gpt-image · 101-skills — Claude Code skill: generate and edit images with GPT
Skill that connects OpenAI GPT-Image-2 to your AI coding agent. Text-to-image, reference-based editing, mask inpainting via inference.sh CLI.
npx skills add 101-skills/superpowers --skill gpt-image -a claude-codeWhat gpt-image does
The skill gives your agent access to GPT-Image-2 from OpenAI — a model that generates images from text, edits them using reference images, and performs mask-based inpainting. You can generate up to 10 images at once, set any resolution in 32px increments (256 to 4096), choose quality (low, medium, high) and format (PNG, JPEG, WebP). For editing, pass a reference image and a prompt — the model changes background, objects or style. Inpainting requires a mask in PNG.
Runs through belt, the inference.sh CLI. Without belt installed, the skill will fail: the agent calls belt app run openai/gpt-image-2 with JSON parameters, and if the command is missing, you get an error. Pricing depends on quality: low around $0.006 per image, medium $0.024, high $0.21. Larger resolutions cost more.
What it helps with
- generate an image from a text description
- swap background or objects in a photo using a reference
- replace a region in an image using a mask
- create multiple visual variants in one request
- get an image at a custom resolution
Example prompts for the agent
- Generate an image: sneakers on white background, studio lighting, high quality
- Change the background in this photo to a beach at sunset
- Create 4 logo variants for a coffee shop
How to install gpt-image
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill gpt-image -a claude-code2Manually: put the files in place
~/.claude/skills/gpt-image/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/gpt-image/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill gpt-image -a codex2Manually: put the files in place
.agents/skills/gpt-image/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill gpt-image -a cursor2Manually: put the files in place
.cursor/skills/gpt-image/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install 101-skills/superpowers gpt-image2Manually: put the files in place
.github/skills/gpt-image/SKILL.md1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill gpt-image -a gemini-cli2Platform CLI
gemini skills install https://github.com/101-skills/superpowers --path tools/image/gpt-imageManually: put the files in place
.devin/skills/gpt-image/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill gpt-image -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about gpt-image
How to install gpt-image in Claude Code?
Run npx skills add 101-skills/superpowers --skill gpt-image -a claude-code, then open ~/.claude/skills/gpt-image/SKILL.md.
What does gpt-image do?
The skill connects your agent to OpenAI GPT-Image-2: text-to-image generation, reference-based editing, mask inpainting, batch generation up to 10 images.
What does gpt-image require to work?
The belt CLI from inference.sh must be installed. The skill calls belt app run openai/gpt-image-2 — without that command, it will not work.
Similar skills
All skills for AI and ML →gpt-image by other authors comes first.
