AgentAsk: Multi-Agent Systems Need to Ask

Authors: Bohan Lin, Kuo Yang, Zelin Tan, Yingchuan Lai, Chen Zhang, Guibin Zhang, Xinlei Yu, Miao Yu, Xu Wang, Yudong Zhang, and Yang Wang.

AgentAsk identifies edge-level error propagation as a major failure mode in multi-agent systems and introduces a lightweight clarification module that intervenes at message handoffs. Across five benchmarks, it improves accuracy while keeping latency and cost overhead low.

Links: ACL Anthology · arXiv · DOI