Agent
A model that works toward a goal in a loop — calling tools, reading results, and deciding the next step until the task is done.
An agent uses a model not for a single answer but as the decision-maker in a loop: it calls a tool, reads the result, decides what to do next, and repeats until it judges the task complete. Investigating a case, researching a question, or completing a multi-step workflow are all agent-shaped tasks.
Agents are largely a system prompt plus a set of well-described tools plus a loop with a limit. Much of the reliability work is prompting — telling the model when to use a tool, never to invent arguments, and when to stop — backed by code that caps the steps and keeps the model away from irreversible actions.