ToolEmu

ToolEmu is a benchmark dataset used to evaluate the effectiveness of poisoning attacks on graph-based agent memory systems.

Papers using ToolEmu (3)

SafeMCP: Proactive Power Regulation for LLM Agent Defense via Environment-Grounded Look-Ahead Reasoning2026ShadowMerge: A Novel Poisoning Attack on Graph-Based Agent Memory via Relation-Channel Conflicts2026Check Yourself Before You Wreck Yourself: Selectively Quitting Improves LLM Agent Safety2025