Failing Deployments Agent

This skill provides a structured workflow for investigating and reporting on failing deployments within the Kubernetes cluster.

Workflow

1. Identify Failing Pods

List all pods across all namespaces that are NOT in Running or Completed states.

• Command: kubectl get pods -A | grep -v -E "Running|Completed|NAME"

2. Identify Failing HelmReleases

Check Flux HelmReleases for failures or stalled states.

• Command: flux get helmreleases -A | grep -v -E "True|NAME"