A self-contained introduction to AI agents — the building blocks of modern agentic systems.
An agent is a system that observes its environment, reasons about what to do, and takes actions to achieve a goal. Unlike a simple chatbot that responds once, an agent can loop: act, observe the result, and act again.
Think of it as a junior engineer with a toolkit. You give them a task, they pick tools, execute steps, check their work, and iterate until done.
Which best describes an AI agent?
For a deeper dive, read Building Effective Agents by Anthropic.