grilling
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill defines a persona for stress-testing designs. It does not attempt to bypass system safety filters. However, it exhibits a surface for indirect prompt injection: 1. Ingestion points: Codebase files (referenced in SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: File system exploration (referenced in SKILL.md). 4. Sanitization: Absent.
- [COMMAND_EXECUTION]: The skill allows the agent to explore the local codebase. This is a legitimate function for the skill's purpose but involves the use of file-reading tools on project data.
Audit Metadata