Image Generation Skill

Image Generation Skill

Generate or edit images for the current project (website assets, game assets, UI and product mockups, wireframes, logos, photorealistic images, infographics) using onyx-cli image. Generation runs server-side with whatever provider the admin configured at /admin/configuration/image-generation (OpenAI, Gemini, or Azure) — no API key is needed here.

When a provider isn't configured

If onyx-cli image … exits with "no image generation provider is configured", stop and tell the user: image generation is unavailable until an admin configures a provider at /admin/configuration/image-generation. Do not try to work around it with another tool.

When to use

• Generate a new image (concept art, product shot, hero, texture, sprite).
• Generate a new image guided by reference images (style, composition, mood).
• Edit an existing image (background replacement, object removal, lighting/weather change, compositing, inpainting).
• Produce many assets or variants for one task.

When not to use

• Extending or matching an existing SVG/vector icon set, logo system, or illustration library already in the repo — edit those directly.
• Simple shapes, diagrams, wireframes, or icons better produced as SVG / HTML/CSS / canvas.
• A small project-local asset edit when the source already exists in an editable native format.
• Any task where the user clearly wants deterministic code-native output, not a generated bitmap.

Decision tree

1. Intent — new image or edit of an existing image?
   • Modify an existing image while preserving parts of it → image edit.
   • Images supplied only as references for style/composition/mood, or no images → image generate.
2. Execution — one asset or many?
   • One asset → a single command.
   • Many distinct assets → one command per asset (do not use -n for distinct assets; -n produces variants of one prompt).

Assume the user wants a new image unless they clearly ask to change an existing one.

Usage