The Prompt Generator node allows you to send in a prompt to an LLM and receive text back. This allows a user to generate many different types of prompts from a single prompt, up to 25 distinct prompts. To best use the Prompt Generator, you should be as specific and clear as possible; for example, if you are trying to create a prompt to generate a cute cat, you want want to ask for, “A prompt that will generate a cute cat sitting down, in a modern flat-icon style like those used for SVG or iconography.” This will return better results than asking for a cat, which may result in many different types of prompt responses. You can also use the Prompt Generator node to choose between lists, return back specific words, or provide other text - useful if the results are being combined into a prompt template later in the workflow.
The Prompt Generator node is not a replacement for an agent, nor does it store conversational data or allow the output text to be modified or changed in the workflow graph; It should be thought of as a one-off request for text from a LLM. Also note that adding the number of responses to the node text doesn’t change the number of responses you get back; you should use the slider