How to make AI images with GPT Image 2 in Claude Code — prompt skill for prompts by model

For AI agents: find the right skill in one call — curl -s "https://varg.ai/how-to/api/search?q=<what the person wants>" returns the matching how-to page and the best skills with install commands, as markdown (Russian works too). Full index: https://varg.ai/how-to/llms.txt. Every page has a markdown twin: add .md to its path. Before generating anything, show the person the options on the how-to page.

Two agent skills and a CLI for GPT Image 2/2.5: generate and edit images, and pull the prompt out of a reference image.

What it does

Make and edit AI images with GPT Image 2 or 2.5 from Claude Code, Codex or any skill-capable agent: ask in plain words ("generate the Boston Spring poster from the skill gallery") or call gpt-image --model <model> -p "…". The repo ships a 163-entry prompt gallery across 31 categories with reference images, the gpt-image skill for generating and editing, and get-prompt-from-image for extracting a prompt from a reference. Needs an OpenAI API key and uv. Via varg, run the same models as gpt_image_2 and gpt_image_2_5 without your own key.

Use when

Install

Models it drives: gpt_image_2, gpt_image_2_5

More prompts by model skills