ai-analytics-saas
Pass
Audited by Gen Agent Trust Hub on Apr 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill documents best practices for AI integration within a SaaS platform. Findings include:
- Data Isolation: The code samples demonstrate mandatory multi-tenant scoping (e.g., enforcing
franchise_idin SQL and vector storage) to prevent cross-tenant data exposure. - Injection Mitigations: The provided PHP services use boundary markers (delimiters) and explicit 'user input as data' instructions to mitigate prompt injection.
- Least Privilege: The skill recommends the use of a dedicated read-only MySQL user (
ai_readonly) and includes regex-based SQL validation to ensure only SELECT statements are executed. - Privacy Controls: Explicit instructions are included to avoid sending PII to external LLMs by using aggregated metrics.
Audit Metadata