ue-test-authoring
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive and technically accurate instructions for writing Unreal Engine automated tests using various frameworks (Automation, CQTest, Functional, etc.).
- [SAFE]: Integration with JetBrains Rider is handled through documented MCP tool patterns. The skill correctly instructs the agent to discover namespaced tools (e.g.,
mcp__<key>__execute_tool) using a standard search query, which is the established method for handling dynamic MCP server keys. - [SAFE]: The skill promotes security best practices by explicitly discouraging the use of general shell commands (like
grep,sed,cat) in favor of specialized, higher-level IDE tools when available. - [SAFE]: While the skill includes capabilities for project building (
build_solution_start) and running Python scripts within the Unreal Editor (ue_execute_python), these are standard functionalities for the intended developer workflow and are documented with appropriate usage constraints. - [SAFE]: The skill inherently possesses an indirect prompt injection surface because it reads and processes external source code from the project workspace. However, this surface is essential to its primary purpose as a test authoring assistant and does not constitute a malicious pattern.
Audit Metadata