The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes official Azure CLI (az) and docker commands to perform diagnostic operations, such as streaming logs, checking registry settings, and investigating provisioning states. These actions are consistent with the skill's primary purpose of production troubleshooting and do not involve unauthorized privilege escalation.
[EXTERNAL_DOWNLOADS]: The skill references the installation of the resource-graph Azure CLI extension. As this is an official extension from Microsoft (a trusted vendor), the reference is safe and follows standard cloud administration workflows.
[DATA_EXFILTRATION]: Network operations are limited to health checks via curl targeting official Azure subdomains (azurecontainerapps.io). No patterns were found indicating the transmission of sensitive credentials or telemetry to untrusted third-party servers.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it analyzes external data. Ingestion points: Resource logs and metrics accessed via az containerapp logs show and mcp_azure_mcp_monitor. Boundary markers: None specified in the provided files. Capability inventory: Execution of CLI commands and KQL queries for system analysis. Sanitization: No explicit content filtering is implemented for log data. This surface is characteristic of log analysis tools, and no malicious override instructions were detected.

azure-diagnostics