Thank you for installing {{ .Chart.Name }} ({{ .Chart.Version }}, app {{ .Chart.AppVersion }}).

* Service:    {{ include "power-hmc-mcp.fullname" . }}.{{ .Release.Namespace }}.svc.cluster.local:{{ .Values.service.port }}
* Transport:  {{ .Values.mcp.transport }}  (path={{ .Values.mcp.http.path }})
{{- if .Values.mcp.http.probesPath }}
* Probes:     enabled at {{ .Values.mcp.http.probesPath }} (livenessProbe + readinessProbe wired in Deployment)
{{- else }}
* Probes:     disabled (Deployment falls back to TCP-socket health checks)
{{- end }}

NEXT STEPS

1. Make sure HMC credentials are present:
{{- if .Values.hmc.existingSecret }}
   - existingSecret={{ .Values.hmc.existingSecret }} must contain HMC_HOST / HMC_USERNAME / HMC_PASSWORD keys.
{{- else }}
   - The chart created Secret/{{ include "power-hmc-mcp.fullname" . }}-hmc from values.yaml. For
     anything beyond a smoke test, switch to .Values.hmc.existingSecret and manage credentials
     out-of-band (sealed-secrets, External Secrets Operator, etc.).
{{- end }}

2. Authentication / TLS termination is intentionally NOT in this chart.
   See examples/deploy/{nginx.conf,envoy.yaml} for the reverse-proxy
   templates the maintainers use; the MCP server itself binds plain HTTP
   on the cluster network and trusts the gateway in front.

3. Health and readiness:
{{- if .Values.mcp.http.probesPath }}
   - kubectl port-forward svc/{{ include "power-hmc-mcp.fullname" . }} {{ .Values.service.port }}:{{ .Values.service.port }}
   - curl http://127.0.0.1:{{ .Values.service.port }}{{ if eq .Values.mcp.http.probesPath "/" }}/healthz{{ else }}{{ trimSuffix "/" .Values.mcp.http.probesPath }}/healthz{{ end }}
{{- else }}
   - Probes disabled. Verify with `kubectl get pods -l app.kubernetes.io/name={{ include "power-hmc-mcp.name" . }}`
     and `kubectl logs` on the pod. To enable HTTP probes, set
     mcp.http.probesPath to "/" (or a prefix) and re-`helm upgrade`.
{{- end }}
