A Model Context Protocol (MCP) server for managing a Kubernetes cluster through llm clients like Claude, Ollama.
Kai MCP server provides a bridge between large language models (LLMs) clients and your Kubernetes clusters, enabling users to interact with their Kubernetes resources through natural language. The server exposes a comprehensive set of tools for managing clusters, namespaces, pods, deployments, services, and other Kubernetes resources.
- Pods - Create, list, get, delete, and stream logs
- Deployments - Create, list, describe, and update deployments
- Services - Create, get, list, and delete services
- Cluster Management - Connect, list, switch, and monitor clusters
- Namespaces - Create, list, update, and delete namespaces
- Ingress - HTTP/HTTPS routing and TLS configuration
- ConfigMaps & Secrets - Configuration and secret management
- Jobs & CronJobs - Batch workload orchestration
- Nodes - Node monitoring, cordoning, and draining
- Utilities - Port forwarding, events, and API exploration
- Persistent Volumes - Storage management and claims
- RBAC - Role-based access control
- Custom Resources - CRD and custom resource operations
The server will by default connect to your current context. Make sure you have:
Access to a Kubernetes cluster configured for kubectl (e.g. minikube, Rancher Desktop, EKS, GKE, etc.)
To install the Kubernetes MCP server, run:
go install github.com/basebandit/kai/cmd/kaicode ~/Library/Application\ Support/Claude/claude_desktop_config.json
Add the server to your Claude for Desktop configuration by editing claude_desktop_config.json:
{
"mcpServers": {
"kubernetes": {
"command": "/path/to/kai binary"
}
}
}