<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article
PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.4 20190208//EN"
       "JATS-journalpublishing1.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" article-type="research-article" dtd-version="1.4" xml:lang="en">
 <front>
  <journal-meta>
   <journal-id journal-id-type="publisher-id">Scientific and analytical journal «Vestnik Saint-Petersburg university of State fire service of EMERCOM of Russia»</journal-id>
   <journal-title-group>
    <journal-title xml:lang="en">Scientific and analytical journal «Vestnik Saint-Petersburg university of State fire service of EMERCOM of Russia»</journal-title>
    <trans-title-group xml:lang="ru">
     <trans-title>Научно-аналитический журнал &quot;Вестник Санкт-Петербургского университета ГПС МЧС России&quot;</trans-title>
    </trans-title-group>
   </journal-title-group>
   <issn publication-format="online">2218-130X</issn>
  </journal-meta>
  <article-meta>
   <article-id pub-id-type="publisher-id">128595</article-id>
   <article-id pub-id-type="doi">10.61260/2218-130Х-2026-2-111-119</article-id>
   <article-categories>
    <subj-group subj-group-type="toc-heading" xml:lang="ru">
     <subject>ИНФОРМАТИКА, ВЫЧИСЛИТЕЛЬНАЯ ТЕХНИКА И УПРАВЛЕНИЕ</subject>
    </subj-group>
    <subj-group subj-group-type="toc-heading" xml:lang="en">
     <subject>INFORMATICS, COMPUTER ENGINEERING AND CONTROL</subject>
    </subj-group>
    <subj-group>
     <subject>ИНФОРМАТИКА, ВЫЧИСЛИТЕЛЬНАЯ ТЕХНИКА И УПРАВЛЕНИЕ</subject>
    </subj-group>
   </article-categories>
   <title-group>
    <article-title xml:lang="en">ANALYSIS OF ARCHITECTURAL COMPONENTS  OF MULTI-AGENT SYSTEMS</article-title>
    <trans-title-group xml:lang="ru">
     <trans-title>АНАЛИЗ АРХИТЕКТУРНЫХ КОМПОНЕНТОВ  МУЛЬТИАГЕНТНЫХ СИСТЕМ</trans-title>
    </trans-title-group>
   </title-group>
   <contrib-group content-type="authors">
    <contrib contrib-type="author">
     <contrib-id contrib-id-type="orcid">https://orcid.org/0000-0003-4810-8734</contrib-id>
     <name-alternatives>
      <name xml:lang="ru">
       <surname>Саратова</surname>
       <given-names>Татьяна Евгеньевна </given-names>
      </name>
      <name xml:lang="en">
       <surname>Saratova</surname>
       <given-names>Tatyana Evgen'evna </given-names>
      </name>
     </name-alternatives>
     <email>saratova@mirea.ru</email>
     <xref ref-type="aff" rid="aff-1"/>
    </contrib>
    <contrib contrib-type="author">
     <name-alternatives>
      <name xml:lang="ru">
       <surname>Апухтин</surname>
       <given-names>Александр Михайлович</given-names>
      </name>
      <name xml:lang="en">
       <surname>Apuhtin</surname>
       <given-names>Aleksandr Mihaylovich</given-names>
      </name>
     </name-alternatives>
     <email>gtmaleksa@gmail.com</email>
     <xref ref-type="aff" rid="aff-2"/>
    </contrib>
   </contrib-group>
   <aff-alternatives id="aff-1">
    <aff>
     <institution xml:lang="ru">МИРЭА – Российский технологический университет</institution>
     <country>Россия</country>
    </aff>
    <aff>
     <institution xml:lang="en">MIREA – Russian Technological University</institution>
     <country>Russian Federation</country>
    </aff>
   </aff-alternatives>
   <aff-alternatives id="aff-2">
    <aff>
     <institution xml:lang="ru">МИРЭА – Российский технологический университет</institution>
     <country>Россия</country>
    </aff>
    <aff>
     <institution xml:lang="en">MIREA – Russian university of technology</institution>
     <country>Russian Federation</country>
    </aff>
   </aff-alternatives>
   <pub-date publication-format="print" date-type="pub" iso-8601-date="2026-07-01T00:00:00+03:00">
    <day>01</day>
    <month>07</month>
    <year>2026</year>
   </pub-date>
   <pub-date publication-format="electronic" date-type="pub" iso-8601-date="2026-07-01T00:00:00+03:00">
    <day>01</day>
    <month>07</month>
    <year>2026</year>
   </pub-date>
   <volume>2026</volume>
   <issue>2</issue>
   <fpage>111</fpage>
