k8s/monitoring: add prometheus grafana loki observability

- Loki log aggregation (MinIO backed, 10-day retention)
- Promtail daemonset (pod + talos journal logs)
- Prometheus + kube-state-metrics
- Grafana dashboards (6-row template per service)

k8s/monitoring/dashboards/services/forgejo.descriptor.yaml

@@ -0,0 +1,39 @@
+service: forgejo
+display_name: "Forgejo"
+namespace: forgejo
+service_type: web-app
+folder: "Forgejo"
+jobs: [forgejo]
+metrics:
+  prefix: forgejo
+  up_selector: 'job="forgejo"'
+  rate:
+    metric: forgejo_http_request_total
+    labels: [method, status]
+  duration:
+    metric: forgejo_http_request_duration_seconds
+    type: histogram
+    unit: s
+  error:
+    metric: forgejo_http_request_total
+    filter: 'status=~"5.."'
+  domain:
+    - metric: forgejo_repositories_total
+      kind: gauge
+    - metric: forgejo_users_total
+      kind: gauge
+    - metric: forgejo_git_operations_total
+      kind: counter
+      labels: [operation_type]
+    - metric: forgejo_runner_tasks_total
+      kind: counter
+      labels: [status]
+correlation_ids: []
+logs:
+  loki_namespace_selector: 'namespace="forgejo"'
+alerts:
+  - name: HighErrorRate
+    expr_template: rate_error_ratio
+    threshold: 0.05
+    for: 10m
+    severity: warning