# Generate Pruna Videos from Text and Images

Creating video model commands requires choosing compatible inputs and settings. This skill guides Pruna P-Video and WAN generation through the inference.sh CLI.

## Install

```bash
npx skillstore add inferen-sh/p-video
```

## Metadata

- Status: approved
- Slug: inferen-sh-p-video
- Skillstore revision: r2
- Version status: missing
- Tree hash: aa69f487212cbb4c0989d605faf61dbabb6ad68e46741f38820d45591b3c0c1e
- Author: inferen-sh
- GitHub username: inferen-sh
- License: MIT
- Repository: https://github.com/inferen-sh/skills/tree/main/tools/video/p-video/
- 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: 78
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/inferen-sh-p-video
- Manifest: https://skillstore.pages.dev/api/skills/inferen-sh-p-video/manifest

## Capabilities

- Runs P-Video text-to-video and image-to-video generation through the belt CLI.
- Uses WAN-T2V for text-based clips at supported resolutions.
- Uses WAN-I2V to animate a supplied still image.
- Adds a supplied audio URL to supported P-Video requests.
- Configures duration, resolution, and draft mode in generation requests.
- Lists available Pruna applications through the inference.sh CLI.

## Use Cases

- Create Short Visual Concepts: Turn a written scene into a short draft or high-resolution video for creative review.
- Animate Campaign Images: Convert an existing product or campaign image into a motion clip with directed movement.
- Prototype Media Workflows: Compare Pruna video models and settings before integrating remote generation into a larger content process.

## Prompt Templates

### Generate a Basic Video

```
Create a five-second 720p video of [scene]. Use the most suitable Pruna text-to-video model and show the generation result.
```

### Animate an Image

```
Animate the image at [public image URL]. Add [movement], keep motion natural, and generate a five-second 720p clip.
```

### Create a Video with Audio

```
Generate a P-Video clip using [scene prompt] and the audio at [public audio URL]. Match visible motion to the audio.
```

### Compare Production Options

```
Compare P-Video and WAN for [goal], [source type], [resolution], and [budget]. Recommend settings, run the selected model, and summarize tradeoffs.
```

## Limitations

- Requires the inference.sh belt CLI, authentication, network access, and an available account balance.
- Generation speed, pricing, model availability, and output quality depend on the remote service.
- Image and audio inputs must use URLs accessible to the inference service.
- The skill does not edit, review, or locally render generated videos.

## Best Practices

- Confirm the model, resolution, duration, and expected cost before starting generation.
- Use specific scene, camera, lighting, and motion descriptions in the prompt.
- Provide only media URLs that the remote service can access and that are safe to share.

## Anti Patterns

- Do not place secrets, private tokens, or confidential details in prompts or media URLs.
- Do not assume every model supports 1080p, audio, or image input.
- Do not start repeated paid generations without reviewing the first result and current pricing.

## Security Audit

- Audited at: 2026-08-06T10:51:28.45\+00:00
- Summary: All 34 static findings are false positives caused by Markdown formatting, fixed CLI examples, placeholder media URLs, and visible documentation links. The skill contains no executable scripts, prompt injection, dynamic shell construction, or concealed network destinations.

## Stats

- Views: 169
- Downloads: 6
- Favorites: 0
- Popularity score: 0