SamuelSSan28/my-manga-recap ? reverse-engineered prompt

Build me a Python tool that takes a folder full of manga chapter images and turns it into a narrated recap video automatically.

I want it to read the pages with OCR, clean up the images if needed, write a short script that feels natural in Portuguese, turn that script into audio, and then combine everything into an MP4 with transitions, titles, and credits. It should work even if I don’t set up any AI keys, and if possible it should use better quality AI when an OpenAI key is available. If one method fails, it should fall back to another one so the process still finishes.

Please include a simple command line version and an interactive menu, plus support for processing more than one chapter at a time. Make it easy to configure with a .env file, and keep it organized so the OCR, script writing, audio, and video parts are separate. If you need to check current docs for any libraries, go ahead and look them up online.