summarize

Summarize videos, articles, PDFs, EPUBs, and podcasts into structured Obsidian notes.

Updated Apr 12, 2026
One-click install
npx skills add https://github.com/apairoflevis/ai-skills --skill summarize-apairoflevis
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: summarize
Source: https://github.com/apairoflevis/ai-skills/tree/main/summarize
Command: npx skills add https://github.com/apairoflevis/ai-skills --skill summarize-apairoflevis

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Users need a reliable way to turn diverse source material—videos, articles, PDFs, ebooks, and podcasts—into well-structured, searchable Obsidian notes without manual transcription, layout cleanup, or excessive reference-note noise.

Core Features & Use Cases

  • Universal extraction: Handles YouTube and podcasts (yt-dlp + transcription), web articles (defuddle), PDFs including scanned/multi-column pages (marker-pdf with pdftotext fallback), EPUBs/DOCX (pandoc), and pasted or vault content.
  • Structured Obsidian output: Produces notes with frontmatter, TLDR callouts, section-by-section breakdowns, timestamps for audio/video, and selective [[wikilinks]] for key people and concepts.
  • Modes and automation: Supports quick/full/deep depth modes, saves transcripts for audio, selectively creates People and References notes, and runs parallel summarization agents for long content.

Quick Start

Summarize a URL, file path, or pasted text into an Obsidian summary note using the summarize skill.

Frequently Asked Questions about summarize

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I convert a YouTube video into an Obsidian note with timestamps?

To convert YouTube videos into Obsidian notes, this tool uses yt-dlp to extract audio and generates transcripts with timestamps. It then creates a structured note featuring section-by-section breakdowns and TLDR callouts.

Can I extract text from scanned or multi-column PDFs into markdown?

Yes, you can extract text from scanned or multi-column PDFs into markdown using marker-pdf, with pdftotext as an automatic fallback. This process handles complex layouts to produce clean, readable reference notes.

Does this transcription method support podcasts and EPUB files?

Podcast transcription and EPUB extraction are fully supported. Audio sources are processed via yt-dlp for transcripts, while EPUB and DOCX files are converted into structured markdown using pandoc.

What is the best way to summarize long articles into Obsidian wikilinks?

The best way to summarize long articles into Obsidian wikilinks is using the deep mode, which deploys parallel subagents. It selectively generates People and References notes to organize key concepts without manual linking.

What are the limitations when converting web articles to markdown?

When converting web articles to markdown, limitations depend on the source website's structure, though defuddle is used to reliably extract main content. Output depth is constrained by the selected quick, full, or deep extraction modes.

Do I need to manually configure my Obsidian vault path for note generation?

You need to set a configurable vault path to direct the note generation output into your specific Obsidian vault. Once configured, summaries and extracted transcripts are saved automatically to the correct directories.