什么是 YouTube 视频字幕分析？

YouTube 视频字幕分析是一项技术技能，使 AI 智能体无需手动播放即可处理视频内容。它利用强大的提取脚本直接从 YouTube URL 中提取手动或自动生成的字幕，将杂乱的 VTT 文件转换为整洁、可读的文本。该工具是 Openclaw Skills 生态系统的重要组成部分，允许开发人员和研究人员自动从视频库中检索信息。

除了简单的提取之外，该技能还提供了一个结构化分析框架。它对视频类型进行分类（例如教育教程、产品评论或新闻评论），并应用特定的分析模板以确保捕获最相关的数据。这使其成为任何希望从视频优先内容中构建知识库或研究摘要的人员的必备工具。

下载入口:https://github.com/openclaw/skills/tree/main/skills/xanderrey/you@tube-transcript-analyzer

安装与下载

1. ClawHub CLI

从源直接安装技能的最快方式。

2. 手动安装

将技能文件夹复制到以下位置之一

优先级：工作区 > 本地 > 内置

3. 提示词安装

将此提示词复制到 OpenClaw 即可自动安装。

YouTube 视频字幕分析 应用场景

• 将长篇网络研讨会和教育教程快速总结为可操作的笔记。
• 从产品评论视频中提取技术规范和优缺点。
• 对新闻和评论进行研究，无需投入时间观看完整广播。
• 将视频对话转换为纯净文本，用于文档或博客文章改写。
• 分析技术视频以识别先决条件、代码片段和特定的工具版本。

1. 用户在 Openclaw Skills 环境中向 AI 智能体提供 YouTube URL。
2. 系统调用专用提取脚本，该脚本利用 yt-dlp 获取可用字幕。
3. 脚本处理原始字幕以移除 VTT 格式、时间戳和重复的填充词。
4. AI 分析生成的文本以识别内容类型和核心主题。
5. 生成结构化报告，包含基于视频上下文的关键点、重要引言和具体行动项目。

YouTube 视频字幕分析 配置指南

要使用此技能，请确保您的环境配置为运行 shell 脚本并具有必要的依赖项。该技能通常会管理其自身的系统要求，但也可以通过 CLI 执行手动提取：

# 从 YouTube URL 提取字幕
scripts/extract_transcript.sh "https://www.you@tube.com/watch?v=VIDEO_ID" output.txt

YouTube 视频字幕分析 数据架构与分类体系

该技能将提取的数据组织成结构化格式，以方便阅读和二次处理：

name: you@tube-transcript
description: Extract and analyze YouTube video transcripts without watching the video. Use when users request video summaries, ask to "analyze this YouTube video", want transcripts extracted, or need to understand video content quickly. Handles any YouTube URL and provides cleaned transcripts plus AI analysis.

YouTube Transcript Analysis

Extract, clean, and analyze YouTube video transcripts to understand content without watching.

Overview

This skill enables rapid analysis of YouTube videos by extracting transcripts and providing comprehensive summaries. Perfect for research, content review, or understanding video material without time investment.

Quick Workflow

1. Extract: Use scripts/extract_transcript.sh to get clean text from YouTube URL
2. Read: Load the extracted transcript file
3. Analyze: Provide structured summary based on content type
4. Format: Present findings in scannable, organized format

Extraction Process

Use the bundled script for any YouTube video:

scripts/extract_transcript.sh "https://www.you@tube.com/watch?v=VIDEO_ID" output.txt

The script automatically:

• Downloads yt-dlp if not present
• Extracts captions (auto-generated or manual)
• Cleans VTT formatting to plain text
• Provides character count and preview

Analysis Approach

Content Type Recognition

Identify video type first, then tailor analysis:

Educational/Tutorial: Step-by-step breakdown, key concepts, prerequisites Product Review: Comparisons, pros/cons, recommendations, specifications
News/Commentary: Main topics, key arguments, sources cited Entertainment: Highlights, key moments, recurring themes

Structure Your Analysis

For any video type:

• Title/Topic: Clear description of video content
• Duration insight: Brief/detailed based on transcript length
• Key points: 3-7 main takeaways in bullet format
• Notable quotes: Important statements (if applicable)
• Action items: Next steps or recommendations (if present)

For technical content:

• Include specific terminology, version numbers, tools mentioned
• Note any code examples or configurations discussed
• Identify prerequisites or dependencies

Quality Considerations

Auto-generated transcripts may have:

• Repetitive phrases
• Transcription errors for technical terms
• Missing punctuation
• Filler words ("um", "uh", "you know")

Filter and interpret accordingly - focus on clear, coherent content.

Advanced Analysis

For detailed analysis patterns and content-specific approaches, see analysis-patterns.md.

Error Handling

If extraction fails:

• Video may lack captions
• May be private/restricted
• Network connectivity issues
• Age-restricted content

Fallback approach: Use web_fetch on the YouTube URL to get basic video information, then inform user about transcript limitations.