arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.23193cs.AIcs.SE
收藏

基于大语言模型的自主代理的事件溯源:ESAA

ESAA: Event Sourcing for Autonomous Agents in LLM-Based Software Engineering

Elzo Brito dos Santos Filho

AI总结:

ESAA架构通过事件溯源技术提升自主代理在软件工程中的状态管理和任务执行能力,支持多代理并发调度和异构LLM的协同工作。

摘要:

基于大语言模型(LLMs)的自主代理已经从反应式助手发展为能够规划、通过工具执行操作并迭代环境观察的系统。然而,它们仍然受到结构性限制的制约:缺乏原生状态、长期时间范围内上下文退化以及概率生成与确定性执行需求之间的差距。本文提出了ESAA(自主代理的事件溯源)架构,该架构将代理的认知意图与项目状态的突变分离,受事件溯源模式的启发。在ESAA中,代理仅发出经过验证的结构化意图(agent.result或issue.report);一个确定性的调度器验证、将事件持久化到只追加的日志(activity.jsonl)中,应用文件写入效果,并投影一个可验证的材料化视图(roadmap.json)。该提案结合了边界合同(AGENT_CONTRACT.yaml)、元提示文件(PARCER)和回放验证与哈希(esaa verify),确保完成任务的不可变性和取证可追溯性。两个案例研究验证了该架构:(i)一个着陆页项目(9个任务,49个事件,单代理组成)和(ii)一个临床仪表盘系统(50个任务,86个事件,4个并发代理跨越8个阶段),均以run.status=success和verify_status=ok结束。多代理案例研究展示了使用异构LLM(Claude Sonnet 4.6、Codex GPT-5、Antigravity/Gemini 3 Pro和Claude Opus 4.6)的现实并发调度,提供了架构在单代理场景之外的可扩展性实证证据。

英文摘要:

Autonomous agents based on Large Language Models (LLMs) have evolved from reactive assistants to systems capable of planning, executing actions via tools, and iterating over environment observations. However, they remain vulnerable to structural limitations: lack of native state, context degradation over long horizons, and the gap between probabilistic generation and deterministic execution requirements. This paper presents the ESAA (Event Sourcing for Autonomous Agents) architecture, which separates the agent's cognitive intention from the project's state mutation, inspired by the Event Sourcing pattern. In ESAA, agents emit only structured intentions in validated JSON (agent.result or issue.report); a deterministic orchestrator validates, persists events in an append-only log (activity.jsonl), applies file-writing effects, and projects a verifiable materialized view (roadmap.json). The proposal incorporates boundary contracts (AGENT_CONTRACT.yaml), metaprompting profiles (PARCER), and replay verification with hashing (esaa verify), ensuring the immutability of completed tasks and forensic traceability. Two case studies validate the architecture: (i) a landing page project (9 tasks, 49 events, single-agent composition) and (ii) a clinical dashboard system (50 tasks, 86 events, 4 concurrent agents across 8 phases), both concluding with run.status=success and verify_status=ok. The multi-agent case study demonstrates real concurrent orchestration with heterogeneous LLMs (Claude Sonnet 4.6, Codex GPT-5, Antigravity/Gemini 3 Pro, and Claude Opus 4.6), providing empirical evidence of the architecture's scalability beyond single-agent scenarios.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