audit-repository-acrazie
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates on a read-only basis for repository inspection. It explicitly restricts the agent from performing record writes, implementation, or installation without distinct and immediate user approval.
- [SAFE]: Instructions include strict prohibitions against sensitive data exposure, requiring the agent to quote only necessary evidence and recommend dedicated security audits if vulnerabilities are incidentally discovered.
- [SAFE]: The skill defines a clear 'Project loading boundary' to prevent unauthorized invocation of dependencies or unactivated skills, ensuring that each step of the process is gated by user consent.
- [SAFE]: The included Python test script performs standard static analysis on instruction files using built-in libraries, with no evidence of remote code execution or unsafe dynamic behavior.
Audit Metadata