auth-patterns
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No instructions attempting to bypass safety filters or override agent behavior were detected.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials or secrets were found. All examples correctly use environment variables for sensitive data like API keys and session secrets.
- [EXTERNAL_DOWNLOADS]: The skill references official GitHub API endpoints for legitimate OAuth2 flows, which are well-known and trusted services.
- [REMOTE_CODE_EXECUTION]: No patterns for remote script execution or arbitrary code evaluation were identified.
- [DATA_EXFILTRATION]: No evidence of unauthorized data transfer was found. Network operations are scoped to standard authentication protocols.
- [COMMAND_EXECUTION]: The skill does not contain any shell command execution or subprocess spawning patterns.
Audit Metadata