GlossaryCore Concepts
Agent
An AI system that autonomously perceives its environment, reasons about goals, takes actions, and adapts based on feedback — typically over multiple turns without explicit step-by-step human direction.
Definition
An agent in the context of AI systems is a program that:
- Perceives its environment (reads input, retrieves from memory, receives tool results).
- Reasons about what action to take to progress toward a goal.
- Acts — executes tools, generates outputs, or delegates to sub-agents.
- Adapts — updates its understanding based on feedback and repeats.
This observe-reason-act loop continues until the agent reaches a terminal condition: goal achieved, error limit exceeded, or explicit stop.
Agent vs. Chatbot
A chatbot responds to a single user message with a single response. An agent pursues a multi-step goal autonomously, potentially taking dozens of actions before surfacing a result to the user. The distinction is one of agency (ability to take consequential actions) and autonomy (doing so without per-step human instruction).
Types of Agents
- Reactive agents — respond to the current state without maintaining internal goal representations.
- Deliberative agents — maintain a model of the world and plan sequences of actions.
- Hybrid agents — combine reactive and deliberative components.
In LLM contexts, most modern agents are deliberative: the language model acts as the planner, and tools provide grounding in the real world.