Canary (no metrics stack yet)

The platform does not ship kube-prometheus-stack, ServiceMonitors, or AnalysisTemplates. Prod Rollouts still shift traffic 25→50→75 with a 60s pause. Staging is a single replica and jumps to 100%.

When you add monitoring later, wire /metrics → scrape → an AnalysisTemplate and attach it under strategy.canary.analysis. Until then, do not add those objects — they would point at a Prometheus that is not there.