EC Visual Skill
A Codex skill that turns a product photo and a reference into a plan of five main images and ten product-page screens.
- 113 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill plans and checks images and writes files only into a folder it creates on the desktop.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/ec-visual-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill plans and checks images and writes files only into a folder it creates on the desktop.
This is third-party code. Review the repository files before installing.
What it does
The `ec-visual` skill reads a product photo and references, takes only the visual language from the reference (light, type attitude, density, rhythm) and does not copy scenes. It then builds a single approval table: five main images and ten vertical product-page screens, each with a statement of what the picture says alone, a presentation pattern and a draft buyer-voice copy. Once the table is approved, the skill writes a prompt per image, checks results against rules (the picture works without captions, no icons, no repeated scenes, no borrowed colours) and redoes failures. Output goes to a separate timestamped folder on the desktop.
Who it is for. Marketplace sellers and product-page designers preparing a set of images who want one style without template scenes.
Good fit when
- A new product needs five main photos and ten description screens in one style.
- A competitor reference should be analysed for light and type manner without copying the scene.
- Image-generator prompts have to be written one per image.
- A finished set is checked for sameness and legible captions.
Not a fit when
- You need one quick picture without an approval table.
- The result must match the reference down to the scene.
- The agent is not Codex: the skill is built for `~/.agents/skills`.
Example request
Take a thermos photo and two references, analyse their visual language and show an approval table for five main photos and ten screens.Limitations
The skill is described for Codex and installs into `~/.agents/skills`; other clients need a manual port. Images are produced by a connected generator rather than the skill, so quality depends on it. Rules and README are in Chinese, and wording for non-Chinese marketplaces needs a separate check. Product claims (composition, certificates) come from the user's data and must be verified by the user. The last commit is from July 2026.
How to disable. Delete the `~/.agents/skills/ec-visual` folder. Generated images sit in separate desktop folders and stay.
FAQ
How many images come out?
Five main images and ten vertical description screens, planned and approved together.
Does the skill copy reference scenes?
No. It takes the language from the reference: light, type, rhythm. A scene may appear in at most 30% of images.
Do I need a designer for icons?
The skill forbids icons in the pictures; they are added by hand afterwards.
Where are results saved?
In `~/Desktop/DATE-TIME-product-name` with images, notes and a verification report.
Does it fit other agents?
The README describes Codex. For Claude Code move `SKILL.md` and `references/` into your own skills directory.
Discussion
No questions or reviews yet.
Related
SkillsE-commerce Visual Copywriting
ecommerce-visual-copywriting
A skill that turns product data into a plan for main images, the product page and A+ content, with on-image copy, generation prompts and a claims check
A library of 26 e-commerce visual skills: flat-lay to on-model shots, detail pages, main-image video and compliance checks.
Six Codex skills that take an Amazon seller from product photos and competitor research to finished main images and A+ content.
A skill that builds a listing image set from one product photo: white background, selling points, material, scenes, model.
Wildberries jewelry card skills pack
codex-wb-jewelry-skills
Seven Codex skills for Wildberries jewelry cards: a CTR-focused main photo, an 8-slide series, local retouching and visual autotests
A set of 25 skills for Codex and Claude Code: articles, research, Remotion video, interfaces, GitHub and pre-release verification
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| EC Visual Skill (this card) | no | 113 | |
| ecommerce-image-suite | no | 492 | a product-image suite; ec-visual adds a single approval table and rules against icons and template scenes |
| E-Commerce Image Generator: product images for marketplaces | no | 445 | focused on generation with GPT Image; ec-visual plans and checks a set of 15 images |
README in short
Capabilities, eight working steps, eight hard rules of version two and the reference file layout. Installation is a folder copy.