azure-diagnostics

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • Log Ingestion and Analysis: The skill is designed to retrieve and analyze application logs and metrics using Kusto Query Language (KQL) and CLI commands. This process involves ingesting data that may originate from untrusted external sources (such as HTTP request headers or error messages). \n
  • Ingestion points: Data is pulled from AppExceptions and AppRequests tables via mcp_azure_mcp_monitor and from container logs via az containerapp logs show. \n
  • Boundary markers: The instructions do not specify explicit delimiters or "ignore" instructions for the ingested log content. \n
  • Capability inventory: The skill possesses the capability to modify Azure resources across its referenced guides, such as updating replica counts (az containerapp update) and setting registry configurations (az containerapp registry set). \n
  • Sanitization: No specific sanitization or escaping logic is described for the log data before it is presented to the agent. \n- Diagnostic Command Execution: The skill provides a variety of az CLI commands for troubleshooting Container Apps and monitoring resource health. These commands are standard administrative tools used within the Azure ecosystem for management and diagnostics. \n- Automated Command Generation: The skill uses MCP tools like mcp_azure_mcp_extension_cli_generate to create CLI commands from user intent. This allows for flexible diagnostic querying using the platform's native command generation capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 01:21 AM