How to build an AI agent with MCP, ADK, and A2A on Google Cloud

Author: Google for Developers - Bewertung: 44x - Views:391

Explore the powerful tools and protocols for AI agent development on Google Cloud. In this video, we dive into how to connect AI models to resources using Model Context Protocol (MCP), streamline agent creation with the Agent Development Kit (ADK), and enable seamless communication between agents with the Agent2Agent (A2A) protocol. By the end, you will have a working AI agent system, while understanding its underlying architecture.

Resources:
Getting Started with MCP, ADK and A2A → https://goo.gle/48scfNV
Build with Vertex AI: Deploy Agents with Agent Development Kit (ADK), MCP, and A2A → https://goo.gle/48LAwjm
ADK samples GitHub → https://goo.gle/3Y5OC96

Chapters:
0:00 - Introduction
0:38- Understanding ADK, A2A, and MCP
2:02 - Setting up your Google Cloud environment
3:09 - Enabling Google Cloud APIs
5:04 - Building the MCP server for tools
6:57 - Deploying the MCP server to Cloud Run
8:42 - Creating your AI agent with ADK
10:21 - Exposing your agent via A2A
12:28 - Testing the A2A communication
13:29 - Cleaning up your Google Cloud project
14:02 - Conclusion: building with AI agents

Subscribe to Google for Developers → https://goo.gle/developers

Speaker: Julia Wiesinger
Products Mentioned: Google AI, Agent Development Kit (ADK)