技能 video-edit

📦

video-edit

Name: video-edit
Author: gencraft-labs

内容修订 r1 安全 ⚙️ 外部命令🌐 网络访问📁 文件系统访问

Edit Videos with Intent-Based RunComfy Routing

Different video edits need different models and input settings. This skill selects a RunComfy route and supplies model-specific prompts for restyling, background changes, or motion transfer.

支持: Claude Codex Code(CC)

🥉 77 青铜

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "video-edit" from https://skillstore.io/skills/gencraft-labs-video-edit.md and its manifest at https://skillstore.io/api/skills/gencraft-labs-video-edit/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/gencraft-labs-video-edit.md 已签名 manifest GET /api/skills/gencraft-labs-video-edit/manifest 已签名 lockfile GET /api/skills/gencraft-labs-video-edit/lockfile

测试它

正在使用“video-edit”。 Replace a six-second presenter video's background with a neutral office and keep the original speech.

预期结果:

Illustrative edit plan: Wan 2.7 Edit-Video.
Preservation goals: retain the speaker's face, pose, and lip movement.
Requested change: neutral office background.
Audio setting: preserve source audio.

正在使用“video-edit”。 Apply motion from a dance reference to my character image and keep the reference audio.

预期结果:

Illustrative edit plan: Kling 2.6 Pro Motion Control.
Inputs: target character image and motion reference video.
Character orientation: image, with output capped at 10 seconds.
Sound setting: retain original reference audio.

正在使用“video-edit”。 Change the subject's outfit to business attire without changing the face or movement.

预期结果:

Illustrative edit plan: Lucy Edit Restyle.
Requested change: business attire.
Preservation goals: retain face and motion.
Output constraints: 720p with the source aspect ratio.

安全审计

安全

v1 • 2026/10/8 打开版本化报告

All 67 static findings are false positives involving Markdown formatting, illustrative URLs, legitimate CLI workflows, or documented token storage. No evidence found of malicious intent, credential theft, or prompt injection in SKILL.md. Remote processing and local downloads are disclosed, but external CLI behavior and security claims were not independently verified.

已扫描文件

226

分析行数

复核项

已忽略误报

风险因素

⚙️ 外部命令 (47)

🌐 网络访问 (10)

SKILL.md:17 SKILL.md:23 SKILL.md:81 SKILL.md:93 SKILL.md:94 SKILL.md:132 SKILL.md:133 SKILL.md:171 SKILL.md:182 SKILL.md:213

📁 文件系统访问 (9)

SKILL.md:81 SKILL.md:93 SKILL.md:94 SKILL.md:132 SKILL.md:133 SKILL.md:171 SKILL.md:182 SKILL.md:221 SKILL.md:221

最新一次完成的静态与语义审计未检测到已确认的安全发现。这并不能证明该 Skill 没有副作用。

审计者: codex

分享与引用此报告

分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。

打开版本化报告

复制报告链接

https://skillstore.io/skills/gencraft-labs-video-edit/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/gencraft-labs-video-edit/security.svg)](https://skillstore.io/skills/gencraft-labs-video-edit?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/gencraft-labs-video-edit?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/gencraft-labs-video-edit/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/gencraft-labs-video-edit.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

学术引用 (APA · BibTeX · CFF)

APA 引用

gencraft-labs. (2026). video-edit security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/gencraft-labs-video-edit/audits/1

BibTeX 引用

@techreport{gencraft-labs-gencraft-labs-video-edit-2026,
  author = {gencraft-labs},
  title = {video-edit security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/gencraft-labs-video-edit/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "video-edit security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "gencraft-labs"
date-released: "2026-10-08"
url: "https://skillstore.io/skills/gencraft-labs-video-edit/audits/1"
identifiers:
  - type: other
    value: "skillstore:gencraft-labs-video-edit:audit:1"
    description: "Skillstore immutable audit report identifier"

