Multi-agent systems (MAS) powered by large language models have shown strong performance in code generation, mathematical reasoning, and question answering. Existing MAS automation methods usually operate at the task level, producing a single fixed workflow per benchmark and applying it uniformly to all queries. This assumption breaks down in realistic settings because query difficulty varies widely within a task and real workloads often mix heterogeneous task types. OOPMAS introduces a training‑free framework that generates both the agent set and the coordination workflow at the granularity of individual queries. Agents are expressed as object‑oriented class definitions with dedicated roles, tools, and persistent state, while workflows are written as executable main functions that operate on these agent objects. A dynamic skill library continuously accumulates structured lessons from execution feedback across optimization rounds, enabling in‑context improvement without any gradient updates or fine‑tuning. On a mixed‑task benchmark covering code, math, and QA queries, OOPMAS achieves 89.6% accuracy, outperforming the strongest baseline by 18.1 percentage points. A model‑swap study across four LLM backbones shows consistent scaling, reaching 92.4% with the strongest model.
Review: OOPMAS leverages object‑oriented abstraction and a dynamic skill repository to achieve query‑level flexibility, markedly boosting MAS performance in multi‑task environments and offering a fresh direction for large‑model‑driven system design.