The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected.
[COMMAND_EXECUTION]: The skill uses the Bash tool but restricts execution specifically to the gemini command using the allowed-tools configuration in the frontmatter. This follow best practices by preventing the agent from executing arbitrary or dangerous shell commands.
[DATA_EXFILTRATION]: The skill is designed to send prompt context to Google Gemini for processing. This is consistent with its stated purpose. No patterns of sensitive file harvesting or exfiltration to unknown external domains were found.

using-gemini