What problem does it solve? After agentic coding tools generate or modify a codebase, developers often struggle to understand what was actually built. This Skill turns codebase comprehension into an interactive quiz, progressively deepening understanding of execution flow, design decisions, and data structures across sessions. ## Core Features & Use Cases - Progressive Quizzing: Asks 6 targeted questions per session covering execution flow, design decisions, data structures, entry points, module relationships, and error handling. - Difficulty Scaling: Automatically raises difficulty from level 1 to 5 as session count grows, moving from surface-level questions to deep architecture and edge cases. - Persistent Progress: Stores session count, topics covered, and scores in a progress.json file inside the target codebase so questions never repeat. - Use Case: After an AI agent scaffolds a new web service, run /codebase-quizzer on the project directory to verify you understand how requests flow through the code before modifying it. ## Quick Start Invoke /codebase-quizzer with the path to your codebase, for example by asking to be quizzed on the project at ./my-app.
