Skip to content

Deployment

Deploy your agent to a development environment or production with a CI/CD pipeline.

Prototype to Production

Infrastructure vs Deployment

Infrastructure (agents-cli infra) provisions the cloud resources your agent needs — service accounts, IAM bindings, APIs, telemetry buckets, and Terraform state. It sets the stage but doesn't run your agent.

Deployment (agents-cli deploy) takes your agent code and puts it on the provisioned infrastructure — building a container, pushing it to a registry, and starting the service.

The typical flow: provision infrastructure first, then deploy on top of it.

Deploy to a Dev Environment

The simplest path to a running deployment:

1. Set your dev project:

gcloud config set project YOUR_DEV_PROJECT_ID

2. Deploy the agent:

agents-cli deploy

The command reads your deployment_target from pyproject.toml and dispatches to the right flow:

deployment_target What happens
agent_runtime Agent Runtime deployment (fully managed)
cloud_run gcloud beta run deploy (container on Cloud Run)
gke Terraform + Docker build + kubectl apply

The deployment target is set when you create your project:

agents-cli create my-agent -d cloud_run    # or agent_runtime, gke

To change the deployment target for an existing project, use scaffold enhance:

agents-cli scaffold enhance -d cloud_run

Run agents-cli scaffold enhance --help to see all available options.

Tip

To enable observability features (prompt-response logging, content logs), run agents-cli infra single-project after deploying. Terraform provisions the telemetry resources and updates your service to use them. See the Observability Guide for details.

Verify it works:

agents-cli deploy --list    # List deployments
agents-cli deploy --status  # Check deployment status

Deployment Targets

Agent Runtime

Selected with agents-cli create my-agent -d agent_runtime or deployment_target = "agent_runtime" in pyproject.toml.

Fully managed — no containers or infrastructure to manage:

agents-cli deploy --project my-gcp-project --region us-east1

Check on an async deployment:

agents-cli deploy --no-wait     # Start and return immediately
agents-cli deploy --status      # Check progress later

Cloud Run

Selected with agents-cli create my-agent -d cloud_run or deployment_target = "cloud_run" in pyproject.toml.

Builds a container from source and deploys as a Cloud Run service:

agents-cli deploy --project my-gcp-project --region us-east1

Override resource limits:

agents-cli deploy --memory 8Gi --port 8080

Deploy a pre-built image instead of building from source:

agents-cli deploy --image gcr.io/my-project/my-agent:v1

Tip

If you need more advanced Cloud Run deployment features not exposed via agents-cli flags, use --dry-run (or -n) to print the full gcloud command. You can then copy it and add additional arguments as needed.

GKE

Selected with agents-cli create my-agent -d gke or deployment_target = "gke" in pyproject.toml.

Deploys to a GKE cluster using Terraform and kubectl:

agents-cli deploy --cluster-name my-cluster --project my-gcp-project

Next Steps