#llm security
topicllm security
2 papers · 1 filter
cs.CR2026
Agent Harness Distillation: Inference-Time Harness Extraction and Exploitation in Autonomous Multi-Agent Systems
Yu Cui, Wuli Yang, Yirui Shi +4
The paper presents Agent Harness Distillation (AHD), a method for extracting and replicating inference-time coordination mechanisms (harnesses) from autonomous multi-agent systems…
cs.CR2026
FlowGuard: From Signals to Evidence for MCP Security Detection
Baichao An, Pei Chen, Geng Hong +2
The paper introduces FlowGuard, a system that detects security risks in Model Context Protocol (MCP) interactions between LLM agents and external tools by combining semantic risk a…