Skillget

Kubeast

mcp
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 8 GitHub stars
Code Pass
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

Open-source multi-cluster Kubernetes console with an AI assistant that works within each user's permissions. Alert investigation, secret masking, audit log, MCP server, one-line Helm install.

README.md

Kubeast

English | 한국어

A multi-cluster Kubernetes operations platform with a built-in AI assistant. In the web UI you can view and manage resources across multiple clusters (Workloads, Network, Storage, RBAC, GPU, Helm, etc.), and use LLM-based chat to view, diagnose and change resources.

Asking the floating AI assistant why the pod on screen keeps crashing

Project site (feature clips): https://junginho0901.github.io/Kubeast/

Key features:

  • AI assistant with multiple LLMs (OpenAI, Anthropic, Gemini, Azure OpenAI, Ollama) that operates clusters through tool calling
  • When an alert from Alertmanager or Grafana arrives, the AI investigates it with read tools and leaves a report
  • Multi-cluster management, RBAC per user and per cluster, and temporary access requests with approval
  • Helm release management (rollback, upgrade, uninstall, test)
  • GPU and DRA monitoring, Prometheus metrics, log search (Loki), topology visualization, Node Shell
  • Audit log (hash chain) for every write and sensitive operation, and terminal session recording
  • Passwords, keys and tokens in what the AI's tools read are masked before they reach the model
  • MCP server for external AI agents
  • One-line install from public container images on ghcr.io, for Kubernetes or Docker
  • Korean and English support

Quick start

helm install kubeast oci://ghcr.io/junginho0901/charts/kubeast --version 0.7.0 \
  --namespace kubeast --create-namespace

The install script, Docker Compose, uninstalling, finding the admin password, connecting a cluster and turning on AI are covered in Installation and first setup.

Documentation

Document Contents
Installation and first setup 3 install methods, uninstall, first access, connecting a cluster, turning on AI
Features at a glance AI assistant, multi-cluster with roles and permissions, resource management, Helm, GPU, observability, authentication and audit
AI and alerts MCP server, alert → AI investigation, Alerts inbox, reading traces and alerts, AI analysis evaluation
Security and audit API keys, access review, dormant account lockout, audit sinks, audit log integrity, session recording
Operations features Policy reports, cluster hygiene check, Argo CD guard, dashboard quick actions, log files, log search, trace links, trusted CAs
Cluster requirements What the clusters you manage must have (RBAC, metrics-server, Prometheus, OpenCost, etc.) and how to check
SSO (OIDC) setup Connecting Keycloak, Google, Entra ID or Okta
Architecture Services, container images, resource usage
Configuration (values.yaml) Chart values explained
Database: TLS, backup and restore DB connection TLS, backup and restore, retention
Console metrics Kubeast's own /metrics
Supply chain Image scanning, SBOM, provenance attestation, signing
Audit log plan Audit action catalog and permission mapping
Development Directory layout, kind dev loop, E2E, per-service development, tech stack

License

MIT License

Reviews (0)

No results found