Kubernetes
The Kubernetes MCP server connects AI agents to a Kubernetes cluster via kubectl and Helm. Agents can get, describe, create, apply, patch, scale, and delete resources, read pod logs, manage contexts and rollouts, port-forward, run arbitrary kubectl commands, install and upgrade Helm charts, and clean up failing pods.
Tools
cleanup_pods
Clean up pods that are stuck in error or crash states.
explain_resource
Explain the schema and fields of a Kubernetes resource type.
install_helm_chart
Install a Helm chart with a custom values configuration.
kubectl_apply
Apply a YAML manifest to the connected Kubernetes cluster.
kubectl_context
Manage and switch between the available kubectl contexts.
kubectl_create
Create a new Kubernetes resource within the cluster.
kubectl_delete
Delete an existing Kubernetes resource from the cluster.
kubectl_describe
Describe a Kubernetes cluster resource in full detail.
kubectl_generic
Execute an arbitrary kubectl command in the cluster.
kubectl_get
Get or list Kubernetes resources within the cluster.
kubectl_logs
Retrieve the logs from a Kubernetes pod container.
kubectl_patch
Update one or more fields of a Kubernetes resource.
kubectl_rollout
Manage the rollout status of a Kubernetes deployment.
kubectl_scale
Scale a Kubernetes deployment or replica set up or down.
list_api_resources
List all of the API resources available in the cluster.
port_forward
Port-forward from localhost to a pod or a service.
uninstall_helm_chart
Uninstall an existing Helm chart release from the cluster.
upgrade_helm_chart
Upgrade an existing Helm chart release in the cluster.