deep-linking

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses standard shell commands like grep, find, and xcrun via the Bash tool to inspect the local project for existing deep linking configurations in Info.plist and entitlements files. These operations are restricted to the local project context.
  • [EXTERNAL_DOWNLOADS]: Includes a diagnostic curl command example to verify the availability and headers of the Apple App Site Association (AASA) file from a server. This is a standard validation step for Universal Links.
  • [PROMPT_INJECTION]: The skill instructions were evaluated for behavior override patterns, safety filter bypasses, or role-play injections; no such patterns were identified.
  • [DATA_EXFILTRATION]: No evidence of hardcoded credentials, sensitive file access (e.g., SSH keys), or unauthorized network operations was found. Placeholder values like yourapp.com and myapp:// are used appropriately for templating.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 01:18 PM
Security Audit — agent-trust-hub — deep-linking