Multi-cloud discovery
Azure subscriptions, AWS accounts (multi-account SSO) and OpenShift connections. Pick the clusters you want and they show up grouped by provider.
AI With your Claude Code or Copilot CLI
Multi‑cloud Kubernetes, zero drama.
The desktop IDE that finds your AKS, EKS and OpenShift clusters, connects you to private ones without tunnel gymnastics and flags expiring certificates, upgrade-blocking PDBs and orphan disks. And when something doesn’t add up, it explains it with AI.
Screenshots show the app’s Spanish UI.
Free
macOS · Apple Silicon
Never touches your ~/.kube/config
14 configuration warningsDead routes, hosts with no Service…
“Explain it” in one clickClaude Code · GitHub Copilot CLI
Speaks your stack’s language
Azure AKSAmazon EKSRed Hat OpenShiftAzure Red Hat OpenShiftkubeconfigAzure Bastionkubelogin · Entra IDAWS SSOIstioArgo Rolloutscert-managerKyvernoKarpenterHelmPrometheusGrafanaClaude CodeGitHub Copilot CLI
Let’s be honest
Kubernetes isn’t hard. Everything around it is.
Twelve terminal tabs, four kubeconfigs, an SSH tunnel that dies at 3 a.m. and a certificate nobody knew was expiring. Sound familiar?
The kubeconfig hairball
An az aks get-credentials here, an aws eks update-kubeconfig there, and your ~/.kube/config now reads like an ancient scroll.
AureliaK8s discovers them for you and keeps them separate.
The private cluster
Bastion, SSH hop, kubelogin, local port… five commands to look at one pod. And again tomorrow.
Set it up once. Then one click: tunnel, credentials, done.
The 3 a.m. surprise
The upgrade hangs on a PDB, the certificate expired yesterday and you’re paying for disks nobody uses.
Checks that warn you first, with AI to point the way.
Multi-cloud
One jellyfish. Many tentacles. Each one a cluster.
That’s why it’s called Aurelia. Azure, AWS, OpenShift in your datacenter or your good old kubeconfig: all attached to the same body.
- Azure AKS: walks your subscriptions, tells public from private and assigns a Bastion profile to each.
- AWS EKS: multi-account SSO or single identity, with a managed kubeconfig.
- OpenShift / ARO: via Bastion, or through a bridge pod when the firewall won’t let you in.
- Fully isolated: discovered kubeconfigs live in their own folder. Deleting a context never deletes your cloud.
30-second tour
See what your clusters weren’t telling you.
All your Istio traffic at a glance: external and internal ingresses, gateways, domains, routes and every egress. Filter, expand and verify against Envoy.
A doctor for your service mesh: missing revisions, down control planes, routes pointing to nowhere. Every finding comes with its “how to fix it”.
Orphan volumes, certificates, PDBs and probes: checks that sweep the whole cluster and tell you only what matters.
Every AI analysis is saved as Markdown: read it again, keep the conversation going or continue it in your terminal.
Built-in AI
A second pair of eyes on every error. It never gets tired.
Not a chatbot glued to a corner. AI lives inside every view: on a failing pod, on an Istio warning, on an expired certificate. One click and it tells you what’s going on, why, and what to check next.
- Automatic context: YAML, events, logs and resource state, no copy-pasting.
- Bring your own AI: Claude Code or GitHub Copilot CLI, whichever you already have.
- Markdown history: every analysis is saved; export it or continue it in the terminal.
- Prompt catalog that updates without reinstalling the app.
Checks
Find the fire while it’s still a spark.
Read-only checks that sweep the whole cluster and show you only what matters. No endless dashboards: one list, the cause, and what to do.
- Expiring certificates in TLS Secrets and cert-manager, with warnings at 30, 15 and 7 days.
- PDBs that block a drain and would leave your AKS upgrade hanging.
- Orphan volumes: PVCs with no pod, loose PVs and how many GiB you’re paying for nothing.
- Readiness / Liveness: measure a Deployment’s real startup and compare it with your probes.
Service Mesh + Progressive Delivery
Istio and Argo Rollouts, finally readable.
Stop reading 400-line YAML to figure out where a request goes. AureliaK8s draws the path and points at where it breaks.
- Control plane: revisions, versions, skew and sidecar coverage.
- North-south and egress flow: gateways, domains, routes and egress, with “Verify in Envoy”.
- Rollouts: strategy, step, traffic weight and the Deployment they manage, with real pods.
