A your AI Code skill that reviews codebases top-down: WHY (goal) → WHAT (architecture) → HOW (implementation) → improvements ranked by leverage.
The antidote to surface-level LLM nitpick reviews.
1 in the Claude listing