structural-physics

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill contains no instructions designed to override AI safety filters, extract system prompts, or bypass behavioral constraints.
  • [DATA_EXFILTRATION]: No network calls or sensitive file system access patterns were detected. The skill processes local geometry and physics data without external transmission.
  • [EXTERNAL_DOWNLOADS]: Dependency usage is limited to the well-known and trusted 'three' library. All other script imports are relative and local to the skill package.
  • [REMOTE_CODE_EXECUTION]: There are no patterns for remote script fetching or dynamic execution (e.g., eval or Function constructors).
  • [COMMAND_EXECUTION]: The scripts do not interact with the operating system shell or spawn subprocesses.
  • [CREDENTIALS_UNSAFE]: No hardcoded API keys, secrets, or authentication tokens were found in the codebase or metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 09:47 AM