profiling
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
unreal.SystemLibrary.execute_console_commandAPI to run engine-level performance commands such asstat dumpframe. While this allows execution of arbitrary Unreal console commands, it is used here for standard performance diagnostic functions within the engine environment. - [INDIRECT_PROMPT_INJECTION]: The
analyse()method reads engine logs and trace files to generate performance summaries. This represents an attack surface where untrusted data (such as actor names or network strings logged by the engine) could be ingested into the agent's context. - Ingestion points:
unreal.PerformanceService.analyse("logs")reads content from the Unreal Engine project logs (SKILL.md). - Boundary markers: None identified; logs are processed as raw diagnostic data.
- Capability inventory: The skill has the ability to execute Python code and Unreal console commands via
unreal.SystemLibrary.execute_console_command(SKILL.md). - Sanitization: No specific sanitization or filtering of log content is mentioned before it is parsed for aggregates.
Audit Metadata