EKS Troubleshooting

EKS Troubleshooting Skill

The /eks-investigate command and eks-investigator agent run the automated investigation orchestrator:

bash skills/eks-troubleshooting/run-investigation.sh "<issue-type>" "<resource-name>" "<context>" "<namespace>"

Supported Issue Types

• deployment - Pod status, rollout status, events, restart counts, resource usage
• networking - Service details, endpoints, DNS resolution, network policies
• node - Node status, conditions, resource pressure, taints/tolerations
• ingress - Ingress configuration, backend services, certificate status

Usage

1. Run /eks-investigate command
2. Agent asks for issue type, resource name, kubectl context, and namespace
3. Agent executes the investigation script
4. Wait for results

Example Scripts

• examples/investigate-deployment.sh - Queries deployment status, events, pod logs
• examples/investigate-networking.sh - Queries service endpoints, DNS, network policies