feat(grafana): wire Prometheus datasource to main Grafana instance (#24)

## Summary

The main Grafana CR has been login-functional but had no datasources, so
the home dashboard is empty. Add a \`GrafanaDatasource\` pointing at the
kube-prometheus-stack Prometheus, selected onto the main instance via
the matching \`dashboards: grafana\` label (the \`status-grafana\`
instance already has its own analogous datasource).

Loki / Tempo / Alertmanager / Pyroscope are not installed — those
datasources are out of scope here and can be added once the backends
are.

## Test plan

- [x] After merge: \`kubectl -n grafana get grafanadatasource
grafana-prometheus\` is Synced
- [x] After merge: in Grafana UI → Connections → Data sources,
"Prometheus" appears as the default
- [x] After merge: a panel with \`up\` query shows live data

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Author: xnoto

workloads/grafana/datasource.yaml

@@ -0,0 +1,23 @@
+---
+# Prometheus datasource for the main Grafana instance, sourced from
+# kube-prometheus-stack's in-cluster Prometheus. Selected onto the main
+# Grafana via the matching `dashboards: grafana` label.
+apiVersion: grafana.integreatly.org/v1beta1
+kind: GrafanaDatasource
+metadata:
+  name: grafana-prometheus
+  namespace: grafana
+  annotations:
+    argocd.argoproj.io/sync-wave: "1"
+spec:
+  instanceSelector:
+    matchLabels:
+      dashboards: "grafana"
+  datasource:
+    name: Prometheus
+    type: prometheus
+    access: proxy
+    isDefault: true
+    url: http://kube-prometheus-stack-prometheus.monitoring.svc:9090
+    jsonData:
+      timeInterval: 30s

workloads/grafana/kustomization.yaml

@@ -3,6 +3,7 @@ apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 resources:
   - grafana.yaml
+  - datasource.yaml
   - status-grafana.yaml
   - status-monitoring.yaml
   - status-datasource.yaml