The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions for the agent. It does not include scripts, commands, or automated actions that interact with the host system or external networks.
[SAFE]: No patterns of prompt injection, data exfiltration, or obfuscation were detected. Use of terms like 'CRITICAL' refers to the importance of design standards (e.g., anti-patterns) rather than security or safety bypasses.
[SAFE]: While the skill is designed to process external code (presenting an indirect prompt injection surface), it lacks the tools or capabilities (such as file writing, shell execution, or network access) to execute malicious instructions contained within that data.

audit