技能 generating-videos

📦

generating-videos

Name: generating-videos
Author: supercmohq

v0.5.0 内容修订 r1 高风险 ⚙️ 外部命令

Generate Cohesive AI Videos from Any Brief

Creating consistent AI video across multiple clips requires careful planning and model selection. This skill coordinates prompts, references, generation, and stitching.

支持: Claude Codex Code(CC)

⚠️ 38 差

安装前请停下并请求确认。

请审查安装计划，并在更改文件前获得用户明确同意。

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 "generating-videos" from https://skillstore.io/skills/supercmohq-generating-videos.md and its manifest at https://skillstore.io/api/skills/supercmohq-generating-videos/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

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/supercmohq-generating-videos.md 已签名 manifest GET /api/skills/supercmohq-generating-videos/manifest 已签名 lockfile GET /api/skills/supercmohq-generating-videos/lockfile

测试它

正在使用“generating-videos”。 Animate a product photo into a six-second vertical reveal.

预期结果:

A vertical video opens on the supplied image, adds a controlled reveal motion, and preserves the product appearance.

正在使用“generating-videos”。 Create a thirty-second story with one presenter across three scenes.

预期结果:

An approved concept defines the presenter, setting, script, and three connected scenes.
Three clips reuse the same presenter reference and visual direction.
The completed clips are joined in story order.

安全审计

高风险

v1 • 2026/8/13 打开版本化报告

All 32 static findings are false positives caused by ordinary prose or Markdown inline code, with no shell execution or system reconnaissance present. However, the model-routing guidance explicitly recommends switching providers after safety, NSFW, or intellectual-property refusals, creating a high-confidence safety-filter circumvention risk.

已扫描文件

435

分析行数

复核项

已忽略误报

已确认安全问题 (1)

高

SKILL.md:57-64 SKILL.md:99-101

Safety Filter Circumvention Through Provider Switching

The skill recommends grok-imagine-video for an "edgy or sensitive subject Seedance would refuse" and after NSFW or IP refusals. This routes blocked content to another provider instead of respecting the refusal.

The routing table and edge-case instructions explicitly connect safety, NSFW, and IP refusals with switching to another model.

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/supercmohq-generating-videos/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/supercmohq-generating-videos/security.svg)](https://skillstore.io/skills/supercmohq-generating-videos?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/supercmohq-generating-videos?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/supercmohq-generating-videos/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/supercmohq-generating-videos.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

supercmohq. (2026). generating-videos security audit report (audit version 1) [Author version 0.5.0]. Skillstore. https://skillstore.io/skills/supercmohq-generating-videos/audits/1

BibTeX 引用

@techreport{supercmohq-supercmohq-generating-videos-2026,
  author = {supercmohq},
  title = {generating-videos security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/supercmohq-generating-videos/audits/1},
  note = {Author version 0.5.0}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "generating-videos security audit report (audit version 1)"
version: "0.5.0"
type: report
authors:
  - name: "supercmohq"
date-released: "2026-08-13"
url: "https://skillstore.io/skills/supercmohq-generating-videos/audits/1"
identifiers:
  - type: other
    value: "skillstore:supercmohq-generating-videos:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 评分

评分依据证据置信度: 中

架构

100

可维护性

内容

社区

规范符合性

你能构建什么

Create Social Videos

Turn a campaign brief and product images into a formatted short video.

Animate Creative Assets

Bring a still image to life while preserving its subject and composition.

Produce Multi-Clip Stories

Plan recurring subjects, generate connected scenes, and join them into one video.

试试这些提示

Generate a Simple Clip

Create a [duration] video of [subject and action] in [setting], formatted for [destination].

Animate a Still Image

Animate this image for [duration]. Preserve [subject details], add [motion], and deliver it in [aspect ratio].

Use Reference Media

Create a video using [media] for [motion, style, or timing]. Keep [elements] and change [elements].

Build a Multi-Clip Story

Produce a [total duration] story about [concept]. Maintain [recurring subjects] across clips, include [dialogue], and deliver in [format].

最佳实践

Provide clear durations, destinations, and aspect ratios.
Use clean reference images for recurring people, products, or props.
Approve the multi-clip concept before starting paid generation.

避免

Do not combine a start frame with reference media in one request.
Do not restart a pending job because it can create duplicate charges.
Do not regenerate an unsatisfactory clip without user approval.

常见问题

Can it create a video from text only?

Yes. It can plan and request a text-to-video clip without reference media.

Can it animate an existing image?

Yes. It can use the image as a start frame or visual reference.

Can it create videos longer than one model clip?

Yes. It divides the story into clips, generates them, and stitches them in order.

Does it preserve people and products across scenes?

It reuses reference images and descriptions to improve continuity, but exact consistency is not guaranteed.

Can it synchronize motion to existing audio?

Yes, when the selected model accepts audio references.

Does it generate separate music or voice files?

No. Separate audio assets require a dedicated audio generation workflow.

开发者详情

作者

supercmohq

许可证

Apache-2.0

作者版本

v0.5.0

Skillstore 修订

仓库

https://github.com/supercmohq/supercmo-skills/tree/d7094d329f15c7e6c73f1db7f83665c684461672/skills/generating-videos

引用

68046dfc1c93b48a04e43344644ad77d3a895012

维护新鲜度

2026/8/13

使用量

0 次下载 · 0 次浏览

文件结构

📁 evals/

📄 eval_cases.json

📁 references/

📄 direct-route.md

📄 storyboard-route.md

📄 SKILL.md