rails-upgrade-4-to-5
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The instructions focus strictly on Rails upgrade patterns without attempting to override agent behavior or safety protocols.
- [DATA_EXPOSURE_EXFILTRATION]: The skill references standard Rails credential management (
credentials.yml.enc) which is a secure practice. No attempts to access sensitive system files or exfiltrate data to remote servers were found. - [REMOTE_CODE_EXECUTION]: No patterns for downloading and executing untrusted scripts (e.g.,
curl | bash) are present. External links point to official documentation and reputable industry blogs. - [INDIRECT_PROMPT_INJECTION]: The skill acts on user-provided code snippets related to Rails upgrades. While it lacks explicit boundary markers, the scope is limited to code migration guidance, presenting minimal risk.
Audit Metadata