This validates and manages OpenShift Grafana dashboards with a sharp focus on operator cognitive load. It enforces a three-tier information architecture where L1 shows only critical tenant-impact predictors, L2 groups platform services by domain, and L3 holds deep dives. The workflow is export, edit JSON, validate all PromQL against Thanos, then apply with sync verification. The validation scripts catch bad queries before they hit production. What makes this opinionated is the insistence that every red panel must imply an action and that noise (like certain ArgoCD alerts) gets filtered explicitly. Good for SRE teams running multi-tenant platforms where dashboard sprawl has already caused alert fatigue.
npx skills add https://github.com/boshu2/agentops --skill grafana-platform-dashboard