What problem does it solve? Explaining an unfamiliar codebase, whitepaper, or complex system well is hard: explanations drown learners in jargon, skip the big picture, and vanish when the chat session ends. This Skill runs a structured walkthrough that builds genuine understanding and persists it as a reusable reference explainer. ## Core Features & Use Cases - Guided walkthroughs: Scopes the learner's goal, presents a big-picture map, then explains one segment at a time using plain language, analogies, and concrete examples before abstraction. - Post-walkthrough understanding checks: Asks a few targeted questions after exposure to find shaky segments and re-explains them from a different angle until they land. - Persistent OKF knowledge bundles: Saves the explainer as an Open Knowledge Format bundle with source fingerprints pinned (git HEAD or sha256), so later sessions can detect drift and refresh diff-aware. - Use Case: Onboard a new engineer onto a repository: the Skill tours one real code path end-to-end with file:line citations, surfaces design rationale, and writes a standalone explainer into docs/knowledge/ that stays fresh-checkable as the code evolves. ## Quick Start Ask the agent to walk you through this repository using the feynman-walkthrough skill and save the explainer for future sessions.
