AI agent
A program that pairs a language model with tools, such as terminal commands and files, to carry out a task in several steps: it reads its context, acts, looks at the result and repeats.
Unlike a chat, an agent doesn't only answer: it does things. It can read files in your folder, write and run code, and use what it sees to choose its next step.
Example: you ask a terminal agent "what share of our customers ordered more than once?". It opens the orders CSV, writes a pandas script, runs it, reads the number and reports it back, all in one session. You can see every command it ran.
What it doesn't mean: an agent that acts isn't therefore right. The script can run without errors and count the wrong thing, such as orders instead of customers. A code error shows; a wrong decision about what to count doesn't. That's why an agent needs someone who knows how to check its work.
Where the line is: a chat that suggests code for you to copy and run isn't an agent. An agent runs the code itself and decides what to do next from what it saw.
Learn more
Related terms
- Agent harness
The program that wraps a language model and makes it an agent: it gives the model tools, asks for your permission, keeps the session and decides what goes into its context.
- Context window
Everything a language model has in front of it at a given moment: your instruction, the files it has read and the results of its tools. What isn't in the context, the model can't use.
- Grain (of a table)
What one row of a table stands for: a person, a transaction, a planet, a person per language. Counting rows counts what you want only if you know the grain.
- Terminal
A terminal is a window where you type commands for your computer instead of clicking. You type a line, press Enter, and the computer answers in text.