Skills ai-image-generation
🎨

ai-image-generation

Low Risk ⚙️ External commands🌐 Network access

Generate AI images with 50+ models

Also available from: inference-sh

Create professional AI images without learning multiple platforms. Access FLUX, Gemini, Grok, Seedream, and Reve through one unified CLI interface.

Supports: Claude Codex Code(CC)
🥉 72 Bronze
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "ai-image-generation". Professional product photo of a coffee mug, studio lighting

Expected outcome:

High-quality product image showing a coffee mug with professional studio lighting, clean background, suitable for e-commerce listings

Using "ai-image-generation". Cyberpunk city at night, neon lights, rain-slicked streets

Expected outcome:

Atmospheric cyberpunk cityscape with glowing neon signs, reflective wet streets, and futuristic architecture in 16:9 aspect ratio

Security Audit

Low Risk
v1 • 2/25/2026

Static analysis flagged 47 patterns in SKILL.md documentation file. All findings are FALSE POSITIVES - the file is markdown documentation containing CLI usage examples, not executable code. Backtick patterns are markdown code block formatting. URLs are documentation links. The skill requires restricted Bash(infsh *) access for the inference.sh CLI, which is a legitimate external command pattern with no command injection risk.

1
Files scanned
140
Lines analyzed
3
findings
1
Total audits
Low Risk Issues (1)
Markdown Documentation Patterns Detected as Code
Static analyzer detected shell command patterns in markdown documentation. SKILL.md is a documentation file containing CLI usage examples in markdown code blocks. The backtick characters are markdown formatting syntax, not executable Ruby/shell code. No actual code execution occurs from this file.
Audited by: claude

Quality Score

38
Architecture
100
Maintainability
87
Content
50
Community
88
Security
91
Spec Compliance

What You Can Build

Marketing and Social Media Content

Create eye-catching visuals for social media posts, ads, and marketing campaigns without hiring a designer.

Product Mockups and Prototyping

Generate product visualization concepts and mockups for client presentations and design reviews.

Concept Art and Storyboarding

Rapidly visualize characters, environments, and scenes for games, films, and creative projects.

Try These Prompts

Basic Image Generation
Generate an image of a cat astronaut floating in space with stars in the background
Product Photography Style
Create a professional product photo of a coffee mug on a wooden table with studio lighting and shallow depth of field
Cinematic Landscape
Generate a cinematic landscape with mountains and a lake at golden hour, photorealistic style, 4K quality, dramatic lighting
Text-in-Image Design
Create a poster design that displays the text HELLO WORLD in bold modern typography with a minimalist background

Best Practices

  • Write detailed prompts including subject, style, lighting, and composition for best results
  • Choose the appropriate model for your use case: FLUX for versatility, Seedream for quality, Reve for text
  • Start with lower-resolution models for testing, then upscale final selections with Topaz

Avoid

  • Avoid vague prompts like make it look good without specifying what good means for your use case
  • Do not use the service for generating copyrighted or trademarked content without proper authorization
  • Avoid expecting photorealistic human faces for commercial use due to potential uncanny valley effects

Frequently Asked Questions

How do I install the inference.sh CLI?
Run curl -fsSL https://cli.inference.sh | sh in your terminal, then run infsh login to authenticate with your account.
Which model should I use for my project?
Use FLUX Dev for general purpose generation, Seedream 4.5 for highest quality 4K images, Reve for images with text, and Grok Imagine for multiple aspect variations.
Can I use generated images commercially?
Check the inference.sh terms of service and the specific model license. Most models allow commercial use but restrictions may apply.
How do I upscale low-resolution images?
Use the Topaz upscaler model by running infsh app run falai/topaz-image-upscaler with your image URL as input.
What image formats are supported?
The service typically outputs PNG or JPEG formats. Input images for editing or upscaling should be in common web formats like JPG or PNG.
How much does image generation cost?
Pricing varies by model and resolution. Check the inference.sh pricing page for current rates. Some models offer free tier usage with limits.

Developer Details

File structure

📄 SKILL.md