Post-incident: queue_liveness triggered 2026-07-19T18:45:02.764Z

{
  "invariant": "queue_liveness",
  "severity": "SEV-2",
  "evidence": {
    "planningValidated": 8,
    "inFlight": 1,
    "maxInFlight": 15,
    "recentDispatchCount": 0,
    "windowMinutes": 15,
    "lastSuccessfulDispatchAt": "2026-07-19T18:44:34.152Z"
  },
  "slack": {
    "channelId": "C0BJH90REE4",
    "ts": "failed-1784486702750"
  }
}

---
<!-- guardian-reclassify -->
**Auto-canceled by Forge reliability audit 2026-07-24:** stale Post-incident ticket. Slack posting from PipelineHealthProbe was broken during this window (ts:failed-* in body). Originating stall has since recovered. AWAITING_HUMAN is not a valid resting state for probe-generated incidents; superseded by Layer-4.6 runbook-required handling.