MinimalKuberneteslayoutwereuseacrossmostclusters

We standardised on a compact set of manifests per service so any engineer can add a workload in under an hour and a new joiner can navigate the repo without a guided tour.

Uniformity across clients is intentional: on-call playbooks and upgrade paths stay similar, which matters more than per-cluster novelty.

Additional complexity is added only when observability or traffic data shows the baseline is insufficient.