SafeHarness is a lifecycle-integrated security architecture for LLM agents that cuts unsafe behavior rate by 38% and attack success rate by 42% via four coordinated layers while keeping task utility intact.
label": <int>, // -1=FAILED, 0=SAFE, 1=PARTIALLY_UNSAFE, 2=FULLY_UNSAFE
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SafeHarness: Lifecycle-Integrated Security Architecture for LLM-based Agent Deployment
SafeHarness is a lifecycle-integrated security architecture for LLM agents that cuts unsafe behavior rate by 38% and attack success rate by 42% via four coordinated layers while keeping task utility intact.