Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Aliyun Wan Video
v1.0.0Use when generating videos with Model Studio DashScope SDK using Wan video generation models (wan2.6-t2v, wan2.6-i2v-flash, wan2.6-i2v and regional variants)...
⭐ 0· 29·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill's stated purpose (Aliyun Wan video generation) matches the included Python scripts that call the DashScope SDK, so the capability itself is coherent. However, the registry metadata declares no required credentials while the SKILL.md and scripts clearly require DASHSCOPE_API_KEY (or credentials in ~/.alibabacloud/credentials). That mismatch (no declared env/primary credential but runtime code needing an API key) is unexpected and should be justified.
Instruction Scope
SKILL.md validation and output paths are inconsistent with the repository code: the validation command references skills/ai/video/aliyun-wan-video/scripts/generate_video.py (path not present) and SKILL.md states default output 'output/aliyun-wan-video/' while the scripts write to 'output/ai-video-wan-video/'. The instructions also instruct loading .env files and ~/.alibabacloud/credentials — legitimate for auth but broader in scope than the registry metadata indicates. These mismatches can cause validation failures and surprise behavior.
Install Mechanism
There is no install spec in the registry (instruction-only). The SKILL.md recommends installing the 'dashscope' Python package (pip). No remote downloads, installers, or archives are embedded in the skill, so install risk is low aside from the usual risk of third-party Python packages.
Credentials
The runtime expects a DASHSCOPE_API_KEY (env var) or a dashscope_api_key in ~/.alibabacloud/credentials; it also auto-loads .env files from CWD or detected repo root. The registry lists no required env vars or primary credential, so the skill's credential demands are not declared. Loading .env and user credentials can expose secrets from the developer/user environment if not understood.
Persistence & Privilege
The skill is not set to always:true and does not attempt to alter other skills or global agent configuration. It reads local files and writes output files under an output directory. Autonomous invocation is allowed (platform default) but is not combined with broad undeclared credentials or always:true, so no elevated persistence red flags beyond normal runtime I/O.
What to consider before installing
Key things to check before installing or running this skill:
- Authentication: The registry declares no required env vars, but the SKILL.md and scripts require DASHSCOPE_API_KEY (or dashscope_api_key in ~/.alibabacloud/credentials). Provide a dedicated, least-privilege API key for testing and do not reuse high-privilege keys.
- Path mismatches: The SKILL.md validation command and output paths don't match the included scripts' paths and output directory names. Expect the provided validation step to fail unless you adjust paths. Review and fix paths before relying on automated validation.
- .env and credentials loading: The scripts auto-load .env files (from cwd or repo root) and read ~/.alibabacloud/credentials. Audit those files for sensitive values you don't want the skill to pick up and consider running in an isolated environment or container.
- Third-party package: The skill uses the 'dashscope' Python package. Inspect that package (and pin a specific version) before pip installing in production.
- Test safely: Run the scripts in a disposable virtualenv and verify behavior (where files are written, what requests are made) before integrating into a pipeline.
If you want a safer posture, ask the skill author to: (1) declare DASHSCOPE_API_KEY as a required credential in the metadata, (2) fix the validation and output-path inconsistencies, and (3) document exactly which files it reads so you can audit them.Like a lobster shell, security has layers — review code before you run it.
latestvk97exvq6g4emkt495fp8y1bpw18414zc
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
