feat(helm): Add observability support with ServiceMonitor and Grafana dashboard

Add comprehensive observability configuration to Helm chart:

**Helm Values:**
- Add observability configuration section for metrics, tracing, and logging
- Add serviceMonitor configuration (disabled by default)
- Add prometheusRule configuration (disabled by default)

**Templates:**
- Update deployment to include observability environment variables
- Update deployment to expose metrics port (9090)
- Update service to expose metrics port
- Add ServiceMonitor template for Prometheus Operator
- Add PrometheusRule template with critical and warning alerts

**Dashboards:**
- Add comprehensive Grafana dashboard JSON with 6 panels:
  - Request Rate (by method and endpoint)
  - Error Rate (5xx errors percentage)
  - Request Latency (P50/P95 by endpoint)
  - Top MCP Tools (by invocation volume)
  - Nextcloud API Latency (by app)
  - Vector Sync Queue Size
- Add dashboard README with import instructions

**Alert Rules:**
- Critical: Server down, high error rate (>5%), high latency (>1s), dependency down
- Warning: Token validation errors (>1%), vector sync queue high (>100), Qdrant slow (>500ms)

All features are opt-in via values.yaml configuration.

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

Co-Authored-By: Claude <noreply@anthropic.com>

charts/nextcloud-mcp-server/templates/service.yaml

@@ -15,5 +15,11 @@ spec:
       targetPort: http
       protocol: TCP
       name: http
+    {{- if .Values.observability.metrics.enabled }}
+    - port: {{ .Values.observability.metrics.port }}
+      targetPort: metrics
+      protocol: TCP
+      name: metrics
+    {{- end }}
   selector:
     {{- include "nextcloud-mcp-server.selectorLabels" . | nindent 4 }}