Use when you need a live read on service health, want to verify a deploy against pre-deploy baseline, or are actively watching an incident. Fans out to alerts/metrics/logs/k8s surfaces via the watchtower capability contract and escalates to lets-triage-incident on CRITICAL findings.
