liyue-aigc/female-portrait-director — reverse-engineered prompt

Build me a Codex skill that turns a few simple notes into a polished AI image prompt for realistic adult female portraits.

I want the user to be able to give things like style, scene, outfit, mood, face direction, body direction, camera angle, lighting, filter, aspect ratio, and platform use. The skill should keep anything the user clearly asked for, then expand the missing details into one coherent photographed moment instead of just repeating keywords.

It should support styles like clean lifestyle, urban fashion, e commerce clothing model, gufeng fantasy, retro Hong Kong, French relaxed, sporty active, travel vacation, studio retouched, and other portrait directions from the docs. Each request should use one style direction at a time so the prompt doesn’t feel mixed or messy.

The output should be easy to copy and include locked parameters, a short module analysis, the final prompt, and negative constraints. Please include clear safety rules so it only creates fictional adult, non explicit, non harmful portrait prompts.