ExtractWisdom

Extract content-adaptive wisdom from YouTube transcripts, articles, and pasted text into structured markdown.

1|Updated Apr 16, 2026

Search Tags:#markdown #webfetch #content summarization #wisdom extraction #key takeaways #youtube transcript #contrarian insights

Ready to run

npx skills add https://github.com/davdunc/pai-framework --skill extractwisdom-davdunc

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ExtractWisdom
Source: https://github.com/davdunc/pai-framework/tree/main/skills/ExtractWisdom
Command: npx skills add https://github.com/davdunc/pai-framework --skill extractwisdom-davdunc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Extracting insights from videos, podcasts, and articles often produces generic summaries with the wrong headings, so the best ideas get buried or lost.

Core Features & Use Cases

Content-adaptive section building: Reads first, detects which wisdom domains exist, then generates custom, relevant sections instead of static headers.
Depth-controlled outputs: Produces Instant, Fast, Basic, Full (default), or Comprehensive extractions with matching section counts and closing sections.
Opinionated wisdom extraction: Requires contrarian/spicy takes, preserves human moments, and writes in the user’s conversational style.
Multi-source ingestion: Uses fabric -y URL for YouTube transcripts and WebFetch for articles, then extracts wisdom with a consistent final format.

Quick Start

Use ExtractWisdom to extract wisdom from this YouTube or article URL and return a Full-depth, markdown-formatted set of dynamic sections plus references.

Frequently Asked Questions about ExtractWisdom

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

FAQPage Schema

How do I extract key takeaways from a YouTube transcript?▼

You can extract key takeaways from a YouTube transcript by fetching the content via fabric -y and routing it through the Extract workflow to generate dynamic, markdown-formatted sections with contrarian insights.

Can I summarize articles and pasted text using content-adaptive sections?▼

Yes, you can summarize articles and pasted text using WebFetch for URLs or direct text ingestion, which detects relevant insight domains and builds custom sections instead of static headers.

What is content-adaptive wisdom extraction for long-form podcasts?▼

Content-adaptive wisdom extraction reads long-form podcasts first, detects which wisdom domains exist, and generates dynamic, opinionated markdown sections with depth-controlled takeaways rather than generic summaries.

How do I control the depth of content summarization for interviews?▼

You control content summarization depth by selecting Instant, Fast, Basic, Full, or Comprehensive extraction levels, which adjust the number of generated sections and closing outputs for interviews.

Does WebFetch work with markdown formatting for article summarization?▼

Yes, WebFetch works with markdown formatting by fetching article URLs and routing the content through a deterministic workflow to output structured markdown with required closing sections and references.

What is the best way to get contrarian insights from blogs?▼

The best way to get contrarian insights from blogs is using an opinionated extraction process that preserves human moments, requires spicy takes, and writes in your conversational style across dynamic sections.