# Run AI Apps from the inference.sh CLI

AI teams often switch between many model APIs and tools. This skill guides Claude, Codex, and Claude Code through inference.sh CLI tasks.

## Install

```bash
npx skillstore add halt-catch-fire/infsh-cli
```

## Metadata

- Status: approved
- Slug: halt-catch-fire-infsh-cli
- Skillstore revision: r1
- Version status: missing
- Tree hash: eee0f0aa483ac3e8f36357258ce2b44f366fbf8ce202bb10773fbddde39c2697
- Author: halt-catch-fire
- GitHub username: halt-catch-fire
- License: MIT
- Repository: https://github.com/halt-catch-fire/skills/tree/main/tools/infsh-cli/
- Ref: 3e4b6c31a74a3bd1a291c98cf585d720cb9fbc88
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: blocked
- Manual install advisory: allowed\_with\_warning
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: network, env\_access, filesystem, external\_commands
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/halt-catch-fire-infsh-cli
- Manifest: https://skillstore.pages.dev/api/skills/halt-catch-fire-infsh-cli/manifest

## Capabilities

- Guides installation and authentication for the belt CLI.
- Finds hosted apps by category, search term, feature status, or app name.
- Generates sample inputs and runs hosted image, video, audio, text, and search apps.
- Tracks long-running tasks and saves generated results through the CLI.
- Supports local media inputs that the CLI uploads for compatible apps.
- Points users to focused skills for images, video, search, LLMs, and Twitter/X automation.

## Use Cases

- Create Media Assets: Generate images, videos, avatars, audio, and 3D assets through hosted model apps.
- Evaluate Model Workflows: Discover apps, inspect schemas, run samples, and compare model outputs from one CLI workflow.
- Run Approved External Actions: Use search apps or approved Twitter/X workflows after confirming prompts, media, and account actions.

## Prompt Templates

### Find an App

```
Find an inference.sh app for image generation. Show the best options and explain which one fits my prompt.
```

### Run a Simple Task

```
Use inference.sh to generate one image from this prompt. Ask before running any command that spends credits.
```

### Use Local Media Safely

```
Prepare an image-to-video run using my local file. Confirm the exact file path with me before upload.
```

### Orchestrate Multi-Step Work

```
Plan a workflow that searches the web, drafts a media prompt, runs a video model, and tracks the task until completion.
```

## Limitations

- Requires an inference.sh account, credentials, and available account credits.
- Requires the belt CLI to be installed and available in the shell.
- Results depend on remote app availability, schemas, rate limits, and provider behavior.
- Local file uploads and social actions need explicit user approval before execution.

## Best Practices

- Confirm app names, inputs, costs, and account-changing actions before running commands.
- Use sample inputs and app schemas before submitting expensive or long-running jobs.
- Review every local file path before allowing the CLI to upload files.

## Anti Patterns

- Do not pipe remote installer scripts into a shell without verification.
- Do not let agents choose local file paths without user approval.
- Do not automate Twitter/X posting or messaging without a final user preview.

## Security Audit

- Audited at: 2026-07-09T07:37:53.836\+00:00
- Summary: The skill is a community CLI integration that intentionally runs external commands and contacts inference.sh services. The remote installer uses curl piped to sh in multiple files, and local file upload plus Twitter/X automation create high-impact side effects that require explicit user confirmation.

## Stats

- Views: 0
- Downloads: 3
- Favorites: 0
- Popularity score: 0