hexr deploy is the final step in the Hexr deployment workflow. It reads the manifests generated by hexr build, prompts you to select a target Kubernetes cluster, applies them in dependency order — namespace first, then RBAC, then ConfigMaps, then the agent pod — and waits until all containers are running before printing the agent’s SPIFFE identity and A2A endpoint. After hexr deploy returns, your agent is live with mTLS, process-level identity, and (if enabled) an A2A discovery endpoint.
Usage
Path to the build directory containing generated manifests. Defaults to
.hexr in the current working directory.Interactive flow
What gets applied
The deploy command applies manifests in this order to satisfy Kubernetes dependency requirements:namespace.yaml— creates thetenant-{name}namespacerbac.yaml— ServiceAccount, Role, and RoleBindingprocess-contexts/*.json— per-process SPIFFE context as ConfigMapsagent-card.yaml— Agent Card ConfigMap for A2A discovery (if A2A is enabled)agent-pod.yaml— the agent Pod with all four containers and the init container
Cloud deploy
For Hexr Cloud,--cloud deploys directly to the managed GKE cluster:
hexr deploy --cloud requires an active Hexr Cloud session. Run hexr login --key <api_key> first if you haven’t already.Verify deployment
After deploying, usehexr status to confirm the agent is running: