arXiv:2604.15034cs.AI2026-04被引 8

让智能体自动进化,通过协议管理资源与更新流程。

Autogenesis: A Self-Evolving Agent Protocol

论文配图:Autogenesis: A Self-Evolving Agent Protocol
图 1 · 摘自论文原文
  • 将提示、工具等资源统一注册为可版本化管理的协议资源。
  • 构建闭环更新机制,支持改进提案、评估与可回滚提交。
  • 适合需要长期规划和自适应能力的复杂任务系统设计者。

基于大模型的智能体系统在处理复杂、长周期任务方面展现出潜力,但现有协议(如A2A和MCP)在跨实体生命周期管理、上下文维护、版本追踪及安全更新接口方面存在不足,导致代码高度耦合且脆弱。本文提出自演化协议(Autogenesis Protocol, AGP),其资源底座层(RSPL)将提示、智能体、工具、环境和记忆建模为具有明确状态、生命周期和版本接口的协议注册资源;自演化层(SEPL)定义了闭环操作接口,支持改进提案、评估与提交,并具备可审计的演化轨迹与回滚能力。基于AGP,我们构建了自演化多智能体系统(AGS),可在执行过程中动态实例化、检索与优化协议注册资源。我们在多个需长周期规划与异构资源协同的任务基准上评估了AGS,结果表明其持续优于强基线,验证了资源管理与闭环自演化机制的有效性。代码已开源:https://github.com/DVampire/Autogenesis。

原文摘要 · Abstract (English)

Recent advances in LLM based agent systems have shown promise in tackling complex, long horizon tasks. However, existing agent protocols (e.g., A2A and MCP) under specify cross entity lifecycle and context management, version tracking, and evolution safe update interfaces, which encourages monolithic compositions and brittle glue code. We introduce Autogenesis Protocol (AGP), a self evolution protocol that decouples what evolves from how evolution occurs. Its Resource Substrate Protocol Layer (RSPL) models prompts, agents, tools, environments, and memory as protocol registered resources with explicit state, lifecycle, and versioned interfaces. Its Self Evolution Protocol Layer (SEPL) specifies a closed loop operator interface for proposing, assessing, and committing improvements with auditable lineage and rollback. Building on AGP, we present Autogenesis System (AGS), a self-evolving multi-agent system that dynamically instantiates, retrieves, and refines protocol-registered resources during execution. We evaluate AGS on multiple challenging benchmarks that require long horizon planning and tool use across heterogeneous resources. The results demonstrate consistent improvements over strong baselines, supporting the effectiveness of agent resource management and closed loop self evolution. The code is available at https://github.com/DVampire/Autogenesis.

智能体系统自演化资源管理协议设计

Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。