Features

Everything you need to manage Kubernetes clusters — built into a single, fast, native desktop application.

Cluster Management

Multi-cluster, multi-context

Connect to any kubeconfig context with pre-flight reachability and auth validation. Monitor connection health with latency measurement and automatic reconnection with exponential backoff.

  • Multiple clusters connected simultaneously
  • Health monitoring every 10 seconds with latency tracking
  • Auto-reconnect with exponential backoff (up to 5 min)
  • Kubeconfig file watcher — auto-reload on file changes
  • Platform auto-detection: EKS, GKE, AKS, K3s, OpenShift
  • Star favorite clusters for quick access
Connected Clusters
● production-us-east
12ms
● staging-eu-west
45ms
● dev-local
210ms
NAME READY STATUS CPU
api-server-7f8b9d2/2Running14m
worker-crash-loop0/1CrashLoop0m
redis-master-01/1Running45m
webapp-pending0/1Pending-
Showing 4 of 1,247 pods · Virtual scrolling enabled
Resource Browsing

27+ resource types with live updates

All resources update in real-time via the Kubernetes watch API. Virtual scrolling handles thousands of rows. Sort, filter, customize columns, and export to CSV.

  • Real-time watch events — no polling needed
  • Virtual scrolling for 5,000+ rows without jank
  • Column sort, search/filter, show/hide, drag-to-reorder
  • CSV export and batch multi-select operations
  • Custom Resource Definitions auto-discovered
AI Integration

Debug pods with AI

One click assembles complete pod context — status YAML, events, logs, and automated pre-analysis — and launches your preferred AI CLI in an interactive terminal session.

  • Supports Claude Code, Gemini CLI, and ChatGPT Codex
  • Automatic sensitive data redaction before AI context
  • Runs locally — your data never leaves your machine
  • Full interactive PTY session with AI tool
$ AI Context gathered for inventory-worker-crash
✔ Pod status YAML (secrets redacted)
✔ Last 200 log lines
✔ 8 related events
✔ Pre-analysis: CrashLoopBackOff, exit code 137
claude >
The pod is being OOM-killed. The container's memory limit is set to 128Mi but the Java process requires at least 256Mi for the heap allocation. I recommend:
1. Increase memory limit to 512Mi
2. Set -Xmx256m JVM flag
▊
shell: api-server shell: redis-cache + New Tab
root@api-server-7f8b9d:/# ls -la /app
total 24
drwxr-xr-x 3 root root 4096 Mar 10 main.go
drwxr-xr-x 5 root root 4096 Mar 10 pkg/
-rwxr-xr-x 1 root root 8192 Mar 10 server
root@api-server-7f8b9d:/# ▊
Terminal & Logs

Shell into any container

Full PTY exec with multiple tabs, ANSI color support, in-terminal search, and configurable themes. Multiplex sessions across containers.

  • Multiple simultaneous terminal tabs with rename
  • Themes: dark, light, monokai, solarized
  • In-terminal search (Ctrl+F) and copy-on-select
  • Session persistence across view switches
Package Management

Helm, built in

Full Helm lifecycle management powered by the Go SDK — no CLI dependency required. Install, upgrade, rollback, and uninstall releases with a visual interface.

  • Install, upgrade, rollback, uninstall releases
  • Revision history with rollback to any version
  • Chart repository management with search
  • OCI chart references supported
  • Values diff view when upgrading
Helm Releases
● nginx-ingress
rev 12
● cert-manager
rev 5
● prometheus-stack
rev 8 (upgrading)
Pod Security Scan Results
CRITICAL Privileged mode enabled on container "debug"
CRITICAL hostPID: true — shares host PID namespace
WARNING runAsNonRoot not set
INFO readOnlyRootFilesystem not enabled
Security Level: PRIVILEGED
Security

Security scanning & audit trail

Scan pods against K8s Pod Security Standards. Visualize RBAC graphs. Every sensitive action is logged to a persistent audit trail.

  • Pod Security Standards scanning with remediation advice
  • RBAC graph visualization: subject → role → binding
  • Secrets masked by default, per-key reveal with audit logging
  • Network policy visualization (ingress/egress rules)
  • File-backed audit trail with 90-day auto-pruning

And much more

Port Forwarding

Centralized dashboard for active port forwards. Auto-reconnect on pod death. One-click "Open in Browser".

Topology View

Tree visualization of workload ownership: Deployment → ReplicaSet → Pods. Filter by name, namespace, or kind.

GitOps Detection

Auto-detects ArgoCD and Flux installations. Surfaces GitOps CRD resources in a dedicated view.

Deployment Actions

Scale, rolling restart, pause/resume rollouts, rollback to specific revisions, cordon/drain nodes.

Creation Wizards

Step-by-step wizards for Deployments, Services, ConfigMaps, and Secrets. YAML preview before applying.

YAML Templates

Built-in templates for common patterns. Variable interpolation with validation. Save and reuse custom templates.

Metrics Dashboard

CPU/memory metrics from metrics-server. Sparklines in pod detail views. Degrades gracefully without metrics.

Command Palette

Cmd+K to fuzzy-search everything: navigation, resources, actions, cluster switching, namespace filtering.

Persistent Settings

Window position, sidebar width, theme, active route — everything restores exactly as you left it between sessions.