# GPT-Image2-Skill

A prompt gallery, image-prompt library, and two agent skills plus a CLI for generating and editing images with OpenAI's GPT Image 2 and 2.5 models.

GPT-Image2-Skill keeps a curated GPT Image 2 prompt collection and gallery, 163 numbered entries across 31 categories as of this material, alongside GPT Image 2.5 API support and task-specific reference material. Two skills handle image generation and editing, and image-to-prompt extraction, and load into Claude Code, Codex, OpenClaw, Hermes Agent and other skill-capable agent runtimes, alongside a standalone CLI.

For the CLI, a user exports reference PDF pages as PNG, WebP or JPG and attaches them with `-i`; the underlying Image API outputs PNG, JPEG or WebP, and turning a result into editable SVG or PowerPoint shapes needs a separate authoring step. A contributed 'Get Prompt from Image' feature uses a vision-capable agent to extract a prompt from a reference image and pass it to a generator.

It is written in Python and licensed under MIT.

Website: https://github.com/wuyoscar/GPT-Image2-Skill