discover
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of the
evoCLI andgitto manage the optimization lifecycle. This includes initializing the workspace (evo init), managing environment variables (evo env load), configuring runtime settings, and executing experiments (evo run). - [DYNAMIC_EXECUTION]: The skill facilitates the creation and execution of benchmark scripts and gate commands based on repository analysis. These commands are executed by the
evorunner during the optimization loop to evaluate experiment performance. - [INDIRECT_PROMPT_INJECTION]: The skill has a broad ingestion surface, reading repository metadata (READMEs, configuration files, and source code) to propose optimization dimensions. It also supports
[EVO DIRECTIVE]banners for runtime instruction overrides from the plugin host, which are treated as authoritative signals to guide agent behavior.
Audit Metadata