mateogon/pdf-narrator ? reverse-engineered prompt

Build me a desktop app that can turn PDF and EPUB files into audiobooks with a simple, friendly interface.

I want to be able to pick one book or a whole folder, extract the text cleanly, and generate spoken audio with Kokoro TTS. It should handle chapters when possible, skip obvious junk like page numbers and headers, and let me choose a voice, output format, and chunk size so it still works on low resource machines. Please add a progress display, logs, and controls to pause, resume, and cancel while it’s running.

It would also be great if there’s a voice test area where I can try one voice or quickly sample all available voices before I make the audiobook. Save the finished files in an output folder and keep the setup as easy as possible. If you need to check current docs online for Kokoro or any other dependency, go ahead.