realitykit-animation-physics
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No override instructions, safety bypasses, or system prompt extraction attempts were found in the instructional content or metadata.
- [DATA_EXFILTRATION]: No hardcoded credentials, sensitive file path access, or unauthorized network operations were identified. The code snippets use standard RealityKit APIs for local resource management.
- [REMOTE_CODE_EXECUTION]: No remote script downloads, piped shell executions, or unverifiable external dependencies are present. All Swift code examples demonstrate legitimate framework usage.
- [OBFUSCATION]: No Base64 encoding, zero-width characters, homoglyphs, or other techniques to hide malicious intent were detected in any of the 21 files.
- [COMMAND_EXECUTION]: No dangerous shell command execution or use of the dynamic context injection syntax (
!command) was found. - [INDIRECT_PROMPT_INJECTION]: While the skill describes loading authored resources (Behavior Trees, Animation Graphs), these are standard structured RealityKit assets. There are no patterns suggesting the ingestion of untrusted natural language data that could influence agent behavior.
Audit Metadata