# Tool Call Benchmark

| Field | Value |
|-------|-------|
| Date | 2026-04-08 18:27:06 NZST |
| Model | gpt-5.4 |
| Suite | quick |
| Categories | Bash Execution, File Operations |
| Total duration | 1645 ms |
| Result | 10 / 10 passed |
| First-attempt accuracy | 100% |

## Summary

The quick tool-call benchmark completed cleanly. All Bash and file-operation tests passed on the first attempt, with no misfires recorded.

## Category results

### Bash Execution

- `TC-B01` Echo exact string: PASS in 313 ms
- `TC-B02` Python arithmetic: PASS in 342 ms
- `TC-B03` Node JSON output: PASS in 343 ms
- `TC-B04` Pipeline command: PASS in 320 ms

### File Operations

- `TC-F01` Write file: PASS in 1 ms
- `TC-F02` Read file back: PASS in 0 ms
- `TC-F03` Edit file: PASS in 0 ms
- `TC-F04` Verify edit: PASS in 0 ms
- `TC-F05` Glob find: PASS in 0 ms
- `TC-F06` Grep search: PASS in 326 ms

## Artifacts

- `results.json`: raw per-test timings and outcomes
- `PROMPT.md`: benchmark invocation used for this route