   <lpage>119</lpage>
   <history>
    <date date-type="received" iso-8601-date="2026-05-26T00:00:00+03:00">
     <day>26</day>
     <month>05</month>
     <year>2026</year>
    </date>
    <date date-type="accepted" iso-8601-date="2026-06-27T00:00:00+03:00">
     <day>27</day>
     <month>06</month>
     <year>2026</year>
    </date>
   </history>
   <self-uri xlink:href="https://journals.igps.ru/en/nauka/article/128595/view">https://journals.igps.ru/en/nauka/article/128595/view</self-uri>
   <abstract xml:lang="ru">
    <p>Целью работы является анализ архитектурных компонентов мультиагентных систем на основе больших языковых моделей. Метод исследования – сравнительный анализ архитектур MetaGPT, Generative Agents, AutoGen, OrgAgent, MIRIX &#13;
и ChatDev по единому набору критериев (организация ролей, память, координация, структурированные результаты, взаимодействие, проверка качества) с последующей интеграцией компонентов в единую систему и опытной проверкой на задачах двух типов &#13;
в одноагентном и мультиагентном режимах. Установлено, что на рассмотренных задачах мультиагентный режим уступает одноагентному по качеству (оценки 3,0 и 4,5 против 9,0 по десятибалльной шкале). Выявлена архитектурная причина: несоответствие типа задачи набору действий в конфигурации ролей. Обнаружена закономерность: при интеграции трех контуров проверки – детектора отказов, проверки соответствия и контура принятия решений – возникает способность к самодиагностике, не наблюдаемая ни в одной из рассмотренных систем по отдельности. Три контура согласованно определили ненадежность результата и отказались его выдать. Данное свойство отличается от известных подходов к диагностике мультиагентных систем, где анализ выполняется после завершения работы. Результаты могут быть использованы при проектировании мультиагентных систем для задач с высокой ценой ошибки.</p>
   </abstract>
   <trans-abstract xml:lang="en">
    <p>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.</p>
   </trans-abstract>
   <kwd-group xml:lang="ru">
    <kwd>мультиагентные системы</kwd>
    <kwd>большие языковые модели</kwd>
    <kwd>архитектура агентов</kwd>
    <kwd>память агентов</kwd>
    <kwd>координация</kwd>
    <kwd>самодиагностика</kwd>
   </kwd-group>
   <kwd-group xml:lang="en">
    <kwd>multi-agent systems</kwd>
    <kwd>large language models</kwd>
    <kwd>agent architecture</kwd>
    <kwd>agent memory</kwd>
    <kwd>coordination</kwd>
    <kwd>self-diagnosis</kwd>
   </kwd-group>
  </article-meta>
 </front>
 <body>
  <p></p>
 </body>
 <back>
  <ref-list>
   <ref id="B1">
    <label>1.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Wooldridge M. An Introduction to MultiAgent Systems. Wiley, 2009. 484 p.</mixed-citation>
     <mixed-citation xml:lang="en">Wooldridge M. An Introduction to MultiAgent Systems. Wiley, 2009. 484 p.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B2">
    <label>2.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Минцберг Г. Структура в кулаке: создание эффективной организации / пер. с англ. СПб.: Питер, 2004. 512 с.</mixed-citation>
     <mixed-citation xml:lang="en">Mincberg G. Struktura v kulake: sozdanie effektivnoj organizacii / per. s angl. SPb.: Piter, 2004. 512 s.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B3">
    <label>3.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Large Language Model based Multi-Agents: A Survey of Progress and Challenges / T. Guo [et al.] // arXiv:2402.01680. 2024.</mixed-citation>
     <mixed-citation xml:lang="en">Large Language Model based Multi-Agents: A Survey of Progress and Challenges / T. Guo [et al.] // arXiv:2402.01680. 2024.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B4">
    <label>4.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Multi-Agent Collaboration Mechanisms: A Survey of LLMs / K.-T. Tran [et al.] // arXiv:2501.06322. 2025.</mixed-citation>
     <mixed-citation xml:lang="en">Multi-Agent Collaboration Mechanisms: A Survey of LLMs / K.-T. Tran [et al.] // arXiv:2501.06322. 2025.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B5">
