# Generate AI Images with inference.sh Models

Teams need quick access to many image models without memorizing each CLI command. This skill provides ready prompts for generation, editing, upscaling, and image stitching.

## Install

```bash
npx skillstore add inference-sh/skills-ai-image-generation
```

## Metadata

- Status: approved
- Slug: inference-sh-skills-ai-image-generation
- Skillstore revision: r1
- Version status: missing
- Tree hash: 2541dbb4377bfc113c1653482af66a00a1b7668a5e6642d192b77cd7fb85a028
- Author: inference-sh
- GitHub username: inference-sh
- License: MIT
- Repository: https://github.com/inference-sh/skills/tree/main/tools/image/ai-image-generation/
- Ref: a25199bc7d6b82598536822d1738eb5d5f54025b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network
- Quality score: 69
- Public page: https://skillstore.pages.dev/skills/inference-sh-skills-ai-image-generation
- Manifest: https://skillstore.pages.dev/api/skills/inference-sh-skills-ai-image-generation/manifest

## Capabilities

- Shows belt CLI commands for image generation through inference.sh apps.
- Lists supported image models and their common use cases.
- Provides examples for text-to-image, image editing, upscaling, and stitching.
- Uses source image URLs for edit, upscale, and multi-image workflows.
- Points users to inference.sh documentation for running apps and image examples.

## Use Cases

- Create Campaign Visuals: Generate draft images for ads, social posts, and landing pages using model-specific prompts.
- Explore Product Concepts: Produce product mockups, style variations, and edited source images for early design review.
- Prototype Image Workflows: Test generation, upscaling, and stitching commands before adding image steps to an app workflow.

## Prompt Templates

### Beginner Image Prompt

```
Generate an image of [subject] in [style]. Use a model that fits speed and quality. Ask before running the CLI command.
```

### Product Mockup Prompt

```
Create a product image for [product] with [lighting], [background], and [aspect ratio]. Recommend the model and explain the tradeoffs.
```

### Image Editing Prompt

```
Edit this image URL: [image URL]. Change [specific element] while preserving [important details]. Confirm the URL is safe to send.
```

### Advanced Workflow Prompt

```
Plan a workflow to generate [concept], upscale the best result, and stitch it with [second image]. Include model choices and approval steps.
```

## Limitations

- Requires the inference.sh belt CLI and an authenticated account.
- Sends prompts, source images, and options to external inference services.
- Model availability, pricing, output quality, and parameters can change.
- Does not provide legal, copyright, or brand safety review for generated images.

## Best Practices

- Review prompts and source image URLs before any belt command runs.
- Start with low-cost models for exploration, then use higher-quality models for final assets.
- Keep generated assets labeled with the model, prompt, date, and usage rights review status.

## Anti Patterns

- Do not send confidential images or unreleased product designs without approval.
- Do not assume every model supports the same options or image sizes.
- Do not install related skills or run extra commands without explicit user consent.

## Security Audit

- Audited at: 2026-07-05T17:39:50.94\+00:00
- Summary: The static Ruby backtick findings are false positives caused by Markdown code fences and inline formatting. No prompt injection language was found in SKILL.md. The meaningful risks are the broad belt CLI permission, remote image service calls, and optional related skill installation examples.

## Stats

- Views: 57
- Downloads: 8
- Favorites: 0
- Popularity score: 0