analyze-unrecognized-apps
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill performs network requests to Apple's official iTunes Lookup API (itunes.apple.com) to retrieve application metadata associated with host bundle IDs.
- [EXTERNAL_DOWNLOADS]: Instructions include fetching Apple App Site Association (AASA) files from external domains (e.g., https:///.well-known/apple-app-site-association) to resolve universal link fallbacks. While the target domains are dynamic, the activity is restricted to retrieving structured JSON configuration data.
- [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface for indirect prompt injection as it processes untrusted data from external sources, including Google Analytics reports and remote AASA files, which then inform code updates to the project's source files.
- Ingestion points: Google Analytics bundle ID lists (screenshots or text) and external AASA JSON files.
- Boundary markers: No specific delimiters or "ignore" instructions are present for the external data processing.
- Capability inventory: The skill can read and modify local source files (VivaDicta/VivaDictaApp.swift, SKILL.md) and perform network GET requests.
- Sanitization: No explicit sanitization logic is provided beyond the instruction to research and verify the fetched schemes before application.
Audit Metadata