Question: capturing and captioning device screen recordings from a script

Not a bug report, more a question for anyone still building tooling on top of libimobiledevice.

I have a small pipeline that pulls screen recordings off a connected device and post-processes them before they go to testers. The part I keep re-inventing is burning readable text onto the clips. I ended up handing that step to an AI caption generator instead of maintaining my own ffmpeg and whisper glue, which collapsed the whole stage into one call.

Two things I would like a second opinion on:

1. Is there a recommended way to capture screen recordings programmatically these days, or is looping idevicescreenshot still the practical answer on current iOS?
   
2. Has anyone found a reliable way to read the device-side timestamp, so captions can be aligned to the original session rather than to the exported file?

Happy to write up whatever I settle on if it would be useful to others.