capture-for-sketch

name	capture-for-sketch
description	This skill should be used when the user wants to sketch content from a URL, YouTube video, or external source. Enables fetching and extracting content before creating a visual sketch.
version	1.0.0

Capture Content for Sketching

When the user wants to sketch external content (URLs, videos, articles), first fetch and extract the content, then generate the visualization.

Activation Triggers

Use this skill when the user mentions:

• "sketch this YouTube video"
• "create a diagram from this article"
• "visualize https://..."
• "sketch the architecture from this GitHub repo"
• "draw a diagram of this URL"
• "make a visual summary of this video"

Workflow

1. For YouTube Videos

Step 1: Fetch video content

Use yt-dlp to get the transcript (if available):

yt-dlp --write-auto-sub --sub-lang en --skip-download -o "%(title)s" "<URL>"

Or fetch video metadata:

yt-dlp --print title --print description --skip-download "<URL>"

Step 2: Extract key information

• Video title and description
• Main topics from transcript
• Sequence of ideas/steps
• Key takeaways

Step 3: Generate diagram

• Flowchart for tutorials/processes
• Mind map for conceptual content
• Timeline for historical content

2. For Articles/Web Pages

Step 1: Fetch content

Use the WebFetch tool to retrieve and process the page:

WebFetch(url, prompt="Extract the main topics, key arguments, and structure of this article")

Step 2: Extract key information

• Main thesis/argument
• Supporting points
• Relationships between concepts
• Conclusions

Step 3: Generate diagram

• Hierarchical diagram for structured arguments
• Concept map for interconnected ideas
• Comparison chart for reviews/analyses

3. For GitHub Repositories

Step 1: Fetch README

WebFetch("https://github.com/{owner}/{repo}", prompt="Extract the architecture, main features, and components")

Or if repo is local, read the README.md directly.

Step 2: Extract key information

• Project purpose and scope
• Architecture components
• Dependencies and integrations
• Key features

Step 3: Generate diagram

• Architecture diagram showing components
• Dependency graph
• Feature mind map

Content Extraction Patterns

Video Content Structure

Title: {video title}
Main Topic: {core subject}
Key Points:
  1. {point 1}
  2. {point 2}
  3. {point 3}
Flow: {step1} → {step2} → {step3}
Takeaways:
  - {takeaway 1}
  - {takeaway 2}

Article Content Structure

Title: {article title}
Source: {URL}
Main Argument: {thesis}
Supporting Points:
  - {point 1}
  - {point 2}
Key Concepts:
  - {concept 1} relates to {concept 2}
Conclusions: {summary}

Repository Content Structure

Project: {repo name}
Purpose: {description}
Components:
  - {component 1}: {description}
  - {component 2}: {description}
Dependencies:
  - {component 1} → {component 2}
Tech Stack: {technologies}

Example Workflows

Example 1: YouTube Video

User: "/sketch https://youtube.com/watch?v=abc123"

1. Detect URL is YouTube
2. Fetch video info:

   yt-dlp --print title --print description --skip-download "https://youtube.com/watch?v=abc123"
   

3. If transcript available, extract key topics
4. Generate Excalidraw diagram:
   • Title box with video name
   • Main topic boxes
   • Arrows showing flow/relationships
5. Save to ${workbench_path}/sketches/

Example 2: Article

User: "/sketch https://example.com/article-about-microservices"

1. Detect URL is web article
2. Fetch with WebFetch:

   WebFetch(url, "Extract main topics and structure")
   

3. Parse response for key concepts
4. Generate concept map diagram
5. Save to ${workbench_path}/sketches/

Example 3: GitHub Repo

User: "/sketch https://github.com/facebook/react"

1. Detect URL is GitHub
2. Fetch README content
3. Extract architecture and components
4. Generate architecture diagram showing:
   • Core packages
   • Dependencies
   • Data flow
5. Save to ${workbench_path}/sketches/

Integration Notes

Tool Detection

Before using yt-dlp, check if it's available:

command -v yt-dlp >/dev/null 2>&1

If not available, offer alternatives:

• WebFetch to get video page metadata
• Ask user to provide content manually

Fallback Behavior

If content extraction fails:

1. Inform user of the issue
2. Offer to create sketch from manual description
3. Suggest installing missing tools (yt-dlp)

Output Location

Sketches are saved to ${workbench_path}/sketches/ (shared with jot plugin).

Read workbench_path from .claude/jot.local.md:

---
workbench_path: ~/workbench
---

Default: ~/workbench if not configured.