Skip to content
Kure Monitor

Quick Start

Five minutes from zero to a working dashboard.

1. Install the chart

Section titled “1. Install the chart”
Terminal window
helm repo add kure-monitor https://nan0c0de.github.io/kure-monitor/
helm repo update


helm install kure-monitor kure-monitor/kure \
  --namespace kure-system \
  --create-namespace \
  --set postgresql.password="$(openssl rand -hex 24)"

2. Open the dashboard

Section titled “2. Open the dashboard”
Terminal window
kubectl port-forward svc/kure-monitor-frontend 8080:8080 -n kure-system

Visit http://localhost:8080. On the very first visit, the dashboard prompts you to create the initial admin account.

3. Configure an LLM provider

Section titled “3. Configure an LLM provider”

In the Admin panel:

  1. AI Configuration → pick a provider
  2. Enter your API key
  3. Pick a model
  4. Test Connection → Save

For air-gapped clusters, point at an in-cluster Ollama and Kure data never leaves your network. See LLM Providers.

4. Watch a failure roll in

Section titled “4. Watch a failure roll in”

In another terminal:

Terminal window
# trigger a CrashLoopBackOff
kubectl run crash-test --image=busybox --restart=Never -- sh -c 'exit 1'

Within a few seconds the dashboard shows the failure with an AI-generated explanation, the pod manifest, recent events, and last container logs. Click Test Fix to deploy a mirror pod with the AI-suggested fix applied.

5. Trigger a security scan

Section titled “5. Trigger a security scan”

Open the Security tab. The scanner audits all pods on a schedule and on demand. Click any finding for severity, affected resource, and an AI-generated remediation.

What’s next

Section titled “What’s next”