What problem does it solve? Deploying Google ADK agents to production involves coordinating Terraform infrastructure, container builds, service accounts, secrets, and CI/CD pipelines across multiple GCP targets. This Skill guides the full deployment workflow so agents reach Agent Runtime, Cloud Run, or GKE with correct IAM, scaling, and rollback configuration. ## Core Features & Use Cases - Deployment Target Selection: A decision matrix compares Agent Runtime, Cloud Run, and GKE across scaling, networking, session state, and cost so you pick the right target. - Guided Deploy Workflow: Covers agents-cli deploy flags, sizing parameters (CPU, memory, workers, concurrency), Secret Manager integration, and human approval gates before deployment. - CI/CD and Operations: Reference files detail GitHub Actions or Cloud Build pipelines with WIF authentication, Terraform custom infrastructure patterns, rollback procedures, and troubleshooting for common 403/503 errors. - Use Case: You have finished an ADK chemistry agent and need to ship it. Use this Skill to choose Cloud Run, configure secrets, deploy with agents-cli deploy, and set up a staging-to-production CI/CD pipeline. ## Quick Start Ask the assistant to deploy your ADK agent to Cloud Run using agents-cli with the appropriate project, region, and secrets configuration.
