Competence-aware guardrails for AI-assisted coding
Automatically evaluates code change complexity against your demonstrated competence level.
Generates questions that probe understanding of why code works, not just what it does.
Persistent YAML model tracks your skills across Python concepts and patterns.
3-level difficulty adaptation from conceptual questions to direct hints.
Claude Code calls a tool. VibeCheck intercepts before the change is applied.
The knowledge gate assesses complexity against your competence model.
Complex changes trigger an interactive Q&A that tests your understanding.
Your competence model updates based on your demonstrated knowledge.