# Run Cloud AI Apps with inference.sh

Teams need a consistent way to discover and run hosted AI apps from agent workflows. This skill gives Claude, Codex, and Claude Code practical infsh CLI guidance for discovery, execution, tasks, and authentication.

## Install

```bash
npx skillstore add inf-sh/agent-tools
```

## Metadata

- Status: approved
- Slug: inf-sh-agent-tools
- Skillstore revision: r1
- Version status: missing
- Tree hash: 49f0cd2c0c1b64e7a777d2a41ef01f2c2c755add79322419c2bdfbb9545e98d0
- Author: inf-sh
- GitHub username: inf-sh
- License: MIT
- Repository: https://github.com/inf-sh/skills/tree/main/skills/agent-tools/
- Ref: 30c73eac2afe762f6aa9c4553158769369d47351
- 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/inf-sh-agent-tools
- Manifest: https://skillstore.pages.dev/api/skills/inf-sh-agent-tools/manifest

## Capabilities

- Install and authenticate the inference.sh CLI for local use.
- List, search, filter, and inspect available inference.sh apps.
- Run hosted AI apps with inline input or saved input files.
- Generate sample input files before running an app.
- Track long-running tasks and save task results.
- Reference app categories for image, video, LLM, search, 3D, and Twitter/X workflows.

## Use Cases

- Run model experiments from the terminal: Use infsh commands to find apps, prepare input, run jobs, and collect results for model testing.
- Create media assets with hosted generators: Discover image, video, audio, and 3D apps, then run generation jobs without managing GPUs.
- Compare search and LLM providers: Use catalog search and app details to select hosted search or LLM tools for a workflow.

## Prompt Templates

### Find an app

```
Use the inference.sh skill to find apps for image generation. Show the safest discovery steps before running anything.
```

### Prepare a run

```
Help me prepare an infsh run for a video generation app. Generate the sample input first and explain what I must review.
```

### Track a long task

```
Submit a long-running inference.sh app only after my approval. Then show how to check task status and save the result.
```

### Compare provider options

```
Compare available LLM and search apps for my workflow. Include authentication needs, data-sharing risks, costs to verify, and recommended commands.
```

## Limitations

- Requires an inference.sh account, CLI access, and valid authentication.
- Remote apps may send prompts, files, and results to external providers.
- Installer examples need manual verification before use.
- Twitter/X actions can affect public accounts and require careful approval.

## Best Practices

- Confirm the app name, provider, input data, and expected cost before running remote work.
- Use sample input files and review them before submitting prompts or files.
- Require explicit approval before installation, related-skill setup, or Twitter/X actions.

## Anti Patterns

- Do not pipe remote installer scripts into a shell without verification.
- Do not paste secrets, private files, or regulated data into remote app inputs without approval.
- Do not automate posts, messages, follows, likes, or reposts without final human review.

## Security Audit

- Audited at: 2026-07-05T15:12:03.322\+00:00
- Summary: The skill is legitimate CLI documentation for inference.sh, but it includes confirmed remote shell installer instructions and many external infsh commands. Most hardcoded documentation links and API-key mentions are false positives, while cloud execution and Twitter/X automation remain meaningful operational risks. No prompt injection text was found in the reviewed files.

## Stats

- Views: 260
- Downloads: 10
- Favorites: 0
- Popularity score: 0