# Create AI Videos with Inference Models

Choosing and invoking video models can require extensive platform knowledge. This skill provides model guidance and adaptable belt CLI workflows.

## Install

```bash
npx skillstore add infsh-skills/ai-video-generation
```

## Metadata

- Status: approved
- Slug: infsh-skills-ai-video-generation
- Skillstore revision: r2
- Version status: missing
- Tree hash: 6ff6ed94f18f8582cacfb9c699a6aa7e0de770ca8d98cf92d388d2a9a0b258ea
- Author: infsh-skills
- GitHub username: infsh-skills
- License: MIT
- Repository: https://github.com/infsh-skills/skills/tree/main/tools/video/ai-video-generation/
- Ref: 4121de961d1b6f2ffca856260e239505c302452c
- 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: 50
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/infsh-skills-ai-video-generation
- Manifest: https://skillstore.pages.dev/api/skills/infsh-skills-ai-video-generation/manifest

## Capabilities

- Recommends model identifiers for text, image, reference, avatar, and editing workflows.
- Prepares belt CLI commands for hosted video generation applications.
- Configures prompts, duration, resolution, generated audio, and media references.
- Guides lipsync and talking-head generation from portrait and audio URLs.
- Supports video upscaling, foley generation, and clip merging workflows.

## Use Cases

- Produce Campaign Clips: Select a model and prepare short promotional video generation inputs for social channels.
- Animate Product Images: Turn an existing product image into a controlled motion sequence with optional audio.
- Build Talking Avatars: Combine authorized portrait and narration assets for lipsynced presenter videos.

## Prompt Templates

### Choose a Video Model

```
Recommend one model for a [duration] video showing [subject and action]. Prioritize [speed, quality, cost, or audio]. Explain the choice.
```

### Create a Text-to-Video Workflow

```
Prepare a belt workflow for a [duration] [resolution] video. Scene: [description]. Include camera movement, lighting, style, and audio requirements.
```

### Animate a Reference Image

```
Use [authorized image URL] to create a video with [motion]. Preserve [identity or product details]. Add [audio requirement] and recommend settings.
```

### Design a Multi-Step Video Pipeline

```
Plan a pipeline that generates [scenes], adds [foley or narration], upscales output, and merges clips. State model choices, dependencies, privacy checks, and approval points.
```

## Limitations

- Requires the belt CLI, an inference.sh account, authentication, and network access.
- Model availability, output quality, processing time, and pricing depend on external providers.
- Examples use remotely accessible media URLs and do not explain local file upload.
- The skill does not verify consent, copyright, privacy, or acceptable-use compliance.

## Best Practices

- Confirm the selected provider, estimated cost, and command before running any generation.
- Use only media with documented consent and rights, especially portraits and voice recordings.
- Start with short, low-resolution tests before requesting longer or higher-resolution outputs.

## Anti Patterns

- Do not submit confidential media through public URLs or unknown hosting services.
- Do not install unpinned skills or packages without reviewing their source and version.
- Do not generate realistic people or voices without consent, attribution, and policy review.

## Security Audit

- Audited at: 2026-08-06T11:53:25.81\+00:00
- Summary: The skill contains genuine external command and network behavior through documented belt and npx workflows. Most static backtick matches are Markdown fences or model identifiers, but command examples, third-party installation, remote image loading, and external media processing remain relevant risks.

## Stats

- Views: 110
- Downloads: 11
- Favorites: 0
- Popularity score: 0