adk-engineer

The 'adk-engineer' skill is designed to assist software engineers in creating production-ready Agent Development Kit (ADK) agents. It focuses on ensuring that the agent code is maintainable, testable, and deployable with best practices in code structure, testing, and deployment automation. This skill is ideal when engineers are tasked with building, refactoring, or deploying ADK agents or applications. By providing clear guidelines and ensuring compliance with industry standards, this skill streamlines the process of engineering agents for real-world use cases. It is triggered by relevant phrases such as 'build ADK agent', 'create agent code', or 'engineer ADK application', making it highly responsive to specific developer needs.

The skill's primary capabilities include designing and implementing ADK agent code with a clear structure, robust test coverage, and operational safeguards. Users can expect a structured approach to designing agent workflows, integrating tools incrementally, and ensuring that deployment processes are well-defined. Furthermore, it emphasizes error handling and provides actionable plans for debugging build, runtime, and deployment issues. It offers resources such as a detailed playbook, repo standards, and references to relevant documentation. This makes the 'adk-engineer' skill a comprehensive toolset for engineers seeking to optimize their ADK-based applications and deployments.

This skill is especially useful for developers working with ADK agents in environments like Python, Java, or Go, particularly when aiming to deploy multi-agent systems or refactor existing ones. Its target users are software engineers, DevOps teams, and anyone involved in creating or maintaining ADK agents in production settings. Typical use cases include building agent systems with clean architecture, adding testing frameworks, and deploying these agents to cloud environments, specifically Google Cloud Platform (GCP).