# Extend Videos with Google Veo 3.1

Short clips can end before a scene or narrative is complete. This skill continues Veo 3.1 clips through RunComfy with focused prompting guidance.

## Install

```bash
npx skillstore add agentspace-so/video-extend
```

## Metadata

- Status: approved
- Slug: agentspace-so-video-extend
- Skillstore revision: r2
- Version status: missing
- Tree hash: d7e59a00358183000fb8c525eee8084f47f52de8a4db25e5ac65cbbdcfea26ca
- Author: agentspace-so
- GitHub username: agentspace-so
- License: MIT
- Repository: https://github.com/agentspace-so/runcomfy-agent-skills/tree/main/video-extend/
- Ref: 181fdefcafd96b041926e61c4b2e306ca7e7820e
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, filesystem
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/agentspace-so-video-extend
- Manifest: https://skillstore.pages.dev/api/skills/agentspace-so-video-extend/manifest

## Capabilities

- Selects standard or Fast Veo 3.1 Extend based on quality and cost priorities.
- Continues an existing Veo clip from a user-provided video URL.
- Creates focused continuation prompts for camera motion and one narrative beat.
- Supports chained extensions by using each generated clip as the next input.
- Downloads generated clips to a selected local output directory.

## Use Cases

- Complete a Product Shot: Continue a short product clip with a controlled camera move and one final action.
- Build a Narrative Sequence: Chain several focused extensions to develop a scene one beat at a time.
- Draft Motion Variations: Use Fast Extend for lower-cost iterations before generating the final continuation.

## Prompt Templates

### Continue One Action

```
Extend this video from [video URL]. The subject completes [single action]. Keep the camera static and preserve the existing lighting.
```

### Add a Camera Move

```
Continue [video URL] as the camera [movement]. The subject [single action]. Preserve framing, identity, lighting, and background motion.
```

### Choose a Cost Mode

```
Extend [video URL] with [scene beat]. Use Fast Extend for drafts and standard Extend for the final output in [directory].
```

### Plan a Chained Sequence

```
Create [number] consecutive extensions from [video URL]. Define one action and one camera instruction per segment. Use each output as the next input.
```

## Limitations

- Requires the RunComfy CLI, authentication, and access to the Veo 3.1 endpoints.
- Accepts an existing video URL and does not create the initial clip from a still image.
- Identity drift can increase after several chained extensions.
- Source video URLs and prompts are processed by RunComfy.

## Best Practices

- Describe only what happens next because the source clip already defines the scene.
- Use one main action and one explicit camera instruction per extension.
- Review every result before using it as input for another extension.

## Anti Patterns

- Do not combine several unrelated actions in one continuation prompt.
- Do not submit private media or signed URLs without approval for external processing.
- Do not use long extension chains without checking identity and motion drift.

## Security Audit

- Audited at: 2026-07-23T09:20:35.982\+00:00
- Summary: All 52 static findings are false positives caused by Markdown formatting, expected documentation links, a documented credential path, and unrelated phrase matches. However, the documented command embeds user-controlled URL and prompt values in single-quoted JSON without a shell-safe construction method, creating a plausible command-injection path. The workflow also transfers user media to RunComfy for external processing.

## Stats

- Views: 37
- Downloads: 4
- Favorites: 0
- Popularity score: 0