uloop-simulate-mouse-ui
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection through the processing of untrusted data from the Unity environment.
- Ingestion points: The skill ingests external data from user-provided $ARGUMENTS and UI element metadata (Label, Name, and Path) retrieved via the uloop screenshot command.
- Boundary markers: There are no delimiters or explicit instructions to the agent to disregard instructions potentially embedded in UI element names or labels.
- Capability inventory: The uloop simulate-mouse-ui tool can trigger interactive events and bypass raycasting to target internal Unity hierarchy paths, which could be misused if influenced by malicious input.
- Sanitization: No evidence of validation, filtering, or escaping is provided for the data interpolated into the agent's context from the Unity hierarchy.
Audit Metadata