k13d¶
██╗ ██╗ ██╗██████╗ ██████╗ ██║ ██╔╝███║╚════██╗██╔══██╗ █████╔╝ ╚██║ █████╔╝██║ ██║ ██╔═██╗ ██║ ╚═══██╗██║ ██║ ██║ ██╗ ██║██████╔╝██████╔╝ ╚═╝ ╚═╝ ╚═╝╚═════╝ ╚═════╝
k**ube**a**i**d**ashboard = **k + 13 letters + d = k13d
What is k13d?¶
k13d is a comprehensive Kubernetes management tool that combines:
k9s-style TUI - Fast terminal dashboard with Vim keybindings
kubectl-ai Intelligence - Agentic AI that actually executes kubectl commands
Modern Web UI - Browser-based dashboard with real-time streaming
It bridges the gap between traditional cluster management and natural language AI, helping you manage, debug, and understand your Kubernetes cluster with unprecedented ease.
-
Quick Start
Get up and running in minutes
-
TUI Dashboard
k9s-style terminal interface with Vim bindings
-
AI Assistant
Natural language Kubernetes management
-
Enterprise Security
RBAC, audit logging, access requests
Key Features¶
TUI Dashboard¶
| Feature | Description |
|---|---|
| k9s Parity | Vim-style navigation (h/j/k/l), quick switching (:pods, :svc) |
| Deep Resource Support | Pods, Deployments, Services, Nodes, Events, ConfigMaps, Secrets, RBAC... |
| Interactive Operations | Scale, Restart, Port-Forward, Delete with confirmation |
| Real-time Updates | Live resource watching with instant refresh |
| Smart Autocomplete | Dropdown suggestions with custom alias support |
| Plugin System | Extend TUI with external tools via plugins.yaml |
| Model Switching | Switch AI profiles on the fly with :model command |
| Configurable Aliases | Custom resource shortcuts via aliases.yaml |
Web Dashboard¶
| Feature | Description |
|---|---|
| Modern Interface | Responsive design with resizable panels |
| SSE Streaming Chat | Real-time AI responses with live cursor |
| Pod Terminal | Interactive xterm.js shell in browser |
| Log Viewer | Real-time logs with ANSI color support |
| Topology Tree | Hierarchical resource ownership visualization |
| Applications | App-centric view by app.kubernetes.io/name labels |
| Validate | Cross-resource validation with severity levels |
| Healing | Auto-remediation rules with event history |
| Helm Manager | Release management, rollback, uninstall |
| Metrics Dashboard | Cluster health cards with CPU/Memory bars |
| 5 Color Themes | Tokyo Night, Production, Staging, Dev, Light |
| Trivy Scanner | CVE vulnerability scanning with auto-download |
| Multi-Cluster | Context switcher for multiple kubeconfig clusters |
| RBAC Viewer | Visual subject→role relationship viewer |
| Net Policy Map | Network policy ingress/egress visualization |
| Event Timeline | Events grouped by time with warning stats |
| GitOps | ArgoCD/Flux application sync status |
| Templates | One-click deploy for common K8s patterns |
| Backups (Velero) | Backup and schedule management |
| Resource Diff | Side-by-side YAML diff |
| Notifications | Slack/Discord/Teams webhook alerts |
| AI Troubleshoot | One-click AI cluster diagnosis |
| kubectl Plugin | Install as kubectl k13d |
Agentic AI Assistant¶
┌─────────────────────────────────────────────────────────────┐
│ You: Show me pods with high memory usage in production │
├─────────────────────────────────────────────────────────────┤
│ AI: I'll check the pods in the production namespace. │
│ │
│ 🔧 Executing: kubectl top pods -n production --sort-by=mem │
│ │
│ Here are the top memory consumers: │
│ NAME CPU MEMORY │
│ api-server-7d4f8b... 250m 1.2Gi ⚠️ High │
│ worker-processor-... 100m 890Mi │
└─────────────────────────────────────────────────────────────┘
| Feature | Description |
|---|---|
| Tool Execution | AI directly runs kubectl/bash commands (not just suggests) |
| MCP Integration | Extensible tools via Model Context Protocol |
| Safety First | Dangerous commands require explicit approval |
| Deep Context | AI receives YAML + Events + Logs for analysis |
Quick Install¶
Current support status
k13d currently recommends local single-binary usage for both the TUI and Web UI. Docker, Docker Compose, Kubernetes, Helm, and other in-cluster deployment paths are still Beta / in preparation and are not officially supported yet.
Download the matching asset from Release v1.0.1. You can also browse every available asset on the full GitHub Releases page.
- macOS Apple Silicon:
k13d_v1.0.1_darwin_arm64.tar.gz - macOS Intel:
k13d_v1.0.1_darwin_amd64.tar.gz - Linux amd64:
k13d_v1.0.1_linux_amd64.tar.gz - Linux arm64:
k13d_v1.0.1_linux_arm64.tar.gz - Windows amd64:
k13d_v1.0.1_windows_amd64.zip
Supported LLM Providers¶
| Provider | Tool Calling | Best For |
|---|---|---|
| Upstage Solar |  | Recommended - Best balance of quality & cost |
| OpenAI | | Production use, best tool support |
| Anthropic | | Claude models, strong reasoning |
| Google Gemini | | Multimodal, fast responses |
| Azure OpenAI | | Enterprise deployments |
| AWS Bedrock | | AWS-hosted models |
| Ollama | | Recommended local/private inference with a tools-capable model such as gpt-oss:20b |
License¶
k13d is released under the MIT License.
Built with 
for the Kubernetes Community