skills.rest
anbeime
anbeimeCOMMUNITY
Explore all skills by this author →
GitHub

gpt-image-2-prompt-engine

Generates structured GPT-Image-2 prompts from JSON templates and atomic schemas.

6.0k|578|Updated Feb 2, 2026
Search Tags:#poster#image generation#infographic#Prompt Engineering#json templates#e-commerce design#GPT-Image-2
One-click install
npx skills add https://github.com/anbeime/skill --skill gpt-image-2-prompt-engine
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gpt-image-2-prompt-engine
Source: https://github.com/anbeime/skill/tree/main/skills/gpt-image-2-prompt-engine
Command: npx skills add https://github.com/anbeime/skill --skill gpt-image-2-prompt-engine

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Vague image requests produce inconsistent GPT-Image-2 results with garbled text, wrong aspect ratios, and unstable layouts. This Skill converts fuzzy requirements into structured, reusable, batch-friendly prompts using an atomic schema and industrial JSON templates.

Core Features & Use Cases

  • Atomic Schema Decomposition: Breaks prompts into seven dimensions (Subject, Composition, Material, Typography, Lighting, Style, Constraints) for precise control.
  • 20+ Industrial JSON Templates: Covers UI screenshots, infographics, posters, e-commerce hero images, brand identity, commercial photography, character IP, narrative illustration, and classical Chinese styles.
  • Four-Step Workflow: Select category, retrieve reference cases via the query script, fill template variables, then generate and iterate with image_generation.
  • Use Case: An e-commerce designer needs a product hero image with exact headline text and 9:16 ratio. The Skill fills the ecommerce-hero JSON template, locks the text and ratio constraints, and outputs a ready-to-use prompt.

Quick Start

Ask the AI to use the gpt-image-2-prompt-engine skill to build a GPT-Image-2 prompt for a perfume e-commerce hero image with the headline text locked and a 9:16 aspect ratio.

Frequently Asked Questions about gpt-image-2-prompt-engine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write GPT-Image-2 prompts for e-commerce product images?▼

Use the ecommerce-hero JSON template and fill in product name, material, color, key feature, scene, composition, lighting, and text overlay fields. Add constraints like accurate product rendering, readable text, and the target aspect ratio before calling image_generation.

How to generate infographics with accurate text using GPT-Image-2?▼

Select the infographic-engine template and limit the layout to 3-5 modules with short labels. Write the exact text content in the prompt and add a constraint forbidding garbled or placeholder text, since text locking is the key stability technique.

Does this skill generate images directly?▼

No, it only constructs structured prompts. The actual image output still requires calling the separate image_generation tool with the generated prompt, and reference images can be passed for image-to-image iteration.

Why does GPT-Image-2 output the wrong aspect ratio?▼

The model defaults to 1:1 or 9:16 when the ratio is buried in the prompt. Place the aspect ratio at the very beginning of the prompt or as the first item in the constraints field to enforce it reliably.

When should I not use this prompt engine?▼

Skip it for casual image generation, simple landscape photos, or any request that does not need precise control over text, layout, or composition. Those cases should go directly to basic image_generation without the template workflow.

How do I search the template and case library?▼

Run scripts/query_templates.py with --category for template lookup, --keyword for case search, or --list to see all categories. If the upstream repository is not cloned locally, the script falls back to online raw GitHub links.

Related Skills

Explore More Design & Creative Skills→
Pika-LabsPika-Labs
38

build-a-brand

Build a complete brand identity and 15-page brand guidelines PDF.

Official
Advanced
iterationlayeriterationlayer
2

remove-background-and-generate-product-card

Remove photo backgrounds and generate branded product listing card PNGs.

Official
Intermediate
QuestNova502QuestNova502
4

image-enhancer

Upscale and sharpen images while reducing compression artifacts.

Community
Intermediate
CUexterCUexter
2

stable-diffusion-image-generation

Generate images from text prompts using Stable Diffusion via HuggingFace Diffusers.

Community
Advanced