AI Agents

These terms describe how agentic AI systems break tasks into loops, use tools, and keep context.

Tool Calling

Tool calling is when an AI system triggers external functions or APIs to fetch data or take actions as part of solving a task.

An agent loop is the repeated cycle of planning, taking an action (often via tools), and checking results until a goal is met.

A planner is a component that turns a goal into an ordered set of steps, sometimes delegating subtasks to specialized tools or modules.

In agent systems, memory is stored context (like notes or retrieved documents) used to stay consistent across multiple steps.