| Agent | Model | Reward | Runtime | Tools | Classification | |
|---|---|---|---|---|---|---|
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 8 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 9 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 7 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 9 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 8 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 7 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 10 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 9 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 9 | GOOD_SUCCESS | view → |
| claude-code | claude-haiku-4-5 | ✓ resolved | — | 10 | GOOD_SUCCESS | view → |
| codex | gpt-5.5 | ✓ resolved | — | 30 | GOOD_SUCCESS | view → |
| codex | gpt-5.5 | ✓ resolved | — | 28 | GOOD_SUCCESS | view → |
| codex | gpt-5.5 | ✓ resolved | — | 22 | GOOD_SUCCESS | view → |
| codex | gpt-5.5 | ✓ resolved | — | 32 | GOOD_SUCCESS | view → |
| codex | gpt-5.5 | ✓ resolved | — | 18 | GOOD_SUCCESS | view → |
The task as the agent saw it and the verifier graded it. The reference solution/ is included below for transparency.