Upfront: I did not build this. It is an open-source project on GitHub (feichanggege/ecommerce-visual-copywriting-skill, MIT). It is listed here on star count, maintenance activity and how well it fits day-to-day storefront operations.
What it does
- Copy and layout together — it does not just suggest composition; it binds the message to the layout, which is the genuinely hard part of a product image.
- Explicit criteria — what the main image must lead with, how to tier information, how much type contrast to use, all as rules rather than vague taste.
- References and examples included — the repo ships
referencesandexamplesfolders, and adapting beats starting from a blank canvas. - Tool-agnostic — standard SKILL.md, so Claude Code, Codex and Cursor all read it.
How to use it
The skills CLI is a one-liner; you can also copy it into Codex's skills folder by hand:
# option 1: skills CLI
npx skills add feichanggege/ecommerce-visual-copywriting-skill
# option 2: manual install for Codex
git clone https://github.com/feichanggege/ecommerce-visual-copywriting-skill.git
mkdir -p ~/.codex/skills/ecommerce-visual-copywriting
cp -r ecommerce-visual-copywriting-skill/SKILL.md \
ecommerce-visual-copywriting-skill/references \
ecommerce-visual-copywriting-skill/examples \
~/.codex/skills/ecommerce-visual-copywriting/
Hand the agent your product details and source photos and ask it to work the SOP. The repo ships both Chinese and English SKILL.md files.
When to use it, and when not to
This governs what goes on the image and how it is arranged. For the image processing itself, see Background Removal JS or ComfyUI-RMBG.


