unity-perception

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill module provides diagnostic tools for Unity scene and project analysis. The majority of functions are read-only (scene_analyze, scene_summarize, scene_health_check).
  • [COMMAND_EXECUTION]: While the module provides an interface to execute Unity-specific analysis skills through a hypothetical API, it does not invoke arbitrary shell commands or accessible subprocesses on the host system.
  • [EXTERNAL_DOWNLOADS]: No external URLs or third-party package managers (npm, pip, etc.) are referenced or utilized.
  • [DATA_EXFILTRATION]: There are no network operations or external data transmissions. The 'scene_export_report' and 'scene_dependency_analyze' skills write report files to the local project disk ('savePath'), which is standard behavior for a reporting tool.
  • [METADATA_POISONING]: The skill contains a 'Before calling any skill' warning and descriptions in both English and Chinese, but these are instructional rather than attempts to override system prompts or bypass safety guardrails.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 12:59 AM