expert at AI - specifically creating images with OpenAPI image and other LLM tools

We have created a youtube channel that involves a male/female vlogger couple who time travel through history and vlog about it. We have a very well define workflow that involves creating scripts using the help of claude or chatgpt, defining scnes, dialogue, etc., and then creating detailed start frames for each scene using either chatgpt image creation, Nano Banana, or whatever model you feel can do the job. We will present you with reference photos of our wo main characters and describe the scenes. Now if you are able to take this further and also come up with script ideas and once we approve, create the full scripts and images together, that would be great. We have a proven workflow to produce the videos usiing a combination of Claude Cowork and Higgsfield and of course the scripts and the start frames, so that might be something we can outsource too, but for now we're just looking for help on the start frames, but open to more.