    <label>5.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework / S. Hong [et al.] // Proc. ICLR. 2024. arXiv:2308.00352.</mixed-citation>
     <mixed-citation xml:lang="en">MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework / S. Hong [et al.] // Proc. ICLR. 2024. arXiv:2308.00352.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B6">
    <label>6.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">ChatDev: Communicative Agents for Software Development / C. Qian [et al.] // Proc. ACL. arXiv:2307.07924. 2024.</mixed-citation>
     <mixed-citation xml:lang="en">ChatDev: Communicative Agents for Software Development / C. Qian [et al.] // Proc. ACL. arXiv:2307.07924. 2024.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B7">
    <label>7.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">OrgAgent: Organize Your Multi-Agent System like a Company / Wang [et al.] // arXiv:2604.01020. 2026.</mixed-citation>
     <mixed-citation xml:lang="en">OrgAgent: Organize Your Multi-Agent System like a Company / Wang [et al.] // arXiv:2604.01020. 2026.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B8">
    <label>8.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Generative Agents: Interactive Simulacra of Human Behavior / J.S. Park [et al.] // Proc. ACM UIST. arXiv:2304.03442. 2023.</mixed-citation>
     <mixed-citation xml:lang="en">Generative Agents: Interactive Simulacra of Human Behavior / J.S. Park [et al.] // Proc. ACM UIST. arXiv:2304.03442. 2023.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B9">
    <label>9.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">MIRIX: Multi-Agent Memory System for LLM-Based Agents / Wang [et al.] // arXiv:2507.07957. 2025.</mixed-citation>
     <mixed-citation xml:lang="en">MIRIX: Multi-Agent Memory System for LLM-Based Agents / Wang [et al.] // arXiv:2507.07957. 2025.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B10">
    <label>10.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation / Q. Wu [et al.] // arXiv:2308.08155. 2023.</mixed-citation>
     <mixed-citation xml:lang="en">AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation / Q. Wu [et al.] // arXiv:2308.08155. 2023.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B11">
    <label>11.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Why Do Multi-Agent LLM Systems Fail? / M. Cemri [et al.] // Proc. NeurIPS (Datasets and Benchmarks Track). arXiv:2503.13657. 2025.</mixed-citation>
     <mixed-citation xml:lang="en">Why Do Multi-Agent LLM Systems Fail? / M. Cemri [et al.] // Proc. NeurIPS (Datasets and Benchmarks Track). arXiv:2503.13657. 2025.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B12">
    <label>12.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Role-Based Access Control Models / R. Sandhu [et al.] // IEEE Computer. 1996. Vol. 29. № 2. P. 38–47.</mixed-citation>
     <mixed-citation xml:lang="en">Role-Based Access Control Models / R. Sandhu [et al.] // IEEE Computer. 1996. Vol. 29. № 2. P. 38–47.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B13">
    <label>13.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">Towards Self-Improving Error Diagnosis in Multi-Agent Systems / Zhang [et al.] // arXiv:2604.17658. 2026.</mixed-citation>
     <mixed-citation xml:lang="en">Towards Self-Improving Error Diagnosis in Multi-Agent Systems / Zhang [et al.] // arXiv:2604.17658. 2026.</mixed-citation>
    </citation-alternatives>
   </ref>
   <ref id="B14">
    <label>14.</label>
    <citation-alternatives>
     <mixed-citation xml:lang="ru">DiLLS: Interactive Diagnosis of LLM-based Multi-agent Systems via Layered Summary / R. Sheng [et al.] // Proc. CHI. arXiv:2602.05446. 2026.</mixed-citation>
     <mixed-citation xml:lang="en">DiLLS: Interactive Diagnosis of LLM-based Multi-agent Systems via Layered Summary / R. Sheng [et al.] // Proc. CHI. arXiv:2602.05446. 2026.</mixed-citation>
    </citation-alternatives>
   </ref>
  </ref-list>
 </back>
</article>
