Features
This page is a workflow map, not just a feature dump. Use it to understand what Krust replaces in a normal operating session and which capabilities matter for day-to-day debugging.
Core Highlights
Lightning fast
Rust core with zero-copy FFI. Sub-millisecond filtering across thousands of resources.
Full-featured
Core Kubernetes resources, Gateway API, CRDs, terminal exec, logs, Helm, YAML, metrics, Prometheus-backed pod charts, and multi-cluster work — all in one window. Pro adds focused incident-time and local networking workflows plus an Experimental managed-GKE preview for invited Google test accounts; it does not remove the everyday kubeconfig baseline.
Privacy first
Cluster connections remain direct from your device, and Krust does not collect in-app usage analytics. Optional native crash diagnostics on macOS are off by default and require explicit consent; Krust disables Sentry tracing, profiling, session tracking, breadcrumbs, and failed-request capture.
Native macOS
Built with SwiftUI and AppKit. Feels like a first-party Apple app with full keyboard navigation and system integration.
How the Features Group Together in Practice
Most users do not adopt Krust for one isolated feature. They adopt it because the evidence-to-action loop stays in one place: inspect resource state, events, logs, metrics, and topology; ask the Agent to investigate; approve a safe action; then verify the result without breaking context.
Feature List
Connection Center & Kubeconfig Free
Use the contexts already in your kubeconfig or import additional kubeconfig files. Select one or multiple contexts, browse their resources through native Kubernetes APIs, and keep the same local workflow whether the cluster is GKE, EKS, AKS, k3s, kind, or another conforming Kubernetes distribution.
Managed Google Cloud & GKE Pro · Experimental
Invited Google test accounts can sign in, enter a project ID, discover its GKE clusters, select an available DNS, public, or internal control-plane endpoint, then add the connection to Krust. Google OAuth is currently in Testing and authorization may expire after seven days. Existing and imported kubeconfig contexts remain Free. Read the Google Cloud and GKE setup guide →
Resource Management Free
Browse and operate on core workloads, networking, storage, RBAC, Gateway API resources, CRDs, and Custom Resources with real-time watchers. Supports delete, scale, restart, suspend, and evict operations. Inspector sidebar shows full resource details. Context menus and bulk operations reduce the need to drop back to CLI for routine actions.
Terminal & Exec Free
Exec into any pod with automatic shell detection (bash/ash/sh fallback). Local shell with kubectl context auto-injected. Multi-tab support with full TTY emulation. This matters when the issue is faster to confirm from inside the container than from resource status alone.
Log Viewer Free
100K line buffer with fast single-pod search, deep history search, log level filtering, regex search, bookmarks, export, and container picker for multi-container pods. This is the baseline logging workflow for day-to-day inspection.
Multi-Pod Log Aggregation Pro
Stream logs from all pods in a workload simultaneously. Pod name prefix for easy identification. Pod picker to toggle individual streams. This is where deployment-level debugging becomes materially faster than repeating `kubectl logs` across replicas.
Structured Log Parsing Free
JSON compact mode collapses structured logs into a readable summary with field discovery/toggles. This now stays in the free baseline log workflow.
Log Inspector Free
Click any log line to inspect parsed fields and raw content in the inspector panel. This stays free as part of the baseline single-pod debugging workflow.
Log Export & Bookmarks Free
Export logs to file for offline analysis. Bookmark important lines for quick reference during debugging sessions. These stay free because they support the basic single-pod workflow.
Helm Free
List all Helm releases across namespaces. View release history, rollback to any revision, and inspect values YAML. Reads release data directly from Kubernetes secrets. Useful when release drift and rollback validation matter more than raw chart authoring.
YAML Editor Free
Standalone editor windows with syntax highlighting. Apply changes directly to the cluster. Revision diff shows exactly what changed between versions. This keeps resource edits reviewable instead of opaque. Learn more →
Cross-Cluster Resource Diff Pro
Compare resources across different contexts/clusters side-by-side with unified diff. Same-context resource diff remains free; cross-context incident comparison is gated.
AI Diagnostic Agent Current
Chat sidebar activated with Cmd+Shift+A. The agent uses tools to inspect, list, get, edit resources and view logs. Human-in-the-loop approval for write operations. Persistent memory across sessions. Bring your own API key (BYOK). This is currently available without an enforced Pro gate.
Resource Topology Free
Visualize resource relationships as an interactive ArgoCD-style horizontal tree. Starting from any Deployment, StatefulSet, DaemonSet, Service, Pod, Job, or CronJob, Krust discovers related resources (ReplicaSets, Pods, Services, Endpoints, ConfigMaps, Secrets, ServiceAccounts, PVCs) and renders them as a draggable graph with expandable nodes. Learn more →
Metrics Free
CPU and memory usage per pod and workload. Table columns poll metrics-server every 30 seconds. Optional Prometheus configuration adds lightweight CPU and memory charts inside the pod inspector through the Kubernetes API Service proxy, without requiring a manual port-forward. Graceful degradation keeps resource browsing usable when either backend is unavailable. Learn more →
Security Audit & Trivy Free
Audit workload security context, image hygiene, resources, node health, RBAC, and service accounts from live Kubernetes state. Optional locally installed Trivy adds cluster and image vulnerability evidence without being confused with Krust's built-in rules. Learn more →
Stable Service URLs for Port Forwarding Free baseline
Krust can keep Kubernetes service-style hostnames usable on macOS during port-forward workflows. Scoped Packet Tunnel DNS handles svc.cluster.local and cluster.local lookups, while public DNS stays on the normal macOS resolver path. Free includes 1 active DNS mapping; Pro unlocks multiple concurrent mappings for multi-service local workflows. Learn more →
Multi-Cluster Free
Toggle multiple contexts on simultaneously. Side-by-side views with per-context watchers. Synchronized namespace filtering across clusters. This is most valuable when comparing staging vs production or checking rollout drift across regions.
Revision Diff Free
Deployment and StatefulSet revision diff in rollback/history flows is free, because it is part of the production safety baseline.
Navigation Free
Command palette (Cmd+K) for instant access to any resource or action. k9s-style single-key shortcuts (S shell, L logs, E edit, ? help). Full keyboard navigation, events timeline, macOS notifications, and automatic kubeconfig file watching for live context updates.
What This Page Is Not Trying to Answer
- For overall tool selection, use Kubernetes dashboard alternative.
- For Lens-specific tradeoffs, use Krust vs Lens.
- For raw performance numbers, use Performance.