Before you build agent graphs or operate Hermes, lift the lid just enough: an agent is a model using tools in a loop.

Underneath, an agent is simpler than it looks: it is a model using tools in a loop. It takes your goal, decides on an action, uses a tool, looks at what happened, and repeats, thinking, acting, observing, and going again until the job is done or it needs you. The reason it can adapt is that at each step it sees the real result of its last action and chooses the next one based on that, rather than following a fixed script. That loop is the entire heart of every agent, from a calculator demo in n8n to the most sophisticated systems in production.

This is also the honest bridge to the Engineering Bootcamp. Operating an agent means directing that loop from the outside. Building an agent means writing the loop, the tool definitions, and the orchestration from the inside, which is exactly what the Engineering course teaches. Whether you go there or not, you leave Week 3 no longer mystified by the word "agent," which is a real thing to own.