The MCP Kubernetes Server provides a bridge between Large Language Models and Kubernetes clusters through a set of specialized tools. Built with Python using the FastMCP framework, it wraps kubectl commands to enable operations like retrieving pod information, managing deployments, switching contexts, and troubleshooting cluster issues. The server requires a configured kubectl environment and can be integrated with Claude Desktop for seamless natural language interactions with Kubernetes resources. It's particularly valuable for DevOps workflows that benefit from conversational interfaces to Kubernetes, allowing users to create, scale, and monitor resources without remembering exact kubectl syntax.
No reviews yet. Be the first to review!
Sign in to join the conversation
Highlight AICreate a new deployment with specified name, image, namespace, and number of replicas.
Update the specified deployment to a new version in the given namespace.
Scale the specified deployment to a desired number of replicas in the specified namespace.
Retrieve all pods in the specified namespace.
Retrieve all namespaces in the Kubernetes cluster.
Retrieve all nodes in the Kubernetes cluster.
Retrieve all services in the Kubernetes cluster.
Retrieve all deployments in the Kubernetes cluster.
Retrieve all jobs in the Kubernetes cluster.
Retrieve all cronjobs in the Kubernetes cluster.
Retrieve all statefulsets in the Kubernetes cluster.
Retrieve all daemonsets in the Kubernetes cluster.
Retrieve the current context of the Kubernetes cluster.
List all contexts available in the Kubernetes cluster.
Switch to the specified context in the Kubernetes cluster.
Retrieve the logs of the specified pod in the given namespace.
Retrieve all events in the specified namespace.
Add an annotation to the specified pod in the given namespace.
Remove an annotation from the specified pod in the given namespace.
Add a label to the specified pod in the given namespace.
Remove a label from the specified pod in the given namespace.
Expose the specified deployment in the given namespace on a specified port.
Port-forward to the specified resource (pod, deployment, or service) in the given namespace to a local port.
Delete the specified resource (pod, deployment, service, job, cronjob, statefulset, daemonset) in the given namespace.
