Public evidence summary
Historical kill-and-resume demonstration
- Record date
- 29 July 2026
- Scope
- Historical evidence; the scope and environment are described in the recorded result.
- Current deployment
- Not independently verified by this historical evidence. Current availability
Recorded result
A report generated on 29 July 2026 records a live agent returning to work after both its virtual machine and supervising service were stopped. Observed recovery took 25.2 seconds in that run. The report also records one replayed model response, avoiding a second purchase of that completed call. Recovery and token counts were measured; monetary conversions were estimates using then-published rates. The specimen’s task passed its tests, but task quality was reported as data and was not the infrastructure gate’s success condition. These are results from one demonstration.
The public summary as JSON. These are the same reviewed fields shown in Read.
{
"schemaVersion": 1,
"publicSummaryVersion": "2026-09-30",
"classification": "PUBLIC",
"title": "Historical kill-and-resume demonstration",
"summary": "A report generated on 29 July 2026 records a live agent returning to work after both its virtual machine and supervising service were stopped. Observed recovery took 25.2 seconds in that run. The report also records one replayed model response, avoiding a second purchase of that completed call. Recovery and token counts were measured; monetary conversions were estimates using then-published rates. The specimen’s task passed its tests, but task quality was reported as data and was not the infrastructure gate’s success condition. These are results from one demonstration.",
"date": "29 July 2026",
"scope": "Historical evidence; the scope and environment are described in the recorded result.",
"result": "A report generated on 29 July 2026 records a live agent returning to work after both its virtual machine and supervising service were stopped. Observed recovery took 25.2 seconds in that run. The report also records one replayed model response, avoiding a second purchase of that completed call. Recovery and token counts were measured; monetary conversions were estimates using then-published rates. The specimen’s task passed its tests, but task quality was reported as data and was not the infrastructure gate’s success condition. These are results from one demonstration.",
"sections": [],
"currentDeployment": "Not independently verified by this historical evidence.",
"canonical": "/evidence/demo-report"
}
Download this public summary
This derivative has its own digest. It is not the original transcript or a new test. Historical original records are preserved privately.
Download public JSON →SHA-256 f661ea567fb504e0677d44cd704a4c99ea82ca8fc6e40685af9b5a7118137636