data-warehouse-integration

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and links to several well-known technology services and official documentation, including Snowflake, Google Cloud BigQuery, Fivetran, Airbyte, Hightouch, Census, and dbt. These references are for documentation and service integration purposes.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to handle user inquiries regarding data architecture. It incorporates safety-oriented instructions for the agent, such as advising against running heavy analytics on production databases and recommending the use of managed ingestion tools for better reliability and security.
  • [CREDENTIALS_UNSAFE]: The skill provides templates for secret management using standard patterns, such as dbt variables for PII salts (e.g., {{ var("pii_salt") }}). It correctly avoids hardcoding any real credentials and encourages the use of environment-managed secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 03:48 AM
Security Audit — agent-trust-hub — data-warehouse-integration