grill
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill provides instructions for a structured decision-making and documentation workflow. It emphasizes gathering facts from the repository before prompting the user.
- [COMMAND_EXECUTION]: The skill mentions running local formal checkers if provided by the target project. This refers to executing existing project-specific tools according to their local documentation, rather than external or arbitrary code.
- [DATA_EXPOSURE]: The skill instructs the agent to read repository files, issues, and documentation to inform the interview process. This is the intended functionality of the skill to ensure informed decision-making and does not involve exfiltration to external domains.
- [INDIRECT_PROMPT_INJECTION]: The skill processes repository content (code, docs, issues) which could theoretically contain malicious instructions. However, the skill treats this data as 'discoverable facts' to inform an interview and uses it to update internal project documentation (ADRs, glossaries) rather than as executable code. No specific vulnerabilities were identified.
Audit Metadata