google-analytics-automation

Pass

Audited by Gen Agent Trust Hub on Feb 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill processes untrusted data from the Google Analytics API. (1) Ingestion points: Data entering via GOOGLE_ANALYTICS_RUN_REPORT and GOOGLE_ANALYTICS_LIST_ACCOUNTS. (2) Boundary markers: No explicit delimiters are used to separate GA4 data from agent instructions. (3) Capability inventory: The skill utilizes GA4 reporting and management tools; it does not request shell or file-write access. (4) Sanitization: No sanitization of API responses is defined.
  • [Data Exposure & Exfiltration] (SAFE): Sensitive data access is handled through authenticated OAuth connections managed via RUBE_MANAGE_CONNECTIONS. No unauthorized exfiltration or hardcoded credentials were found.
  • [Remote Code Execution] (SAFE): The skill consists of instructional content for existing tools and does not perform script execution or package installation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 20, 2026, 08:19 AM