Chat with Your Kubernetes Cluster Using AI

Stop switching between your AI assistant and terminal every time you need to check pod status, deploy applications, or troubleshoot cluster issues. kubectl-mcp-server bridges the gap by letting Claude, Cursor, Windsurf, and other AI assistants directly interact with your Kubernetes clusters through natural language.

The Problem: Context Switching Kills Productivity

You're debugging a production issue in Claude. You need to check pod logs, examine deployment status, and maybe scale a service. Currently, that means:

1. Leave your AI conversation
2. Open a terminal
3. Run kubectl commands
4. Copy-paste results back to your AI
5. Lose conversation context and momentum

kubectl-mcp-server eliminates this workflow friction entirely.

What This Actually Does

This MCP server implements 26 specialized tools that let AI assistants execute Kubernetes operations directly:

• Core Operations: List pods, deployments, services, get logs, check cluster health
• Helm Management: Install, upgrade, and uninstall charts without leaving your AI chat
• Security Auditing: RBAC validation, security context analysis, permission checks
• Diagnostics: Cluster troubleshooting, error pattern detection, resource validation
• Monitoring: Resource usage tracking, event monitoring, pod health checks

Instead of explaining what you want to do, you just ask: "Scale the nginx deployment to 5 replicas and check if the pods are healthy." The AI handles the kubectl commands and gives you structured results.

Real-World Scenarios

Production Debugging:

"The payment service is throwing 500 errors. Check the pod status, 
get recent logs, and see if there are any related events."

Deployment Validation:

"I just deployed the new user-service. Verify the deployment is healthy, 
check resource usage, and confirm the service endpoints are working."

Security Audit:

"Audit the RBAC permissions for the api-gateway service account 
and check if any pods are running with excessive privileges."

Cluster Maintenance:

"Show me which nodes are under resource pressure and identify 
any pods that might need to be rescheduled."

Installation: Two Minutes to Running

Install from PyPI:

pip install kubectl-mcp-tool

Configure your AI assistant by adding this to their MCP settings:

{
  "mcpServers": {
    "kubernetes": {
      "command": "python",
      "args": ["-m", "kubectl_mcp_tool.mcp_server"],
      "env": {
        "KUBECONFIG": "/path/to/your/.kube/config"
      }
    }
  }
}

That's it. Your AI can now execute kubectl operations directly in conversation.

Why This Matters for DevOps Teams

Faster Incident Response: Debug production issues without breaking conversation flow with your AI assistant. Get cluster insights, check logs, and identify problems in a single chat session.

Natural Language Operations: Instead of remembering kubectl syntax, describe what you want. "Check if the database pods are ready" becomes a simple request rather than constructing kubectl get pods -l app=database -o jsonpath='{.items[*].status.conditions[?(@.type=="Ready")].status}'.

Context-Aware Troubleshooting: The server maintains namespace context and can perform complex multi-step operations. Ask about a pod's health and automatically get related events, resource usage, and dependency status.

Security-First Design: Built-in RBAC validation and security auditing tools help you maintain cluster security while operating at AI speed.

Advanced Features That Set This Apart

Mock Mode: Test and demo cluster operations without needing a real Kubernetes cluster. Perfect for training or development scenarios.

Multi-Transport Support: Works with stdio, SSE, and HTTP transport methods, ensuring compatibility across different AI assistant architectures.

Intelligent Error Handling: When operations fail, you get actionable recovery suggestions, not just error messages.

Cross-Namespace Operations: Seamlessly work across different namespaces without manually switching context.

The server exposes a comprehensive toolkit that covers real-world DevOps scenarios, from basic resource management to complex security audits and performance monitoring.

Getting Started

1. Install: pip install kubectl-mcp-tool
2. Configure: Add the server to your AI assistant's MCP settings
3. Use: Start asking your AI to perform Kubernetes operations naturally

The included installation script can automatically configure Claude Desktop, Cursor, and Windsurf for you.

With 634+ GitHub stars and active development, kubectl-mcp-server is becoming the standard way to integrate AI assistants with Kubernetes workflows. Stop context-switching and start operating your clusters at the speed of thought.