Generates a single poster, key art, or editorial illustration per turn using whatever image model your project specifies (gpt-image-2, Flux, Imagen, etc.). It takes your brief, builds a structured prompt covering composition, lighting, palette, and camera feel, then dispatches through the unified media contract and saves a PNG or JPEG straight to your project folder. The workflow is tight: no artifacts, no extra variations unless you ask, and it respects the aspect ratio in your metadata so you don't waste a rerender. If you need one hero image and you already know roughly what it should look like, this is faster than assembling the prompt yourself.
npx skills add https://github.com/nexu-io/open-design --skill image-poster