Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Cn Video Gen
v1.0.1国产AI视频生成技能。支持通义万相(Wan2.6文生视频/图生视频)和可灵(Kling V2)。适合创意视频制作、图生视频、动态内容生成等场景。当用户要求生成视频、将图片生成动态视频、或基于描述生成视频片段时触发。
⭐ 1· 627·3 current·4 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
high confidencePurpose & Capability
The skill claims support for both 通义万相 (Wan2.6) and 可灵 (Kling V2) in SKILL.md, but the provided scripts (generate.py, upload_image.py) only implement calls to DashScope/Aliyun (Wan2.6). There is no code to create KLING JWTs or call Kling endpoints despite SKILL.md mentioning KLING_ACCESS_KEY/KLING_SECRET_KEY. Registry metadata also lists no required env vars, but the runtime needs an API key for DashScope — these are inconsistent with the declared manifest.
Instruction Scope
Runtime instructions tell the agent to upload local images to an external image host and to run included scripts that perform network calls. The SKILL.md references a TOOLS.md for credentials that is not present. The instructions and scripts will transmit user image files and request data to external endpoints (dashscope.aliyuncs.com and www.imgurl.org), which is expected for this use case but raises data-exposure concerns, especially given the embedded credentials in the upload script.
Install Mechanism
There is no install spec (instruction-only install), and the skill bundle includes two Python scripts. No external installer or downloaded archives are used, which lowers supply-chain risk; however, included scripts will be executed locally and make outbound HTTP requests, so reviewing the scripts before running is required.
Credentials
SKILL.md documents required environment variables (DASHSCOPE_API_KEY, KLING_ACCESS_KEY, KLING_SECRET_KEY) but the registry metadata claims none. generate.py uses DASHSCOPE_API_KEY (not declared in registry). upload_image.py contains a hard-coded IMGURL_TOKEN and default UID in source code — embedding a secret in the repo is a security concern (it may be valid and could allow an attacker or the service operator to access uploaded images). Requiring multiple credentials is reasonable for multi-provider support, but missing declaration plus the embedded token is disproportionate and inconsistent.
Persistence & Privilege
The skill is not always-enabled, does not request elevated platform privileges, and does not modify other skills or global agent settings. Autonomous invocation is allowed (platform default) but is not combined with any extra privilege in this bundle.
What to consider before installing
This skill contains useful scripts to call Wan2.6 video-generation, but there are several red flags you should address before installing or running it:
- Missing/undeclared credentials: The registry lists no required env vars, yet generate.py requires DASHSCOPE_API_KEY and SKILL.md documents additional Kling keys. Confirm which credentials are actually required and only provide them if you trust the endpoints.
- Hard-coded token: upload_image.py embeds IMGURL_TOKEN and UID defaults. Treat this as suspicious — the token may be valid and will send any uploaded images to a third-party account. Do not run the upload script with sensitive images until you remove or replace the hard-coded credentials. Prefer setting IMGURL_... via environment variables and rotate any tokens you own if they were exposed.
- Kling support mismatch: SKILL.md mentions Kling (Kling V2) but there is no implementation in the provided scripts. If you need Kling, request the author provide code or remove the claim.
- Missing TOOLS.md: the instructions reference a TOOLS.md for credentials/config that isn't included; ask the author for it or for clearer runtime requirements.
- Data exposure: both scripts make outbound network requests (to DashScope/Aliyun and ImgURL). If privacy of input images or generated videos matters, review the remote services' policies and avoid uploading sensitive content.
Recommended actions:
1) Inspect and remove the embedded IMGURL_TOKEN from upload_image.py; replace with reading from environment and require the user to set it.
2) Confirm the validity and least-privileged scope of any API keys used (rotate them if they were accidentally leaked).
3) Ask the publisher to update the skill manifest to declare required env vars and to either implement or remove Kling support.
4) Only run the scripts in a controlled environment and avoid uploading private images until you are satisfied with the destination and credential handling.Like a lobster shell, security has layers — review code before you run it.
latestvk97aksms3ycfxfw8nqs9m2j74d828bb2
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
