MyAG: A Graph-Based Framework for Designing and Analyzing Composable LLM Agent Systems
For developers of LLM agent systems, MyAG provides a structured approach to system design and analysis, but the contribution is incremental as it formalizes existing practices.
MyAG introduces a graph-based framework for designing and analyzing composable LLM agent systems, enabling flexible reuse of components and hierarchical composition. Experiments demonstrate support for flexible design and analysis of performance-efficiency tradeoffs.
We present MyAG, a graph-based framework for designing and analyzing composable LLM agent systems. Our framework separates agent system construction into three graph abstractions: a component graph for agents, environments, and modules; a workflow graph for execution control; and a search graph for runtime execution. This separation allows users to flexibly reuse the same components with different strategies. We further support hierarchical composition through recursive system nodes and provide monitoring and visualization tools for inspecting agent execution. Experiments on representative agent applications show that our framework supports flexible agent system design and helps analyze performance-efficiency tradeoffs. Our framework is publicly available and fully open-source.