# 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.

Навык · Бесплатно · Лицензия: MIT · Версия: 2026.9.22-cab634b

Установка через CLI NotHoly: `notholy skills install zhaoxuya520/llm-security`

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