[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85261-en":3,"doc-seo-85261-105":29,"detail-sidebar-cat-0-en-105":82},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},85261,1374391974585,"Genevieve","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",6,"Technology","LOGOS A Living Logic for AI Agent Teams That Evolve With Humans","AI agents are shifting from one-off answer systems to persistent teams that use tools, delegate tasks, learn from experience, and modify the artifacts driving their behavior. Deployment safety becomes a governance question: who controls what the system is allowed to become. Logos introduces a pluggable self-evolution and governance layer that compiles multimodal inputs into versioned agent packs, emits auditable event traces, and enforces fail-closed verification with evidence and human authorization to enable verifiable human-agent loop engineering.","arXiv :2607 . 10878v 1 [ cs .AI] 12 Jul 2026  \nLOGOS: A Living Logic for AI Agent Teams That Evolve With Humans  \nYuma Ichikawa 1,2 , Yamato Arai 1,3 , Kosaku Kimura 1 , Akira Sakai 1,4 , Hiromichi Kobashi 1  \nAI agents are evolving from answer engines into persistent teams that use tools, delegate work, learn from experience, and modify the artifacts that shape their future behavior. The defining question for deployment is no longer merely what agents can do, but who controls what they are allowed to become. We introduce logos, a pluggable layer for self-evolution and governance that strengthens existing multiagent frameworks rather than replacing them. logos compiles heterogeneous multimodal inputs, including documents, images, audio, tables, databases, APIs, and human instructions into versioned agent packs containing agents, tools, knowledge, tests, permissions, and policies. During operation, it transforms agent activity into portable, auditable event traces and applies fail-closed verification across frameworks and backends. Every learned prompt, memory, skill, tool, role, or workflow remains an untrusted release candidate until held-out execution evidence, human-controlled policy, and explicit authorization permit its promotion. This architecture enables “verifiable human-agent loop engineering”: agents can act, ask, learn, and propose improvements, while humans can steer objectives, permissions, approvals, and irreversible actions without interrupting continuous operation. logos provides a living logic for accountable automation. Agents may evolve at machine speed, but only evidence and human authority can close the loop.  \n1 Introduction  \nAI systems are becoming persistent actors rather than passive answer engines. Modern agents use tools over long horizons, retain state, coordinate across agents and organizations, and adapt through evaluation and experience [OpenAI, 2025 , Google, 2025] . Once they can revise prompts, memories, skills, tools, roles, and workflows, adaptation becomes a software release problem. The key deployment question is no longer only what can the model do? It is also who decides what the deployed system may become?  \nThis question matters because persistence, permissions, credentials, memory, delegation, and access to production state turn model capabilities into operational consequences [National Institute of Standards and Technology, 2023 , OWASP Gen AI Security Project, 2025 , 2026] . A model should not both propose and approve its own updates. A better prompt, useful memory, or a higher-scoring workflow does not, by itself, establish deployment safety or utility. Our governing principle is, therefore, simple: proposal isnot promotion. Agents may propose changes, but release must depend on independently defined evidence, policy, and authority.  \nWe introduce Logos, a pluggable self-evolution and governance layer for multi-agent systems. Rather than replacing existing frameworks, Logos adds the compile–operate–evolve lifecycle shown in Figure 1.1 Compilation converts bounded, heterogeneous inputs, including documents, source code, images, audio, tables, databases, API specifications, and human instructions, into a versioned Agent Pack containing agents, tools, knowledge, memory, routing policies, permissions, verifiers, validation probes, and provenance. Operation normalizes framework-specific behavior into typed, auditable events. Evolution keeps every learned prompt, memory, skill, tool, role, routing policy, or workflow isolated until held-out execution evidence, declared safety and regression constraints, human-owned policy, and required authorization permit promotion.  \nA root policy remains outside of ordinary self-evolution. It fixes the human-defined objectives, evaluators, and gate versions, protected holdouts, credential boundaries, permission-expansion rules, approval  \n1 Fujitsu Limited, 2 RIKEN Center for AIP, 3 The University of Tokyo, 4 Tokai University Correspondence: Yuma Ich","cbCainWlMo4STO8c","https://ap.wps.com/l/cbCainWlMo4STO8c","pdf",9473413,1,58,"English","en",105,"# Introduction\n## Agent persistence and governance problem\n## Proposal is not promotion\n## Logos compile–operate–evolve lifecycle","[{\"question\":\"How does Logos enable accountability during agent operation and evolution?\",\"answer\":\"Logos normalizes framework-specific behavior into portable, auditable event traces and uses fail-closed verification across frameworks and backends, with rollback and safeguards for external effects.\"}]",1784202139,146,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":77,"head_meta":79,"extra_data":81,"updated_unix":27},"logos-a-living-logic-for-ai-agent-teams-that-evolve-with-humans","",{"@graph":35,"@context":76},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/logos-a-living-logic-for-ai-agent-teams-that-evolve-with-humans/85261/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70],{"name":71,"@type":72,"acceptedAnswer":73},"How does Logos enable accountability during agent operation and evolution?","Question",{"text":74,"@type":75},"Logos normalizes framework-specific behavior into portable, auditable event traces and uses fail-closed verification across frameworks and backends, with rollback and safeguards for external effects.","Answer","https://schema.org",{"og:url":51,"og:type":78,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":80,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":83},[84,88,92,96,101,104,109,114,119,122,126],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":102,"slug":103},50,"technology",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},7,"Healthcare",40,"healthcare",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},8,"Research & Report",30,"research-report",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":45,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general"]