skills/microsoft/rayfin/dax-authoring/Gen Agent Trust Hub

dax-authoring

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Command Execution: The skill instructs the agent to execute queries using the npx fabric-app-data CLI command. This involves running shell commands where dynamically generated query strings are passed as arguments, requiring careful handling to prevent command injection.\n- External Downloads: The skill utilizes npx to download and execute the fabric-app-data package from the npm registry. While this is a standard distribution method for developer tools, it involves fetching code from an external source at runtime.\n- Indirect Prompt Injection: The skill ingests results from external Power BI semantic models, which could potentially contain malicious instructions if the data source is untrusted.\n
  • Ingestion points: Results are ingested from the output of the npx fabric-app-data query command (documented in SKILL.md).\n
  • Boundary markers: No specific delimiters or instructions to ignore embedded content are provided for the tool output.\n
  • Capability inventory: The skill allows for command execution via npx and network access through the CLI tool.\n
  • Sanitization: The skill does not specify sanitization or escaping procedures for data returned from the queries before it enters the agent's context.\n- External References: The skill references official Azure CLI documentation and installation scripts from well-known Microsoft domains, which is a standard and expected practice.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 09:58 PM