Skip to main content
GET
/
api
/
v1
/
aiinsights
List AI Insights
curl --request GET \
  --url https://api.example.com/api/v1/aiinsights
{
  "apiVersion": "v1",
  "kind": "AIInsightList",
  "metadata": {
    "totalCount": 2,
    "page": 1,
    "pageSize": 20
  },
  "items": [
    {
      "name": "insight-inc-20260319-001",
      "namespace": "production",
      "issueRef": "INC-20260319-001",
      "provider": "anthropic",
      "model": "claude-sonnet-4-20250514",
      "analysis": "The payment-service deployment is experiencing OOMKill events. Root cause: memory leak in the webhook handler that fails to close HTTP response bodies on validation errors, causing heap accumulation. The leak correlates with a recent deployment (revision 42) that introduced the payment notification webhook handler.",
      "confidence": 0.91,
      "recommendations": [
        "Increase memory limit from 512Mi to 1Gi as immediate mitigation",
        "Restart pods to reclaim leaked memory",
        "Review webhook handler code for unclosed response bodies"
      ],
      "suggestedActions": [
        {
          "name": "Increase memory limit",
          "action": "AdjustResources",
          "description": "Increase memory limit to prevent further OOMKills while the code fix is deployed",
          "params": {
            "memoryLimit": "1Gi",
            "deployment": "payment-service",
            "namespace": "production"
          }
        },
        {
          "name": "Restart deployment",
          "action": "RestartDeployment",
          "description": "Restart pods to reclaim leaked memory immediately",
          "params": {
            "deployment": "payment-service",
            "namespace": "production"
          }
        }
      ],
      "logAnalysis": "1 Go panic stack trace detected; 12 critical error patterns (OOMKilled, memory allocation failure); Primary exception: runtime: out of memory",
      "gitOpsContext": "Helm release 'payment' chart=payment-service version=2.3.1 status=deployed revision=42",
      "cascadeAnalysis": "Cascade chain: production/payment-service(root_cause) -> production/api-gateway(victim). The api-gateway is receiving 503 responses from payment-service.",
      "blastRadiusPrediction": "AdjustResources: Low risk. Affects 3 pods, 1 service (payment-svc). No PDB violations.",
      "generatedAt": "2026-03-19T14:32:00Z",
      "creationTimestamp": "2026-03-19T14:31:00Z"
    }
  ]
}
​
namespace
string
Filter by Kubernetes namespace
​
issue
string
Filter by incident name (e.g., INC-20260319-001)
​
page
integer
default:"1"
Page number for pagination
​
pageSize
integer
default:"20"
Items per page (max 100)

​
What Are AI Insights?

Every time the AIOps platform detects an incident, an AIInsight resource is created containing the AI’s complete analysis. This includes:
  • Root cause analysis with confidence score (0-100%)
  • Recommendations in human-readable format
  • Suggested actions with concrete parameters (e.g., “ScaleDeployment to 5 replicas”)
  • Enrichment data: log analysis, Prometheus metrics, GitOps context, cascade analysis, and blast radius prediction
This endpoint is essential for understanding how the AI reasoned about each incident, especially when incidents are escalated to human operators. 
{
  "apiVersion": "v1",
  "kind": "AIInsightList",
  "metadata": {
    "totalCount": 2,
    "page": 1,
    "pageSize": 20
  },
  "items": [
    {
      "name": "insight-inc-20260319-001",
      "namespace": "production",
      "issueRef": "INC-20260319-001",
      "provider": "anthropic",
      "model": "claude-sonnet-4-20250514",
      "analysis": "The payment-service deployment is experiencing OOMKill events. Root cause: memory leak in the webhook handler that fails to close HTTP response bodies on validation errors, causing heap accumulation. The leak correlates with a recent deployment (revision 42) that introduced the payment notification webhook handler.",
      "confidence": 0.91,
      "recommendations": [
        "Increase memory limit from 512Mi to 1Gi as immediate mitigation",
        "Restart pods to reclaim leaked memory",
        "Review webhook handler code for unclosed response bodies"
      ],
      "suggestedActions": [
        {
          "name": "Increase memory limit",
          "action": "AdjustResources",
          "description": "Increase memory limit to prevent further OOMKills while the code fix is deployed",
          "params": {
            "memoryLimit": "1Gi",
            "deployment": "payment-service",
            "namespace": "production"
          }
        },
        {
          "name": "Restart deployment",
          "action": "RestartDeployment",
          "description": "Restart pods to reclaim leaked memory immediately",
          "params": {
            "deployment": "payment-service",
            "namespace": "production"
          }
        }
      ],
      "logAnalysis": "1 Go panic stack trace detected; 12 critical error patterns (OOMKilled, memory allocation failure); Primary exception: runtime: out of memory",
      "gitOpsContext": "Helm release 'payment' chart=payment-service version=2.3.1 status=deployed revision=42",
      "cascadeAnalysis": "Cascade chain: production/payment-service(root_cause) -> production/api-gateway(victim). The api-gateway is receiving 503 responses from payment-service.",
      "blastRadiusPrediction": "AdjustResources: Low risk. Affects 3 pods, 1 service (payment-svc). No PDB violations.",
      "generatedAt": "2026-03-19T14:32:00Z",
      "creationTimestamp": "2026-03-19T14:31:00Z"
    }
  ]
}