This project provides a structured prompt-generation system for AI image creation, specifically targeting detailed and coherent female portraits. It serves AI artists and power users by offering an extensible framework that expands minimal inputs into rich, copy-ready prompts, ensuring visual consistency and stable generation while preserving user-defined parameters.

How It Works

The Female Portrait Prompt Director Skill operates as a modular framework within the Codex ecosystem. It routes each request through an on-demand style file, allowing for specific visual aesthetics while locking explicit user parameters. The system parses distinct modules—face, body, outfit, scene, camera, pose, lighting, and filters—to expand short inputs into a coherent, detailed photographic moment. This approach avoids conflicting style keywords and fuses expanded modules into natural prompts with a photography-director's intent, preserving priorities like clothing display or identity visuals.

Quick Start & Requirements

Installation is streamlined via npx for Codex users: npx skills@latest add liyue-aigc/female-portrait-director -g -a codex -y A manual installation via git clone into the Codex skills directory is also supported. This requires Node.js with npx. Refer to parameter_schema.md for the full input schema and usage_examples.md for usage examples.

Highlighted Details

• Supports a wide array of styles including clean lifestyle, curve-focused, urban fashion, Gufeng fantasy, e-commerce, retro Hong Kong, French relaxed, New Chinese oriental, sporty, travel, studio-retouched, oriental voluptuous, cold xianxia, and bright luxury gufeng.
• Core capabilities include locking and stabilizing user parameters, routing requests through style files, parsing visual modules, expanding short inputs into detailed prompts, and fusing modules into natural outputs.
• Preserves clothing-display priority for e-commerce images and authorized selfie identity or product core visuals for direct reference-image generation.
• Outputs include locked parameters, module analysis, a final prompt, and negative constraints.

Maintenance & Community

The project is authored by Li Yue (李岳) and is currently at version V1.4.1. No specific community links (e.g., Discord, Slack) or roadmap details are provided in the README.

Licensing & Compatibility

This project is licensed under the permissive MIT License, allowing for use, copying, modification, distribution, sublicensing, and selling. The responsible-use safety boundaries do not alter the MIT License terms, making it compatible with commercial use.

Limitations & Caveats

The system defaults to fictional, clearly adult subjects. Strict safety boundaries are enforced, prohibiting the generation of content involving sexualized minors, explicit nudity, non-consensual images, deceptive identity, harassment, defamation, privacy violations, or other unlawful purposes. These are responsible-use guidelines.