3d-virtual-tour

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional and does not contain executable code, scripts, or automated dependency management.
  • [EXTERNAL_DOWNLOADS]: Contains references to official Three.js documentation and a public GitHub repository (MengTo/seijaku) for architectural patterns. These are documented for reference and do not involve automated execution or piping to a shell.
  • [PROMPT_INJECTION]: No attempts to override agent behavior, bypass safety guidelines, or extract system prompts were detected. The instructions follow a standard technical guidance format.
  • [DATA_EXFILTRATION]: No sensitive file access, network exfiltration commands (like curl/wget), or hardcoded credentials were found.
  • [OBFUSCATION]: No hidden content, Base64 encoding of commands, zero-width characters, or homoglyph attacks were detected in the files.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 04:45 PM