yt-notes ✦

YouTube videos/playlists → extract chapters/timestamps → auto-generate Markdown "notes" with TOC (table of contents), title, description, and links.
Bonus: Audio-only mode for podcasts, watch mode for channel RSS feeds.

Features

• 📹 Smart Downloads: Single videos, playlists, or channels via yt-dlp (1,000+ sites supported)
• 📝 Chapter Magic: Auto-parses timestamps → clickable Markdown TOC (- [Intro](video.mp4#t=0s))
• 🎧 Audio Mode: Extract MP3s for podcasts (no video bloat)
• 👀 Watch Mode: Poll RSS feeds → auto-download new uploads
• 🎨 Pro UX: Rich progress + Loguru logs
• ⚡ Lightweight: Minimal deps — runs fast, outputs clean notes.md

Installation

Requires Python 3.10+ and ffmpeg (for audio/merges — install via brew/apt/choco).

pip install git+https://github.com/iminierai-aig/yt-notes.git

Or install from source:

git clone https://github.com/iminierai-aig/yt-notes.git
cd yt-notes
pip install -e .

Quick Start & Usage

Download video + generate notes

yt-notes https://youtube.com/watch?v=abc123

Playlist batch (saves to ./downloads/)

yt-notes https://youtube.com/playlist?list=PLxyz -o ./downloads

Audio-only (MP3 + notes)

yt-notes https://youtube.com/watch?v=abc123 --audio

Notes only (no download, from metadata)

yt-notes https://youtube.com/watch?v=abc123 --notes-only

Watch channel (auto-download new vids every 5 min)

yt-notes --watch "https://youtube.com/feeds/videos.xml?channel_id=UC123" -o ./auto --interval 300

Quiet mode

yt-notes big-playlist --quiet

See yt-notes --help for more options.

Demo

<iframe src="https://www.loom.com/embed/e352c38ea8e244c093d81f7d300c8d1f" frameborder="0" webkitallowfullscreen mozallowfullscreen allowfullscreen style="position: relative; top: 0; left: 0; width: 100%; max-with: 700px: height: 450px;"></iframe>

Development

1. Clone: git clone https://github.com/iminierai-aig/yt-notes.git
2. Install: pip install -e .
3. Run: python -m yt-notes --help
4. Build: python -m build && twine upload dist/*

Built with yt-dlp in ~24 hours. PRs for subtitle integration or AI summaries welcome!

License

MIT — see LICENSE.