The Agent Skills Directory

[SAFE]: The skill operates entirely through natural language instructions and structured workflows. It contains no executable scripts (Python, Node.js) and does not request any tool permissions in the frontmatter.
[SAFE]: External links provided in references/empirical-foundations.md point to well-known academic, government (FTC), and research (RAND, PubMed) institutions. These are used solely for educational purposes to support the methodology.
[SAFE]: The workflow includes explicit 'High-stakes escalation' rules that advise users to seek human review and independent primary sources for medical, financial, or legal decisions, which is a security and safety best practice.
[SAFE]: The skill instructs the agent to avoid common pitfalls like assessing a speaker's body language or inferring intent, focusing instead on objective verifiable evidence.
[SAFE]: No prompt injection or obfuscation patterns were detected in the instructions or the evaluation files.

de-spin