skills/earthtojake/text-to-cad/urdf/Gen Agent Trust Hub

urdf

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes the cadgen package from PyPI for its core validation functionality and leverages the Playwright framework to install the Chromium browser for rendering robot snapshots. These resources are either vendor-owned or well-known services.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and parse URDF XML and various mesh assets to perform validation and rendering. This activity is central to the skill's purpose.
  • Ingestion points: Robot description files and geometry assets are read from the project workspace.
  • Boundary markers: The instructions do not define specific delimiters when reading these files, as they are processed via structured XML/mesh parsers.
  • Capability inventory: The environment includes structural validation via cadgen and headless browser rendering via Playwright.
  • Sanitization: Content is validated against URDF structural rules and physical plausibility gates (inertial tensors, axis normalization) using the bundled validator.
  • [DYNAMIC_EXECUTION]: To support accurate physical modeling, the documentation provides templates for generating Python scripts that compute inertial properties from mesh data. This facilitates accurate mathematical derivation rather than guesswork.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 12:54 PM
Security Audit — agent-trust-hub — urdf