What problem does it solve? Foundational image and video models like Midjourney, Sora, Runway, and DALL-E default to systemic stereotypes, clone faces, gibberish cultural text, and geographically inaccurate settings when depicting diverse people. This Skill writes structured prompts with explicit constraints that counter those biases and produce dignified, authentic representation. ## Core Features & Use Cases - Annotated Prompt Architectures: Builds prompts systematically across Subject, Action, Context, Camera, Style, and explicit negative constraints. - Negative-Prompt Libraries: Blocks AI artifacts such as clone faces, fake cultural symbols, nonsensical non-English text, and physics glitches in video. - Video Physics Definitions: Specifies temporal consistency for clothing, hair, and mobility aids in Sora and Runway generations. - Use Case: A marketing team needs a video of a Black female executive leading a meeting in Nairobi. The Skill produces a prompt specifying natural 4C hair, accurate architecture, skin-tone-appropriate lighting, and negative constraints against stock-photo tropes and cloned background actors. ## Quick Start Ask the agent to write an inclusive, bias-resistant image or video prompt for your creative brief, specifying the subject, setting, and target platform.
