Anemoi: A Semi-Centralized Multi-agent System Based on Agent-to-Agent Communication MCP server from Coral Protocol
Xinxing Ren, Caelum Forder, Qianbo Zang, Ahsen Tahir, Roman J. Georgio, Suman Deb, Peter Carroll, Önder Gürcan, Zekun Guo
TL;DR
Anemoi introduces a semi-centralized multi-agent system that uses an Agent-to-Agent (A2A) communication MCP server to enable direct, structured collaboration among agents. By distributing coordination between a planning agent and multiple specialized workers, and allowing real-time plan updates and iterative refinement, Anemoi reduces reliance on a single planner and mitigates token-heavy context passing. On the GAIA benchmark, Anemoi achieves 52.73% accuracy with a small LLM planner (GPT-4.1-mini), outperforming the open-source baseline OWL (43.63%) by 9.09 percentage points under the same settings. The work demonstrates the practical viability of a thread-based, A2A coordination framework and provides a public implementation, highlighting improvements in scalability, robustness, and inter-agent dialogue while outlining avenues for future refinements in multi-agent collaboration.
Abstract
Recent advances in generalist multi-agent systems (MAS) have largely followed a context-engineering plus centralized paradigm, where a planner agent coordinates multiple worker agents through unidirectional prompt passing. While effective under strong planner models, this design suffers from two critical limitations: (1) strong dependency on the planner's capability, which leads to degraded performance when a smaller LLM powers the planner; and (2) limited inter-agent communication, where collaboration relies on prompt concatenation rather than genuine refinement through structured discussions. To address these challenges, we propose Anemoi, a semi-centralized MAS built on the Agent-to-Agent (A2A) communication MCP server from Coral Protocol. Unlike traditional designs, Anemoi enables structured and direct inter-agent collaboration, allowing all agents to monitor progress, assess results, identify bottlenecks, and propose refinements in real time. This paradigm reduces reliance on a single planner, supports adaptive plan updates, and minimizes redundant context passing, resulting in more scalable execution. Evaluated on the GAIA benchmark, Anemoi achieved 52.73% accuracy with a small LLM (GPT-4.1-mini) as the planner, surpassing the strongest open-source baseline OWL (43.63%) by +9.09% under identical LLM settings. Our implementation is publicly available at https://github.com/Coral-Protocol/Anemoi.