Gemini Media

cs7-claude-plugins

A Claude Code plugin marketplace providing AI-powered image and video generation using Google's Gemini API.

Plugins

gemini-media

Generate and edit images and videos directly from Claude Code using Google's Gemini models — commonly known as Nano Banana (image generation) and Veo (video generation).

Skills:

Skill	Trigger	Models
generate-image	"generate an image", "edit this image", "refine the image"	Nano Banana (Gemini 3.1 Flash, Gemini 3 Pro, Gemini 2.5 Flash)
generate-video	"generate a video", "animate this image", "extend the video"	Veo 3.1, Veo 3

Image Generation

• Text-to-image generation
• Image editing with up to 14 reference images
• Multi-turn conversational editing with session persistence
• Google Search grounding for real-world subjects
• Automatic thinking mode for complex prompts
• Aspect ratios from 1:1 to 21:9, resolutions up to 4K
• PNG or JPEG output with configurable compression

Video Generation

• Text-to-video generation
• Image-to-video animation
• Frame interpolation between two images
• Video extension (up to 20 extensions per chain)
• Reference images for style/content guidance (up to 3)
• Native audio synthesis
• 720p to 4K, 4-8 second clips
• Async generation with background submission and polling

Installation

Prerequisites

• Claude Code v1.0.33+
• Python 3.7+ (scripts use only stdlib — no pip install needed)
• A Gemini API Key from Google's AI Studio

Add the marketplace

claude plugin marketplace add christian-schlichtherle/cs7-claude-plugins

Install the plugin

claude plugin install gemini-media@cs7-claude-plugins

Set your API key

export GEMINI_API_KEY='your-key-here'

Or create a .env file in the directory where you run Claude Code:

GEMINI_API_KEY=your-key-here

Usage

Once installed, the skills activate automatically when you ask Claude Code to generate media:

> Generate an image of a steampunk cityscape at sunset in 16:9

> Edit the image — add a dirigible in the sky

> Generate a video of a camera slowly flying through the city

The skills handle model selection, prompt optimization, and output management. Generated files are saved to ./generated-images/ and ./generated-videos/ in your working directory and auto-opened on macOS.

Local Development

To test the plugin locally without installing from a marketplace:

claude --plugin-dir ./gemini-media

Or run the scripts directly:

# Image generation
python3 gemini-media/skills/generate-image/scripts/generate_image.py generate \
  --prompt "a steampunk cityscape" \
  --output-dir "./generated-images"

# Video generation
python3 gemini-media/skills/generate-video/scripts/generate_video.py generate \
  --prompt "a camera flies through a steampunk city" \
  --output-dir "./generated-videos"

License

MIT