youtube-summarizer

The youtube-summarizer skill extracts transcripts from YouTube videos and generates comprehensive, detailed summaries using a STAR + R-I-S-E analysis framework. It solves the problem of documenting video content, lectures, tutorials, and talks without rewatching them, producing verbose, structured Markdown that prioritizes completeness over brevity and captures insights, arguments, terminology, and mentioned resources. It targets users who need thorough reference documentation from educational or informational videos.

The workflow validates a YouTube URL and extracts the video ID across common URL formats, checks that the video is accessible and has captions, extracts the transcript via the third-party youtube-transcript-api Python library (preferring the original or Portuguese language, falling back to English), and then generates a sectioned summary with an executive synthesis, detailed breakdown, key insights, concepts and terminology, mentioned resources, and a conclusion. It manages its own dependency by detecting whether youtube-transcript-api is installed and offering to pip install it after asking the user, and it shows a visual progress gauge across five steps. Flexible save options let the user store the summary only, summary plus raw transcript, or transcript only. The skill explicitly does not download video files; only text transcripts are retrieved.

Use cases include documenting lectures and conference talks, analyzing technical tutorials, capturing research findings, and building language-learning notes. Limitations are clearly stated: it cannot handle private, unlisted, age-restricted, or caption-less videos. The only system actions are running python3 and an optional user-approved pip install, both benign.