Bold cartoon style, thick black outlines, flat bright colours with one hard shadow tone, glossy highlights, playful underwater palette. Draw layers for a generative PFP collection. All files 1024x1024 RGBA PNG. 1) artifacts/base.png: the base character, A cool cartoon fish character with a big round fish head, large expressive eyes, big lips, shiny orange scales, a sturdy neck and broad rounded shoulders like a fish-headed person

, facing straight at the viewer, both eyes level and looking at the viewer, head centred left to right; framed head and shoulders: the top of the head near y=210, the neck meeting the shoulders near y=740, both shoulders and the upper chest in the picture and running off the bottom edge; the arms are down and out of view, nothing in front of the chest; extra detail from the creator: looking straight at the viewer with a confident, laid-back expression, ON THIS TEMPLATE: crown (top of the head) at y=210, eye line at y=425, chin at y=665, narrowest point of the neck at y=700, head 350px wide at the cheekbones centred on x=512, bare shoulders running off the bottom edge about 900px wide. Natural proportions for what it is: a human head or skull is about 1.3x taller than wide with a domed cranium and a jaw that narrows, eyes near the vertical midpoint of the head; never squashed or oval. Transparent background; eyes open; a simple neutral closed mouth drawn on the base, matching the style (there is no separate mouth layer in this collection); no hat or headwear (hats are a separate layer); no glasses or eyewear (a separate layer); NO CLOTHING AT ALL: bare neck and shoulders, no collar, no scarf, no coat, no shirt (clothing is a separate layer drawn on top, and anything on the base shows through behind it); no chains or necklaces (a separate layer). 2) 12 full-canvas opaque backgrounds: flat solid colours: deep ocean blue, aqua, coral pink, sandy yellow, seaweed green, plus a rare bubbly gradient. Also write artifacts/base_b.png and artifacts/base_c.png: two more takes of the SAME character on the SAME template and pose, each a SEPARATE, FRESHLY GENERATED image (run the image tool again for each one; never copy, edit, recolour or re-save base.png), so that if one comes out too big, too small or off-centre another may fit. Write artifacts/base.png, artifacts/background_01.png .. background_12.png, and artifacts/traits.json: {"backgrounds":[{"file","name","weight"}...]} with weights summing to 100.