Shopify 电商图片生成与编辑

Shopify Ecommerce Image Generation and Editing

Build a coherent image system across a Shopify storefront instead of isolated product renders. The same product should remain accurate across PDP, collection, homepage, email and paid social placements while composition changes for each channel.

Brand and product lock

Gather product angles, packaging, variants, dimensions, materials, brand colors, typography direction, audience, approved claims and current storefront examples. Define non-negotiables: product shape, logo, color, included items and any visual identity that must remain consistent.

Storefront image architecture

Scenarios and commands

1. PDP gallery system

python3 "$SKILL_PATH/scripts/imagegen.py" generate \
  --prompt 'Create a 5-image Shopify PDP gallery using the reference product: clean primary image, material close-up, scale-in-hand image, real use scene and packaging contents. Preserve exact product shape, color, logo and accessories across every image. Use one coherent lighting system and do not invent dimensions, claims or included items.' \
  --image /path/to/product-front.png \
  --image /path/to/product-side.png \
  --batch 5

2. Collection-card consistency

python3 "$SKILL_PATH/scripts/imagegen.py" generate \
  --prompt 'Create a Shopify collection-card image for the supplied SKU. Match the reference collection angle, subject scale, background tone and shadow; replace only the product with this accurate SKU. Keep packaging and logo unchanged, with enough crop tolerance for responsive storefront cards.' \
  --image /path/to/new-sku.png \
  --image /path/to/collection-style.png

3. DTC homepage hero

python3 "$SKILL_PATH/scripts/imagegen.py" generate \
  --prompt 'Wide Shopify homepage hero for a minimalist home brand. Place the accurate reference product in a credible morning routine, build a strong focal point on the right and reserve calm negative space on the left for approved headline and button. Preserve product details; no generated offer, testimonial or claim.' \
  --image /path/to/product.png

4. Meta Ads creative directions

python3 "$SKILL_PATH/scripts/imagegen.py" generate \
  --prompt 'Generate 3 paid-social creative hypotheses for the same DTC product: problem-in-context, hands-on demonstration and material proof. Keep product and brand system consistent, but change the customer reason to stop. Reserve safe copy space and do not invent a discount, review, result or certification.' \
  --image /path/to/product.png \
  --batch 3

5. Multi-market lifestyle adaptation

python3 "$SKILL_PATH/scripts/imagegen.py" generate \
  --prompt 'Adapt the reference lifestyle image for a Japanese urban apartment while preserving the exact product and brand color. Change only the environment, styling and available negative space; keep the use case realistic and culturally neutral. Remove old campaign copy and leave clean space for approved Japanese localization.' \
  --image /path/to/source-campaign.jpg

QA and asset handoff

• Product, logo, packaging, variants and included items remain consistent.
• PDP images answer different questions instead of repeating the same beauty shot.
• Collection cards align in angle, scale, background and crop behavior.
• Responsive crops retain the product and intended copy space.
• No generated offer, review, claim, certification or legal text.
• Deliver filenames by product handle, placement, market and version; verify current Shopify theme and ad-channel requirements.

Run

pip3 install requests
python3 "$SKILL_PATH/scripts/imagegen.py" init --skill-name shopify-ecommerce-image-generation-editing
python3 "$SKILL_PATH/scripts/imagegen.py" task --task-id 

The fixed image model is public_model_nano_banana_pro. generate supports repeatable --image, --batch, model --param, routing, output directory and submit-only mode. Use live AI Hive pricing and parameters.