dax-authoring

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • Command Line Tool Execution: The skill utilizes command-line interfaces, specifically the fabric-app-data tool (via npx) and the Azure CLI (az login). These tools are used to authenticate, list, and query data models. While this involves executing commands on the host system, it is restricted to the specific functionality of interacting with the data platform as described in the skill's purpose.
  • Authentication Management: Instructions are provided for the agent to guide users through the az login process. This is the standard method for establishing a secure identity context when interacting with Azure-hosted resources like Microsoft Fabric.
  • External Software Installation: The documentation references official installation paths for necessary CLI tools (such as the Azure CLI). These references point to established vendor domains and are part of the standard environment setup required for the skill to operate.
  • Query Execution Surface: The skill translates natural language requests into DAX queries executed via the CLI. This pattern inherently processes user-influenced logic into code. The skill mitigates risks in this area by providing a deep reference of syntax rules, must/prefer/avoid guardrails, and specific query patterns to ensure generated queries remain within valid DAX boundaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 02:31 AM
Security Audit — agent-trust-hub — dax-authoring