analytics-tracking

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [PROMPT_INJECTION] (SAFE): The skill instructions are purely task-oriented and lack any attempts to bypass safety filters or override system instructions.
  • [DATA_EXFILTRATION] (SAFE): The documentation repeatedly emphasizes privacy compliance and the avoidance of collecting PII (Personally Identifiable Information). While it reads a local file for context, this is a standard behavior for localized agent customization.
  • [COMMAND_EXECUTION] (SAFE): No shell commands, system calls, or unauthorized script executions are present in the skill files.
  • [REMOTE_CODE_EXECUTION] (SAFE): JavaScript snippets for GA4 and GTM are provided as templates for the user's web implementation and are not executed by the agent itself.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 06:42 AM