What problem does it solve? Deploying AI agents to Google Cloud involves coordinating Terraform, Docker, service accounts, secrets, and CI/CD pipelines, which is error-prone when done manually. This Skill guides the full deployment workflow through the agents-cli tool, covering target selection, sizing, secrets, rollback, and troubleshooting. ## Core Features & Use Cases - Multi-target deployment: Deploy agents to Agent Runtime, Cloud Run, or GKE with a decision matrix comparing scaling, networking, session state, and cost trade-offs. - CI/CD pipeline setup: Configure staging-to-production pipelines with GitHub Actions or Cloud Build using Workload Identity Federation authentication. - Production operations: Manage secrets via Secret Manager, size deployments (CPU, memory, concurrency), bind agents to Agent Gateways, and handle rollback and common 403/timeout failures. - Use Case: After building an ADK agent, ask the assistant to deploy it to Cloud Run with secrets configured, then set up a CI/CD pipeline that promotes builds from staging to production with manual approval. ## Quick Start Deploy my agent to Cloud Run in project my-gcp-project using agents-cli with min-instances set to 1.
