security_audit
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): The skill serves as a set of instructional guidelines for security audits.
- [DATA_EXPOSURE] (SAFE): While it identifies sensitive file paths (guards, workflows), it does so to instruct the agent not to modify them. No sensitive information or credentials are hardcoded.
- [COMMAND_EXECUTION] (SAFE): The allowed tools (Read, Grep, Glob) are limited to file system read operations. No shell execution, subprocess spawning, or write capabilities are requested.
- [EXTERNAL_DOWNLOADS] (SAFE): No remote resources, external scripts, or package installations are present.
- [PROMPT_INJECTION] (SAFE): No instructions to bypass safety filters or override system instructions were detected. The language is purely instructional for performing a task.
Audit Metadata