 admin
|
658e78d495
|
docs(audit): no-mercy full-system campaign report (2026-07-06)
Overnight unattended stress/regression on the demo node (9201). Platform
resilience + safety strong: all fault-recovery + reboot/hard-reset recovery
passed (58-126s; sysrq hard reset 103s), 2.5h soak 0 breaks, DR restore proven
sound. Raw 25 FAIL triaged: ~10 collateral of one app finding (F1 vaultwarden
SMTP crash-loop), ~8 harness mis-scores of correct product refusals
(single-flight 409, needs-confirmation), rest harness-calibration incl. the P9
volid-format bug (DR mechanism itself works - recovered 9201 manually from R1).
Genuine findings: F1 vaultwarden SMTP_HOST (HIGH), F2 decommission agent-403
(MED), F3 gitea-outage sync recovery (MED). No production code changed; demo
left running+healthy at R1, notifications restored.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PSK5g6qYLknKj8u3QAFEr6
|
2026-07-06 12:23:12 +02:00 |
|