比较变体

6 个可安装变体

每位作者的技能都保持为独立的可安装项。推荐变体按 Skillstore 证据排序。

为什么此变体排名第一

Skillstore 使用量更高

作者 / 来源 Skillstore 评分证据置信度 Skillstore 使用量已更新

runcomfy-com 推荐

runcomfy-com-video-edit

Skillstore 评分 78

证据置信度高

Skillstore 使用量 52

已更新

2026-09-09

agentspace-so

agentspace-so-video-edit

Skillstore 评分 78

证据置信度高

Skillstore 使用量 46

已更新

2026-09-09

gencraft-labs 当前

gencraft-labs-video-edit

Skillstore 评分 77

证据置信度中

Skillstore 使用量 1

已更新

2026-10-11

doany-ai

doany-ai-video-edit

Skillstore 评分 50

证据置信度高

Skillstore 使用量 40

已更新

2026-09-09

skills-collective

skills-collective-video-edit

Skillstore 评分 38

证据置信度中

Skillstore 使用量 38

已更新

2026-09-09

genmedia-labs

genmedia-labs-video-edit

Skillstore 评分 38

证据置信度中

Skillstore 使用量 31

已更新

2026-09-09

Skillstore 评分

评分依据证据置信度: 中

架构

可维护性

内容

社区

规范符合性

你能构建什么

Refresh Presenter Backgrounds

Change a short presenter video's background with Wan while requesting preservation of face, pose, lip movement, and source audio.

Preview Packaging Designs

Use a reference image to replace packaging in a product clip while requesting consistent framing and hand movement.

Transfer Character Motion

Apply motion from a reference video to a target character through Kling with an appropriate orientation and duration.

试试这些提示

Simple Outfit Change

Use Lucy to change the outfit in [video URL] to [outfit]. Preserve face and motion. Save results in [output directory].

Presenter Background Replacement

Use Wan for [video URL]. Preserve face, pose, and lip movement. Replace the background with [scene]. Keep source audio. Save to [output directory].

Reference-Based Packaging Edit

Use Wan to replace packaging in [video URL] using [reference image URL]. Preserve framing and hand movement. Save results in [output directory].

Controlled Motion Transfer

Use Kling to transfer motion from [reference video URL] to [character image URL]. Use image orientation and preserve original sound. Keep output within 10 seconds. Place the character on [frame side]. Save to [output directory].

最佳实践

State preservation goals before one specific change, then inspect the result for identity and motion drift.
Check input duration, format, size, and orientation before submitting a model request.
Use authorized media and confirm remote processing is acceptable; keep authentication tokens out of prompts and shared logs.

避免

Combining several unrelated changes in one request or expecting multi-model blending.
Using Lucy for complete subject replacement, custom aspect ratios, or output above 720p.
Submitting private media without permission or including account tokens in editing prompts.

常见问题

Which model is selected by default?

Wan 2.7 Edit-Video is the default. Kling handles motion transfer, while Lucy handles localized outfit and style changes.

What must be installed before use?

Install the RunComfy CLI and authenticate with an account. Provide a source video URL that meets the selected model's limits.

Can the original audio be preserved?

Wan offers the origin audio setting. Kling offers keep_original_sound. This skill does not document an equivalent Lucy audio option.

Does video processing happen locally?

No. The CLI submits requests to RunComfy, and model servers fetch input media. Generated outputs are downloaded to the selected directory.

Does the skill guarantee identity preservation?

No. It supplies preservation prompts and model selection guidance. Review generated videos for face, movement, framing, and lip synchronization changes.

Can one request combine several model routes?

No. The skill selects one model per call. Each route retains its own duration, resolution, and input restrictions.

开发者详情

作者

gencraft-labs

许可证

MIT

Skillstore 修订

版本提示

作者未声明版本。

仓库

https://github.com/gencraft-labs/skills/tree/5d625f59c710d7a33316b50b302b7158cdee5971/video-edit

引用

e406bf64b5db6e514879399e87b346fa7ff7c2fb

维护新鲜度

2026/10/11

使用量

0 次下载 · 0 次浏览

文件结构

📄 SKILL.md

gencraft-labs 的更多技能

查看全部

📦

使用 RunComfy 创建 AI 数字人视频

由 runcomfy-com

创建口型同步的数字人视频通常需要选择合适的模型和输入格式。此技能会将请求路由到 RunComfy 模型，并提供可直接使用的 CLI 调用。

设计与创意查看

video-edit

使用我的智能体安装

Agent 可读资源

测试它

安全审计

风险因素

复制报告链接

Markdown 徽章

HTML 徽章

嵌入卡片

APA 引用

BibTeX 引用

CITATION.cff

比较变体

Skillstore 评分

你能构建什么

Refresh Presenter Backgrounds

Preview Packaging Designs

Transfer Character Motion

试试这些提示

最佳实践

避免

常见问题

开发者详情

gencraft-labs 的更多技能

image-to-video

ai-image-generation

相关技能

video-extend

controlnet-pose

ace-step

elevenlabs-music-generation

image-inpainting

ai-avatar-video