Glossary / AI agent
What is an AI agent?
An AI agent is an AI system that works toward a goal on its own: it plans steps, uses tools such as a browser, a terminal or a database, checks the results, and keeps going until the job is done.
A chatbot answers one message at a time. An agent takes a task, like "add a booking calendar to this app", and breaks it into steps: read the existing code, write new files, run the app, see an error, fix it, run it again.
What makes it an agent is the loop. It acts, looks at what happened, and decides the next action, rather than giving you a single answer and stopping.
Agents are only as good as the tools and limits they are given. A well-built agent can run code, read logs and search, but asks before doing anything costly or hard to undo.
In an AI app builder, the agent is the part that actually builds. You describe the outcome; the agent does the many small steps in between.
Examples
- A coding agent that writes, tests and fixes code
- A research agent that searches and summarises sources
Questions
What is the difference between an AI agent and a chatbot?
A chatbot replies to messages. An agent works on a task across many steps, using tools and checking its own results, until the task is finished or it needs your input.
Can an AI agent make mistakes?
Yes. Agents can misread a request or fix one thing and break another. That is why good tools show what the agent did, keep checkpoints you can roll back to, and let you test before publishing.