← ClaudeAtlas

packaging-mockup-generationlisted

Generate packaging mockups and label visualisations, and handle the fact that models cannot render text reliably. Use whenever the user wants a packaging mockup, a label visualisation, a box or bottle or pouch render, or to see artwork applied to a product.
apimageorg/apimage-skills · ★ 0 · AI & Automation · score 72
Install: claude install-skill apimageorg/apimage-skills
# Packaging Mockups Packaging is the hardest thing to generate, for one reason: **packaging is mostly text, and models render text badly.** That doesn't make mockups impossible. It means the workflow is different from every other product image: generate the *form*, apply the *artwork* separately. ## The rule **Never let a model generate your packaging copy.** It produces text that looks like text — right typeface weight, right layout, plausible letterforms — and reads as gibberish or, worse, as almost-correct wrong words. On a product mockup that's: - Embarrassing if anyone reads it - A compliance failure if it's replacing an ingredients list, allergen declaration or warning - Useless as an approval asset, because nobody can sign off copy that isn't the copy ## The workflow that works ``` 1. Generate or photograph the blank form — the bottle, box, pouch, tube, unlabelled 2. create_brand_asset(type="product") free — reusable blank 3. Apply the real artwork in an editor — as a warped overlay following the form's surface 4. edit_image relight to integrate it — so the label picks up the scene's light ``` The artwork comes from your design files, at full fidelity, with the real copy. The generation supplies the object it sits on. ## Generating the blank form ``` generate_image( model="flux-2-pro", prompt="A blank matte white 250ml cosmetic pump bottle, no label, " "no text, no markings. Standing upright on pure white, " "straight