Kubernetes

Infrastructure integration · 67 node(s).

00Overview

Operate a Kubernetes cluster from a flow — roll out and restart deployments, scale workloads, read pod logs, run and trigger jobs, manage namespaces, services, ingresses, config maps and secrets, and cordon or drain nodes. Anything you would reach for kubectl to do, a flow can do on a schedule, on approval, or in response to something happening elsewhere in your stack.

Every field below is exactly what you see in the Flomation editor. Fields marked ● live picker let you choose from a list pulled live from your account — no IDs to look up.

01Connecting Kubernetes

1. Find your cluster's API Server URL — run kubectl cluster-info, or copy it from your cloud provider's console. It looks like https://your-cluster:6443.
2. Choose how to authenticate with the Authentication dropdown: Service Account Token (recommended), Client Certificate for mutual TLS, or Kubeconfig if you would rather paste an existing config file.
3. For a token, create a service account in the cluster and grant it a role covering only the resources and namespaces you want to automate — give it no more than it needs. Then mint a token with kubectl create token <serviceaccount> -n <namespace>. These tokens expire (an hour by default), so pass --duration for a longer-lived one and diarise the renewal.
4. In Flomation, add the token as an environment secret (e.g. kubernetes_token), then pick it in each node's Service Account Token field.
5. If your cluster uses a private certificate authority, paste its certificate into Cluster CA Certificate (PEM) — the CA is public information, so it is safe to paste literally. Leave it blank for a cluster with a publicly-trusted certificate. Only use Allow Insecure TLS against a throwaway self-signed cluster.
6. The Kubeconfig method takes a whole kubeconfig file, pasted as a secret; its current context is the one used. Note that the live pickers (namespaces, deployments, pods and the rest) only fill themselves when you authenticate with a Service Account Token held as an environment secret, and, on a cluster with a private certificate authority, only when the CA is pasted in as literal PEM text — a picker cannot resolve a reference. With the other methods the same fields still work; you just type the name yourself.

02Apply

Kubernetes Apply Manifest

infrastructure/kubernetes/apply_manifest · Action

Advanced: apply any Kubernetes object from a YAML or JSON manifest via Server-Side Apply — including CRDs and kinds with no dedicated action. Creates or updates in place.

Returns: id, result, kind, applied, tool_result, success, error

03Configmap

Kubernetes Create ConfigMap

infrastructure/kubernetes/configmap_create · Action

Create a ConfigMap in a namespace from a map of key/value pairs, with optional labels.

Returns: id, result, tool_result, success, error

Kubernetes Delete ConfigMap

infrastructure/kubernetes/configmap_delete · Action

Permanently delete a ConfigMap from a namespace. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get ConfigMap

infrastructure/kubernetes/configmap_get · Action

Fetch a single ConfigMap by name, including its data keys and values.

Returns: id, result, tool_result, success, error

Kubernetes List ConfigMaps

infrastructure/kubernetes/configmap_list · Action

List the ConfigMaps in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Update ConfigMap

infrastructure/kubernetes/configmap_update · Action

Update a ConfigMap's data — merge the supplied keys into the existing data, or replace the whole data map.

Returns: id, result, tool_result, success, error

04Cronjob

Kubernetes Create CronJob

infrastructure/kubernetes/cronjob_create · Action

Create a CronJob that runs a container on a cron schedule — set the image, an optional command, and the concurrency policy.

Returns: id, result, tool_result, success, error

Kubernetes Delete CronJob

infrastructure/kubernetes/cronjob_delete · Action

Permanently delete a CronJob, and by default the Jobs and pods it owns. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get CronJob

infrastructure/kubernetes/cronjob_get · Action

Fetch a single CronJob by name — its schedule, suspend state, job template, and last/next run times.

Returns: id, result, tool_result, success, error

Kubernetes List CronJobs

infrastructure/kubernetes/cronjob_list · Action

List the CronJobs in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Resume CronJob

infrastructure/kubernetes/cronjob_resume · Action

Un-pause a suspended CronJob so it resumes scheduling Jobs. Missed runs are not backfilled.

Returns: id, result, tool_result, success, error

Kubernetes Suspend CronJob

infrastructure/kubernetes/cronjob_suspend · Action

Pause a CronJob so it stops scheduling new Jobs. Jobs already running are left alone.

