Newest papers

Teaching Code Agents to Ask Without Leaking Private Code

Wenjia Jiang, Jieke Shi, Zhensu Sun, Chengran Yang, Shidong Pan, Peixin Zhang, Yifan Jia, James Hoang, Xiwei Xu, Zhenchang Xing, David Lo, Zhou Yang · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.24824 · cited by 0 works (OpenAlex) · open-access PDF

Russian

Местные агенты кодирования могут расширить свой охват, консультируя более сильную модель облака. Такие консультации, однако, часто раскрывают частные идентификаторы, пути файлов и исходный код репозитория. Маскировка на стороне хоста скрывает признанные идентификаторы, но оставляет окружающий код и формулировку агента неизменными, в то время как обучение для меньшего раскрытия рискует подавлением консультации или выработкой вопросов, слишком неясных для ответа. Мы предлагаем уточнение с верификацией полезности (UVA), которое учит агента описывать частные субъекты по их роли и типу, сохраняя при этом технические детали, необходимые эксперту. От решенных траекторий ремонта мы переписываем каждый вопрос в эту форму и сохраня

Machine translation (Meta M2M100); the English original follows.

Abstract

Local coding agents can extend their reach by consulting a stronger cloud model. Such consultations, however, often expose the repository's private identifiers, file paths, and source code. Host-side masking hides recognized identifiers but leaves the surrounding code and the agent's phrasing unchanged, while training for less disclosure risks suppressing consultation or producing questions too vague to answer. We propose utility-verified abstraction (UVA), which teaches the agent to describe private entities by their role and type while retaining the technical details an expert needs. From solved repair trajectories, we rewrite each question into this form and keep the rewrite only if it discloses fewer private entities and the repair, replayed from the same execution state with the repair trajectories. Across two benchmarks, the trained agent consults more often than before training. On repository-level issue repair, it solves 37.6% of issues, against 25.3% for the same model working alone, a gain from the training as a whole rather than from consultation alone. Its questions disclose 3.7-fold fewer private entities and 1.4-fold less verbatim code than the untrained agent's questions to the same expert.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Identifier, Coding (social sciences), Ask price, Computer security, Code (set theory)

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.24824, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange