# Benchmark run 20261015T171401Z-suite-tool_selection-baseline-a1426d

Status `success`, agent `baseline`, commit `2925aeab2262`.

| suite | tasks | pass rate | 95% CI | incidents as fail | incident rate | tokens in / out | cost |
|---|---|---|---|---|---|---|---|
| tool_selection | 70 | 0.943 | [0.862, 0.978] (wilson) | 0.943 | 0.000 | 0 / 0 | — |

- tool_selection: failures wrong_tool 4

Rendered from summary.json; every number there is recomputable from results.jsonl.