Automated Composition of Agents: A Knapsack Approach for Agentic Component Selection
2510.16499v1
cs.CL, cs.AI, cs.LG
2025-10-22
Авторы:
Michelle Yuan, Khushbu Pahwa, Shuaichen Chang, Mustafa Kaba, Jiarong Jiang, Xiaofei Ma, Yi Zhang, Monica Sunkara
Abstract
Designing effective agentic systems requires the seamless composition and
integration of agents, tools, and models within dynamic and uncertain
environments. Most existing methods rely on static, semantic retrieval
approaches for tool or agent discovery. However, effective reuse and
composition of existing components remain challenging due to incomplete
capability descriptions and the limitations of retrieval methods. Component
selection suffers because the decisions are not based on capability, cost, and
real-time utility. To address these challenges, we introduce a structured,
automated framework for agentic system composition that is inspired by the
knapsack problem. Our framework enables a composer agent to systematically
identify, select, and assemble an optimal set of agentic components by jointly
considering performance, budget constraints, and compatibility. By dynamically
testing candidate components and modeling their utility in real-time, our
approach streamlines the assembly of agentic systems and facilitates scalable
reuse of resources. Empirical evaluation with Claude 3.5 Sonnet across five
benchmarking datasets shows that our online-knapsack-based composer
consistently lies on the Pareto frontier, achieving higher success rates at
significantly lower component costs compared to our baselines. In the
single-agent setup, the online knapsack composer shows a success rate
improvement of up to 31.6% in comparison to the retrieval baselines. In
multi-agent systems, the online knapsack composer increases success rate from
37% to 87% when agents are selected from an agent inventory of 100+ agents. The
substantial performance gap confirms the robust adaptability of our method
across diverse domains and budget constraints.
Ссылки и действия
Дополнительные ресурсы: