unity-validation
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection surface identified through project data processing. 1. Ingestion points: Project hierarchy and asset metadata via validate_scene and validate_project_structure (SKILL.md). 2. Boundary markers: Absent in skill definition. 3. Capability inventory: File system exploration, folder deletion (validate_cleanup_empty_folders), and component modification (validate_fix_missing_scripts). 4. Sanitization: None defined for ingested project strings.
- [NO_CODE]: No executable code or scripts are provided with this skill definition, which limits the risk of direct malicious execution.
Audit Metadata