skills

Build agent-to-agent (A2A) communication systems using Google's A2A Protocol Python SDK. Use when working with multi-agent systems, agent orchestration, or implementing agent communication patterns including: (1) Creating A2A-compliant server agents, (2) Building A2A clients to communicate with agents, (3) Sending messages and handling responses, (4) Managing tasks and task lifecycles, (5) Implementing streaming responses, (6) Creating agent cards and capabilities, (7) Multi-turn conversations between agents.

Build AI agents, chatbots, and autonomous assistants with tools, memory, and multi-agent systems using Python. Use when: (1) Creating agents/chatbots/assistants that use tools (Google Search, APIs, custom functions, databases), (2) Building multi-agent systems with specialized agents that collaborate, (3) Agents that remember conversation history or maintain state, (4) Deploying agents to Cloud Run, Vertex AI Agent Engine, or GKE, (5) Evaluating agent performance with test sets, (6) User mentions 'agent', 'chatbot', 'assistant', 'autonomous', 'multi-agent', 'ADK', or 'agentic'. Choose this over vertex-ai skill when building complete agents with tools and orchestration, not just making model API calls.

Generate text, analyze images/PDFs/videos, create structured JSON output, and build AI features using Google's Gemini models or Anthropic's Claude models on Vertex AI. Use when: (1) Generating text with AI (summaries, content, responses), (2) Vision tasks (analyze/describe/caption images, extract text from PDFs/screenshots), (3) Multimodal processing (image+text, video analysis, audio transcription), (4) Structured output (JSON, Pydantic schemas, data extraction), (5) User mentions Vertex AI, Gemini, Claude, or GCP AI, (6) Direct model API calls needed (vs full agent framework). NOT for building agents with tools - use google-adk skill instead.

Complete end-to-end examples demonstrating Google's AI agent ecosystem. Shows how to combine vertex-ai SDK, google-adk, vertex-agent-engine, and a2a protocol to build real-world agent systems. Use when: (1) Learning the full agent development workflow, (2) Understanding how technologies work together, (3) Building production multi-agent systems, (4) Need reference implementations for common patterns.

Deploy and manage AI agents on Vertex AI Agent Engine using the Agent Development Kit (ADK) and google-cloud-aiplatform SDK. Use when: (1) Building AI agents with function calling and tool use, (2) Deploying agents to managed production infrastructure, (3) Creating multi-agent systems, (4) Integrating agents with LangChain or LangGraph, (5) Working with production-ready agent templates, (6) Querying or managing deployed agents on Vertex AI.