agents-cli

Learn to build multi-agent systems with agents-cli and ADK's SequentialAgent, ParallelAgent, and LoopAgent. Discover deterministic workflows and seamless state passing for powerful agent compositions.

Safely upgrade your existing projects with agents-cli scaffold upgrade. This command merges template updates with your code, preserving customizations.

Troubleshoot common agents-cli deployment failures on Cloud Run and GKE. Learn to fix Dockerfile issues, ignore-file conflicts, environment variable conflicts, and IAM permission gaps quickly.

Synthesize multi-turn evaluation scenarios for your ADK agent using agents-cli eval dataset synthesize. Generate test conversations and grade them with agents-cli eval grade for comprehensive testing.

Learn to implement RAG with agents-cli by scaffolding a project and using adk-samples. Get a complete guide to Retrieval-Augmented Generation with this workflow.

Master secrets and environment variables for agents-cli deployments. Learn a layered approach using .env files, Terraform, CI/CD, and runtime helpers for secure and efficient management across development and production.

Optimize agent prompts using the GEPA framework with agents-cli eval optimize. Iteratively rewrite your agent's system instruction against evaluation metrics for better performance.

Learn how to create and register custom evaluation metrics for agents cli. Define them in eval_config.yaml and run them with flags for seamless integration.

Implement observability in agents-cli with Cloud Trace and logging. Automatically wire OpenTelemetry and send logs to GCS/BigQuery using environment variables.

Easily set up CI/CD pipelines with agents-cli infra cicd. Provision GitHub repos, Terraform backends, and build runners like GitHub Actions or Cloud Build in one go.

Configure deployment targets like agent runtime, Cloud Run, or GKE in agents-cli-manifest.yaml. Learn how to set the deployment_target field for your agent deployments.

Discover how lazy command loading in agents-cli defers heavy module imports until command invocation optimizing performance through its Click-based architecture.