Mock API online Simulated data

KubePilot Demo API

A simulated production cluster: 8 nodes, about 60 pods across 10 namespaces, live metrics, real-looking logs, and incidents that rotate every 30 minutes. Use it to explore CoPilot, Pilot and AutoPilot before connecting your own cluster.

Simulated cluster

kp-demo-prod-eu-west-2
Nodes ready
–
Pods running
–
Failing pods
–
CPU
–
Memory
–
Next incident
–

Active incidents

invented
Loading…

Recent events

refreshes every 10s
  • —Loading…

What the simulation does

Always on: one checkout-api replica crash-loops after a bad config change, with kubelet back-off timing and restart counts that keep climbing.

Rotating every 30 minutes: an OOMKilled indexer, a rollout stuck on ImagePullBackOff, a node under disk pressure, or a worker node going NotReady.

Live: node CPU and memory drift over time, logs stream new lines, pods roll weekly, and resolved incidents stay in RCA history. Every replica of this API shows the same cluster at the same moment.

Your own cluster

Connect KubePilot to your Kubernetes clusters

The demo can't see your cluster, and your cluster's data never passes through demo.kubepilot.org. KubePilot runs next to your cluster, and the app connects to it directly.

01

Run KubePilot against your cluster

Install the Helm chart into the cluster, or run the binary with your kubeconfig.

git clone https://github.com/bwalia/kubepilot
helm upgrade --install kubepilot \
  kubepilot/charts/kubepilot \
  -n kubepilot --create-namespace
02

Make it reachable — securely

Use a port-forward to try it, or an Ingress with TLS and auth for your team. Keep it off the public internet without auth.

kubectl -n kubepilot port-forward \
  svc/kubepilot 8080:8080
03

Point the app at your server

In the KubePilot app, choose Add server (not Try demo), enter your KubePilot URL and credentials, and tap Connect. Add more clusters any time.

https://kubepilot.example.com
This demoYour own KubePilot
DataInvented — the same simulated cluster for everyoneYour real nodes, pods, events and logs
FixesPreviewed only; every action returns “skipped”Applied through policy gates, dry-run, and CR-code approval
AI analysisPre-written root-cause reportsGenerated by your LLM (e.g. Ollama), so data stays in your network
AccessShared demo login, read-onlyYour own auth, RBAC and network policy
Try the mock API

Use it from the app or from a terminal

In the iOS app, tap Try demo on the Welcome screen and the URL and login below are filled in for you. You can also call the API directly — it's the same /api/v1 surface a real KubePilot server exposes.

Quick start

Shared demo login (read-only, synthetic data): apple / review

curl -u apple:review \
  https://demo.kubepilot.org/api/v1/troubleshooting/summary
curl -u apple:review \
  "https://demo.kubepilot.org/api/v1/clusters/pods?namespace=payments"

Endpoints

Basic auth
  • GET
    /clusters/pods pods
  • GET
    /clusters/nodes nodes
  • GET
    /clusters/deployments
  • GET
    /clusters/services
  • GET
    /events timeline
  • GET
    /troubleshooting/summary
  • GET
    /clusters/pods/{ns}/{pod}/logs
  • GET
    /resource/{kind}/{ns}/{name}/yaml
  • GET
    /ai/troubleshoot/{ns}/{pod}
  • GET
    /rca reports
  • GET
    /anomalies
  • GET
    /autopilot decisions

All paths are under /api/v1. Writes are accepted but never change anything. Public: /healthz, /demo/status.json.