gpt-image-2-style-library

What problem does it solve?

It helps you quickly choose the right GPT-Image2 visual style and prompt template for a given image-generation request, then assemble a structured prompt with concrete constraints (layout, text, aspect ratio, and negative details).

Core Features & Use Cases

• Template + Style selection from a style library: Matches your request to a template category, then refines using visual style tags and scene tags.
• Production-grade prompt construction: Produces a final prompt composed of subject/task, composition, materials/style, required text/labels, output format, and constraints/negative details.
• Higher control via repository-backed reference: Uses references/style-library.md (generated from data/style-library.json) to prioritize correct template names, categories, cover/style tags, and scene tags.
• Chinese/English handling: Automatically outputs the final prompt in Chinese or English to match the user request.
• When ambiguity exists, it offers options: If multiple templates fit, it proposes 2–3 candidates with short reasons for the user to choose.

Quick Start

Use the gpt-image-2-style-library skill to generate a production-ready GPT-Image2 prompt for a city life system map in Chinese.