azure-policy-advisor

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill implements legitimate Azure Policy auditing logic using standard Azure CLI tools and official Microsoft documentation.
  • [EXTERNAL_DOWNLOADS]: Fetches policy definitions and compliance guidance from Microsoft Learn (learn.microsoft.com). As this targets a well-known and official service for technical documentation, it is considered a safe operation.
  • [COMMAND_EXECUTION]: Uses Azure CLI commands to query the current state of policy assignments and definitions. These operations are read-only ('list' commands) and are necessary for the skill's primary function of assessing compliance.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from external files to determine its assessment logic.
  • Ingestion points: Reads compliance preferences from the '## Compliance & Azure Policy' section of 'copilot-instructions.md' and parses user-provided ARM templates.
  • Boundary markers: Uses a specific markdown header to scope the ingestion of instructions.
  • Capability inventory: Performs Azure CLI queries and Microsoft documentation searches based on the ingested configuration.
  • Sanitization: No explicit sanitization of the ingested text is described, though the output is restricted to audit recommendations.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 08:30 PM