The Agent Skills Directory

Resource Modification Capabilities: The skill provides templates and instructions for executing Azure CLI and PowerShell commands to remediate compliance findings. These commands include modifying storage account settings, updating Key Vault configurations, and altering network settings. While these are intended for security hardening, their execution within an environment requires appropriate oversight to ensure they align with organizational policies.
Indirect Prompt Injection Surface: The assessment workflow involves analyzing external data sources, specifically Excel reports generated by the azqr tool and metadata from Key Vault resources. As with any skill that ingests external content, there is a potential for indirect prompt injection if the scanned resources contain malicious data designed to influence the agent's logic.
Ingestion points: azqr scan results (Excel sheets such as 'ImpactedResources' in references/azure-quick-review.md) and Key Vault metadata (expiresOn, subject in references/azure-keyvault-expiration-audit.md).
Boundary markers: The instructions do not explicitly define boundary markers for external data processing.
Capability inventory: The skill utilizes MCP tools for Azure resource management and provides instructions for generating CLI commands to modify infrastructure via references/azqr-remediation-patterns.md.
Sanitization: Explicit sanitization or validation of the ingested scan data is not detailed in the provided instructions.
Authentication Best Practices: The skill correctly identifies and recommends the use of Managed Identities for production environments and DefaultAzureCredential for local development in references/auth-best-practices.md, which aligns with industry-standard security protocols for Azure resource access.

azure-compliance