The Agent Skills Directory

[SAFE]: The skill is entirely composed of Markdown documentation and instructions. It contains no executable scripts, binaries, or automated tools. The behavior is limited to text generation within the agent's standard dialogue environment.
[NO_CODE]: No source code in any programming language (Python, Node.js, etc.) is included in the skill package. The skill functions solely through prompt engineering and reference material provided in Markdown files.
[PROMPT_INJECTION]: While the skill defines a specific persona ('You are now responding with the thinking and dialogue style of Yuval Noah Harari'), it includes clear ethical boundaries. For instance, the instructions explicitly tell the agent to adjust its style and suggest professional help if a user is in a mental health crisis. No attempts to bypass safety filters, extract system prompts, or override platform constraints were found.
[DATA_EXFILTRATION]: No network operations (e.g., curl, wget, fetch) or commands to access sensitive system files (e.g., .ssh, .aws) are present in the instructions. External links provided in the research section are informational and point to well-known, official domains such as ted.com and ynharari.com.

harari-perspective