google-adk

Develop AI agents using Google's Agent Development Kit.

1|Updated Mar 12, 2026

Search Tags:#workflow #mcp #ai-agent #adk #agent-development #vertex-ai

Ready to run

npx skills add https://github.com/lifeodyssey/opencode-team-config --skill google-adk-lifeodyssey

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: google-adk
Source: https://github.com/lifeodyssey/opencode-team-config/tree/main/skills/google-adk
Command: npx skills add https://github.com/lifeodyssey/opencode-team-config --skill google-adk-lifeodyssey

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Google ADK development guidance helps teams build and evaluate AI agents using Google's Agent Development Kit.

Core Features & Use Cases

Agent Types: LlmAgent, SequentialAgent, ParallelAgent, LoopAgent, Custom Agents for extending BaseAgent.
Tools & MCP: FunctionTool, McpToolset, built-in tools for search and code execution.
Sessions & Deployment: InMemorySessionService and DatabaseSessionService for development and production persistence; CLI commands for dev, deploy, eval, and web.
Use Case: Create an agent that can search docs, run code, and orchestrate multiple sub-agents to complete a task.

Quick Start

Set up a new LlmAgent using a sample tool and launch the local ADK development server.

Frequently Asked Questions about google-adk

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build AI agents using Google's Agent Development Kit?▼

To build AI agents with Google's Agent Development Kit, you define agent types like LlmAgent, orchestrate sub-agents using SequentialAgent or ParallelAgent, equip them with FunctionTool or McpToolset, and launch the local dev server via CLI commands.

What types of agents can I create with Google ADK?▼

Google ADK supports multiple agent types including LlmAgent for core logic, SequentialAgent and ParallelAgent for orchestration, LoopAgent for iterative tasks, and Custom Agents created by extending the BaseAgent class for specialized workflows.

How do I integrate MCP tools into an ADK-based agent?▼

You integrate MCP tools into an ADK-based agent by using the McpToolset component, which connects your agent to external Model Context Protocol servers, alongside built-in tools for search and code execution.

Can I use Google ADK to deploy agents to Vertex AI?▼

Yes, Google ADK provides CLI commands for deployment to platforms like Vertex AI, allowing you to transition agents from local development using InMemorySessionService to production environments backed by DatabaseSessionService.

What is the best way to evaluate AI agents built with Google ADK?▼

The best way to evaluate AI agents built with Google ADK is by using the framework's built-in CLI eval commands, which test agent behaviors and tool integrations across development and deployment workflows.

Does Google ADK support persistent sessions for production agents?▼

Yes, Google ADK supports persistent sessions for production agents by providing DatabaseSessionService to replace the InMemorySessionService used during local development, ensuring state is maintained across deployments.