authentication
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface by processing authentication responses and callback URLs from external services.
- Ingestion points: Credentials such as identity tokens and authorization codes are received through ASAuthorizationControllerDelegate methods in SKILL.md and references/passkeys.md, and OAuth redirection is handled via ASWebAuthenticationSession.
- Boundary markers: The documentation explicitly instructs that client-side data must not be trusted and must be validated server-side against Apple's JWKS.
- Capability inventory: The skill includes patterns for network communication to servers and persistent storage in the system Keychain using SecItemAdd.
- Sanitization: Data processing relies on native iOS framework types and standard parsing logic without additional sanitization layers.
Audit Metadata