Skills ai-image-generation
📦

ai-image-generation

Safe ⚙️ External commands🌐 Network access

Generate AI images with FLUX, Gemini, Grok, and 50+ models

Also available from: inference-sh-9,infsh-skills,inference-sh-8,skillssh,inferen-sh,inference-sh,tool-belt,toolshell,tul-sh

Create stunning visuals without switching between AI tools. Access 50+ professional image models through a single CLI interface optimized for Claude, Codex, and Claude Code workflows.

Supports: Claude Codex Code(CC)
📊 70 Adequate
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". A minimalist logo design for a tech startup, blue and white color scheme, clean geometric shapes

Expected outcome:

Generated a 1024x1024 image featuring a modern geometric logo with interlocking blue shapes on white background, suitable for branding and marketing materials

Using "ai-image-generation". Fantasy castle on a cliff overlooking the ocean at sunset, dramatic clouds, epic scale

Expected outcome:

Created a 1920x1080 landscape showing a medieval castle silhouette against orange and purple sunset sky, waves crashing below, cinematic composition with depth and atmosphere

Using "ai-image-generation". Product mockup of a smartphone displaying a fitness app interface on a marble surface

Expected outcome:

Generated a professional product photo showing an iPhone on white marble with natural shadows, screen displaying a fitness dashboard with charts and metrics, studio quality lighting

Security Audit

Safe
v1 • 3/12/2026

Static scanner flagged 44 potential issues (external_commands, network URLs, weak crypto). Evaluation reveals all are false positives: backticks are documentation examples, URLs are legitimate service endpoints, and 'Seedream' triggered crypto detector. This is a pure documentation skill with no executable code.

1
Files scanned
141
Lines analyzed
2
findings
1
Total audits
Audited by: claude

Quality Score

38
Architecture
100
Maintainability
87
Content
23
Community
100
Security
91
Spec Compliance

What You Can Build

Product Marketing Visuals

Create professional product mockups and marketing images for e-commerce listings, social media posts, and advertising campaigns using high-quality models.

Concept Art and Prototyping

Rapidly visualize design concepts, character designs, and scene mockups for games, films, and creative projects with text-to-image generation.

Content Creation and Illustration

Generate custom illustrations, blog headers, and social media graphics with specific styles and text rendering capabilities.

Try These Prompts

Basic Text-to-Image
Generate a professional product photo of a coffee mug on a wooden table with studio lighting, warm tones
Styled Portrait Generation
Create a cinematic portrait of a woman at golden hour, dramatic lighting, shallow depth of field, 85mm lens aesthetic
Text Rendering Poster
Design a poster that displays 'WELCOME' in bold modern typography, neon colors, cyberpunk background with city lights
Advanced Scene Composition
Generate a photorealistic landscape with mountains reflecting in a calm lake at sunrise, misty atmosphere, wide angle perspective, National Geographic style

Best Practices

  • Use specific, descriptive prompts with clear subject, style, lighting, and composition details for better results
  • Start with faster models like FLUX.2 Klein for iterations, then use high-quality models like Seedream 4.5 for final outputs
  • Include aspect ratio parameters early in the prompt to avoid regenerating images with different dimensions

Avoid

  • Avoid vague prompts like 'make it pretty' - specify concrete visual elements and styles instead
  • Do not expect perfect text rendering on all models - use Reve or Seedream 3.0 when text accuracy is critical
  • Avoid sending sensitive or private information in prompts as images are processed by external AI services

Frequently Asked Questions

How do I install the inference.sh CLI required for this skill?
Run `npx skills add inference-sh/skills@agent-tools` to install the CLI. This provides authentication setup and the `infsh` command needed for image generation.
Which model should I use for the highest quality images?
Seedream 4.5 and ImagineArt 1.5 Pro offer the highest quality with 4K resolution support. Use Seedream 3.0 when accurate text rendering is required.
Can I edit existing images with this skill?
Yes. Use the Reve model for natural language image editing, or provide an image URL with modification instructions for image-to-image generation.
What aspect ratios are supported?
Most models support common ratios like 16:9, 4:3, 1:1, 3:4, and 9:16. Specify the aspect ratio in your prompt parameters for best results.
Are there rate limits for image generation?
Rate limits depend on your inference.sh account tier. Free accounts have daily generation limits while paid plans offer higher quotas and priority processing.
Can I use generated images for commercial purposes?
Commercial usage rights depend on the specific model and your inference.sh subscription. Check the model documentation and your account terms for licensing details.

Developer Details

File structure

📄 SKILL.md