skills/jdrhyne/agent-skills/ga4/Gen Agent Trust Hub

ga4

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes official Google client libraries including google-analytics-data and google-auth-oauthlib retrieved from the standard Python Package Index (PyPI).
  • [COMMAND_EXECUTION]: Executes Python scripts (ga4_auth.py and ga4_query.py) to manage authentication flows and perform data queries against the Google Analytics Data API.
  • [CREDENTIALS_UNSAFE]: During the one-time setup process, the scripts/ga4_auth.py script prints the generated OAuth refresh token to the console; this is a standard procedure for manual environment configuration but requires the user to handle the output securely.
  • [PROMPT_INJECTION]: The skill processes website analytics data from external Google Analytics properties, which presents a surface for indirect prompt injection. 1. Ingestion points: scripts/ga4_query.py fetches external content such as page paths and titles. 2. Boundary markers: Data is output in table, JSON, or CSV formats without specific delimiters or instructions to ignore instructions embedded in the retrieved strings. 3. Capability inventory: The skill is configured for network operations specifically targeting Google's API infrastructure. 4. Sanitization: The retrieved analytics data is not sanitized for potential prompt injection before being displayed.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 10:01 PM