What problem does it solve? Onboarding to an unfamiliar repository is slow and shallow when developers only skim files or read stale documentation. This Skill turns an AI agent into a structured tutor that teaches any real codebase from the whole-system map down to module internals, with evidence-grounded explanations and mastery checks. ## Core Features & Use Cases - Whole-System Mapping: Builds a stable architecture map covering entry points, async flows, data stores, deployment units, and cross-cutting concerns before diving into any module. - Grounded Teaching Reports: Produces layered visual learning reports with diagrams, real code citations (repo, ref, file, line range), and explicit evidence labels separating confirmed facts from inference. - Quality and Gap Analysis: Compares the code against current best practices across security, privacy, reliability, performance, and maintainability using a three-way verdict table, and maintains a defect and risk ledger. - Adaptive Assessment: Tests transfer and reasoning rather than trivia, retests only missed competencies, and tracks mastery states with spaced review. - Use Case: A new engineer joins a team owning a multi-service system. The Skill maps the system, teaches each module in dependency order, answers free-form questions against live code, and quizzes the engineer until they can trace flows and predict failure behavior. ## Quick Start Use the codebase-teacher skill to teach me this repository from the whole-system map down to its module internals.
