The Kubernetes IDE BFG.

• See the cluster first

  Version, node readiness, allocatable capacity against live usage, pods by phase and recent warning events in one overview.

• Browse every resource type

  Discovery supplies the tables, so CRDs appear without a plugin. Filter on live fields, choose namespaces, add columns from fields or labels, act on a selection, and save a filtered table under a name.

• Inspect and edit the live object

  Metadata, conditions, events and YAML in one drawer. Monaco completes from the cluster's OpenAPI. Apply carries the version you read, so an object that changed underneath you is refused rather than overwritten.

• Operate workloads

  Scale, restart, suspend or trigger a CronJob, cordon a node, or review a drain plan before moving its pods. Protected clusters require their name before destructive work.

• Go back a revision

  A Deployment, StatefulSet or DaemonSet lists what it rolled out and which revision is running. Diff any two revisions, then undo to one; the cluster is asked first, and a refusal greys the button out with its reason.

• Work with Helm releases

  Read releases from Helm storage, search configured repositories and OCI registries, install behind a dry run, preview an upgrade as a manifest diff, then roll back or uninstall.

• Turn findings into a work list

  128 checks cover security, reliability and waste. The view leads with the findings; scope, baselines, muted rules and export sit behind one Configure control. Issues rank live failures and clear as the cluster recovers.

• Answer for a framework

  Read the checks by control for the CIS Kubernetes Benchmark, PSS and NSA/CISA, including the controls nothing reading a live cluster can answer, and why. Export findings as CSV, JSON or SARIF.

• Find what is reserved and unused

  Every namespace and workload's requests against what it actually uses, ranked by what could be given back. Where nothing measured a workload, the view says so instead of showing a zero.

• Open a shell, logs or a forward

  Exec into a container, add an ephemeral debug container to a distroless image, stream logs, or forward a pod or service to your machine without leaving the view.

• See metrics and relationships

  Live CPU and memory sit in the tables. Prometheus supplies history, topology draws ownership and routing edges, and Cilium Hubble can supply the traffic graph.

• Keep several clusters open

  Each open cluster is a tab in one strip, keeping its own state, health and history. Name, colour and group the tabs, and choose which reopen next time. Opening a cluster says how long it will wait and lets you stop, and if one is already open it asks whether to replace that tab or keep both. Fleet views combine checks, issues, search, Helm releases and GitOps across every open cluster.

• Notice when a cluster goes quiet

  A banner names the cluster, how long it has been silent and why: timed out, connection refused, certificate not trusted. The rows you already had stay on screen, marked with when they were read, and one control reconnects.

• Compare what you authored

  Diff one object or a whole resource type across two clusters. Status, managedFields and cluster-assigned spec fields are removed before the comparison, and the result names the fields it left out.

• One cluster, shared

  Preview cluster mode serves one cluster to a team from the same binary and Helm chart. Sign in through OIDC or an authenticated proxy, with viewer, editor and admin roles. Every change lands on an audit trail admins can export.

Every capability, group by group · local and cluster modes