How to build a real AI agent using

Google's new agent development kit in 60

seconds. First, import and define the

agent with LLM agent. Set the model, the

agents name, and its behavior profile.

Next, give the agent a tool, a simple

Python function it can call here, map

countries to their capitals, then

register it. You can also store and

access memory, perfect for past

preferences and context, making the

agent smarter across multiple

interactions. Now run the agent with a

natural language goal like what's the

capital of Japan? The agent parses the

query, calls the right tool, and returns

the answer. That's it. A goal- driven

agent with tools, memory, and structured

instructions, all in Python, all local.

This is how real agents are built.