fix(grafana): move nl2sql.json into provisioning folder and fix dashboard mount path

docker-compose.prom.yml

@@ -1,5 +1,3 @@
-version: "3.8"
-
 services:
   prometheus:
     image: prom/prometheus:v2.55.0
@@ -10,16 +8,16 @@ services:
       - ./prometheus/prometheus.yml:/etc/prometheus/prometheus.yml:ro
       - ./prometheus/rules.yml:/etc/prometheus/rules.yml:ro
     ports:
-      - "9090:9090"          # Prometheus UI
+      - "9090:9090"                  # Prometheus UI
     networks:
-      - observability
+      - observability               # attach to custom network
     restart: unless-stopped
 
   grafana:
     image: grafana/grafana:latest
     container_name: nl2sql-grafana
     ports:
-      - "3000:3000"          # Grafana UI
+      - "3000:3000"
     depends_on:
       - prometheus
     environment:
@@ -27,15 +25,16 @@ services:
       - GF_SECURITY_ADMIN_PASSWORD=admin
       - GF_USERS_ALLOW_SIGN_UP=false
       - GF_AUTH_ANONYMOUS_ENABLED=false
-      # Optional hardening:
+      # Optional: set home dashboard to our JSON inside the mounted folder
       - GF_DASHBOARDS_DEFAULT_HOME_DASHBOARD_PATH=/etc/grafana/provisioning/dashboards/nl2sql.json
     volumes:
-      # Provisioning (datasource + dashboard providers)
+      # Provisioning folders (read-only)
       - ./grafana/provisioning/datasources:/etc/grafana/provisioning/datasources:ro
       - ./grafana/provisioning/dashboards:/etc/grafana/provisioning/dashboards:ro
-
-      # The actual dashboard JSON (mounted into the same dashboards dir)
-      - ./prometheus/grafana_dashboard.json:/etc/grafana/provisioning/dashboards/nl2sql.json:ro
     networks:
       - observability
     restart: unless-stopped
+
+networks:
+  observability:
+    driver: bridge                   # define the custom network

grafana/provisioning/dashboards/nl2sql.json

@@ -0,0 +1,68 @@
+{
+  "title": "NL2SQL Copilot - Observability",
+  "editable": true,
+  "panels": [
+    {
+      "type": "timeseries",
+      "title": "Stage p95 Latency (ms)",
+      "targets": [
+        {
+          "expr": "nl2sql:stage_p95_ms",
+          "legendFormat": "{{stage}}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "milliseconds",
+          "decimals": 0
+        }
+      },
+      "id": 1
+    },
+    {
+      "type": "timeseries",
+      "title": "Pipeline Success Ratio",
+      "targets": [
+        {
+          "expr": "nl2sql:pipeline_success_ratio",
+          "legendFormat": "success ratio",
+          "refId": "B"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "min": 0,
+          "max": 1,
+          "decimals": 2
+        }
+      },
+      "id": 2
+    },
+    {
+      "type": "timeseries",
+      "title": "Safety & Verifier Events",
+      "targets": [
+        {
+          "expr": "rate(safety_blocks_total[5m])",
+          "legendFormat": "safety blocks/min",
+          "refId": "C"
+        },
+        {
+          "expr": "rate(verifier_failures_total[5m])",
+          "legendFormat": "verifier failures/min",
+          "refId": "D"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "min": 0
+        }
+      },
+      "id": 3
+    }
+  ],
+  "schemaVersion": 38,
+  "version": 1,
+  "refresh": "30s"
+}