anashany98/OCR — reverse-engineered prompt

Build me a simple OCR document tool from this repo. The README is empty, so please inspect the existing files first and turn it into something usable.

I want to be able to give it scanned documents or images, have it read the text, and show the extracted text in a clean way. If there are multiple pages, keep the results separated by page. If the text is messy, try to do basic cleanup so it is readable. It should also let me save or copy the extracted result.

Please make it easy to run locally, with clear setup instructions and a simple command to start it. If there is already architecture or analysis written in the Spanish docs, use that as guidance. Keep the code organized, add any missing error handling, and include a small example so I can test it quickly. Look up current OCR docs online if you need to.