📄 中文摘要
现有的代码助手依然依赖于静态代码快照，无法有效建模项目的时间演变过程中的关键信息。这种局限性导致其在应对复杂的代码库问题时表现出僵化的行为逻辑和缺乏自主适应能力。为了解决这一静态与动态之间的矛盾，提出了MemCoder框架。该框架旨在实现人类与人工智能的持续共同进化，通过结构化历史人类经验，提炼出过去成功实践中的潜在意图与代码映射，从而提升代码助手的智能水平和适应能力。
您的代码助手可以与您一起成长，借助结构化记忆
出处: Your Code Agent Can Grow Alongside You with Structured Memory
发布: 2026年3月17日
📄 English Summary
Your Code Agent Can Grow Alongside You with Structured Memory
Existing code agents are tethered to static code snapshots, which limits their ability to model critical information embedded in the temporal evolution of projects. This limitation results in rigid behavioral logic and a lack of autonomous adaptability, hindering their effectiveness in addressing complex repository-level problems. To address this static-dynamic mismatch, the MemCoder framework is proposed. It enables continual human-AI co-evolution by structuring historical human experiences to distill latent intent-to-code mappings from past successful practices, thereby enhancing the intelligence and adaptability of code agents.