incident-management/incident-lifecycle overview

Flagship detection-to-postmortem incident lifecycle with severity-routed policy gates, an adversarial postmortem-completeness gate, and kip incident memory. Single owner for incident handling across the library: detection + severity classification -> incident-commander assignment -> parallel mitigation strands (iterative diagnosis, blast-radius assessment, comms drafting) -> policy-gated production mitigation -> policy-gated status-page/customer comms on a severity-driven cadence -> recovery verification (one bounded re-mitigation loop) -> blameless postmortem with adversarial completeness gate -> policy-gated postmortem publication -> action-item tracking + kip assert at close.