editor-scripting-and-python
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides documentation and code examples for official Unreal Engine editor APIs. All identified patterns are consistent with the stated purpose of developer tooling and automation.
- [COMMAND_EXECUTION]: The skill describes the use of
UnrealEditor-Cmd.exefor running headless Python scripts via commandlets, which is a standard practice in game development CI/CD and automation workflows. - [REMOTE_CODE_EXECUTION]: The skill explains the use of the engine's built-in Python scripting plugin and the
IPythonScriptPluginC++ interface. These are intended features for developer productivity and do not involve execution of code from untrusted remote sources. - [PROMPT_INJECTION]: The reference material details methods for loading project assets (e.g.,
LoadAsset,LoadLevel), which inherently presents a surface for indirect prompt injection if an agent processes untrusted asset metadata. However, these are documented neutrally as standard API functions for editor scripting. \n - Ingestion points:
EditorAssetSubsystem::LoadAssetandLevelEditorSubsystem::LoadLevelinreferences/editor-subsystems.md. \n - Boundary markers: None provided in the documentation snippets. \n
- Capability inventory: Actor spawning, asset deletion, and Python command execution. \n
- Sanitization: Not addressed, as the content is primarily an API reference rather than an implementation of an ingestion tool.
Audit Metadata