AutoRAS: Learning Robust Agentic Systems with Primitive Representations
For researchers building LLM-based agentic systems, AutoRAS addresses the overlooked problem of robustness against adversaries and failures.
AutoRAS proposes a framework for automatically designing robust multi-agent systems by learning sequences of symbolic primitives that encode structure and behavior, achieving best performance in both vanilla and adversarial settings with minimal degradation under attacks.
The automated design of agentic systems offers a promising pathway for scaling large language models (LLMs) beyond single-agent reasoning. While prior work has advanced task performance through handcrafted or automatically generated multi-agent workflows, robustness is often treated as an afterthought, leaving systems vulnerable to external adversaries and internal failures. We propose AutoRAS, a framework for the Automated design of Robust Agentic Systems. AutoRAS formulates system design as generating a sequence of symbolic primitives that jointly encode structural connectivity and behavioral actions, and learns to optimize this sequence using execution-derived safety signals and flow-based sequence-level objectives. Extensive experiments show that AutoRAS achieves the best performance in both vanilla and adversarial settings, with the smallest performance degradation under attacks. Further analyses demonstrate strong transferability, stable optimization behavior, stability across primitive sets, and favorable cost trade-offs. Our code is available at $\href{https://github.com/guohezuy/AutoRAS}{\text{this https URL}}$.