Prompt
Also known as: User Prompt
A prompt is the text (token sequence) you give a language model as input before it generates a continuation. In the simplest case it is a short prefix such as Rain falls. The model then predicts the next token, then the next, conditioned on that prefix plus whatever it has already produced.
In chat products the prompt is usually packaged with roles (system prompt, user, assistant) via a chat template. Underneath, it is still tokens into a causal model. Prompt engineering is the craft of writing those inputs; it does not change the trained weights.