Video Automation Lab - PROMPTS

Prompt to convert script to JSON format:

Adapt the attached YouTube script to a structured JSON output that consists of a series of SCENES. Each SCENE will include two fields: a “script” and a “visual”. The “script” field should consist of the text associated with the narrator – HOST. HOST’s text should be copied verbatim from the YouTube script. The "script" field in each SCENE should contain one to two sentences or phrases for the HOST to speak. Segment the “script” naturally into 4–8s beats, with priority on keeping the HOST’s delivery smooth and natural. The “visual” field in each SCENE should consist of a text only description of a single suitable static image (no video, no animation) that complements the "script" in the same SCENE. This text only description for the "visual" field should be suitable for inclusion into a larger prompt that will be fed to an image AI.


###EXAMPLE OUTPUT FORMAT###
{
  "scenes": [
    {
     "script": "The airline loyalty game changed while you weren't looking - and right now, you're playing by rules that were designed to take your money, not reward your time.",
      "visual": "Close-up of HOST at desk with vintage airline loyalty cards spread out, concerned expression."
    },
    {
      "script": "At our age, comfort isn't a luxury - it's about arriving healthy, rested, and actually enjoying the trip we've earned.",
      "visual": "Split-screen or side-by-side comparison showing the stark contrast between uncomfortable and comfortable travel for seniors."
    },
    {
      "script": "In the next 30 minutes, I'll show you exactly why elite status became a $28,000 scam, expose why upgrades don't work on the flights that matter most, and give you a step-by-step system to fly business class for way less - guaranteed seat, no waiting list, no games.",
      "visual": "Large, alarming "$28,000" in bold red text with a prohibition symbol (circle with diagonal line) overlaid."
    }
  ]
}