A plug‑in layer for AI agents: safe memory, reusable skills, and traceable decisions
Universal Prompt Security Standard - Python Implementation
