GitHarness：为持续的用户需求初始化你的工作记忆

中文摘要 AI 辅助

基于大语言模型的智能体越来越多地与用户在长期任务上协作，通过广泛的搜索、推理和执行来积累证据、代码和草稿。当用户检查这些结果时，他们可能会提供缺失的信息（需求补全）、引入新需求（需求 elicitation）或修改现有需求（需求转移）。这些变化通常只影响已积累工作的一部分，但智能体可能会沿用过时的信息，或将局部修订变成全局重写。现有方法澄清当前意图而不确定先前工作应如何改变，或在固定目标下重用执行历史。我们通过将动态需求协作表述为联合需求跟踪和局部更新来解决这一差距。我们引入了GitHarness，一个可插拔的Git风格框架，将需求状态及其对应的工作状态组织成可分支的版本历史。一个可训练的Git Agent解决需求变化并选择一个语义兼容的历史状态。一个统一的版本接口随后恢复该状态并创建新分支，使底层框架能够排除过时信息、继承兼容工作，并将执行集中在受影响的部件上。Git Agent通过接口级黑盒强化学习进行训练，同时保持下游框架和任务执行模型固定。我们还构建了MTAgentBench，一个保留验证器的基准，涵盖数学推理、文本到SQL、智能体搜索、软件工程和研究综合。实验表明，在有效需求跟踪、保留有效工作和高效执行的同时，任务性能表现强劲。

英文摘要

LLM-based agents increasingly collaborate with users on long-horizon tasks, accumulating evidence, code, and drafts through extensive search, reasoning, and execution. As users inspect these results, they may supply missing information requirement completion, introduce new requirements requirement elicitation, or revise existing ones requirement shift. These changes often affect only part of the accumulated work, yet agents may carry forward obsolete information or turn local revisions into global rewrites. Existing approaches clarify current intent without determining how prior work should change, or reuse execution histories under a fixed objective. We address this gap by formulating dynamic-requirement collaboration as joint requirement tracking and local update. We introduce GitHarness, a pluggable Git-style framework that organizes requirement states and their corresponding harness work states into a branchable version history. A trainable Git Agent resolves requirement changes and selects a semantically compatible historical state. A unified version interface then restores that state and creates a new branch, enabling the underlying harness to exclude obsolete information, inherit compatible work, and focus execution on affected parts. The Git Agent is trained through interface-level black-box reinforcement learning, with downstream harnesses and task-execution models kept fixed. We also construct MTAgentBench, a verifier-preserving benchmark covering mathematical reasoning, text-to-SQL, agentic search, software engineering, and research synthesis. Experiments demonstrate strong task performance alongside effective requirement tracking, preservation of valid work, and efficient execution.

发表机构

• Peking University（北京大学）
• National Engineering Research Center of Software Engineering, Peking University（软件工程国家工程研究中心）

机构由 AI 辅助整理，请以论文原文为准。