cclank/lanshu-create-ai-presenter-video ? reverse-engineered prompt

Build me a simple workflow that takes a script or topic and an authorized presenter photo, then turns it into a finished AI presenter video.

I want it to handle the whole process for me, including organizing the script, generating or syncing the voice, creating the presenter video, adding accurate subtitles, highlighting important keywords, and exporting a polished final version plus a QA report. Make the default format a clean vertical or horizontal social video, with a clear opening, a few content beats, and a short closing.

Please also include basic checks so it warns me if the image rights are not confirmed, the presenter is not clearly an adult, or the input is missing something important. If you need to look up current docs online or use the local tools available in the environment, go ahead. Keep it provider neutral so it can work with whatever video and voice tools are available.