Skip to content
Back to skills

Image Gen

ASecurity

Generate or edit images with Google Gemini and place them in the site repo. Use when the owner asks to generate, create, or illustrate an image, icon, hero graphic, or artwork for the site.

  • 5 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added September 2, 2026
ai-agentspythongobashnodeapi

Works with

  • api

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 2, 2026

npx -y skills add vibhasjain/agent-keyboard --skill image-gen --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Image Gen?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Image Gen
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vibhasjain-image-gen/badge)](https://www.skillsdirectory.com/skills/vibhasjain-image-gen)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: image-gen
description: Generate or edit images with Google Gemini and place them in the site repo. Use when the owner asks to generate, create, or illustrate an image, icon, hero graphic, or artwork for the site.
---

# image-gen

Generate images with Gemini and drop them into the site checkout. Requires the
`GEMINI_API_KEY` environment variable — it is already in your environment on
deployments that configured it (it is a Fly secret; it never lives in this
repo). If it is unset, say so plainly: "Image generation isn't configured on
this deployment — the owner needs to set the GEMINI_API_KEY secret."

## How

```bash
node ~/.claude/skills/image-gen/generate.mjs "<prompt>" <output.png> [input-image ...]
```

- `<prompt>` — what to generate. Optional input images (paths) are sent along
  for editing/variation.
- `<output.png>` — where the PNG lands. Generate to a neutral temp name first
  (e.g. `/tmp/gen-<random>.png`), then move/rename it into the repo.

After generating: move the image into the repo, reference it from the page,
compress if large (a hero image should be < 300 KB — use `python3` + whatever
is available, or request smaller dimensions in the prompt), commit, push.

## Privacy rules (always apply)

The API provider can see request content. Treat every call as semi-public:

1. **Strip identifying context from prompts.** No names, domains, brand or
   project names, or recognizable proper nouns. Describe the *visual content
   only*: "minimalist illustrated bowl of greens, flat design, white
   background" — not "hero image for <the site>".
2. **Generate to a neutral temp filename** (random id), then rename into the
   repo after generation.

Files in this skill

  • SKILL.md1.7 KB
  • generate.mjs2 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.