An AI agent combines a model with instructions, context, memory, tools, and an execution loop. It can break a goal into steps, gather information, act in software, inspect results, and adjust its next action rather than producing only one static response.
Agents need clearly bounded authority because tool use can change durable state or affect other people. Good designs define permissions, verification, stopping conditions, budgets, monitoring, and approval requirements according to the consequences of the task.
ELI5
An agent is an AI system that can work through a goal in several steps. It observes what is available, chooses an action, uses an allowed tool, checks the result, and decides what to do next.
For example, a travel-planning agent could compare train times, organize suitable options, and prepare an itinerary. It should not purchase a ticket unless it has permission, because taking action is different from merely giving advice.
