Nano Skill Plugin

A Claude Code plugin that generates images using Google's Gemini API (Nano Banana models).

Features

• Text-to-image generation - Describe what you want, get an image
• Multiple models - Fast (flash) or high-quality (pro)
• Aspect ratios - 1:1, 16:9, 9:16, 4:3, 3:4
• Resolution options - Up to 4K with pro model
• Context-aware output - Places images where they belong in your project

Setup

1. Get a Gemini API key at https://aistudio.google.com/api-keys

2. Set the environment variable:

   export GEMINI_API_KEY="your-key-here"

Installation

From Local Path

/plugin marketplace add /path/to/nano-skills/.claude-plugin/
/plugin install nano-skill

From GitHub

/plugin marketplace add livelabs-ventures/nano-skills
/plugin install nano-skill

Plugin Structure

nano-skill/
├── .claude-plugin/
│   └── plugin.json           # Plugin manifest
├── skills/
│   └── gemini-image-generator/
│       ├── SKILL.md          # Skill definition
│       └── scripts/
│           └── generate_image.py
└── README.md

Usage

Once installed, the skill activates when you ask Claude to:

• "Generate an image of..."
• "Create an icon for..."
• "Make a banner..."
• "Design a logo..."

Claude will use the bundled script to generate images and save them to appropriate locations in your project.

Direct Script Usage

# Basic usage
python scripts/generate_image.py "A cute robot" --output ./robot.png

# With aspect ratio
python scripts/generate_image.py "Website banner" --aspect 16:9 --output ./banner.png

# High quality
python scripts/generate_image.py "Detailed art" --model pro --size 4K --output ./art.png

Options

Option	Values	Default	Description
--model, -m	flash, pro	flash	Model selection
--aspect, -a	1:1, 16:9, 9:16, 4:3, 3:4	1:1	Aspect ratio
--size, -s	1K, 2K, 4K	-	Resolution (pro only)

License

MIT