tool invocation

LLM Agents

Diagnosing Tool Failures in Multi-Agent LLM Systems

New research introduces a systematic framework for identifying why LLM agents fail to invoke tools correctly, addressing a critical reliability gap in multi-agent AI systems.