment to avoid conflicts with other Python packages:```bash
# Create and activate a virtual environment
python3 -m venv youtube-mcp-env
source youtube-mcp-env/bin/activate  # On Windows: youtube-mcp-env\Scripts\activate

# Install the package
pip install youtube-insights-mcp

> Troubleshooting: If you see ModuleNotFoundError: No module named 'pip', your system Python is misconfigured. Create a virtual environment first (as shown above) — this gives you a clean, working pip.

Quick Start

Claude Desktop Configuration

After installing, find the absolute path to the command:```bash
which youtube-insights-mcp


Then add to your Claude Desktop MCP configuration (`claude_desktop_config.json`):```json
{
  "mcpServers": {
    "youtube-insights-mcp": {
      "command": "/absolute/path/to/youtube-mcp-env/bin/youtube-insights-mcp",
      "env": {
        "LOG_LEVEL": "INFO"
      }
    }
  }
}

Replace /absolute/path/to/youtube-mcp-env/bin/youtube-insights-mcp with the output from which youtube-insights-mcp.

> Important: Claude Desktop does not use your shell's PATH or virtual environment. You must provide the full absolute path to the command. Using just "command": "youtube-insights-mcp" will fail unless it's installed globally on your system PATH.

Starting the Server

With Claude Desktop — no manual start needed. Claude Desktop automatically launches and manages the server process using the configuration above. Just restart Claude Desktop after adding the config.

Manual start (for testing or other MCP clients):```bash

Start the MCP server (stdio transport)

youtube-insights-mcp

Or using Python module directly

python -m youtube_insights_mcp


The server communicates via stdin/stdout using the MCP JSON-RPC protocol. It is designed to be started by an MCP client (like Claude Desktop) rather than run interactively.

### CLI Usage```bash
# Health check
youtube-insights-mcp cli ping

# Fetch a transcript
youtube-insights-mcp cli get-transcript "https://www.youtube.com/watch?v=VIDEO_ID"

# List available transcript languages
youtube-insights-mcp cli list-transcripts "https://www.youtube.com/watch?v=VIDEO_ID"

# Extract insights (pipe transcript from stdin)
youtube-insights-mcp cli get-transcript VIDEO_ID --no-segments | \
  jq -r '.full_text' | \
  youtube-insights-mcp cli extract-insights - --focus "entrepreneurial,investment"

# List all focus areas and categories
youtube-insights-mcp cli list-focus-areas

MCP Tools

get_transcript

Fetch transcript for a YouTube video with intelligent language fallback.

Parameter Type Default Description
video_url string required YouTube URL or video ID
language string "en" Preferred language code

Fallback order: requested language > auto-generated version > any English variant > first available.

list_available_transcripts

List all available transcript languages for a video. Useful for checking what's available before fetching.

Parameter Type Default Description
video_url string required YouTube URL or video ID

extract_insights

Prepare a transcript for structured insight extraction by Claude. Returns an extraction prompt, categories, and chunked transcript data.

Parameter Type Default Description
transcript string required Full transcript text (min 100 chars)
focus_areas string "general" Comma-separated focus areas
video_id string "" Optional video ID for tracking
max_insights string "10" Maximum insights to extract

list_focus_areas

List all available focus area presets and their category definitions.

Focus Areas

Preset Categories
general Key points, action items, notable quotes, context
entrepreneurial Business strategies, growth tactics, lessons learned, mistakes to avoid
investment Market trends, opportunities, risks, recommendations
technical Technologies, tools, best practices, pitfalls
youtube-channel Channel strategy, content ideas, audience growth, production tips
ai-learning AI concepts, AI tools, mental models, practical applications

Use "all" to extract across every category, or combine presets: "entrepreneurial,investment".