Features
Everything you do with kubectl. And what you wish it did.
Pods, deployments, logs, Helm, terminal and events, polished. And on top, tools built for real platform work.
Private, painlessly
Built-in Bastion and SSH hop, kubelogin with Entra ID handled. No ssh or sshpass to install.
AI right where you need it
An “Explain it” button on resources, warnings and checks, with the context already gathered.
Preventive checks
Certificates, PDBs, orphan volumes and probes, in one prioritized list.
Istio and Argo Rollouts
Summary, traffic flow, analysis and a readable view of every resource. Promote or abort a rollout without memorizing the plugin.
Debug “hardened” pods
Distroless, no shell, read-only? An assistant builds the right kubectl debug for what you want to do.
And also…
- Force-delete pods stuck in Terminating, with the cause detected
- Combined logs for a Deployment with a container filter
- VS Code-style tabs and right-click everywhere
- Kyverno, Karpenter, Helm and Grafana
- Automatic updates
Pricing
Free. For real. Aura is optional.
Every feature, for everyone. If AureliaK8s saved your Friday afternoon, you can return the favor… and farm some aura while you’re at it.
Free
$0 forever
All of AureliaK8s. No fine print, no cluster limit, no “14-day trial”.
- Every feature
- Unlimited clusters · AKS, EKS, OpenShift
- Built-in AI (with your Claude Code or Copilot)
- Automatic updates
Main character energy
Supporter ✨
$5 / month
Exactly the same as the free plan. But with +1000 aura. Science can’t explain it yet, but your pods can feel it.
- Everything in Free (yes, everything)
- +1000 aura, credited instantly
- Your deploys land on the first try* (*not guaranteed, but it feels that way)
- You keep an independent project — and its dev — alive and caffeinated
Cancel anytime. You keep the aura.
Aura is non-transferable, not publicly traded and won’t fix CrashLoopBackOff. But it helps.
For companies
Does your team live in Kubernetes? Sponsor the tool that takes the drama out.
If AureliaK8s saves your team even a couple of headaches a month, sponsoring it is one of the easiest lines to justify in your platform budget. And it buys more uptime than a post-mortem meeting.
Your support pays for development time, code signing, update infrastructure and the features your team will need tomorrow.
No contracts or lock-in · The free version stays free
PlanktonFor startups and small teams that want to chip in.
+10k auraOctopusYour logo on this site. Eight arms of gratitude.
+100k auraWhaleFeatured logo and a direct line to tell us what your platform needs.
+1M auraFounding sponsors wanted. The first ones in get the best spot on this page and an aura that never expires.
FAQ
Frequently asked questions
If it’s free, what’s the catch?
There isn’t one. No data selling, no locked features. The project runs on donations and sponsorships; the Supporter plan unlocks nothing extra except +1000 aura and our eternal gratitude.
Which AI does it use? Where does my data go?
AureliaK8s doesn’t ship its own AI: it uses the CLI you choose, Claude Code or GitHub Copilot CLI, with your own account. Resource context is sent to that provider only when you ask for an analysis. If you configure none, AI features simply don’t show up.
Does it touch my ~/.kube/config?
No. Clusters discovered in Azure, AWS or OpenShift are stored in their own kubeconfigs inside the app’s private folder. Your local kubeconfig is read as-is, never modified.
What do I need installed?
Just what you already use: the az CLI (with kubelogin) for Azure and the aws CLI for EKS. The SSH hop for private clusters is built in. For AI, Claude Code or GitHub Copilot CLI.
Does it work with private clusters?
Yes. For private AKS it uses Azure Bastion with a built-in SSH hop. For OpenShift behind a firewall it can use a bridge pod inside a cluster you can reach.
Why do you need my email?
To send you an access code, so there are no passwords to remember. We store your email and an anonymous identifier for your computer (with its name, OS and app version) only to manage your access. Your cloud credentials stay on your machine, handled by your own CLIs (az, aws).
Which platforms does it run on?
Today on macOS with Apple Silicon, with automatic updates. Windows and Linux are on the roadmap.
What exactly is aura?
A scientifically imprecise unit measuring how good you feel about supporting independent software. Studies (ours) show +1000 aura reduces pre-deploy anxiety by a percentage we’d rather not make up.
Your clusters are waiting. Go see them.
Download AureliaK8s, connect your first cloud in a couple of minutes and find out what your clusters have been trying to tell you.
macOS · Apple Silicon · Windows and Linux on the roadmap