Returns: id, result, tool_result, success, error

Kubernetes Trigger CronJob

infrastructure/kubernetes/cronjob_trigger · Action

Run a CronJob immediately, off-schedule, by creating a one-off Job from its template — the same as kubectl create job --from=cronjob.

Returns: id, result, tool_result, success, error

05Daemonset

Kubernetes Delete DaemonSet

infrastructure/kubernetes/daemonset_delete · Action

Permanently delete a DaemonSet, stopping its pod on every node. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get DaemonSet

infrastructure/kubernetes/daemonset_get · Action

Fetch a single DaemonSet by name, including its rollout status and the number of nodes it is scheduled and ready on.

Returns: id, result, tool_result, success, error

Kubernetes List DaemonSets

infrastructure/kubernetes/daemonset_list · Action

List the DaemonSets in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Restart DaemonSet

infrastructure/kubernetes/daemonset_restart · Action

Trigger a rolling restart of a DaemonSet — replaces its pod on each node in turn, honouring the update strategy.

Returns: id, result, restarted_at, tool_result, success, error

06Deployment

Kubernetes Delete Deployment

infrastructure/kubernetes/deployment_delete · Action

Permanently delete a Deployment and the pods it manages. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Deployment

infrastructure/kubernetes/deployment_get · Action

Fetch a single Deployment by name — its spec, replica counts, image, and rollout status.

Returns: id, result, tool_result, success, error

Kubernetes List Deployments

infrastructure/kubernetes/deployment_list · Action

List the Deployments in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Restart Deployment

infrastructure/kubernetes/deployment_restart · Action

Trigger a rolling restart of a Deployment — replaces its pods one batch at a time, with no downtime.

Returns: id, result, restarted_at, tool_result, success, error

Kubernetes Deployment Rollout Status

infrastructure/kubernetes/deployment_rollout_status · Action

Check whether a Deployment has finished rolling out — optionally waiting until it completes, fails, or times out.

Returns: id, result, status, complete, tool_result, success, error

Kubernetes Scale Deployment

infrastructure/kubernetes/deployment_scale · Action

Set the number of replicas a Deployment runs — scale up for load, or down to zero to pause it.

Returns: id, result, replicas, tool_result, success, error

07Event

Kubernetes List Events

infrastructure/kubernetes/event_list · Action

List cluster Events in a namespace or across all of them — the diagnostic for why a pod will not start, an image will not pull, or a volume will not mount.

Returns: results, count, continue_token, has_more, tool_result, success, error

08Hpa

Kubernetes Delete Horizontal Pod Autoscaler

infrastructure/kubernetes/hpa_delete · Action

Delete a HorizontalPodAutoscaler. Autoscaling stops; the workload is left at its current replica count. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Horizontal Pod Autoscaler

infrastructure/kubernetes/hpa_get · Action

Read one HorizontalPodAutoscaler — its target, its current and desired replica counts, and the live metrics driving it. Needs a Kubernetes 1.23+ cluster (autoscaling/v2).

Returns: id, result, tool_result, success, error

Kubernetes List Horizontal Pod Autoscalers

infrastructure/kubernetes/hpa_list · Action

List the HorizontalPodAutoscalers in a namespace, or across every namespace, optionally filtered by label or field. Needs a Kubernetes 1.23+ cluster (autoscaling/v2).

Returns: results, count, continue_token, has_more, tool_result, success, error

09Ingress

Kubernetes Delete Ingress

infrastructure/kubernetes/ingress_delete · Action

Permanently delete an Ingress, withdrawing the external route it defined. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Ingress

infrastructure/kubernetes/ingress_get · Action

Retrieve one Ingress by name — its host rules, path backends, TLS blocks, and assigned load-balancer address.

Returns: id, result, tool_result, success, error

Kubernetes List Ingresses

infrastructure/kubernetes/ingress_list · Action

List the Ingresses in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

10Job

Kubernetes Create Job

infrastructure/kubernetes/job_create · Action

Create a Job that runs a container image to completion, with retry, TTL cleanup, and an optional command and environment.

