Kure Monitor

Kure Monitor is an open-source, real-time Kubernetes health agent designed to instantly detect cluster failures—such as pod crashes, image pull errors, and stuck pending states—and use artificial intelligence to explain exactly why they occurred [1, 2]. Rather than just alerting you that something broke, it analyzes context directly from your pod events, logs, and manifests to generate plain-language troubleshooting guidance and concrete fixes using your choice of LLM, including fully local, air-gapped models like Ollama for high-security environments [2-4]. Beyond AI-powered root-cause diagnosis, Kure provides a unified web dashboard that features an interactive live topology graph of your workloads, allows you to safely test fixes by deploying isolated "mirror pods," and continuously runs over 50 proactive security misconfiguration checks [3-5]. Ultimately, it acts as a dedicated failure diagnosis and security posture assistant that perfectly complements, rather than replaces, your existing general observability platforms like Prometheus, Grafana, or Datadog [6, 7].