This paper presents AgentFlow, a MAS-based framework for programmable distributed systems in heterogeneous cloud-edge environments. It introduces logistics objects and abstract agent interfaces to enable dynamic service flows and modular orchestration. AgentFlow supports decentralized publish-subscribe messaging and many-to-many service elections, enabling decision coordination without a central server. It features plug-and-play node discovery, flexible task reorganization, and highly adaptable fault tolerance and substitution mechanisms. AgentFlow advances scalable, real-time coordination for resilient and autonomous mission-critical systems.
@article{arxiv.2505.07603,
title = {AgentFlow: Resilient Adaptive Cloud-Edge Framework for Multi-Agent Coordination},
author = {Ching Han Chen and Ming Fang Shiu},
journal= {arXiv preprint arXiv:2505.07603},
year = {2025}
}