---
name: one-creator-sol
description: Create a consistent fictional UGC creator brief, 50 ad scripts, and reference-led image prompts from a product photo and a one-line description. Use when the user wants multiple ads with the same creator and product appearance.
---

# CreatorFrame — GPT-6 Sol

Create a reusable campaign pack. Keep one creator identity, outfit, setting, lighting direction, product appearance, and voice direction across the pack. Change hooks, angles, actions, and framing rather than the person's identity.

## Inputs

Use the attached product photo and the user's one-line product description. Ask for these if missing. Use an attached creator reference when provided; otherwise define one fictional adult creator appropriate to the audience. Infer a reasonable audience and tone when unspecified and briefly state your assumptions. Do not invent product benefits, ingredients, prices, statistics, endorsements, or personal experience.

## Shared continuity brief

Write a compact master block describing the creator's adult age range, face shape, skin tone, eye color, hair color and texture, and a few stable distinguishing features. Specify one outfit, one ordinary lived-in setting, window position and light direction, phone-camera style, and voice direction. Preserve the product's packaging shape, colors, scale, label and cap from the supplied image. Mark unreadable label text as unknown rather than guessing it.

Create one anchor portrait prompt. If image tools are available and the user requests images, generate the anchor first and reuse its actual image as the identity reference for all later shots. Text descriptions alone cannot guarantee the same face. If tools are unavailable, deliver usable prompts and clearly label them as prompts, not completed images.

## Campaign pack

Unless the user chooses another count, write 50 distinct 15–25 second scripts, grouped into 10 angles with 5 hooks each. For each include:
- Number and angle.
- Spoken hook, short body, and a relevant call to action.
- Shot description and a scene-specific image prompt.

Use the shared continuity brief verbatim as a prefix to every fully assembled generation prompt; a numbered scene can reference the brief in the written pack to keep the document readable. Include explicit instructions to use both the anchor creator image and the original product photo. Vary camera crop, hand action and expression while retaining identity, outfit and light. Keep demonstrations physically plausible. Frame invented dialogue as proposed ad copy, not a real customer testimonial.

For realistic imagery use visible pores, ordinary skin variation, a few stable facial details, hair flyaways, mild shadow grain and normal phone-camera perspective. Avoid airbrushed skin, beauty filters, studio rim lights, cinematic grading, warped hands, altered product labels and inconsistent facial features. Do not add arbitrary blemishes that change between images.

## Review and handoff

Check the full pack for duplicate hooks, unsupported claims, product drift and changes to the continuity brief. When reviewing generated images, compare face shape, eyes, hair, product proportions, labels, hands and lighting with the anchor. Identify drift honestly; do not promise perfect identity locking or identical generated voices. Voice direction is a brief, not an audio-cloning capability.

Deliver the shared brief, anchor prompt and all scripts with their scene prompts. Image and video generation require suitable tools separately. Do not install tools, make purchases, publish ads or launch campaigns from this skill. Follow the user's requested scope if they ask for fewer scripts or only images.
