verifying-external-behavior
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides best-practice guidelines for integration testing and verifying external system behaviors.
- [COMMAND_EXECUTION]: The skill includes examples of using
uv,curl, anddocker. These are standard tools used within a development context for their intended purposes, such as package management, network probing, and running containers. - [EXTERNAL_DOWNLOADS]: The skill references a public GitHub repository and uses package managers to fetch libraries for throwaway probes. These are standard development activities involving well-known services.
- [DYNAMIC_EXECUTION]: The skill demonstrates monkeypatching for debugging purposes (specifically intercepting
requests.Session.request). This is a common debugging technique used to inspect outbound traffic during local testing and does not involve execution of untrusted remote code.
Audit Metadata