anthropic-skilljar-extractorlisted

# Anthropic Skilljar Course Extractor ## Step 1: Run the Extraction Script ```bash python3 scripts/extract_course.py \ <skilljar_url> <output_dir> ``` This produces: ``` <course-slug>/ ├── README.md ├── 01-<section>/ │ ├── 01-<lesson>.md │ └── ... └── assets/images/ ← downloaded .webp diagrams ``` ## Step 2: Describe Diagrams For each full-size `.webp` in `assets/images/` (not `-medium.webp`): - Use the `Read` tool to view it visually - Add a detailed alt-text block and embed with `` in the relevant lesson markdown Assign images to lessons based on topic — architecture diagrams go in intro lessons, code screenshots in hands-on lessons, summary slides in review lessons. ## Step 3: Enrich Lesson Markdown The script embeds bullet-point notes. Expand each lesson file: - Add proper `##` headings per concept - Format code examples in fenced code blocks with language tags - Convert "Key: value" lines to `**Key:** value` bold format - Add tables for comparisons (e.g., message types, primitives) - Keep the frontmatter block (Section, URL, Lesson Type) ## Step 4: Enhance README Add to the auto-generated README: - Full course description paragraph - Key concepts reference table - Quick-reference code snippets - Diagram index with one-line descriptions per image ## How the Script Works (for debugging) - Fetches page HTML; extracts `window.__chatData` (LLM notes embedded in every Skilljar page) - Parses sidebar `<ul class="dp-curricul