一句话：提出基于提交工件的四级 AI 成熟度模型 RAMP，发现编码智能体带来普遍提速，但缺少提交式 AI 配置的仓库质量代价约翻倍。
编码智能体提升开发速度，同时也带来技术债；既有研究只报告采用者的平均效应，掩盖了团队间的巨大差异。本文提出 RAMP（Repository AI Maturity Profile），一个以版本控制中提交的 AI 工具配置工件为依据的四级累积成熟度模型，涵盖行为规则与编码规范、命名智能体定义到多智能体编排，观测实践主要集中在前三级。在 441 个仓库上，各等级表现为累积量表，独立人工标注在留出样本上以 97% 的一致性复现 RAMP 的仓库级标签。采用呈累积、单向且“设置后即忘”的特点：73.8% 的工件只提交一次便不再修改。作者在既有智能体采用面板中按成熟度分层重估，发现无论成熟度高低智能体都加速开发（提交数增加 28–38%），但质量出现分化：在 agent-first 仓库中，未提交 AI 配置者认知复杂度增幅约为两倍（+53% 对 +27%），静态分析告警增加 1.7 倍。由于成熟度为观测性变量，相关工程纪律或模型能力可能部分解释该差距；作者将结论定位为假设生成性，并将 RAMP 作为可复用工具发布。
Coding agents increase development velocity but also technical debt. Prior work reports only average effects across adopters, hiding wide differences between teams. We introduce RAMP (Repository AI Maturity Profile), a four-level cumulative maturity model grounded in version-controlled artifacts that teams commit to configure AI tools. RAMP runs from behavioral rules and coding standards through named agent definitions to multi-agent orchestration, with observed practice concentrated in the first three levels. Across 441 repositories the levels behave as a cumulative scale, and independent human annotation reproduces RAMP's repository-level labels on 97% of a held-out sample. Adoption is cumulative, forward-only, and set-and-forget: 73.8% of artifacts are committed once and never modified. Re-estimating an existing agent-adoption panel within each stratum, agents accelerate development regardless of maturity (28-38% more commits), but quality diverges: among agent-first repositories, where the contrast is identified, those without committed AI configuration show roughly twice the increase in cognitive complexity (+53% versus +27%) and 1.7× the increase in static-analysis warnings. Because maturity is observational, correlated engineering discipline or model capability may explain part of the gap; we present these findings as hypothesis-generating and release RAMP as a reusable instrument.
@inproceedings{paperbot12321,
title = {A Few Pages of Markdown: Committed AI Configuration and Lower Quality Cost after Coding-Agent Adoption},
author = {Yegor Denisov-Blanch and Shyam Agarwal and Pavel Azaletskiy and Hao He and Rylan Schaeffer and Brando Miranda and Bogdan Vasilescu and Sanmi Koyejo},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3837546}
}