ck:google-adk-python

Build multi-agent workflows in Google ADK by structuring Python agents into coordinator, sequential, parallel, or loop-based patterns. This Skill provides reference-driven architecture guidance for orchestrating these workflows, handling A2A communication, and routing tasks between specialist sub-agents.

Yes, you can integrate MCP servers and custom Python tools with Google ADK agents. The Skill covers tool context support and reference integration patterns for connecting external Model Context Protocol servers and built-in functions to your multi-agent application.

Manage session state and memory in Python AI agents using Google ADK's built-in state, memory, and artifact handling features. This Skill provides architectural guidance for preserving session context across multi-agent interactions and customer support routing scenarios.

Yes, Google ADK supports deployment on Vertex AI, Cloud Run, and GKE for production agents. This Skill provides reference-driven guidance for cloud deployment scenarios, ensuring your multi-agent Python application ships with evaluation coverage and observability.

Evaluate and test AI agents built with Google ADK Python by leveraging the framework's evaluation and callback features. This Skill covers production readiness patterns including testing, observability, and evaluation coverage to validate multi-agent workflows before deployment.

Yes, you need Python agent entry points to build multi-agent apps with Google ADK. This Skill requires Python entry points and tool context support to provide architecture, state handling, and deployment guidance for production-ready AI agents.