push-notifications
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data from push notification payloads (userInfo) in multiple locations, which could be exploited via indirect prompt injection to influence application flow or network activity. \n- Ingestion points: Ingestion occurs in references/notification-patterns.md (BackgroundNotificationHandler), references/rich-notifications.md (NotificationService), and SKILL.md (NotificationDelegate and handleNotificationTap). \n- Boundary markers: The skill does not implement delimiters or specific instructions to ignore embedded instructions in the payload. \n- Capability inventory: The code demonstrates capabilities for network requests (URLSession), file writing (temporaryDirectory), and application navigation (DeepLinkRouter). \n- Sanitization: No sanitization or validation of the payload content is performed beyond basic Swift type casting.
Audit Metadata