The Agent Skills Directory

[SAFE]: The skill is entirely composed of documentation and RAG (Retrieval-Augmented Generation) artifacts. It contains no executable scripts (Python, JavaScript, or Shell).
[SAFE]: All instructions to the agent are persona-driven, focusing on business efficiency, cost control, and strategic philanthropy. There are no attempts to bypass safety filters, escalate privileges, or access sensitive system data.
[SAFE]: The external URLs provided in the reference and workspace files are purely for knowledge grounding and link to reputable educational, historical, and media institutions such as the Library of Congress, Wikipedia, PBS, and Carnegie.org.
[SAFE]: No obfuscation, hidden content, or dynamic context injections (!command syntax) were identified in any of the skill files.

andrew-carnegie