MCP Kubernetes Server

Example Use Cases

• 1

  Create, update, scale, or delete deployments and other workloads (pods, jobs, cronjobs, statefulsets, daemonsets) in a Kubernetes cluster using natural language.

• 2

  Inspect cluster state and resources such as pods, services, nodes, namespaces, contexts, logs and events without directly invoking kubectl.

• 3

  Modify resource metadata (labels, annotations), perform port-forwarding or expose services, and manage context switching across clusters.

Description

This MCP server exposes common kubectl operations as MCP tools, allowing language models to create, update, scale, inspect, and delete Kubernetes resources. Functions are annotated with @mcp.tool(), enabling type-safe, documented interactions between LLMs and the cluster. It simplifies cluster management for conversational agents while preserving context and offering basic security guidance (requires kubectl-configured access).