vinayaklatthe/microsoft-security-skills — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a Microsoft Security skills plugin that I can plug into coding agents like Cursor, Claude Code, Copilot, Codex, and Gemini. I want it to give much better, more specific guidance than generic security answers, using curated skills across Defender, Sentinel, Entra, Purview, Intune, Security Copilot, and Azure security scenarios.

Make it feel like a packaged capability layer for agents, not just a prompt list. Each skill should help with real decision making, recommended workflows, and guardrails, and it should stay grounded in public Microsoft Learn content only. Cover common areas like threat protection, identity and access, compliance and privacy, device management, cloud security, and AI security.

Please include a simple install experience, clear README instructions, a way to verify it is working with a few example prompts, and some kind of validation setup so I can test whether the skills improve the model’s answers. Keep it practical and usable by someone who wants reliable Microsoft security guidance inside their coding assistant.

Want more depth? Deep Reverse