Summary

skills/openai-image-gen/scripts/gen.py currently accepts any --background (GPT models) and --style (dall-e-3) value and forwards them directly to the Images API.

Invalid values fail only at request time with API errors, which is less clear for CLI users and causes avoidable failed requests.

Proposed improvement

• Validate --background against transparent, opaque, and auto for GPT image models.
• Validate --style against vivid and natural for dall-e-3.
• Normalize accepted values to lowercase and fail fast with clear CLI errors for invalid input.

Implemented by PR: #36762