# JEWEL AUDIT 1.28

The campaign contains **163 jewels across 45 levels**.

The audit now treats every jewel as authored content rather than filler. It checks bounds, solid embedding, plausible support/jump geometry, moving-platform proximity, hazard-route placement and secret-route placement.

After the 1.28 adjustments the structural audit reports:

- **163 / 163** jewels pass the physical-placement plausibility checks.
- **0** are out of bounds.
- **0** are substantially embedded in ordinary solid geometry.
- **0** remain in the audit's 'review/unexplained' category.
- **6** are deliberately associated with hidden routes as exploration bait/reward.
- **6** are deliberately positioned as hazard-route treasure.

This is a placement audit, not a mathematical proof of a 100%-jewel run through every dynamic level. Robot Dave remains the dynamic solvability tool; Blind Dave and human playtesting judge whether the collectible route is understandable and enjoyable.

### Human-feedback fixes included

Level 2's two spike-route jewels were moved higher into the natural jump arc so they read as collectible rather than impossible.

AFTER YOU had two jewels that were technically near the top edge of Dave's jump envelope. They were lowered to provide human margin.

THE DROP had one apex-level jewel with similarly poor margin. It was lowered.