defense-in-depth

SKILL.md

Defense in depth applies multiple security layers so that if one fails, others still protect the system. For AI applications, this means validating at every boundary: edge, gateway, input, authorization, data, LLM, output, and observability.

Core Principle: No single security control should be the only thing protecting sensitive operations.

| Defense in Depth | NIST | Multiple validation layers | | Zero Trust | Google BeyondCorp | Every request verified | | Least Privilege | AWS IAM | Minimal permissions | | Complete Mediation | Saltzer & Schroeder | Every access checked |

Úselo al crear canales de IA seguros o reforzar las integraciones de LLM. La defensa en profundidad implementa 8 capas de validación desde el borde hasta el almacenamiento sin ningún punto único de falla. Fuente: yonatangross/skillforge-claude-plugin.

Ver original