interactive learning

Traditional learning often fails because it follows a linear path regardless of what the student already knows. This skill solves that by identifying hidden knowledge gaps before teaching a single concept. It transforms static documents into a dynamic, personalized curriculum that focuses strictly on your blind spots.

What it does

• Blind-spot diagnosis probes for misconceptions using targeted questions rather than simple quizzes.
• Deep-source analysis decomposes reference materials into a granular skeleton with line-by-line citations.
• Adaptive sequencing reorders the learning path to prioritize P0 knowledge gaps over the book's table of contents.
• Socratic instruction uses questioning to lead you to insights rather than lecturing.
• Feynman verification requires you to explain concepts in your own words before unlocking new content.

How it works

1. Material Ingest: Provide a topic or upload source files like .md, .txt, or .pdf for analysis.
2. Diagnosis: Complete a four-question assessment to map out your current understanding and errors.
3. Skeleton Building: The skill runs parallel agents to index the source material with precise line-number evidence.
4. Unit Delivery: Engage with one unit at a time, providing feedback and passing Feynman checks to progress.

Frameworks & tools

This skill utilizes standard file system operations for directory management and includes specific support for convert_docx_to_md.sh and web-based source reconstruction for image-heavy PDFs.

Why this beats prompting it yourself

Generic prompts produce generic summaries that gloss over what you don't understand. This skill enforces a strict "diagnose before teaching" rule and uses parallel processing to ensure every claim is backed by a specific line citation from your source text.

Use cases

• Deep-reading technical books to master complex conceptual systems without skimming.
• Onboarding to new codebases by converting documentation into a structured learning path.
• Mastering counterintuitive subjects where your existing mental models might be holding you back.
• Systematic research into specific documents where precise citations are required for every argument.

Known limitations

Scanned or image-based PDFs require web search reconstruction rather than verbatim extraction. Manual feedback is required to advance between units; it will not auto-generate the next lesson.