技能 clipcat
📦
clipcat
內容修訂 r1 嚴重 ⚙️ 外部命令🌐 網路存取📁 檔案系統存取🔑 環境變數
Research TikTok Products and Create Selling Videos
TikTok Shop research and selling-video production require many separate tools. Clipcat combines commerce data, proven prompts, media generation, and task tracking.
支援: Claude Codex Code(CC)
請勿自動安裝此 Skill。
權威政策要求在執行任何安裝操作前由操作人員審查。
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 "clipcat" from https://skillstore.io/skills/clipcat-ai-clipcat.md and its manifest at https://skillstore.io/api/skills/clipcat-ai-clipcat/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.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「clipcat」。 Research serum selling videos in the United States and identify a useful creative pattern.
預期結果:
- Top pattern: handheld bathroom demonstration with the product visible in the opening two seconds.
- Evidence: the selected source combines strong sales with a close match to the requested audience and setting.
- Next step: adapt its shot order and voiceover while removing unsupported efficacy claims.
正在使用「clipcat」。 Create an eight-second vertical product video from my image using a standard model.
預期結果:
- Recommended settings: Seedance 2, 480p, eight seconds, vertical format.
- The current quote is 160 credits. Generation will start only after explicit approval.
- The image will be uploaded to Clipcat for remote processing.
正在使用「clipcat」。 Check my latest image generation task.
預期結果:
The latest image task is still processing. Check again in about 30 seconds using its saved task identifier.
安全審計
嚴重v1 • 2026/9/10 開啟版本化報告
Most static command findings mistake Markdown backticks and documentation for executable code. However, the skill directly executes mutable remote installers, handles an API key, updates its CLI, accesses local task state, and uploads media to Clipcat.
1
已掃描檔案
518
分析行數
11
複核項目
0
已忽略的誤報
已確認的安全問題 (4)
嚴重
Pipe to shell pattern
curl -fsSL https://clipcat.ai/cli | bash
The installation command pipes an unauthenticated-at-rest response from a mutable remote URL directly into bash. Compromise of the endpoint or transport terminates in arbitrary code execution.
嚴重
Remote PowerShell Download Executed Immediately
The command `irm https://clipcat.ai/cli.ps1 | iex` executes mutable remote content without inspection, pinning, or integrity verification.
The documented PowerShell pipeline directly passes an Internet response to Invoke-Expression. This is an explicit remote code execution pattern.
高
Unattended CLI Self-Update
The skill tells agents to run `clipcat update` after an update notice, replacing executable code without renewed user approval or version pinning.
The update command reruns the installer, and the workflow explicitly directs the agent to run it automatically after a CLI-controlled notice.
中
Local Media Uploaded to External Service
Video and image commands upload user-selected local media through presigned URLs, creating a disclosed but material confidentiality boundary.
The documentation explicitly states that local videos and images are uploaded for remote processing. The behavior is intentional and not merely inferred.
能力複核項目 (11)
這些是真實的本機能力，對此技能而言可能是預期行為，因此需要複核，但不會按已確認的惡意行為計分。
高
Hidden file in home directory
1. Task ID is saved locally to `~/.clipcat/tasks.json` automatically.
SKILL.md line 444 states that the CLI writes or reads ~/.clipcat/tasks.json. This is disclosed task-state storage, but it is real access to a hidden file in the user's home directory.
高
Hidden file in home directory
`clipcat query_task` with no flags automatically reads the latest task from `~/.clipcat/tasks.json`
SKILL.md line 455 states that the CLI writes or reads ~/.clipcat/tasks.json. This is disclosed task-state storage, but it is real access to a hidden file in the user's home directory.
高
Generic API/secret keys
"requires": { "env": ["CLIPCAT_API_KEY"] },
SKILL.md line 9 declares or discusses CLIPCAT_API_KEY as an authentication secret. The access is expected for the service, but the skill handles a sensitive credential.
高
Generic API/secret keys
"primaryEnv": "CLIPCAT_API_KEY",
SKILL.md line 10 declares or discusses CLIPCAT_API_KEY as an authentication secret. The access is expected for the service, but the skill handles a sensitive credential.
高
Generic API/secret keys
Get the key at https://clipcat.ai/workspace?modal=settings&tab=apikeys. Prefer the config file over
SKILL.md line 50 declares or discusses CLIPCAT_API_KEY as an authentication secret. The access is expected for the service, but the skill handles a sensitive credential.
中
Hidden file access
1. Task ID is saved locally to `~/.clipcat/tasks.json` automatically.
SKILL.md line 444 states that the CLI writes or reads ~/.clipcat/tasks.json. This is disclosed task-state storage, but it is real access to a hidden file in the user's home directory.
中
Hidden file access
`clipcat query_task` with no flags automatically reads the latest task from `~/.clipcat/tasks.json`
SKILL.md line 455 states that the CLI writes or reads ~/.clipcat/tasks.json. This is disclosed task-state storage, but it is real access to a hidden file in the user's home directory.
低
Hardcoded URL
curl -fsSL https://clipcat.ai/cli | bash
The URL is an active remote installer endpoint fetched before shell execution. Network trust in this mutable community-controlled endpoint is security-relevant.
