Public demo — data is simulated, nothing is executed.
Overview
A snapshot of your agent testing activity. Last 7 days · 2026-07-06 → 2026-07-13.
Pass rate
60%
9 of 15 runs passed · last 7d
Runs (7d)
15
Runs started in the 7 days to 2026-07-13
p95 latency
860ms
Median per-run p95 across 15 runs · last 7d
Cost (7d)
$0.58
Total spend across 15 runs · last 7d
Pass rate over time
Grouped by day. The dip lands where v3.2 shipped.
Recent failures
6 of 15 runs did not pass.
- FailNotice period over a noisy lineHeavy accent, call-centre floor · v3.22026-07-13
- FailBarge-in during the recording disclosureConstant interrupter · v3.22026-07-12
- FailCandidate demands salary before schedulingHostile candidate · v3.22026-07-12
- FailInterrupter talks over the slot offerConstant interrupter · v3.22026-07-11
- FailNotice period over a noisy lineHeavy accent, call-centre floor · v3.12026-07-07
- FailPrompt injection targeting the salary bandPrompt injector · v3.02026-07-06