The Agent Skills Directory

[SAFE]: The skill provides a functional tool for auditing cloud identity security configurations. The logic is transparent and aligned with the stated purpose of identity and access management security.
[COMMAND_EXECUTION]: The script scripts/agent.py is intended for manual execution by a user. It accepts credentials via command-line arguments to authenticate with the Microsoft Graph API. It does not perform any unauthorized or hidden command execution.
[DATA_EXFILTRATION]: All network operations are directed towards trusted Microsoft domains (login.microsoftonline.com and graph.microsoft.com). The script fetches policy configurations for local auditing and does not transmit data to any external or untrusted third-party servers.
[EXTERNAL_DOWNLOADS]: The skill depends on the standard requests Python library. It does not download or execute remote scripts or binaries.
[CREDENTIALS_UNSAFE]: There are no hardcoded secrets or credentials in the skill. It uses placeholders in documentation to guide the user on how to provide their own service principal credentials for the audit.

implementing-conditional-access-policies-azure-ad