Your Kubernetes workspace.
Explore, manage, compare and troubleshoot all your clusters from one fast desktop app. Production stays protected, problems surface first, and every feature is free.
In active development — the first public release is on its way. Watch the repository to hear when it ships.
Design preview with sample data. Interface may change before the first release.
Built for the engineer who has five clusters open before lunch.
See your clusters clearly
- Select several clusters and browse them as one list, with a cluster column on every row.
- Unhealthy resources sort to the top. Healthy rows stay quiet.
- Thousands of resources stay fast with virtualized tables and live watches.
Investigate problems quickly
- Open any resource in an inspector without losing your place.
- Events, logs, termination reasons and owners are one click away.
- Collect evidence into an investigation you can return to and share.
Operate production confidently
- Production is marked on every screen, never by color alone.
- Protected clusters show the impact of a change and ask you to type the name.
- Viewing and changing are always separate actions.
Pick clusters once. Every view follows.
Clustrail treats a set of clusters as your working context. Pods, Deployments, Nodes and Events all show the selection together, or grouped by cluster when you need to tell them apart.
- Clusters grouped as Production, Staging and Development.
- Namespace is a filter, so it works across clusters with different namespaces.
- A cluster that drops out stays visible with its reason and a fix, instead of silently disappearing.
Mistakes on production should be hard to make.
Each context gets a protection level. It is a safety net inside Clustrail; your cluster’s RBAC still decides what you can do.
Standard
Destructive actions ask once. Good for staging and development.
Protected
Every change and every shell shows its impact, the equivalent kubectl command, and needs the resource name typed. The default for production.
Read-only
Changes and shells are blocked in Clustrail. Logs, YAML and investigations still work.
- A hatched band and a PROD tag follow production into every tab, terminal and dialog.
- Drain shows every pod it would evict, and which ones a PodDisruptionBudget will block, before anything runs.
- When a resource is managed by Argo CD or Flux, Clustrail warns that a direct edit will be reverted.
Facts on one side. Your reasoning on the other.
When a Deployment starts failing, open an investigation. Clustrail gathers what the cluster reports: termination reasons, events, restarts, metrics and the log lines you pin. You write the hypotheses and link them to evidence.
- Every fact names its source and when it was captured.
- Clustrail never guesses a root cause, and it says when Kubernetes no longer keeps the data you’d need.
- Investigations are saved locally and export to Markdown for your incident notes.
Everything you reach for, in one window.
Logs
Merge pods into one stream, search and filter, read the previous container, and pause without losing lines.
Terminal
Shells in tabs and splits, with cluster, pod and container always in view and reconnects handled.
YAML editor
Folding, validation, side-by-side diff and server-side dry run before you apply.
Environment compare
Staging against production: images, limits, env vars and HPA, showing only what differs.
Workloads
Scale, restart and roll back Deployments with rollout history and related resources.
Nodes
Conditions, allocation and taints, plus cordon and a drain preflight that lists every pod.
Helm
Releases, values, manifests and history across clusters, with rollback when you need it.
GitOps
Argo CD and Flux status, drift and errors, wherever a controller is installed.
Your machine, your kubeconfig, your clusters.
Clustrail is a desktop app that talks straight to your API servers with the credentials you already use.
- No account. No sign-in, no email, nothing to register.
- No agents. Nothing is installed in your clusters.
- No copied credentials. Kubeconfig files are read in place and never edited.
- No telemetry. Diagnostics are built locally and only shared if you choose.
- No pushes to git. GitOps views read controller status only.
- No invented diagnoses. What you see is what the cluster reports.
Every feature is free.
No paid tier, no team plan, no locked features. Clustrail is developed in the open by mihnk, and contributions, bug reports and design feedback are welcome.
A native desktop shell around a fast Go core. Lists are virtualized and log streams use bounded buffers, so large clusters stay responsive.
Good to know
Is it really free?
Yes. Clustrail is open source under the Apache 2.0 license, and every feature is available to everyone, with no account or license key.
Do I need to install anything in my clusters?
No. Clustrail uses the Kubernetes API with your existing kubeconfig. Usage graphs appear when metrics-server is available; Helm and GitOps views appear when those tools are present.
Does protected mode replace RBAC?
No. Protection levels add confirmations and blocks inside Clustrail. What you are allowed to do is still decided by your cluster’s RBAC.
Which platforms will it support?
macOS, Windows and Linux. Builds will be published with each release on GitHub.
Which languages does the app speak?
English and Turkish. It follows your system language, and you can switch in Settings.
When can I try it?
Clustrail is in active development. Watch the GitHub repository to be notified about the first release.