Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting

LLM-based agents are increasingly deployed in multi-agent systems (MAS). As these systems move toward real-world applications, their security becomes paramount. Existing research largely evaluates single-agent security, leaving a critical gap in understanding the vulnerabilities introduced by multi-agent design. However, existing evaluation approaches fall short due to lack of unified frameworks and metrics focusing on unique rejection modes in MAS. We present SafeAgents, a unified and extensible framework for fine-grained security assessment of MAS. SafeAgents systematically exposes how design choices such as plan construction strategies, inter-agent context sharing, and fallback behaviors affect susceptibility to adversarial prompting. We introduce Dharma, a diagnostic measure that helps identify weak links within multi-agent pipelines. Using SafeAgents, we conduct a comprehensive study across five widely adopted multi-agent architectures (centralized, decentralized, and hybrid variants) on four datasets spanning web tasks, tool use, and code generation. Our central finding is that MAS security failures are not random—they are predictable consequences of specific, identifiable design choices. We trace failures to three recurring weak links: atomic-instruction delegation that hides harmful intent from sub-agents, missing planner fallbacks that turn refusals into execution, and stratified plans executed without re-evaluation. These results argue for security-aware design rather than post-hoc safeguards. Link to code is https://github.com/microsoft/SafeAgents