imagegen

Use it when:: A project needs a raster asset, new or edited: a photo, illustration, texture, sprite, mockup or transparent cutout.
It does:: Decides between a new image and an edit, labels each input image's role, rewrites your prompt as a labelled spec, generates, inspects and changes one thing at a time.
You get:: The chosen image copied into your workspace, plus the final prompt and which mode made it; an existing file gets a versioned sibling, not an overwrite.
Two modes:: Codex's built-in image_gen tool by default; a bundled Python CLI with generate, edit and batch commands only when you ask for it.
Skips:: Simple shapes, diagrams and icons better made in SVG, HTML/CSS or canvas, and anything that should match an existing icon set or logo system in the repo.

Made for Codex's built-in image_gen tool; in another agent you must ask for its CLI mode, which needs an OPENAI_API_KEY. Prompts and input images go to OpenAI.

TL;DR

OpenAI's image-generation skill, a system skill that the latest Codex installs automatically. The agent first decides whether the request is a new image or an edit (reference images alone mean a new one) and whether it needs one asset or many, then collects the prompt, exact text, constraints and input images, labelling each image as a reference, the edit target or a compositing input. The prompt becomes a short labelled spec: a use case from a fixed list of 16 (product mockup, logo, style transfer, background extraction and others), asset type, subject, style, composition, lighting, palette, verbatim text, constraints and things to avoid. Detail is added only to generic prompts, never extra characters, brands or slogans, and edits restate what must stay unchanged on every pass. After each image it checks subject, style, composition, text accuracy and constraints, then iterates with one targeted change. Only the selected finals are copied into the workspace; previews are just shown inline. The CLI fallback, scripts/image_gen.py, needs the openai Python package and an OPENAI_API_KEY, and writes under output/imagegen/; the agent never switches to it on its own. The openai/skills repository is marked deprecated; its README points to openai/plugins for current Codex skills.

Summarizes the skill at 49f948f.

npx skills add https://github.com/openai/skills --skill imagegen

skills.sh page Source on GitHub All skills from openai/skills

Total Downloads: 6,100 on skills.sh, Nov 9, 2026
Repo stars: 28,602 openai/skills
First seen: Jan 28, 2026 on skills.sh

Weekly Downloads

207

-2% over the 8 weeks to Nov 9

Weekly values

Week to	Installs
Sep 21	211
Sep 28	212
Oct 5	282
Oct 12	187
Oct 19	205
Oct 26	222
Nov 2	226
Nov 9	207

Category	category:media Generate and edit video, images and audio.
Type	type:workflow A multi-step process the agent follows: interview, debug, run a TDD loop, review.
Effect	effect:local Its own steps create or change files, git state or local tools.
Needs	needs:account An account, API key or paid plan with an outside service.
Publisher	by:official A verified publisher on skills.sh, or the company or maintainer behind the product the skill operates.
Topics	image ai-generation
Stack	stack:openai stack:codex

SKILL.md

No copy of its SKILL.md yet.