darshil0/commute-news-audio-ai ? reverse-engineered prompt

Build me a full stack web app called CommuteBrief that turns news articles, web links, and pasted text into short audio briefings for a commute.

I want users to be able to search for current news, pull in an article from a URL, or paste in text, then have it turned into a clean spoken summary with a few different narrator voices to choose from. Make the audio player feel polished, with play, pause, seek, skip, a speed slider, preset speed buttons, volume control, and a sleep timer. Add a simple settings area where the voice styles can be previewed before generating anything.

Please also make it work offline as much as possible by saving articles, playlists, and listening history locally, and add light haptic feedback on taps, skips, and when playback finishes. If there is a reasonable way to support syncing and login between devices, include that too. Use the current docs online if you need to, and keep the whole thing feeling fast, modern, and easy to use.