📦
ai-video-generation
74Generate AI Videos with RunComfy
Choosing a video model and formatting its inputs takes time. This skill selects a suitable RunComfy route and prepares the generation request.
設計與創意 查看
技能 video-edit
📦
video-edit
內容修訂 r2 高風險 ⚙️ 外部命令🌐 網路存取📁 檔案系統存取
Edit Videos with Intent-Routed AI Models
Choosing the wrong video model wastes time and produces inconsistent edits. This skill selects a suitable RunComfy model and builds model-specific instructions.
支援: Claude Codex Code(CC)
安裝前請停止並請求確認。
請審查安裝計畫，並在變更檔案前取得使用者明確同意。
使用我的智慧代理安裝
將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。
智能體請求
Review the Skillstore skill "video-edit" from https://skillstore.io/skills/skills-collective-video-edit.md and its manifest at https://skillstore.io/api/skills/skills-collective-video-edit/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。
Agent 可讀資源
當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。
測試它
正在使用「video-edit」。 Give a presenter video a modern office background while preserving speech and lip movement.
預期結果:
- Selected Wan 2.7 Edit-Video for identity-preserving background replacement.
- The edit instruction prioritizes face, pose, lip movement, framing, and original audio.
正在使用「video-edit」。 Apply a dance clip's motion to a character image and keep the source sound.
預期結果:
- Selected Kling 2.6 Pro Motion Control for reference-driven motion transfer.
- The request uses image orientation and preserves the reference video's sound.
正在使用「video-edit」。 Change a walking subject's outfit to business attire without changing the face.
預期結果:
- Selected Lucy Edit Restyle for a localized outfit change.
- The instruction preserves face and motion while limiting the edit to clothing.
安全審計
高風險v4 • 2026/7/24 開啟版本化報告
All 67 static findings are false positives caused by Markdown formatting, example URLs, or documentation of credential storage. Separate review found shell-construction, unpinned dependency, and third-party media disclosure risks that require clear safeguards.
1
已掃描檔案
226
分析行數
0
複核項目
0
已忽略的誤報
已確認的安全問題 (3)
高
Unsafe Shell Construction for Generated Prompts
The examples embed generated prompt text inside single-quoted JSON passed through a shell. The skill lacks a mandatory serialization rule for arbitrary user-controlled characters.
The command examples directly place prompt text inside shell-quoted JSON, while line 222 asserts safety after shell parsing. Robust argument construction is not required.
中
Unpinned Global CLI Installation
The prerequisite installs the latest RunComfy CLI globally without a version pin. Package compromise or incompatible updates could execute code with user permissions.
Line 49 explicitly uses a global npm installation with no version. This is a clear dependency integrity and update risk.
中
Third-Party Prompt and Media Disclosure
The CLI sends prompts to RunComfy and its servers fetch user-provided media URLs. Sensitive content may leave the user's controlled environment.
Lines 217 and 223 explicitly describe API submission and server-side retrieval of user media. The external data flow is documented and unambiguous.
風險因素
⚙️ 外部命令 (47)
SKILL.md:13 SKILL.md:27-29 SKILL.md:29-37 SKILL.md:37-49 SKILL.md:49-50 SKILL.md:50-51 SKILL.md:51-58 SKILL.md:58-64 SKILL.md:64-65 SKILL.md:65-66 SKILL.md:66-67 SKILL.md:67 SKILL.md:68-69 SKILL.md:69 SKILL.md:70 SKILL.md:71-77 SKILL.md:77-85 SKILL.md:85-89 SKILL.md:89-98 SKILL.md:98-102 SKILL.md:102-104 SKILL.md:104-105 SKILL.md:105-112 SKILL.md:112-120 SKILL.md:120-121 SKILL.md:121-122 SKILL.md:122-123 SKILL.md:123-124 SKILL.md:124 SKILL.md:128-138 SKILL.md:138-143 SKILL.md:143-145 SKILL.md:145 SKILL.md:151-159 SKILL.md:159-160 SKILL.md:160-161 SKILL.md:161 SKILL.md:167-174 SKILL.md:174-178 SKILL.md:178-185 SKILL.md:185-190 SKILL.md:190-200 SKILL.md:200-217 SKILL.md:217 SKILL.md:221 SKILL.md:222-224 SKILL.md:224
🌐 網路存取 (10)
審計單位: codex 檢視審計紀錄 →
分享與引用此報告
分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。
複製報告連結
https://skillstore.io/skills/skills-collective-video-edit/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/skills-collective-video-edit?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/skills-collective-video-edit?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/skills-collective-video-edit/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/skills-collective-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>學術引用 (APA · BibTeX · CFF)
APA 引用
skills-collective. (2026). video-edit security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/skills-collective-video-edit/audits/4BibTeX 引用
@techreport{skills-collective-skills-collective-video-edit-2026,
author = {skills-collective},
title = {video-edit security audit report (audit version 4)},
institution = {Skillstore},
year = {2026},
number = {4},
url = {https://skillstore.io/skills/skills-collective-video-edit/audits/4},
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 4)"
version: "unspecified"
type: report
authors:
- name: "skills-collective"
date-released: "2026-07-24"
url: "https://skillstore.io/skills/skills-collective-video-edit/audits/4"
identifiers:
- type: other
value: "skillstore:skills-collective-video-edit:audit:4"
description: "Skillstore immutable audit report identifier"
比較變體
6 個可安裝變體每位作者的技能都保留為獨立的可安裝項目。推薦變體依 Skillstore 證據排序。
為何此變體排名第一
Skillstore 使用量較高
作者 / 來源 Skillstore 評分 證據置信度 Skillstore 使用量 已更新
skills-collective-video-edit
Skillstore 評分 38
證據置信度 中
Skillstore 使用量 41
已更新
2026-09-09
Skillstore 評分
評分依據 證據置信度: 中 55
架構
85
可維護性
87
內容
65
社群
83
規範符合度
你可以打造什麼
Restyle Talking-Head Clips
Preserve a speaker's face, motion, lip movement, and audio while changing the setting or visual treatment.
