Routes your image edit intent to one of four RunComfy models without making you pick upfront. You say "swap the background" or "rewrite this headline in Japanese" and it sends the job to Nano Banana Edit for batch identity preservation, GPT Image 2 Edit for multilingual text or multi-reference composition, Flux Kontext Pro for single-shot local edits, or Z-Image Turbo Inpaint for mask-driven object removal. The skill bundles each model's prompting patterns so you don't waste runs on the wrong tool. Useful if you edit product shots one day and localized ad variants the next. Needs the RunComfy CLI installed locally.
npx skills add https://github.com/runcomfy-com/skills --skill image-edit