What is it?
An AI agent is a system that uses a model to decide and take actions toward a goal.
An agent may use tools, memory, context, and a control loop. The application defines what actions are allowed and when a person must review the result.
06 · Agents, Tools & MCP · Topic 01
What is an AI Agent?
An AI agent combines a model with tools and a control loop to work toward a goal.
Think of it like...
Think of it like a navigator with a route and approved tools.
The navigator checks the goal, chooses a next step, and updates the route after new information arrives. The analogy describes a control loop, not human-like thought.
How it works
1 · GoalReceive the taskThe system gets an objective and constraints.
2 · PlanChoose a next stepThe model selects an action from the allowed options.
3 · ActUse a toolThe application performs the approved action.
4 · ObserveRead the resultThe agent receives tool output or new context.
5 · FinishContinue or stopThe system returns a result or requests review.
Real-world examples
Everyday example · Research assistant
Input: a request to compare two products
AI / System: searches approved sources and collects evidence
Output: a comparison with sources
Business example · Support workflow
Input: a customer issue
AI / System: checks account data and approved actions
Output: a resolution or human handoff
What it is NOT
AI Agent ≠ Chatbot
AI agent: can plan and take approved actions.
Chatbot: mainly exchanges messages with a user.
AI Agent ≠ LLM
AI agent: is an application system around a model.
LLM: generates model output from input context.
AI Agent ≠ Fully Autonomous System
AI agent: operates within defined tools and limits.
Fully autonomous system: would act without meaningful human or system controls.
Related concepts
Goal → Model Decision → Tool Action → Observation → Next Step
Explore next
Remember this
An AI agent is a controlled system that uses a model and tools to work toward a goal.
Independent explainer