Conventional prompt
average structural self-audit score
→
Optimized prompt
average structural self-audit score
Structural gain
percentage points, use cases
These scores apply the skill's own pre-launch rubric. They measure prompt structure,
not whether a model will solve the task.