The Agent Skills Directory

Command Execution: The skill generates and executes shell commands using the Azure CLI (az graph query).
Evidence: Use of extension_cli_generate to build az commands in SKILL.md.
Context: This is the primary mechanism for the skill's functionality. The commands are intended for read-only resource discovery.
Dynamic Execution: The skill dynamically assembles CLI commands at runtime based on user intent.
Evidence: The workflow in SKILL.md describes using extension_cli_generate to build a command string that is subsequently executed.
Context: This allows the skill to handle a wide variety of resource types without predefined templates for every case.
Indirect Prompt Injection: The skill processes natural language input to generate structured queries, which represents a potential attack surface.
Ingestion points: User requests are processed by extension_cli_generate to build CLI commands.
Boundary markers: The skill does not explicitly define delimiters for untrusted input within the generated command strings.
Capability inventory: The skill utilizes az graph query to retrieve resource metadata across Azure subscriptions.
Sanitization: Safety relies on the underlying agent's command generation logic and the read-only nature of the Azure Resource Graph service.
External Extension Requirement: The skill references an external Azure CLI extension.
Evidence: az extension add --name resource-graph in SKILL.md.
Context: The extension is an official component provided by the platform vendor for Resource Graph queries.

azure-resource-lookup