Hi, I'm wondering how to handle a flow of 4 tasks related to a simple input. My input is a basic object or set of objects and i want GPT to generate a "acceptable" story about it, define a title to the story, some chapter title and a prompt to generate art for each chapter. What would be the best approach ?
1 agent with the role of a novel writer, maybe even graphic designer to who i ask to generate it all
or
1 agent with the same to who i would request each task separately in the logical order ? (story > title > chapter number > chapter title > prompts)
or
multiple agent with a more specific role to each task but still requesting task separably in the same order as before ?
thank you