ai-image-generation

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes and runs remote code via npx (e.g., "npx -y luxin-cli@latest create ...") which will fetch and execute the Luxin CLI that talks to the hosted service at https://api.luxin.sh (and is tied to https://luxin.sh), so runtime execution of externally-fetched code and reliance on that hosted API is evident.