social-creative-weapon

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill explicitly prioritizes user privacy by stripping EXIF/GPS metadata from photos through sharp re-encoding before any assets are committed or published.
  • [SAFE]: Input data from external sources (brand guides and post content) is sanitized using a dedicated escaping function to prevent injection attacks when generating SVG or HTML templates.
  • [SAFE]: Sensitive API keys for third-party services like kie.ai are managed through environment variables, avoiding credential exposure in source code.
  • [EXTERNAL_DOWNLOADS]: Downloads generated video content from the kie.ai API and utilizes rendering tools from the heygen-com GitHub repository, both of which are well-known services for video AI and automation.
  • [COMMAND_EXECUTION]: Executes local rendering tasks via npx hyperframes and standard file system operations for organizing generated social media assets.
  • [DATA_EXFILTRATION]: Communicates with the kie.ai API to send prompts and image URLs for video generation. This behavior is documented and central to the skill's primary function.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 03:23 AM
Security Audit — agent-trust-hub — social-creative-weapon