physicalai-runtime-loading-exported-policies

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected in the skill instructions or referenced documents.
  • [REMOTE_CODE_EXECUTION]: The skill facilitates downloading model artifacts from Hugging Face Hub. Hugging Face is a well-known and trusted service for hosting machine learning models. The instructions promote security best practices, such as using pinned revisions and ensuring that authentication tokens are handled securely.
  • [COMMAND_EXECUTION]: The skill includes standard development commands using uv, a well-known Python package and environment manager, to run unit tests. This is a common and safe practice for validating policy loading logic.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external policy artifacts, including manifest.json files and model binaries. It mitigates potential risks associated with untrusted data by incorporating structured validation against defined schemas and using a registry system for model adapters.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 04:38 PM
Security Audit — agent-trust-hub — physicalai-runtime-loading-exported-policies