The Agent Skills Directory

[SAFE]: The skill consists entirely of markdown documentation and instructional prompt templates intended to guide an AI agent's reasoning during evaluation tasks.\n
No executable files, scripts, binaries, or tool definitions are provided with this skill.\n
Analysis of the content confirms no network operations, data exfiltration attempts, or unauthorized file system access patterns.\n
No evidence of prompt injection, obfuscation, or persistence mechanisms was detected in the skill body or metadata reference files.

advanced-evaluation