This page was translated from Dutch with AI.
Tell your agent when it is finished
There comes a moment in my training sessions when the foundation is ready. The folders are set up, the first skills are built. And then I ask a new question: what do you want your agent to do while you go and eat? Then it is suddenly about something else than asking a question and getting an answer. It is about an agent that continues working on its own.
There is a word for that, loop engineering. You are going to hear that a lot in the coming period. It sounds technical. But the idea behind it is simple, and I use it very often myself.
Feel free to copy the prompt at the bottom of this article, with it you determine together with your agent when something is finished.
What good looks like
You agree beforehand when the work is finished. In jargon they call that 'what good looks like', what a good result looks like. That can be an example of a final result. It can be a screenshot of how you want it to look. Or a document structure, with the titles and sections that must be in it.
That becomes the test for your agent. It works and then checks if the result passes the test. If that is not yet the case, it continues working. Only when everything is correct, it says that it is finished.
In my prompt I just write that out. I provide a kind of checklist, and I add: keep working until you can tick off all these points. And the new models really do that too. I actually think that is the nice thing about it.
Keep reading for free
Register free with your e-mail address and read on. Your account opens the free articles in the learning space; the rest is for subscribers, from Learn Plus up.