The Agent Skills Directory

Use of Official Documentation: The skill strictly uses learn.microsoft.com as the authoritative source for all technical procedures and command syntax. This ensures the information provided is vetted and originates from a trusted vendor repository.
Operational Safety Guardrails: It incorporates explicit 'Safety Rules' that require the agent to ask for user confirmation before performing actions that could impact system availability or data, such as resource deletion, network configuration changes, or installing updates.
Standard Tool Usage: The skill guides the user through standard Azure CLI and Azure Resource Graph (ARG) patterns for infrastructure management. It includes explicit warnings against embedding secrets in commands and recommends using secure alternatives like Key Vault or environment variables.
Least Privilege Principle: The instructions consistently advocate for applying least-privilege access using built-in Azure Role-Based Access Control (RBAC) roles for managing virtual machines and workloads.
No Malicious Patterns Detected: A thorough review found no evidence of prompt injection, data exfiltration, obfuscation, or unauthorized dynamic code execution.

azure-local