Search + Incident Memory

[orchide]

As a user investigating a recurring failure, I want to search previous incident diagnoses by pod, namespace, or root cause, So that I can reuse past resolutions and reduce mean time to recovery.