A GOOD AGENT KNOWS WHEN IT DOESN’T KNOW

A trustworthy Agent is not one that always gives an answer.

It needs to know:

• when data is missing;
• when there is no source;
• when information is contradictory;
• when something falls outside the task scope;
• when to ask the user again;
• when to hand over to a human.

For example:

Instead of filling in a missing figure by itself:

The data needed to confirm this conclusion could not be found.

In many professional workflows:

knowing when to stop is as important as knowing how to answer.