buildwithhanif/hyperframe-pro ? reverse-engineered prompt

Build me a HyperFrames production layer that helps me make short vertical explainer videos people actually watch.

I want a workflow where I start with a topic, turn it into a tight script, then a beat plan, then voice over, then the video, with checks before the final render so I catch problems early. The beat plan should come first, not be an afterthought. I also want a built in lint step that fails when there is dead air, broken cues, missing elements, stretched images, or a cut that goes over the 2:59 limit.

Make the timing smarter too. Cues should lock to words in the voice over instead of timestamps, so the video can retime itself if I change the script. I also want a way to swap one line in a finished take, keep the same voice, and retime everything after it without re recording the whole thing.

If you need to check current HyperFrames or ffmpeg docs online, go ahead and do that.