Russian Federation
UDC 004.75
The aim of this work is to analyze the architectural components of LLM-based multi-agent systems. The research method involves a comparative analysis of MetaGPT, Generative Agents, AutoGen, OrgAgent, MIRIX, and ChatDev architectures against a uniform set of criteria (role organization, memory, coordination, structured outputs, communication, quality control), followed by integration of the components into a single system and experimental evaluation on two task types in single-agent and multi-agent modes. It is found that on the examined tasks the multi-agent mode underperforms the single-agent baseline in quality (scores of 3,0 and 4,5 versus 9,0 on a ten-point scale). The architectural cause is identified: a mismatch between task type and the preconfigured action chain. A pattern is discovered: the integration of three verification loops – a failure detector, a compliance check, and a decision-making loop – produces a self-diagnosis capability not observed in any of the examined systems individually. The three loops jointly identified result unreliability and refused to deliver it. This property differs from known multi-agent systems diagnostic approaches where analysis is performed after task completion. The results may be applied when designing multi-agent systems for tasks where the cost of error is high.
multi-agent systems, large language models, agent architecture, agent memory, coordination, self-diagnosis
1. Wooldridge M. An Introduction to MultiAgent Systems. Wiley, 2009. 484 p.
2. Mincberg G. Struktura v kulake: sozdanie effektivnoj organizacii / per. s angl. SPb.: Piter, 2004. 512 s.
3. Large Language Model based Multi-Agents: A Survey of Progress and Challenges / T. Guo [et al.] // arXiv:2402.01680. 2024.
4. Multi-Agent Collaboration Mechanisms: A Survey of LLMs / K.-T. Tran [et al.] // arXiv:2501.06322. 2025.
5. MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework / S. Hong [et al.] // Proc. ICLR. 2024. arXiv:2308.00352.
6. ChatDev: Communicative Agents for Software Development / C. Qian [et al.] // Proc. ACL. arXiv:2307.07924. 2024.
7. OrgAgent: Organize Your Multi-Agent System like a Company / Wang [et al.] // arXiv:2604.01020. 2026.
8. Generative Agents: Interactive Simulacra of Human Behavior / J.S. Park [et al.] // Proc. ACM UIST. arXiv:2304.03442. 2023.
9. MIRIX: Multi-Agent Memory System for LLM-Based Agents / Wang [et al.] // arXiv:2507.07957. 2025.
10. AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation / Q. Wu [et al.] // arXiv:2308.08155. 2023.
11. Why Do Multi-Agent LLM Systems Fail? / M. Cemri [et al.] // Proc. NeurIPS (Datasets and Benchmarks Track). arXiv:2503.13657. 2025.
12. Role-Based Access Control Models / R. Sandhu [et al.] // IEEE Computer. 1996. Vol. 29. № 2. P. 38–47.
13. Towards Self-Improving Error Diagnosis in Multi-Agent Systems / Zhang [et al.] // arXiv:2604.17658. 2026.
14. DiLLS: Interactive Diagnosis of LLM-based Multi-agent Systems via Layered Summary / R. Sheng [et al.] // Proc. CHI. arXiv:2602.05446. 2026.
