pjecuacion/script-to-video-skill — reverse-engineered prompt
Reverse engineered prompt
Build me a public, reusable script to video skill for HyperFrames. I want someone to be able to copy the folder into their skills setup, open one clear skill file, replace a few local paths, and then generate a timed video composition from a topic, a finished script, an audio file, or a talking head video.
It should explain three simple workflows, normal script to video with voiceover and storyboard timing, a catalog showcase from an existing media file, and a talking cut mode that keeps the original audio in sync while switching between camera footage and graphic cutaways. Include a small set of theme JSON files for visual styles.
Please make the repo safe to publish. No private keys, no personal paths, no real voice samples, and no machine specific examples. Use placeholders where needed. The README should clearly say what tools are required, like HyperFrames, ffmpeg, optional local voice generation, and an optional Pexels key. Include a few example prompts a user can copy. Look up current docs online if you need to.
Want more depth? Deep Reverse