Returns: id, result, tool_result, success, error

Kubernetes Delete Job

infrastructure/kubernetes/job_delete · Action

Permanently delete a Job. Deleting with Background (default) or Foreground removes its pods too; Orphan leaves them running. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Job

infrastructure/kubernetes/job_get · Action

Fetch a single Job by name, including its status — how many pods have succeeded, failed, or are still running.

Returns: id, result, tool_result, success, error

Kubernetes List Jobs

infrastructure/kubernetes/job_list · Action

List the Jobs in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

11Namespace

Kubernetes Create Namespace

infrastructure/kubernetes/namespace_create · Action

Create a namespace, optionally with labels and annotations.

Returns: id, result, tool_result, success, error

Kubernetes Delete Namespace

infrastructure/kubernetes/namespace_delete · Action

Permanently delete a namespace and every resource inside it. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Namespace

infrastructure/kubernetes/namespace_get · Action

Fetch a single namespace by name, including its labels, annotations and phase.

Returns: id, result, tool_result, success, error

Kubernetes List Namespaces

infrastructure/kubernetes/namespace_list · Action

List the namespaces in the cluster, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

12Node

Kubernetes Cordon Node

infrastructure/kubernetes/node_cordon · Action

Mark a Node unschedulable so no new pods land on it. Running pods are untouched; reverse it with Uncordon Node.

Returns: id, result, tool_result, success, error

Kubernetes Drain Node

infrastructure/kubernetes/node_drain · Action

Cordon a Node and evict its pods so they reschedule elsewhere, ready for maintenance. Requires explicit confirmation.

Returns: id, evicted, skipped, evicted_pods, tool_result, success, error

Kubernetes Get Node

infrastructure/kubernetes/node_get · Action

Fetch a single Node by name, including its status, capacity, conditions and labels.

Returns: id, result, tool_result, success, error

Kubernetes List Nodes

infrastructure/kubernetes/node_list · Action

List the Nodes in the cluster, optionally filtered by label or field selector.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Uncordon Node

infrastructure/kubernetes/node_uncordon · Action

Mark a Node schedulable again so the scheduler can place pods on it. Reverses Cordon Node.

Returns: id, result, tool_result, success, error

13Pod

Kubernetes Delete Pod

infrastructure/kubernetes/pod_delete · Action

Delete a Pod. If it is managed by a controller the controller recreates it — the usual way to restart one stuck Pod. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Pod

infrastructure/kubernetes/pod_get · Action

Fetch a single Pod by name, including its status, conditions and per-container state.

Returns: id, result, tool_result, success, error

Kubernetes List Pods

infrastructure/kubernetes/pod_list · Action

List the Pods in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Get Pod Logs

infrastructure/kubernetes/pod_logs · Action

Read the logs of a container in a pod — the last N lines, a recent time window, or the previous crashed instance.

Returns: id, logs, truncated, tool_result, success, error

14Pvc

Kubernetes Delete PVC

infrastructure/kubernetes/pvc_delete · Action

Permanently delete a PersistentVolumeClaim. If the StorageClass reclaim policy is Delete (the common default), the underlying volume and its data are destroyed too. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get PVC

infrastructure/kubernetes/pvc_get · Action

Fetch a single PersistentVolumeClaim by name — its phase, capacity, StorageClass, and bound volume.

Returns: id, result, tool_result, success, error

Kubernetes List PVCs

infrastructure/kubernetes/pvc_list · Action

List the PersistentVolumeClaims in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

15Secret

Kubernetes Create Secret

infrastructure/kubernetes/secret_create · Action

Create a Secret from plain key/value pairs — Kubernetes base64-encodes the values for you.

Returns: id, result, tool_result, success, error

Kubernetes Delete Secret

infrastructure/kubernetes/secret_delete · Action

Permanently delete a Secret. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Secret

infrastructure/kubernetes/secret_get · Action

Read a single Secret. Optionally base64-decode its values into a separate plaintext output.

Returns: id, result, decoded_data, tool_result, success, error

Kubernetes List Secrets

infrastructure/kubernetes/secret_list · Action

List the Secrets in a namespace, or across every namespace, optionally filtered by label or field. Note: each entry includes the secret's base64 data.

