dataflows-authoring-cli

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • Command Execution: The skill facilitates the generation and execution of Bash and PowerShell scripts to perform CRUD operations on Microsoft Fabric resources. These operations utilize the official Azure CLI (az) and curl to interact with Microsoft API endpoints, which is appropriate for its stated function.
  • Data Ingestion Considerations: The skill processes external data, including workspace metadata and query result sets (via Apache Arrow). While this represents a surface for indirect prompt injection, the skill mitigates risk by using structured processing with jq and requiring human-in-the-loop verification of query results.
  • Secure Secret Management: The instructions explicitly caution against the use of plaintext credentials in scripts and provide clear implementation patterns for using passwordReference and secretReference to fetch credentials from secure stores.
  • External Resource References: The skill references public, well-known educational datasets (such as Titanic and Northwind) to provide runnable examples. These external references are used as data sources for the dataflow authoring workflow rather than as sources of executable code.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 02:44 PM
Security Audit — agent-trust-hub — dataflows-authoring-cli