低
Hardcoded URL
irm https://clipcat.ai/cli.ps1 | iex
The URL is an active PowerShell installer endpoint whose response is immediately evaluated. This creates a real network and supply-chain trust boundary.
低
Hardcoded URL
clipcat config --api-key <your-key> --base-url https://clipcat.ai
The command configures clipcat.ai as the API endpoint, so subsequent CLI operations communicate with an external service. This network use is expected but real.
低
Hardcoded URL
Get the key at https://clipcat.ai/workspace?modal=settings&tab=apikeys. Prefer the config file over
The line directs users to clipcat.ai to obtain an API key. This is expected authentication traffic, but it is a real external network dependency.
風險因素
⚙️ 外部命令 (50)
SKILL.md:18 SKILL.md:22 SKILL.md:26-28 SKILL.md:28-32 SKILL.md:32-34 SKILL.md:34-36 SKILL.md:36-40 SKILL.md:40 SKILL.md:46-48 SKILL.md:48-50 SKILL.md:50 SKILL.md:54-67 SKILL.md:67-68 SKILL.md:68-71 SKILL.md:71-72 SKILL.md:72-73 SKILL.md:74 SKILL.md:81 SKILL.md:82-84 SKILL.md:84-85 SKILL.md:85-86 SKILL.md:86-87 SKILL.md:87-88 SKILL.md:88-89 SKILL.md:89 SKILL.md:92-96 SKILL.md:96-97 SKILL.md:97-98 SKILL.md:98-99 SKILL.md:99-100 SKILL.md:100-101 SKILL.md:101-102 SKILL.md:102-103 SKILL.md:103 SKILL.md:104-105 SKILL.md:105 SKILL.md:108 SKILL.md:110 SKILL.md:116 SKILL.md:117-118 SKILL.md:118 SKILL.md:119 SKILL.md:120-123 SKILL.md:123 SKILL.md:124 SKILL.md:126-127 SKILL.md:127-129 SKILL.md:129-134 SKILL.md:134-140 SKILL.md:140
🌐 網路存取 (8)
📁 檔案系統存取 (4)
🔑 環境變數 (3)
偵測到的模式
Pipe to shell pattern
審計單位: codex
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/clipcat-ai-clipcat/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/clipcat-ai-clipcat?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/clipcat-ai-clipcat?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/clipcat-ai-clipcat/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/clipcat-ai-clipcat.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
clipcat-ai. (2026). clipcat security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/clipcat-ai-clipcat/audits/1BibTeX 引用
@techreport{clipcat-ai-clipcat-ai-clipcat-2026,
author = {clipcat-ai},
title = {clipcat security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/clipcat-ai-clipcat/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: "clipcat security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "clipcat-ai"
date-released: "2026-09-10"
url: "https://skillstore.io/skills/clipcat-ai-clipcat/audits/1"
identifiers:
- type: other
value: "skillstore:clipcat-ai-clipcat:audit:1"
description: "Skillstore immutable audit report identifier"
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
78
規範符合度
你可以打造什麼
Validate Product Opportunities
Compare product, seller, creator, and video performance before selecting a TikTok Shop campaign direction.
Adapt Proven Creative
Find a relevant high-GMV video prompt and rewrite its structure for a new product and audience.
Produce Product Media
Generate product images or videos, confirm costs, submit tasks, and monitor results.
試試這些提示
Find Leading Products
Research leading TikTok Shop products in [category] for [region]. Compare sales, GMV, price, and recent trends.
Compare Competitors
Compare [seller or product] with three relevant competitors. Summarize creator activity, selling videos, pricing, and evidence-backed opportunities.
Create a Selling-Video Prompt
Find proven prompts for [product] targeting [audience]. Adapt the strongest match, remove unsupported claims, and cite the source before generation.
Plan and Generate a Campaign Asset
Research [market], select a proven creative structure, propose model settings, quote credits, request approval, generate the asset, and track completion.
最佳實務
- Run command help before unfamiliar operations because supported flags and models can change.
- Quote every paid generation with identical settings, show the cost, and obtain explicit approval before submission.
- Use proven prompts as structure, replace product details, and remove claims that the user cannot support.
避免
- Do not retry empty historical queries when the entity may only exist in current public data.
- Do not guess identifiers, model availability, prices, or category levels.
- Do not upload media, spend credits, increase resolution, or update the CLI without explicit user approval.
常見問題
What do I need before using Clipcat?
You need the Clipcat CLI, a Clipcat account, and a configured API key.
Does every command consume credits?
No. Some account metadata is free, but research and generation commands can consume credits.
Can Clipcat research any TikTok account or video?
Current public data has broad coverage, while historical commerce datasets may not include every entity.
How are paid generations approved?
The skill requests a server quote, shows the exact settings and cost, and waits for explicit approval.
Where are local files processed?
Selected images and videos are uploaded to Clipcat for remote processing through its API.
Can long-running tasks finish in one command?
No. Generation and analysis tasks return identifiers and require later status checks.
開發者詳細資訊
作者
clipcat-ai授權
MIT
Skillstore 修訂
r1
版本提示
作者未聲明版本。
引用
19414c6b8c288f69f8d7863113a691912fa20043
維護新鮮度
2026/9/10
使用量
0 次下載 · 1 次瀏覽
檔案結構
📄 SKILL.md