YouTube Watcher
Fetch and read transcripts from YouTube videos for summarization, question answering, and content extraction.
# YouTube Watcher Fetch and read transcripts from YouTube videos for summarization, question answering, and content extraction. YouTube Watcher retrieves text transcripts from YouTube videos, enabling your AI agent to summarize videos, answer questions about video content, and extract specific information — all without watching the video itself. ## How It Works The skill uses yt-dlp to download closed captions or auto-generated subtitles from any YouTube video. Once the transcript is fetched, your AI agent can process the text to summarize content, search for keywords, or answer user questions based on the video's spoken content. ## Key Features - **Transcript Extraction**: Pulls full text transcripts from any YouTube video with available captions - **Auto-Generated Subtitle Support**: Works with both manually uploaded closed captions and YouTube's auto-generated subtitles - **Video Summarization**: Enables AI-powered summaries of video content via transcript analysis - **Content Search**: Search through video transcripts for specific keywords or topics ## Requirements - **yt-dlp**: Command-line tool for downloading YouTube subtitles. Must be installed and available in PATH - **Python 3**: Required to run the transcript extraction script ## Use Cases - **Video Summarization**: Get a quick summary of a long YouTube video without watching it - **Research**: Extract specific information or quotes from educational or tutorial videos - **Content Analysis**: Analyze video content for keywords, topics, or sentiment - **Accessibility**: Convert video content to searchable text for reference ## Installation Install via: `npx clawhub@latest install youtube-watcher`