K8scortex MCP server

K8sCortex — Kubernetes MCP Server

A production-grade Model Context Protocol (MCP) server that gives developers, AI agents, and automation pipelines a single, secure, natural-language interface to Kubernetes — across any cloud, any cluster, any team.

Features

• Curated multi-domain toolset across workload, deployment, configuration, observability, jobs, networking, GitOps, cluster admin, and generic read paths
• Multi-cluster routing — target any registered cluster with a single cluster parameter
• Entra ID OIDC authentication with per-tool RBAC (5-tier role hierarchy)
• Structured audit logging — every tool call recorded with caller identity, arguments, and outcome
• Dual transport — SSE/HTTP for programmatic clients + stdio for Claude Desktop
• Azure Key Vault integration for dynamic credential management with 5-minute TTL cache
• Dry-run safety — destructive tools default to simulation mode
• Generic read coverage — list/get support for resources outside curated write paths to avoid troubleshooting dead-ends

Install via npx (Fastest)

npx k8scortex-mcp

Or install globally:

npm install -g k8scortex-mcp

Quick Start

Prerequisites

• Node.js ≥ 18
• A Kubernetes cluster (minikube, Rancher Desktop, or AKS)
• kubectl configured with a valid context

Install & Run

# Clone
git clone https://github.com/apatilgtn/k8scortex-mcp.git
cd k8scortex-mcp

# Install
npm install

# Build
npm run build

# Run (local dev mode — auth bypassed)
DISABLE_AUTH=true PORT=3001 npm run dev

Connect with MCP Inspector

npx @modelcontextprotocol/inspector sse http://localhost:3001/mcp

Connect with Claude Desktop

Add to ~/Library/Application Support/Claude/claudedesktopconfig.json:

{
  "mcpServers": {
    "k8scortex": {
      "command": "node",
      "args": ["<path-to>/k8scortex-mcp/dist/stdio.js"],
      "env": {
        "DISABLE_AUTH": "true",
        "KUBECONFIG": "~/.kube/config"
      }
    }
  }
}

Restart Claude Desktop. Ask: "List all pods in the default namespace".

Tools

K8sCortex intentionally keeps writes curated and governed, while allowing flexible read access for diagnostics.

Shortened. The full README is on GitHub.