What problem does it solve? Understanding a large open-source codebase end-to-end is slow and unstructured; this Skill autonomously produces complete teaching documentation covering every source file, from high-level architecture down to individual functions and edge cases. ## Core Features & Use Cases - Goal Loop Orchestration: Autonomously builds a goal backlog (L0 overview, L1 modules, L2 vertical slices, L3 API references, L4 deep dives), dispatches up to 4 parallel subagents, audits outputs, and loops until full source coverage is verified. - Five-Level Teaching System: Produces HTML short lessons and reference documents organized into 00-overview/, module-, slice-, and deep-dive-* topic directories under teach/<project>/. - Quality Gates: Enforces path validation, topic audits via audit_topic.py, sampled teach-review subagent reviews, and a Definition of Done checklist with coverage diff scanning. - Use Case: Point it at a repository like RuoYiVuePlus and ask for a complete walkthrough; it generates a Wiki-style 00-index.md navigation plus dozens of short lessons tracing features like login authentication from route to database. ## Quick Start Ask the agent to generate complete teaching documentation for a specific project directory or repository in the workspace, covering everything from architecture to implementation details.