Introduction: What You’ll Build

The Goal

We’re building a Personal To-Do Agent. It’s a focused agent that does one thing well: manages your to-do list.

Here’s how it works:

1. You type: “I have a meeting with Ali tomorrow at 10, add it”
2. The agent decides to call the addTodo tool
3. The tool adds the task to memory
4. The agent confirms: “I’ve added your meeting with Ali tomorrow at 10”

Why Agents Matter

Most AI applications are just single API calls. You send a message, get a response, done.

Agents are different. They can:

• Take actions - Not just chat, but actually do things
• Remember context - Keep state across multiple turns
• Make decisions - Choose which tools to use
• Loop until done - Keep working until the task is complete

What This Tutorial Covers

• Building a simple agent loop
• Defining and calling tools
• Managing in-memory state
• Creating a basic web UI

What This Tutorial Doesn’t Cover

• Complex planning or reasoning
• External databases (we use in-memory storage)
• Authentication or security
• Production deployment

We’re keeping it simple so you understand the core concepts.

The Final Result

When you’re done, you’ll have a working agent with:

• A chat interface (left side)
• A tools log (right side) showing which tools were called
• A to-do list that updates in real-time

Let’s start building.

---