ops-dashboard · pi-gemini-3.8-flash
pi · started 2026-10-08T14:26:18.665Z · horizon 420s · judge ai (openai:gpt-5.5), 5 frames scored
Area under curve
0.874
headline, 0-1
Tab title/icon
19.9s
browser chrome, not the page
First render
19.9s
anything on screen
First reviewable
1m 2s
worth feedback
Final score
1.00
no regression
Correctness over time
Where the time went
99%
Model thinking 1m 37sTool round trips 0.1sUnaccounted 0.7s
| Phase | Time | Share | Side |
|---|---|---|---|
| Model thinking | 1m 37s | 99.2% | agent |
| Tool round trips | 0.1s | 0.1% | agent |
| Dependency install | 0.0s | 0.0% | toolchain |
| Build | 0.0s | 0.0% | toolchain |
| Dev server boot | 0.0s | 0.0% | toolchain |
| First paint | 0.0s | 0.0% | toolchain |
| Unaccounted | 0.7s | 0.7% | unknown |
| Wall clock (cold start) | 1m 38s | 100% |
Agent or toolchain
99%
Agent (model + tool round trips) 1m 37s
Toolchain (install + build + boot + paint) 0.0s
Unaccounted 0.7s
Accounted for: 99.3% of wall clock.
- Excludes 15.0s of deliberate idle observation (settle window); the curve still covers it.
Iteration: prompt to visible change
Follow-ups were injected into the same live session.
| Edit | First visible change | Correct change | Agent done | Tool calls | After agent | Broken for |
|---|---|---|---|---|---|---|
| Make the page heading blue. | 4.9s | 4.9s | 13.1s | 1 (edit) | -8.2s | -- |
| Add an eleventh depot row for Porto: 6,480 shipments, 93.1% on-time, $19.95 cost per shipment. | 5.8s | 5.8s | 10.9s | 1 (edit) | -5.1s | -- |
Judged frames
A styled dashboard shell with correct heading, but required KPI, table, chart, and footer content are missing.
Polished header and KPI row, but table, chart bars, and footer are missing in the screenshot.
Strong dashboard layout with complete KPIs and table, but chart bars and footer are missing.
Meets nearly all visible requirements; footer is not visible in screenshot.
All requested dashboard elements are visible, accurate, aligned, and deliberately styled.
Caveats
- 1 poll ticks were skipped because a capture overran the interval.
- judge: openai:gpt-5.5 does not accept a sampling temperature, so these frames were scored at the provider's default rather than at 0. Two runs of the same page can be scored differently, and that variance is inside the AUC. result.json records the temperature as null rather than as the value that was asked for, so p2p compare will not pool this run with one that was scored at a fixed temperature.