Example Prompts for Claude Desktop

Once the MCP server is configured, you can use these prompts directly in Claude Desktop:

Transcript & Summary

> Get the transcript for https://www.youtube.com/watch?v=VIDEO_ID and give me a concise summary

> Fetch the transcript for this video and list the 5 most important takeaways: https://youtu.be/VIDEO_ID

Business & Entrepreneurship

> Get the transcript for https://www.youtube.com/watch?v=VIDEO_ID and extract entrepreneurial insights - focus on actionable business strategies and growth tactics

> Analyse this startup pitch and pull out the key business model, revenue strategy, and mistakes to avoid: https://www.youtube.com/watch?v=VIDEO_ID

Investment Analysis

> Fetch the transcript for this market analysis video and extract investment insights including trends, opportunities, and risks: https://www.youtube.com/watch?v=VIDEO_ID

Technical Learning

> Get the transcript for this tech talk and extract the key technologies mentioned, best practices, and common pitfalls: https://www.youtube.com/watch?v=VIDEO_ID

> Summarise this programming tutorial and list the tools and frameworks recommended: https://youtu.be/VIDEO_ID

YouTube Channel Growth

> Analyse this video about growing a YouTube channel and extract content strategy tips, audience growth tactics, and production advice: https://www.youtube.com/watch?v=VIDEO_ID

AI & Machine Learning

> Fetch the transcript for this AI talk and extract the key concepts, tools mentioned, and practical applications: https://www.youtube.com/watch?v=VIDEO_ID

Multi-Focus Analysis

> Get the transcript for https://www.youtube.com/watch?v=VIDEO_ID and extract insights across both entrepreneurial and investment categories - I want business strategies AND market opportunities

Comparing Videos

> Get the transcripts for these two videos and compare their advice on starting a business:
> - https://www.youtube.com/watch?v=VIDEO_ID_1
> - https://www.youtube.com/watch?v=VIDEO_ID_2

Using with Claude Code & Cowork

The project includes a /yt-insights skill for Claude Code that provides an interactive workflow for extracting insights from YouTube videos. It works in both Claude Code (terminal) and Claude Cowork (collaborative mode).

Setup

  1. Clone the repository and add the MCP server to your Claude Code config (.mcp.json):```json
    {
    "mcpServers": {
    "youtube-insights-mcp": {
    "command": "/absolute/path/to/youtube-mcp-env/bin/youtube-insights-mcp",
    "env": {
    "LOG_LEVEL": "INFO"
    }
    }
    }
    }

2. The `/yt-insights` skill is automatically available from the project's `.claude/commands/` directory.

### Usage

Run the skill in Claude Code or Cowork:```
/yt-insights https://www.youtube.com/watch?v=VIDEO_ID

Or just /yt-insights and it will prompt you for the URL.

The skill will:

  1. Ask you to choose focus areas (general, entrepreneurial, investment, technical, youtube-channel, ai-learning, or all)
  2. Fetch the video transcript
  3. Extract structured insights using Claude
  4. Save two markdown files:
    - Insights — structured analysis with key quotes, action items, and categorised insights
    - Transcript — full transcript text for future reference

Cowork Mode

In Cowork, the skill is automatically converted to a Cowork skill. It uses the MCP tools directly (rather than CLI commands), which means it connects to the youtube-insights-mcp server through the MCP protocol for a seamless experience.

Configuration

Proxy Support

If YouTube blocks your IP, configure a proxy via environment variables:json { "env": { "PROXY_URL": "http://user:pass@proxy:8080", "WEBSHARE_USERNAME": "", "WEBSHARE_PASSWORD": "" } }

Priority: Webshare credentials > PROXY_URL > no proxy.

Requirements

  • Python >= 3.10
  • Dependencies: mcp, pydantic, youtube-transcript-api, yt-dlp

License

MIT

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
Youtube Transcript MCP 11460a5b 怎么安装?

在客户端按下面的安装命令接入:python3 --version。装完在客户端配置里引用同一个包或端点。

Youtube Transcript MCP 11460a5b 怎么在 Claude Desktop、Claude 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Youtube Transcript MCP 11460a5b 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Youtube Transcript MCP 11460a5b 需要 API Key 吗?

资料标注为无需认证,装好就能连;如果实际调用时报鉴权错误,按官方文档补上凭据。

Youtube Transcript MCP 11460a5b 支持哪些客户端?

资料标注支持 Claude Desktop、Claude;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Youtube Transcript MCP 11460a5b 能提供哪些能力?

资料显示约 4 个工具;能力标签包括 视频分析、Python、Claude。具体工具清单以官方文档为准。

Youtube Transcript MCP 11460a5b 需要什么运行环境?

运行时为 Python;属于本地进程,需要在本机装好运行环境。

Youtube Transcript MCP 11460a5b 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-05-17。