kube-medic

Name: kube-medic
Availability: InStock
Author: CacheForge

Execute read-only Kubernetes triage to identify and quantify cluster health issues.

10|2|Updated Feb 14, 2026

Search Tags:#kubernetes #diagnostics #kubectl #sre #read-only #incident-response #cluster-triage

Ready to run

npx skills add https://github.com/cacheforge-ai/cacheforge-skills --skill kube-medic-cacheforge-ai

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: kube-medic
Source: https://github.com/cacheforge-ai/cacheforge-skills/tree/main/skills/kube-medic
Command: npx skills add https://github.com/cacheforge-ai/cacheforge-skills --skill kube-medic-cacheforge-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires kubectl, jq, and includes scripts (resource) components.

What problem does it solve?

Kubernetes clusters can drift into unhealthy states that are difficult to diagnose with ad-hoc kubectl commands. kube-medic provides rapid, AI-assisted triage by orchestrating read-only Kubernetes queries and correlating nodes, pods, events, and resources to deliver actionable diagnoses.

Core Features & Use Cases

Full cluster health triage: surface non-running pods, CrashLoopBackOff issues, ImagePullBackOff errors, and warning events with contextual insights.
Pod autopsy: detailed pod view including container statuses, current and previous logs, events, and image mismatches.
Deployment & resource insights: rollout status, history, replicasets, and node-resource pressure across the cluster.
Multi-cluster support: operate across contexts and namespaces to manage large environments.

Quick Start

Ask kube-medic to run a full cluster triage with optional context and namespace settings.

Frequently Asked Questions about kube-medic

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I run Kubernetes cluster triage to surface CrashLoopBackOff and ImagePullBackOff errors?▼

You can run a full cluster health triage by asking the tool to analyze your Kubernetes environment with optional context and namespace settings. It orchestrates read-only queries across nodes, pods, and events to surface critical problems and provide AI-assisted diagnostics.

Can I diagnose pod health across multiple Kubernetes contexts and namespaces?▼

Yes, multi-cluster Kubernetes diagnostics are supported by operating across different contexts and namespaces. This allows you to manage large environments and perform detailed pod autopsies, including container statuses, current and previous logs, and image mismatches.

Do I need kubectl and jq installed to perform read-only Kubernetes diagnostics?▼

Yes, kubectl and jq are required dependencies for performing read-only Kubernetes diagnostics. They gather cluster data, execute safe reads, and process JSON output for downstream AI reasoning during incident response and cluster triage.

What is the best way to analyze node resource pressure and deployment rollout status in Kubernetes?▼

Analyzing node-resource pressure and deployment rollout status requires querying replicasets, rollout history, and node utilization. An AI-assisted triage tool correlates this data to provide deployment insights and identify resource constraints across the Kubernetes cluster.

Does this Kubernetes diagnostics tool modify cluster resources during incident response?▼

No, this Kubernetes diagnostics tool performs strictly read-only operations during incident response. It gathers data on nodes, pods, and events safely without modifying cluster resources, ensuring secure triage and analysis.