domains / families / security / llm-security

llm-security

authentic

Authorized security assessment of LLM applications and AI agents: prompt injection, tool abuse, RAG exposure, memory poisoning, system-prompt extraction, and agent-compliance engineering per OWASP LLM/ASI Top 10.

When to use

• Red-teaming an LLM-based application within an approved scope.
• Mapping agent tool permissions against abuse scenarios.

When not to use

• Model behavior is nondeterministic; findings need repeated trials.
• Provider-side safeguards may change without notice.

Skill page

Source