mukul975/Anthropic-Cybersecurity-Skills ? reverse-engineered prompt

I want to build a big open-source project that's a collection of cybersecurity skills for AI agents. It should contain over 750 structured skills that tools like Claude, Copilot, or Cursor can use to perform security tasks.

The skills should cover a huge range of topics like penetration testing, digital forensics and incident response (DFIR), cloud security for AWS and Azure, malware analysis, and threat intelligence. I want each skill to be a structured markdown file, kind of like a playbook, with clear steps and maybe some metadata at the top so an AI can easily discover and execute it.

The whole collection should be mapped to an industry standard like the MITRE ATT&CK framework. Organize the skills into logical folders based on their domain, like "Cloud Security" or "Network Security". The goal is to create a go-to resource that gives AI agents the kind of practical knowledge senior security pros have.