AgentGit:一个面向可靠且可扩展的 LLM 驱动多智能体系统的版本控制框架
AgentGit: A Version Control Framework for Reliable and Scalable LLM-Powered Multi-Agent Systems
AI总结:
本文提出 AgentGit 框架,通过在 LangGraph 之上引入类 Git 的状态提交、回退和分支机制,支持智能体高效遍历与探索多轨迹,显著降低冗余计算、运行时间与 token 消耗,提升了 LLM 驱动多智能体系统的可靠性与可扩展性。
摘要:
随着大语言模型(LLMs)的快速发展,LLM 驱动的多智能体系统(MAS)正引起学术界和工业界日益浓厚的兴趣。然而,当前的许多 MAS 框架在可靠性和可扩展性方面存在困难,尤其是在处理复杂任务时。我们提出了 AgentGit,这是一个将类 Git 的回滚和分支机制引入 MAS 工作流的框架。AgentGit 作为构建在 LangGraph 之上的基础设施层,支持状态提交、回退和分支,允许智能体高效地遍历、比较和探索多条轨迹。为了评估 AgentGit,我们设计了一项通过选择更优提示词来优化目标智能体的实验。我们在一个真实世界任务(检索并分析论文摘要)上,针对三个基线——LangGraph、AutoGen 和 Agno 运行了多步 A/B 测试。结果表明,AgentGit 显著减少了冗余计算,降低了运行时间和 token 使用量,并支持跨多个分支的并行探索,从而增强了 MAS 开发中的可靠性和可扩展性。这项工作为实现更稳健的 MAS 设计提供了一条实用路径,并在协作式 AI 系统中实现了错误恢复、安全探索、迭代调试和 A/B 测试。
英文摘要:
With the rapid progress of large language models (LLMs), LLM-powered multi-agent systems (MAS) are drawing increasing interest across academia and industry. However, many current MAS frameworks struggle with reliability and scalability, especially on complex tasks. We present AgentGit, a framework that brings Git-like rollback and branching to MAS workflows. Built as an infrastructure layer on top of LangGraph, AgentGit supports state commit, revert, and branching, allowing agents to traverse, compare, and explore multiple trajectories efficiently. To evaluate AgentGit, we designed an experiment that optimizes target agents by selecting better prompts. We ran a multi-step A/B test against three baselines -- LangGraph, AutoGen, and Agno -- on a real-world task: retrieving and analyzing paper abstracts. Results show that AgentGit significantly reduces redundant computation, lowers runtime and token usage, and supports parallel exploration across multiple branches, enhancing both reliability and scalability in MAS development. This work offers a practical path to more robust MAS design and enables error recovery, safe exploration, iterative debugging, and A/B testing in collaborative AI systems.
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。