fabriciq

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • Instructional Metadata Integration: The skill is designed to ingest and prioritize "Custom Instructions" and "Verified Answers" from Power BI semantic models. While this ensures the agent follows business-specific logic, it also creates a surface where external metadata influences the agent's reasoning process.
  • Dynamic Query Generation: The core functionality of the skill involves translating natural language into DAX queries executed against the Power BI environment. This process follows strict formatting and structure rules to ensure queries are valid and focused on the intended data scope.
  • Session Update Verification: The skill includes a mandatory procedure to check for updates at the start of a session, comparing local configuration versions with remote sources. This is a standard practice for maintaining the skill's operational integrity.
  • Restricted Tool Orchestration: The instructions provide specific guidance on using optimized discovery tools rather than broad administrative APIs, aligning with principles of efficiency and least-privilege access during artifact discovery.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 07:54 AM
Security Audit — agent-trust-hub — fabriciq