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
Risk Level
SAFE
Analyzed
Apr 7, 2026, 10:29 AM
Security Audit — agent-trust-hub — auth-patterns