image-gen
Create precise, safe image-generation prompts and iterate against visual requirements.
SKILL.mdskills/design/image-gen/SKILL.md
GitHubImage Gen
Purpose
Create precise, effective prompts for AI image generation models and iterate against visual requirements. The skill ensures generated images match user intent through structured prompt refinement.
When to Use
- The user wants to create AI-generated imagery
- Visual assets are needed for projects
- Prototyping visual concepts
When NOT to Use
- For photography that requires real-world capture
- When stock imagery would be more appropriate
- For content requiring precise technical accuracy
Capabilities
- Prompt writing and refinement
- Style specification (photorealistic, illustrative, abstract, etc.)
- Aspect ratio and resolution selection
- Negative prompt crafting
- Iteration based on feedback
- Multiple model support (DALL-E, Midjourney, Stable Diffusion, etc.)
Inputs
description(required) — what the image should depictstyle(optional) — artistic style preferenceaspect_ratio(optional) — 1:1, 16:9, 9:16, etc.mood(optional) — tone or atmospherereference(optional) — reference image URL or uploadmodel(optional) — preferred generation model
Workflow
- Analyze — Understand user requirements
- Draft — Create initial prompt
- Refine — Add style, quality, and specificity modifiers
- Generate — Submit to image generation API
- Iterate — Refine based on user feedback
- Deliver — Provide image file and prompt used
Tools
- Image generation APIs (DALL-E, Midjourney, Stable Diffusion, etc.)
- Image editing tools for post-processing
Examples
User: "Create a logo for my tech startup" Response:
Prompt: "Minimalist tech startup logo, geometric fox silhouette,
gradient blue to purple, clean lines, modern, professional, white
background, vector style, scalable"
Generating... [██████████] 100%
Output: logo-startup-2026-02-09.png (1024x1024)
Prompt saved for reproducibility.
Safety
- Never generate harmful, violent, or deceptive imagery
- Respect copyright and trademark considerations
- Avoid generating content that could be mistaken for real people
- Label AI-generated images appropriately
Source
Auto-generated from openclawskills.net description.
Notes
Maintained as part of canonical design category.