update

deployments/master/monitoring/grafana/dashboards/neuroforge-master-subagents.json

@@ -9,7 +9,7 @@
   ],
   "timezone": "browser",
   "schemaVersion": 41,
-  "version": 1,
+  "version": 2,
   "refresh": "15s",
   "time": {
     "from": "now-6h",
@@ -1443,6 +1443,45 @@
           "sort": "desc"
         }
       }
+    },
+    {
+      "id": 26,
+      "title": "Durable Job Payload Bytes",
+      "type": "timeseries",
+      "datasource": {
+        "type": "prometheus",
+        "uid": "prometheus"
+      },
+      "gridPos": {
+        "x": 0,
+        "y": 56,
+        "w": 24,
+        "h": 8
+      },
+      "fieldConfig": {
+        "defaults": {
+          "unit": "bytes"
+        },
+        "overrides": []
+      },
+      "options": {
+        "legend": {
+          "displayMode": "list",
+          "placement": "bottom",
+          "showLegend": true
+        },
+        "tooltip": {
+          "mode": "multi",
+          "sort": "none"
+        }
+      },
+      "targets": [
+        {
+          "expr": "neuroforge_job_payload_bytes",
+          "legendFormat": "{{state}}",
+          "refId": "A"
+        }
+      ]
     }
   ]
 }

deployments/master/monitoring/grafana/provisioning/dashboards/neuroforge.yml

@@ -0,0 +1,12 @@
+apiVersion: 1
+
+providers:
+  - name: NeuroForge
+    orgId: 1
+    folder: NeuroForge
+    type: file
+    disableDeletion: true
+    updateIntervalSeconds: 30
+    allowUiUpdates: false
+    options:
+      path: /var/lib/grafana/dashboards

deployments/master/monitoring/grafana/provisioning/datasources/prometheus.yml

@@ -1,4 +1,5 @@
 apiVersion: 1
+
 datasources:
   - name: Prometheus
     uid: prometheus

deployments/master/monitoring/prometheus/alerts.yml

@@ -56,3 +56,19 @@ groups:
         annotations:
           summary: "Knowledge graph remains highly isolated"
           description: "More than 80% of memories are still isolated after 30 minutes."
+
+      - alert: NeuroForgePendingPayloadHigh
+        expr: neuroforge_job_payload_bytes{state="pending"} > 104857600
+        for: 5m
+        labels: {severity: warning}
+        annotations:
+          summary: "NeuroForge pending job payload memory is high"
+          description: "Pending durable job payload/result bytes exceed 100 MiB for 5 minutes; backfill should remain below the v1.6.2 safety budget."
+
+      - alert: NeuroForgeTerminalPayloadHigh
+        expr: neuroforge_job_payload_bytes{state="terminal"} > 67108864
+        for: 10m
+        labels: {severity: warning}
+        annotations:
+          summary: "NeuroForge terminal job payload retention is high"
+          description: "Terminal job payload/result bytes exceed 64 MiB. Completed vector.relink jobs should be compacted automatically in v1.6.2."