# Animate Still Images Into Short Videos

Still images need careful model choices and restrained motion prompts to become usable videos. This skill guides Claude, Codex, and Claude Code through image-to-video workflows with belt CLI examples.

## Install

```bash
npx skillstore add halt-catch-fire/image-to-video
```

## Metadata

- Status: approved
- Slug: halt-catch-fire-image-to-video
- Skillstore revision: r1
- Version status: missing
- Tree hash: db899372352dc8a34b331af6d1d05c586fdbf78cab3861731090ad5615a7bc7f
- Author: halt-catch-fire
- GitHub username: halt-catch-fire
- License: MIT
- Repository: https://github.com/halt-catch-fire/skills/tree/main/tools/video/image-to-video/
- Ref: a06681402992ceae98ba04d54cfd4ab004862696
- 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/halt-catch-fire-image-to-video
- Manifest: https://skillstore.pages.dev/api/skills/halt-catch-fire-image-to-video/manifest

## Capabilities

- Selects image-to-video models for landscapes, portraits, products, fabric, and general animation.
- Provides belt CLI examples for image generation, animation, upscaling, audio, and media merging.
- Explains camera movement prompts such as dolly, pan, tilt, orbit, and static shots.
- Recommends motion prompt structure for subject motion, atmosphere, mood, and pace.
- Gives duration guidance for short clips, looping backgrounds, and stitched sequences.

## Use Cases

- Create Social Video Loops: Turn a still image into a short loop with subtle motion, clear duration guidance, and a suitable model choice.
- Produce Product Motion Shots: Plan product animations with camera movement, reflections, fabric motion, or material flow for campaign assets.
- Build Cinemagraph Concepts: Keep most of a photo still while animating one chosen element for a clean cinemagraph effect.

## Prompt Templates

### Choose a Model

```
I have a still image of [subject]. Recommend the best image-to-video model, motion direction, and clip length. Explain the choice briefly.
```

### Write a Motion Prompt

```
Create an image-to-video prompt for [image description]. Include camera movement, subject motion, atmosphere, mood, and pace. Keep motion subtle.
```

### Plan a Product Animation

```
Design a short product animation workflow for [product]. Choose a model, motion style, duration, and review steps before adding audio.
```

### Build a Multi-Clip Pipeline

```
Plan a multi-clip image-to-video pipeline for [scene]. Include progressive motion, upscaling, audio, stitching, privacy checks, and quality review criteria.
```

## Limitations

- Requires a configured inference.sh belt CLI account for command examples.
- Sends media to external inference services when belt commands are run.
- Does not create video locally without external model access.
- Quality can degrade with long clips, complex action, or low-resolution source images.

## Best Practices

- Start with high-resolution images and keep motion subtle.
- State camera movement, subject motion, atmosphere, and pace in one prompt.
- Use short clips first, then stitch approved segments for longer results.

## Anti Patterns

- Requesting fast action, spinning cameras, or many changes in one clip.
- Uploading sensitive media without confirming external processing is acceptable.
- Choosing a model by popularity instead of matching the subject and motion.

## Security Audit

- Audited at: 2026-07-06T16:42:30.796\+00:00
- Summary: The skill is a legitimate image-to-video guide, but several examples run belt commands that contact third-party inference services and upload local media. Most backtick and reconnaissance static hits are markdown formatting or model names, while actual belt and npx command examples remain confirmed medium risk.

## Stats

- Views: 0
- Downloads: 7
- Favorites: 0
- Popularity score: 0