omniverse-gaussian-splat-to-usd

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the Shell tool for executing conversion commands (usd-convert-gsplat) and running repository build scripts (repo.bat, repo.sh). These actions are consistent with the skill's stated purpose of facilitating 3D asset conversion and development workflows.
  • [EXTERNAL_DOWNLOADS]: The skill contains instructions to install Python packages from local paths (e.g., pip install ./source/python). This is a standard practice for installing dependencies bundled with a repository.
  • [PROMPT_INJECTION]: The skill processes external data in the form of .ply and .spz files.
  • Ingestion points: Untrusted data enters the agent context when reading .ply or .spz source files for conversion.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded content are provided for the input data.
  • Capability inventory: The skill has access to the Shell tool, enabling file system writes and command execution.
  • Sanitization: There is no explicit mention of sanitizing the content of the 3D data files before processing.
  • While these files are structured binary/numeric 3D data formats rather than natural language, they represent a potential attack surface for indirect prompt injection if the parser interprets metadata or comments as instructions. (Internal Severity: LOW)
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 10:59 PM
Security Audit — agent-trust-hub — omniverse-gaussian-splat-to-usd