Returns: results, count, continue_token, has_more, tool_result, success, error

16Service

Kubernetes Create Service

infrastructure/kubernetes/service_create · Action

Create a Service to expose a set of pods on a stable IP and port — ClusterIP, NodePort, LoadBalancer, or ExternalName.

Returns: id, result, tool_result, success, error

Kubernetes Delete Service

infrastructure/kubernetes/service_delete · Action

Permanently delete a Service, removing its stable IP and DNS name. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Service

infrastructure/kubernetes/service_get · Action

Retrieve one Service by name — its type, selector, ports, and any assigned cluster or external IP.

Returns: id, result, tool_result, success, error

Kubernetes List Services

infrastructure/kubernetes/service_list · Action

List the Services in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

17Serviceaccount

Kubernetes Create Service Account

infrastructure/kubernetes/serviceaccount_create · Action

Create a ServiceAccount in a namespace, with optional labels and annotations. Note: since Kubernetes 1.24 this no longer auto-creates a token Secret — use kubectl create token for a short-lived one.

Returns: id, result, tool_result, success, error

Kubernetes Delete Service Account

infrastructure/kubernetes/serviceaccount_delete · Action

Permanently delete a ServiceAccount, invalidating every token issued for it. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get Service Account

infrastructure/kubernetes/serviceaccount_get · Action

Fetch a single ServiceAccount by name — its labels, annotations, imagePullSecrets, and referenced secrets.

Returns: id, result, tool_result, success, error

Kubernetes List Service Accounts

infrastructure/kubernetes/serviceaccount_list · Action

List the ServiceAccounts in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

18Statefulset

Kubernetes Delete StatefulSet

infrastructure/kubernetes/statefulset_delete · Action

Permanently delete a StatefulSet. Its PersistentVolumeClaims are NOT removed — delete them separately to reclaim storage. Requires explicit confirmation.

Returns: id, result, tool_result, success, error

Kubernetes Get StatefulSet

infrastructure/kubernetes/statefulset_get · Action

Fetch a single StatefulSet by name and return its full definition and status.

Returns: id, result, tool_result, success, error

Kubernetes List StatefulSets

infrastructure/kubernetes/statefulset_list · Action

List the StatefulSets in a namespace, or across every namespace, optionally filtered by label or field.

Returns: results, count, continue_token, has_more, tool_result, success, error

Kubernetes Restart StatefulSet

infrastructure/kubernetes/statefulset_restart · Action

Trigger a rolling restart of a StatefulSet — recreates its pods one ordinal at a time, reattaching the same volumes.

Returns: id, result, restarted_at, tool_result, success, error

Kubernetes Scale StatefulSet

infrastructure/kubernetes/statefulset_scale · Action

Set the number of replicas for a StatefulSet — scale it up, down, or to zero.

Returns: id, result, replicas, tool_result, success, error

19Notes & Limitations

Behaviours and constraints worth knowing before you build with these nodes.

• Destructive actions (deletes, node drain) fail closed and will not run unless their confirmation checkbox is ticked, and because it is a normal boolean field you can bind it to a variable or an upstream node's output to gate the deletion at run time rather than deciding it when you build the flow.
• Kubernetes has no trigger node — the cluster cannot call Flomation, so flows must be started by a schedule or another trigger and then poll with the List, Get or Rollout Status actions.
• List Secrets returns each Secret's base64-encoded data in its output, so treat that step's result as sensitive and avoid routing it into logs, chat messages or other downstream systems; the live Secret picker only ever shows names.
• Deleting a StatefulSet deliberately leaves its PersistentVolumeClaims in place, while deleting a PVC destroys the underlying volume and its data whenever the StorageClass reclaim policy is Delete (the common default) — plan storage cleanup as its own, irreversible step.
• The live resource pickers only work on a connection that authenticates with a Service Account Token held as an environment secret, and if your cluster uses a private certificate authority its certificate must be pasted in as literal PEM text (a picker cannot resolve a reference). Every action still works with the other authentication methods — you just type the namespace and resource names by hand.
• The HorizontalPodAutoscaler actions use the autoscaling/v2 API and therefore require a Kubernetes 1.23 or newer cluster; on older clusters they will fail rather than fall back.