Feynman Learning Guide

Problem

Passive reading often creates the illusion of understanding, but the knowledge fades quickly. The missing piece is a feedback loop that exposes where the explanation breaks down. This skill turns a topic into a few core concepts and asks the user to explain them while the AI probes for gaps.

How It Works

The skill follows a closed-loop process:

• Concept: Extract 3–5 core concepts from the topic and order them by dependency.
• Teach: The AI acts as a curious beginner and asks at most 2 questions per round, avoiding answers and flagging unexplained jargon or logical jumps.
• Review: Rate each explanation as STRONG, WEAK, or WRONG, then patch gaps or correct misunderstandings.
• Simplify: Add 1–2 analogies per concept, including similarities, limits, and where the analogy can mislead.
• Loop-back check: The user retells the whole topic without prompts; any stuck concept returns to teaching and review until it becomes clear.

Caveats

This is best for self-checking technical concepts, paper takeaways, or system design rationale. It is not a summarization tool; it requires active explanation from the user. If the goal is only to receive answers, the value is limited.