generate-image · gupsammy — Claude Code skill: create and edit images
generate-image — a Claude Code skill that creates and edits images through generate.py — logos, stickers, product mockups, t2i, i2i and multi-reference composites. It kicks in when you need a finished visual, not markup.
no license ★ 276 updated 2026-10-29
/plugin marketplace add gupsammy/Claudest
/plugin install claude-content@ClaudestWhat generate-image does
The skill runs in Claude Code and handles image work end to end: text-to-image, edits driven by a reference, and frames built from several sources. It calls a Python script, generate.py, through uv, and has read access plus the ability to ask you questions. HTML/CSS mockups, diagrams and coded UI are outside its scope, so those requests pass it by.
Two things have to be in place: the GEMINI_API_KEY environment variable and the uv package manager. The flow is staged. The skill first pins down the mode and parameters — model, aspect ratio, resolution, output path — then writes the prompt, shows it to you for approval, and only then runs the script. For shots where pose or exact cropping matters, it suggests --batch 3 or --batch 4; image generation is stochastic, and a precise directive lands roughly half the time. Anything you want kept out of the frame gets described as a positive alternative, since naming it under a negation pulls the model toward it.
With no path given, output lands in ~/Documents/generated images/, and each image is paired with a .md file recording the prompt and model. Prompts are prose, not tag lists — the model reads a scene with spatial and mood context that comma-separated keywords drop. On hard geometry, adding every available constraint backfires: the skill locks only the attribute that is drifting.
What it helps with
- generate a logo or sticker from a written description
- edit a photo while changing a single element
- compose a shot from several reference images
- produce a product mockup for a listing
- land a precise frame by batching several variants
Example prompts for the agent
- Generate a minimalist logo for a coffee shop
- Replace the background with a cool-neutral gray, leave the person untouched
- Compose a shot: the shirt from Image 2, the pose from Image 3, studio lighting
How to install generate-image
1Plugin marketplace
/plugin marketplace add gupsammy/Claudest/plugin install claude-content@ClaudestThis installs the whole claude-content plugin with all its skills
2Skills CLI (Vercel)
npx skills add gupsammy/Claudest --skill generate-image -a claude-code3Manually: put the files in place
~/.claude/skills/generate-image/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/generate-image/
1Skills CLI (Vercel)
npx skills add gupsammy/Claudest --skill generate-image -a codex2Manually: put the files in place
.agents/skills/generate-image/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add gupsammy/Claudest --skill generate-image -a cursor2Manually: put the files in place
.cursor/skills/generate-image/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install gupsammy/Claudest plugins/claude-content/skills/generate-image2Manually: put the files in place
.github/skills/generate-image/SKILL.md1Skills CLI (Vercel)
npx skills add gupsammy/Claudest --skill generate-image -a gemini-cli2Platform CLI
gemini skills install https://github.com/gupsammy/Claudest --path plugins/claude-content/skills/generate-imageManually: put the files in place
.devin/skills/generate-image/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.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 generate-image
How do I install generate-image in Claude Code?
Add the marketplace with /plugin marketplace add gupsammy/Claudest, then install the plugin: /plugin install claude-content@Claudest.
What does generate-image do?
It creates and edits images through generate.py: t2i, i2i and multi-reference composites. It runs inside Claude Code.
What does the skill need to work?
The GEMINI_API_KEY environment variable and the uv package manager.
Similar skills
All skills for AI and ML →generate-image by other authors comes first.