PetSmart
Rebuilt by Claude Opus 5.5: 125 of 837 hidden tests passed (14.9%)
reference.local/
Referencerebuild.local/
Claude Opus 5.5| Model | Passed | % | Screenshot |
|---|---|---|---|
| 125 | 14.9 | ||
| 125 | 14.9 | ||
| 124 | 14.8 | ||
| 106 | 12.7 | ||
| 93 | 11.1 | ||
| 90 | 10.8 | ||
| 69 | 8.2 | ||
| 21 | 2.5 |