The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py executes Azure CLI commands using subprocess.run. The implementation passes arguments as a list rather than a shell string, which is a recommended security practice to prevent shell injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources via the Azure CLI (e.g., in scripts/agent.py and SKILL.md).
Ingestion points: Data enters the agent via az storage account list and az storage container list outputs.
Boundary markers: The script uses json.loads to process structured output, providing clear data boundaries.
Capability inventory: Subprocess calls are restricted to the az executable.
Sanitization: The script relies on the Azure CLI's own handling of arguments and the Python json module for safe parsing, minimizing the risk of processing malicious data embedded in storage account names or properties.
[EXTERNAL_DOWNLOADS]: The skill includes instructions to test for anonymous access using curl against *.blob.core.windows.net. This targets an official and trusted Microsoft domain, following the established trust scope for well-known services.

detecting-misconfigured-azure-storage