The Agent Skills Directory

[SAFE]: The skill consists entirely of Markdown files for instructions and research documentation. No executable scripts, binaries, or configuration files capable of performing system operations were detected.
[PROMPT_INJECTION]: The skill establishes a specific persona for the agent. It includes explicit instructions for entering and exiting the role (e.g., 'user says exit/back to normal'), which is a security best practice for persona skills to prevent behavioral persistence. It does not attempt to override the platform's underlying safety or ethical guidelines.
[EXTERNAL_DOWNLOADS]: The documentation references multiple external sources for research purposes, including academic papers on arXiv, articles on Medium, and official project repositories on GitHub. All identified URLs are legitimate, clearly visible in plain text, and belong to well-known intellectual or developer platforms. No automated downloads of executable content are performed.

taleb-perspective