AeonDave/malskill ? reverse-engineered prompt

Build me a clean Python based repo for an agent skill library focused on offensive security workflows.

I want each skill to live in its own folder with a simple SKILL.md that teaches an AI agent what to do, what not to do, and how to stay focused on the task. Organize it so there are separate areas for tools, techniques, roles, coding, knowledge, behaviours, AI, and hardware, with the offensive security content front and center.

Make the structure feel practical for real agent use, not like a textbook. The skills should be short, high signal, and easy to load on demand. Include install scripts for Linux and Windows, a logo or basic assets folder, and a README that explains the purpose and how the collection is meant to be used.

If you need to, look up the current AgentSkills style online so the format matches common expectations. I mainly want the repo scaffolded in a way that makes it easy to add more skills later and use them from an agent right away.