ark deploy
Apply resources from a YAML file to a Kubernetes cluster in the correct dependency order.
ark deploy <team.yaml> [flags]
ark deploy is a thin wrapper around kubectl apply that ensures resources are applied in the order their dependencies require:
- ConfigMaps
- ArkSettings
- ArkMemory
- ArkAgents
- ArkTeams
- ArkServices
- ArkEvents
This prevents “not found” errors when an ArkAgent references an ArkSettings that hasn’t been created yet.
Flags
| Flag | Default | Description |
|---|---|---|
-n, --namespace | (from file metadata) | Target namespace. Overrides metadata.namespace in the YAML. |
--dry-run | false | Print the resources that would be applied without applying them. |
--kubeconfig | ~/.kube/config | Path to kubeconfig. |
Examples
# Deploy all resources in a YAML file
ark deploy content-team.yaml -n my-org
# Dry run — see what would be applied
ark deploy content-team.yaml -n my-org --dry-run
# Multi-doc YAML with all resources
ark deploy my-team.yaml
Multi-doc YAML
ark deploy supports multi-document YAML files (separated by ---). A single file can contain ConfigMaps, ArkAgents, ArkTeams, and ArkEvents — ark deploy sorts and applies them in the correct order:
apiVersion: v1
kind: ConfigMap
metadata:
name: researcher-prompt
namespace: my-org
data:
system.txt: "You are a researcher..."
---
apiVersion: arkonis.dev/v1alpha1
kind: ArkAgent
metadata:
name: researcher-agent
namespace: my-org
spec:
model: llama3.2
systemPromptRef:
configMapKeyRef:
name: researcher-prompt
key: system.txt
---
apiVersion: arkonis.dev/v1alpha1
kind: ArkTeam
metadata:
name: research-pipeline
namespace: my-org
spec:
roles:
- name: research
arkAgent: researcher-agent
pipeline:
- role: research
inputs:
prompt: ""
output: ""
See also
- GitOps guide —
ark deployin a GitOps workflow - ark trigger — trigger a team run after deploying
- ark validate — validate before deploying