Creo image generation

Skill oyusypenko/creo/extensions/image-generation/skills/creo-image-generation

Generate marketing images using DALL-E 3 or ComfyUI (Stable Diffusion XL). Batch generation, optimization, cost estimation. Requires image-generation extension. Triggers on: /creo image-generation, generate images, DALL-E, ComfyUI.From its SKILL.md

Install

npx -y skills add oyusypenko/creo --skill creo-image-generation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

reads credentialsReads from 1 credential source: `OPENAI_API_KEY`.
3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
runs commandsInstructs the agent to run 7 commands, including `node index.js generate --estimate` and 6 more.

SKILL.md

3.6 KB, 838 tokens by cl100k_base, as published. Nobody here has run it

Creo Image Generation

Generate, optimize, and manage marketing images using AI (DALL-E 3 or ComfyUI with Stable Diffusion XL).

Commands

Command	Description
`/creo image-generation generate`	Generate all images using DALL-E 3
`/creo image-generation generate --seo-only`	Generate only SEO/OG images
`/creo image-generation generate --optimize-only`	Optimize existing images (resize, compress)
`/creo image-generation estimate`	Show cost estimation without generating
`/creo image-generation optimize`	Optimize all images in output directory
`/creo image-generation comfyui`	Generate images using local ComfyUI (SDXL)
`/creo image-generation comfyui landing`	Generate only landing page images via ComfyUI

How It Works

DALL-E 3 (Cloud)

Reads image configuration from paths.js (or uses defaults)
Loads prompts from a prompts.js file (set via PROMPTS_FILE env var)
Generates images via the OpenAI DALL-E 3 API
Skips images that already exist on disk
Optionally generates SEO overlay images
Optimizes all images (resize, compress with sharp)

ComfyUI (Local)

Requires a running ComfyUI instance (see comfyui/SETUP.md)
Sends workflows to ComfyUI via its REST/WebSocket API
Uses Stable Diffusion XL models for photorealistic output
Supports batch generation by category

Environment Variables

Variable	Required	Description
`OPENAI_API_KEY`	Yes (DALL-E)	OpenAI API key for DALL-E 3 generation
`IMAGE_CONFIG_DIR`	No	Directory containing `paths.js` config
`IMAGE_OUTPUT_DIR`	No	Output directory for generated images
`PROMPTS_FILE`	No	Path to prompts.js with generation prompts
`APP_NAME`	No	Application name for SEO image overlays
`COMFYUI_URL`	No	ComfyUI server URL (default: http://127.0.0.1:8000)

Cost Reference (DALL-E 3)

Quality	Size	Cost per Image
Standard	1024x1024	$0.040
HD	1792x1024	$0.080

Always run --estimate first to review costs before generating.

Library Reference

The extension code lives in the lib/ directory:

lib/ImageGenerator.js -- Main generator class. Handles DALL-E 3 API calls, SEO image generation, and optimization.
lib/PathManager.js -- Manages image path generation from configuration. Supports dynamic config parsing.
lib/ConfigParser.js -- Parses TypeScript image configuration files to extract categories and image types.
lib/utils.js -- Utility functions: file existence checks, cost calculation, file stats, sleep.
comfyui/generate-batch.js -- ComfyUI batch generator. Sends workflows via WebSocket, supports category filtering.

Usage Examples

Generate with cost estimate first

# Check cost
node index.js generate --estimate

# Generate all
OPENAI_API_KEY=sk-... node index.js generate

# Generate HD quality
OPENAI_API_KEY=sk-... node index.js generate --quality=hd

ComfyUI local generation

# Start ComfyUI first, then:
node comfyui/generate-batch.js
node comfyui/generate-batch.js landing
node comfyui/generate-batch.js features audio

Optimize existing images

node index.js generate --optimize-only

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.