AI Agent

An AI agent is an autonomous software system that uses a large language model as its reasoning engine to perceive its environment, plan actions, call tools, and complete tasks with minimal human intervention at each step.

Unlike a simple AI assistant that responds to single queries, an agent maintains state across multiple steps, makes decisions based on intermediate results, and can trigger real-world effects through tool calls and API integrations.

Closely related concepts include A2A Protocol (Agent-to-Agent Protocol), Context Window Poisoning, Exfiltration via Agent, MCP (Model Context Protocol). Each is defined in the Prefactor glossary.