Cinematic Story Image Generator

I produce short, 5–10-minute stories and I need a small tool that can turn each script into a complete sequence of realistic, cinematic images—one image per scene, so the full story is visually covered from start to finish. My ideal flow is simple: I paste the story text, the system (leveraging ChatGPT for prompt generation and any image engine such as DALL·E, Stable Diffusion, or Midjourney) breaks the narrative into the right number of beats, writes precise, camera-ready prompts, then renders the images in high resolution.

Essential requirements
• Realistic, cinematic visual style throughout all images
• Automatic scene detection so every major moment in the story receives its own frame
• Straightforward interface—a lightweight web page, local script, or notebook is fine as long as I can drop in new stories without code edits
• Clear, reusable code with comments so I can expand or retrain later if needed

Acceptance criteria
1. I feed a sample 5-minute story and receive a cohesive set of scene images that match the plot.
2. Visual tone stays consistent and filmic (no cartoon elements).
3. Prompts and generated outputs are stored or displayed so I can reuse or fine-tune them.

If you have experience chaining ChatGPT (for prompt engineering) with your preferred image model and can deliver the above, I’d love to see a quick demo or outline of your approach.