architecture-single-responsibility-principle
Pass
Audited by Gen Agent Trust Hub on Feb 24, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill was evaluated across ten threat categories, including persistence, privilege escalation, and data exfiltration, and no security issues were found.
- [EXTERNAL_DOWNLOADS]: The skill suggests installing 'radon' and 'pylint' for complexity metrics. These are well-known, trusted industry tools from official package registries, which is consistent with the skill's purpose.
- [COMMAND_EXECUTION]: A provided Bash script is used to execute AST searches and metrics calculations locally. This is a legitimate and expected use of command-line tools for static code analysis.
- [PROMPT_INJECTION]: The skill processes untrusted code files, which is an inherent risk surface for indirect prompt injection. Ingestion points: File reading and AST search tools (Grep, Glob, ast-grep). Boundary markers: Absent. Capability inventory: Bash and AST search tools. Sanitization: Absent. The analysis is limited to structural pattern matching, which effectively mitigates the risk of executing embedded instructions.
Audit Metadata