From google-image
Generate images using AI. Use for: create an image, generate a picture, draw, illustrate, visualize, make an image, text to image, ai art, design a logo, render, sketch.
How this skill is triggered — by the user, by Claude, or both
Slash command
/google-image:google-imageThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
**After calling `generate_image`, ALWAYS use the Read tool on the returned `file_path` to display the image inline to the user.**
After calling generate_image, ALWAYS use the Read tool on the returned file_path to display the image inline to the user.
This skill provides the generate_image tool for AI image generation with Google Gemini.
generate_imageGenerate an image from a text prompt using Google Gemini.
Use for:
Parameters:
prompt (required): Image description (5-2000 chars)output_dir (optional): Custom output directory (default: generated-images/)aspect_ratio (optional): 1:1 (default), 16:9, 9:16, 4:3, 3:4image_size (optional): Output resolution — 1K (default, ~1024px), 2K (~2048px), 4K (ultra HD)model (optional): gemini-2.5-flash-image (default, fast) or gemini-3-pro-image-preview (higher quality, reasoning-driven)Latency: ~5-20 seconds
generate_image with the promptfile_pathfile_path to display the image inlineUse generate_image with prompt: "A serene mountain lake at sunset with reflections"
Use generate_image with:
prompt: "Panoramic view of a futuristic city skyline"
aspect_ratio: "16:9"
Use generate_image with:
prompt: "Detailed botanical illustration of a rare orchid"
image_size: "4K"
Use generate_image with:
prompt: "Logo design for a coffee shop"
output_dir: "assets/images"
Use generate_image with:
prompt: "Photorealistic portrait in oil painting style"
model: "gemini-3-pro-image-preview"
Returns JSON:
Requires GEMINI_API_KEY environment variable. Get your key at:
https://aistudio.google.com/apikey
npx claudepluginhub yteruel31/claude-marketplace --plugin google-imageCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.