AgentSkillsCN

learning-content-summarizer

将来自 URL、YouTube 视频或操作指南的技术/AI/数据相关内容，归纳总结为可操作的、循序渐进的学习指南。适用于收到视频、文章或教程 URL 时，将其转化为易于学习的指导手册。

SKILL.md
--- frontmatter
name: learning-content-summarizer
description: Summarizes technical/AI/data content from URLs, YouTube videos, or how-to guides into actionable step-by-step learning instructions. Use when given a video, article, or tutorial URL to convert into a learnable guide.

Learning Content Summarizer

Transforms technical content (videos, articles, tutorials) into optimized step-by-step learning guides for rapid skill adoption.

Capabilities

  • •Extract key concepts from YouTube videos, articles, and documentation
  • •Create structured, actionable step-by-step instructions
  • •Capture screenshots at critical moments for visual reference
  • •Identify prerequisites and dependencies
  • •Highlight common pitfalls and pro tips
  • •Generate practice exercises for reinforcement

Workflow

1. Content Ingestion

When given a URL:

  1. •Use read_web_page to fetch the content
  2. •For YouTube videos, extract video ID and use transcript if available
  3. •Identify the content type: tutorial, explanation, demo, or reference

2. Content Analysis

Extract and organize:

  • •Goal: What will you be able to do after completing this?
  • •Prerequisites: Required knowledge, tools, accounts, or setup
  • •Key concepts: Core ideas explained simply (ELI5 where helpful)
  • •Time estimate: Realistic time to complete hands-on

3. Step-by-Step Guide Generation

Create instructions following this structure:

markdown
# [Topic Title]

## 🎯 Goal
One-sentence outcome statement.

## ⚙️ Prerequisites
- [ ] Tool/software installed
- [ ] Account created
- [ ] Prior knowledge required

## ⏱️ Time Estimate
~X minutes/hours

## 📋 Steps

### Step 1: [Action Title]
**What**: Brief description
**Why**: Reason this step matters
**How**:
1. Specific action
2. Specific action
3. Verify: Expected result

> 💡 **Tip**: Pro tip or shortcut

> ⚠️ **Watch out**: Common mistake to avoid

### Step 2: [Action Title]
...

## ✅ Verification
How to confirm you've succeeded.

## 🔗 Next Steps
What to learn next for deeper understanding.

## 📚 Resources
- Original source link
- Related documentation
- Community/support links

4. Screenshots and Visuals

When processing video content or when visuals are critical:

  1. •Note timestamps where key UI/visuals appear
  2. •Describe what the screenshot should capture
  3. •If user provides screenshots, use look_at to analyze and reference them

For web content:

  1. •Identify diagrams or UI elements worth capturing
  2. •Describe or link to relevant visual assets

Output Guidelines

  • •Be actionable: Every step should be something the user can DO
  • •Be specific: Include exact commands, button names, menu paths
  • •Be sequential: Order matters—dependencies first
  • •Be concise: Cut fluff, keep only what helps learning
  • •Be practical: Focus on "how" over "why" (unless understanding is crucial)

Example Usage

User provides: "Summarize this video: https://youtube.com/watch?v=..."

Response structure:

  1. •Acknowledge the content type
  2. •Fetch/analyze the content
  3. •Generate the structured guide
  4. •Offer to save as a markdown file in the repo

Integration with Learn In Public

When appropriate, offer to:

  • •Save the guide to weeks/YYYY/week-XX/ as a detailed post
  • •Link it from the weekly README
  • •Tag with appropriate category (💻 AI, Data & Programming)