Large Language Models (LLMs) have enabled collaborative Multi-Agent (MA) systems, where interacting agents improve performance through diverse reasoning and iterative refinement. However, these systems remain vulnerable to error propagation, where early-stage information degrades downstream reasoning. To address this, we conduct a systematic analysis of inter-agent communication to identify which information drives MA performance. We find that the absence of reasoning and verification in inter-agent communication significantly degrades performance. Based on these insights, we propose Category-Aware Recovery Augmentation (technique), which enforces the presence of critical information during communication. recovers up to 86.2% of failed cases. Our results highlight the key role of information quality in effective MA collaboration. Our code is available at https://anonymous.4open.science/r/cara_mas

翻译：大型语言模型（LLMs）已赋能协作式多智能体（Multi-Agent, MA）系统，其中交互的智能体通过多样化推理与迭代优化提升性能。然而，这类系统仍易受错误传播影响，即早期阶段的信息会降低后续推理质量。为应对此问题，我们对智能体间通信进行系统性分析，以识别驱动多智能体系统性能的关键信息。研究发现，智能体间通信中缺乏推理与验证环节会显著降低系统性能。基于此洞察，我们提出类别感知恢复增强技术（Category-Aware Recovery Augmentation），该技术通过强制在通信过程中保留关键信息，可恢复高达86.2%的失败案例。研究结果凸显了信息质量在高效多智能体协作中的核心作用。我们的代码已开源至 https://anonymous.4open.science/r/cara_mas