The Agent Skills Directory

[SAFE]: The skill is designed for legitimate security auditing and architecture configuration. No malicious code or instructions were found.\n- [DATA_EXFILTRATION]: The auditing script and documentation examples interact exclusively with trusted Microsoft domains (graph.microsoft.com and cloudappsecurity.com). It retrieves security posture data and creates a local report, which is consistent with its stated purpose.\n- [COMMAND_EXECUTION]: The provided PowerShell and Bash snippets are administrative examples for setting up Conditional Access and CASB policies. These are transparent, educational, and require authorized access to execute.\n- [EXTERNAL_DOWNLOADS]: The tool relies on the standard requests library for Python. There are no patterns indicating the download or execution of remote, untrusted code.

implementing-zero-trust-for-saas-applications