Day to day you must pass -n / --context every time, or trust kubeconfig. That is easy to get wrong across local kind and remote clusters. Environments already know the intended context.
deployah use # show current
deployah use staging # environment -> context (+ ns if set)
deployah use -n payments # namespace only
deployah use --context kind-deployah
Store sticky state in Deployah (.deployah/active or ~/.config/deployah/active.yaml). Optional --kubeconfig-write if you want kubectl to follow. Flags still win.
--context / --namespace
-> DPY_* env
-> deployah active state
-> platform environment.context
-> kubeconfig
Does not replace the platform file. deployah use is sugar, not the source of truth for namespace (#9).
Related: #9, #28.
Day to day you must pass
-n/--contextevery time, or trust kubeconfig. That is easy to get wrong across local kind and remote clusters. Environments already know the intended context.Store sticky state in Deployah (
.deployah/activeor~/.config/deployah/active.yaml). Optional--kubeconfig-writeif you want kubectl to follow. Flags still win.Does not replace the platform file.
deployah useis sugar, not the source of truth for namespace (#9).Related: #9, #28.