What problem does it solve? Choosing the right image generation model and writing effective prompts is error-prone: text rendering fails on some models, transparency is unsupported on others, and vague prompts produce unusable results. This Skill provides model selection rules, a structured prompting framework, and working API call patterns for Gemini and OpenAI image generation. ## Core Features & Use Cases - Model Selection Matrix: Routes tasks to the right model — Gemini for photorealistic scenes, GPT Image 2 for text rendering, batch variations, and multi-reference compositing, GPT Image 1.5 for transparent icons. - 5-Part Prompting Framework: Structures prompts by image type, subject, environment, technical specs, and constraints, with photography parameters (lighting, lens, aperture) in the references guide. - Ready-to-Run API Patterns: Python snippets for Gemini and OpenAI image generation, multi-turn editing with thought signatures, WebP conversion, and vision-model quality checks. - Use Case: When building a landing page that needs a hero image with a headline, an OG share image, and a transparent icon set, use this Skill to pick the correct model for each asset and generate them via API calls. ## Quick Start Generate a hero image for my website using the ai-image-generator skill with a prompt describing a modern office interior in warm natural light.
