The Agent Skills Directory

[SAFE]: The skill implements a rigorous decision-making framework ('What Next') focused on evidence-based planning and validation. It explicitly instructs the agent to avoid speculation, verify claims through runtime behavior or documentation, and maintain a log of concerns. There are no executable scripts, remote downloads, or credential-handling patterns present. The instructions for 'adversarial review' and 'validation spikes' are cognitive frameworks designed to improve task accuracy rather than bypass safety protocols.

what-next