scene-list-opened
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends installing the
unity-mcp-clipackage via the npm registry (npm install -g unity-mcp-cli) to enable tool execution. - [COMMAND_EXECUTION]: The skill operates by executing the
unity-mcp-clibinary in the shell to interact with the Unity Editor. - [INDIRECT_PROMPT_INJECTION]: The skill ingests scene metadata from an external environment (Unity Editor), which represents a potential attack surface if the project files contain malicious naming patterns.
- Ingestion points: Scene names, paths, and instance IDs returned by the
unity-mcp-clitool (SKILL.md). - Boundary markers: None provided to distinguish returned data from agent instructions.
- Capability inventory: Execution of external CLI tools for data retrieval (SKILL.md).
- Sanitization: No evidence of output validation or sanitization for external scene strings.
Audit Metadata