# llm-security

> Use for authorized security assessment of LLM applications and AI agents, including prompt injection, tool abuse, RAG exposure, memory poisoning, and model supply-chain risks.

Skill · Gratis · Licencia: MIT · Versión: 2026.9.22-cab634b

Instalar con la CLI de NotHoly: `notholy skills install zhaoxuya520/llm-security`

Canonical: https://notholy.com/es/skills/zhaoxuya520/llm-security
