Youtube Summarizer

brianrwagner/ai-marketing-claude-code-skills

Drops a YouTube URL, get a structured summary and full transcript in 20 seconds. Works where yt-dlp fails on VPS by using Android client emulation. You get metadata (title, channel, views, date), a consistent thesis/insights/takeaway format, and the full timestamped transcript saved to file and optionally pushed to Telegram. Three modes: quick TL;DR, standard summary, or deep analysis with content repurposing angles. The real win is skipping the manual copy-paste-to-ChatGPT dance and getting searchable archives automatically. One user cut competitor video analysis from 30 minutes to 20 seconds per video, built a database of 160+ summaries in two months.

skills.sh installs: 473
GitHub stars: 455
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add brianrwagner/ai-marketing-claude-code-skills --skill youtube-summarizer --agent claude-code

Installs into .claude/skills of the current project.

Files

YouTube Summarizer Skill

Automatically fetch transcripts from YouTube videos, generate structured summaries, and deliver full transcripts to messaging platforms.

Mode

Detect from context or ask: "Quick TL;DR, full summary, or full summary with content angles?"

Mode	What you get	Best for
`quick`	3-bullet TL;DR + single key takeaway	Fast consumption, sharing a clip
`standard`	Full structured summary: thesis, insights, takeaway	Learning, note-taking, research
`deep`	Full summary + chapter breakdown + content repurposing opportunities	Turning a video into a content asset

Default: standard — use quick if they just want the gist. Use deep if they want to extract the video into usable content.

Why This vs ChatGPT?

Problem with ChatGPT: It can't access YouTube transcripts directly. You have to manually copy/paste captions or use a third-party tool first, then feed the text to ChatGPT. Multi-step, clunky, loses video metadata.

SKILL.md on GitHub