From graphics-studio
Generates, edits, and composes images using Zenmux API with Google GenAI. Supports text-to-image, image editing, collages, aspect ratios (1:1 to 21:9), and resolutions (1K-4K). Python CLI with uv.
How this skill is triggered — by the user, by Claude, or both
Slash command
/graphics-studio:zenmux-image-genThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Image generation, editing, and composition using Zenmux API with Google GenAI protocol.
Image generation, editing, and composition using Zenmux API with Google GenAI protocol.
Required environment variables:
export ZENMUX_API_BASE="your-api-base-url"
export ZENMUX_API_KEY="your-api-key"
export ZENMUX_MODEL_IMAGE="your-model-id"
This skill uses uv for dependency management. Before running any commands:
# Navigate to skill directory
cd /path/to/zenmux-image-gen
# Activate virtual environment
source .venv/bin/activate
# Or use uv run directly (auto-activates venv)
uv run python scripts/image_gen.py --help
If .venv doesn't exist, initialize it:
uv sync
# Activate venv first
source .venv/bin/activate
# Generate image
python scripts/image_gen.py generate "A serene mountain landscape at sunset" --ratio 16:9 --resolution 2K
# Or use uv run (no manual activation needed)
uv run python scripts/image_gen.py generate "A serene mountain landscape at sunset" --ratio 16:9 --resolution 2K
uv run python scripts/image_gen.py generate "A futuristic cityscape" --ratio 16:9 --resolution 2K
uv run python scripts/image_gen.py edit "Remove the background and replace with a beach scene" source.png
uv run python scripts/image_gen.py compose "Blend these images into a collage" image1.png image2.png image3.png
| Ratio | Use Case |
|---|---|
| 1:1 | Social media avatars, icons |
| 2:3 | Portrait photos, posters |
| 3:2 | Landscape photos |
| 4:3 | Presentations, traditional displays |
| 3:4 | Mobile wallpapers |
| 9:16 | Mobile stories, vertical video |
| 16:9 | Widescreen, presentations, banners |
| 21:9 | Ultra-wide banners, cinematic |
| Resolution | Description |
|---|---|
| 1K | Standard quality (~1024px on longest side) |
| 2K | High quality (~2048px on longest side) |
| 4K | Ultra high quality (~4096px on longest side) |
from scripts.image_gen import generate_image
result = generate_image(
prompt="A futuristic cityscape with flying cars",
aspect_ratio="16:9",
resolution="2K",
output_path="./output",
filename="city.png"
)
from scripts.image_gen import edit_image
result = edit_image(
prompt="Change the sky to a dramatic sunset",
source_image="landscape.png",
resolution="2K"
)
Common edit operations:
from scripts.image_gen import compose_images
result = compose_images(
prompt="Create a seamless collage with vintage aesthetic",
source_images=["photo1.png", "photo2.png"],
aspect_ratio="16:9",
resolution="2K"
)
image_YYYYMMDD_HHMMSS.png--output or output_path--filename or filenameAutomatic retry: 1 retry on failure with 1s delay. Returns detailed error message if both attempts fail.
import subprocess
import os
skill_dir = "/path/to/zenmux-image-gen"
# Use uv run for isolated execution
subprocess.run([
"uv", "run", "--directory", skill_dir,
"python", "scripts/image_gen.py", "generate",
"Abstract gradient background",
"--ratio", "21:9",
"--output", "./public/images",
"--filename", "hero-bg.png"
])
image_gen.py generate <prompt> [options]
image_gen.py edit <prompt> <source> [options]
image_gen.py compose <prompt> <sources...> [options]
Options:
--ratio, -r Aspect ratio (1:1, 2:3, 3:2, 4:3, 3:4, 9:16, 16:9, 21:9)
--resolution, -res Resolution (1K, 2K, 4K)
--output, -o Output directory
--filename, -f Output filename
npx claudepluginhub jorben/jorben-skills --plugin graphics-studioGenerates, edits, and composes images using Google's Gemini 3 Pro Image model. Supports text-to-image, multi-image composition, aspect ratio control, and search-grounded generation for data visuals.
Generate and edit images using Google's Gemini API via Python scripts. Supports text-to-image, image editing, multi-turn refinement, custom resolutions, and aspect ratios.
Generates images from text, edits existing images, applies style transfers, composes from multiple references, and supports multi-turn refinement using Google's Gemini API via Python scripts. For logos, stickers, mockups.