The Agent Skills Directory

[SAFE]: The skill instructions provide robust security guidelines, explicitly directing the agent to treat external inputs such as logs, manifests, and documentation as untrusted data and forbidding the printing or committing of sensitive credentials.
[SAFE]: All external references and documentation links target official, well-known, and reputable technology organizations or open-source projects.
[SAFE]: The skill incorporates explicit safety gates, such as requiring user authorization for any infrastructure mutation and forbidding the automatic installation of external tools or packages.

squad-devops