push-notifications

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill provides templates for processing remote notification payloads (APNs), which serves as a significant ingestion point for untrusted data that influences application logic.
  • Ingestion points: Data enters the application context via didReceiveRemoteNotification in AppDelegate (File: references/notification-patterns.md) and the didReceive method in NotificationServiceExtension (File: references/rich-notifications.md).
  • Boundary markers: No explicit boundary markers or sanitization layers are used beyond standard Swift type casting for dictionary values.
  • Capability inventory: The skill enables capabilities based on payload content, including UI navigation via DeepLinkRouter, background data synchronization via DataSyncService, and external media retrieval via URLSession (File: references/notification-patterns.md, references/rich-notifications.md).
  • Sanitization: The implementation lacks formal sanitization or validation for the incoming payload keys (e.g., action, imageUrl, chatId) before they are used to trigger internal services.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 05:20 AM
Security Audit — agent-trust-hub — push-notifications