Update Product Visuals
Replace packaging or backgrounds while retaining original framing, hand movement, and camera motion.
Transfer Character Motion
Apply motion from a reference clip to a target character while maintaining identity and spatial constraints.
試試這些提示
Simple Video Restyle
Restyle [video URL] with [visual change]. Preserve the subject's face and motion.
Background Edit with Audio
Change the background in [video URL] to [new setting]. Preserve face, pose, lip movement, framing, and original audio.
Reference Packaging Transfer
Use [reference image URL] to replace the packaging in [video URL]. Preserve hand movement, camera motion, framing, and original audio.
Advanced Motion Transfer
Transfer motion from [reference video URL] to the character in [target image or video URL]. Keep [sound preference] and maintain [spatial constraints].
最佳實務
- State preservation requirements before describing the requested change.
- Request one visual change per invocation to reduce temporal drift.
- Confirm model limits and obtain approval before sending media to RunComfy.
避免
- Do not combine unrelated background, outfit, character, and motion changes in one request.
- Do not use Lucy Edit Restyle for complete character replacement or precise motion transfer.
- Do not place sensitive media at public URLs or concatenate untrusted text into shell commands.
常見問題
Which model is selected by default?
Wan 2.7 Edit-Video is the default because it supports versatile edits and resolutions up to 1080p.
Can the skill preserve original audio?
Yes. Wan and Kling provide audio preservation options when the selected route supports them.
Can it transfer motion from another video?
Yes. The skill selects Kling 2.6 Pro Motion Control for reference-driven character motion.
Does it accept local video files?
The documented routes require source and reference URLs accepted by RunComfy.
What are the main output limits?
Wan supports 2 to 10 seconds, Kling supports up to 30 seconds, and Lucy is limited to 720p.
Where is media processed?
RunComfy receives the model request, fetches referenced media, performs the edit, and returns output for download.
開發者詳細資訊
授權
MIT
Skillstore 修訂
r2
版本提示
作者未聲明版本。
引用
9e952417e76879bc9d853e1b8b2cd6d6d8d4a1c2
維護新鮮度
2026/7/24
使用量
1 次下載 · 0 次瀏覽
檔案結構
📄 SKILL.md
skills-collective 的更多技能
查看全部📦
ai-music
50Create and Edit AI Music with RunComfy
Choosing a music model and valid parameters can slow production. This skill routes each request and prepares a RunComfy command for generation or audio editing.
設計與創意 查看
📦
image-to-video
50Animate Still Images with the Right RunComfy Model
Choosing an image-to-video model can waste time and credits. This skill routes each request to a suitable RunComfy model and provides focused prompting guidance.
設計與創意 查看
📦
ai-image-generation
38Generate Images with the Right RunComfy Model
Choosing an image model and its parameters is difficult. This skill routes each request to a suitable RunComfy model and prepares the generation command.
設計與創意 查看
相關技能
查看全部📦
video-extend
78使用 RunComfy 延伸影片
由 runcomfy-com
短篇生成片段通常需要更長時間、更俐落的結尾，或串連的故事節拍。這項技能可協助 Claude、Codex 和 Claude Code 使用 RunComfy 延伸已核准的影片。
設計與創意 查看
📦
controlnet-pose
79建立姿勢控制的圖片與影片
由 runcomfy-com
姿勢控制生成通常需要選擇正確的模型和輸入格式。此技能會將圖片和影片請求路由至 RunComfy 工作流程。
設計與創意 查看
📦
ace-step
79在 RunComfy 上生成 ACE Step 音樂
由 runcomfy-com
建立平價的 AI 音樂通常需要端點選擇、結構化歌詞，以及可重複的 CLI 輸入。此技能會引導你透過 RunComfy 進行 ACE Step 生成、inpainting 和 outpainting。
設計與創意 查看
📦
elevenlabs-music-generation
79在 RunComfy 上使用 ElevenLabs 生成音樂
由 runcomfy-com
建立可用於製作的音樂提示詞和 CLI 呼叫可能既耗時又容易出錯。此技能會引導 Claude、Codex 和 Claude Code 在 RunComfy 上進行 ElevenLabs Music 生成，提供清楚的參數與工作流程建議。
設計與創意 查看
📦
image-inpainting
79使用 RunComfy 進行影像修補
由 runcomfy-com
當提示詞、遮罩與模型選擇都以手動處理時，遮罩驅動的影像編輯可能會很慢。此技能會引導 Claude、Codex 或 Claude Code 使用正確的模型與輸入來執行 RunComfy inpainting。
設計與創意 查看
📦
ai-avatar-video
79使用 RunComfy 建立 AI 虛擬人影片
由 runcomfy-com
建立對嘴同步的虛擬人影片通常需要選擇正確的模型與輸入格式。此技能會將請求路由至 RunComfy 模型，並提供可直接使用的 CLI 呼叫指令。
設計與創意 查看