skills.rest
krusemediallc
krusemediallcCOMMUNITY
Explore all skills by this author →
GitHub

chatgpt-image-ad

Generate Meta image-ad creatives via ChatGPT Image 2 through the Arcads external API.

1.5k|357|Updated Apr 1, 2026
Search Tags:#image generation#Marketing#Prompt Engineering#ad creative#Meta ads#GPT-Image-2#arcads api
One-click install
npx skills add https://github.com/krusemediallc/arcads-claude-code --skill chatgpt-image-ad
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: chatgpt-image-ad
Source: https://github.com/krusemediallc/arcads-claude-code/tree/main/skills/chatgpt-image-ad
Command: npx skills add https://github.com/krusemediallc/arcads-claude-code --skill chatgpt-image-ad

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Creating typography-heavy Meta ad creatives (chat threads, comparison tables, fake search results, iOS dialogs) requires precise prompt engineering and repeated manual API calls. This Skill automates the full Arcads gpt-image-2 generation pipeline — presigned reference uploads, prompt safety suffixes, polling, and download — so you get production-ready static ad images from a single brief.

Core Features & Use Cases

  • Locked gpt-image-2 generation: Calls Arcads POST /v2/images/generate with the model hard-locked to gpt-image-2, supporting text-to-image and image_edit modes with up to 5 reference images.
  • Built-in ad safety guards: Automatically appends no-platform-chrome, edge-safe-zone, and glyph-safety suffixes to prevent common failures like garbled text and cropped headlines.
  • Structured workflow: Enforces prompt rewriting from a shared template library, mandatory credit-cost confirmation, and visual QA with up to 2 retries per variant.
  • Use Case: Ask for 3 variants of a ChatGPT-conversation-style ad at 1:1 ratio with your product screenshot as reference; the Skill uploads references, generates, polls, downloads the PNGs, and hands the paths to your Meta ad-builder skill.

Quick Start

Ask the agent to generate a gpt-image-2 Meta image ad from your brief, specifying the aspect ratio, any reference images, and how many variants you want.

Frequently Asked Questions about chatgpt-image-ad

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

FAQPage Schema
How do I generate a ChatGPT Image ad with the Arcads API?▼

Run the generate_image.py script with a rewritten prompt, an aspect ratio (1:1, 16:9, or 9:16), and optional reference images. It handles presigned uploads, submits to POST /v2/images/generate with model gpt-image-2, polls until generated, and downloads the PNGs.

What is the difference between gpt-image-2 and Nano Banana for ad creatives?▼

gpt-image-2 excels at typography-heavy, dense-text, and UI-mimicry ads like chat threads and comparison tables, while Nano Banana is preferred for photoreal and lifestyle creatives. This skill is locked to gpt-image-2; use the nano-banana-image-ad skill for those templates.

What aspect ratios does the Arcads image generation endpoint support?▼

The Arcads /v2/images/generate endpoint only accepts 1:1, 16:9, and 9:16. Templates designed for 2:3, 3:2, or 4:5 must fall back to 1:1 and crop, or use a sibling skill with a backend supporting those ratios.

How many reference images can I use with gpt-image-2?▼

Arcads enforces a hard cap of 5 reference images per gpt-image-2 generation call, returning a 400 error if exceeded. The script validates this limit before submitting and re-uploads references per variant since filePaths are single-use.

Why does my generated ad have garbled text or cropped headlines?▼

Garbled small text is the most common gpt-image-2 failure on dense body copy. The skill appends glyph-safety and edge-safe suffixes by default to prevent this; if defects persist, regenerate with a revised prompt, capped at 2 retries per variant.

Does this skill upload ads to Meta?▼

No, this skill only generates image files and outputs their local paths as JSON lines. Uploading to Meta is handled by a separate Meta ad-builder skill that consumes the generated file paths.

Related Skills

Explore More Marketing & Sales Skills→
gooseworks-aigooseworks-ai
6

create-product-images-higgsfield-product-photoshoot

Generate brand-consistent product stills from Higgsfield photoshoots.

Official
Advanced
fdarkaoufdarkaou
39

meta-ads

Manage Meta Ads campaign monitoring, optimization, creative generation, and upload workflows.

Community
Advanced
vikingokftvikingokft

gws-modelarmor-sanitize-prompt

Sanitize user prompts with a Model Armor template via CLI flags.

Official
Intermediate
hackefellerhackefeller
1

gemini-imagegen

Generate and edit images via the Gemini API with customizable resolutions and aspect ratios.

Official
Intermediate