audit-context-building

A structured analysis skill for the pre-audit phase of code review. It guides Claude through line-by-line function analysis, cross-boundary call tracing, and invariant mapping before any vulnerability hunting begins. The output is a stable, evidence-based mental model of how the system actually works. Manual pre-audit context building is inconsistent and often skips helper functions or external call edge cases, which is exactly where assumptions propagate into bugs — this skill enforces a repeatable checklist that prevents that.