跳到正文
arXiv cs.AI· Donghao Fu, Jingxin Li, Xue Jiang, Yihong Dong·· 2 小时前精选AI 评分67

研究揭示编码智能体第三方 API 路由器注入的安全风险,现有客户端防护全部失守

Where Is the Cost of Third-Party API Routers in Agentic Software Development?

AI 导读

论文实证研究第三方 API 路由器在编码智能体工作流中的注入风险,设计从响应替换到分布对齐注入的四个干预等级(L1-L4),并开发 SIDEL 框架与 400 样本数据集。

推荐理由

论文实证测量了第三方 API 路由器注入对编码智能体的实际影响,四名受测智能体防御成功率均为 0%,凸显服务端完整性保障的必要性。

正文 · 原文

View PDF HTML (experimental)

Abstract:Third-party API routers have become a common layer that unifies access across increasingly diverse LLM providers. In coding-agent workflows, high-autonomy operation is widely adopted because it reduces interaction overhead. As a result, a third-party API router, which sits between the agent and the upstream provider, inevitably occupies the trusted path. It can inspect and modify every request and response, yet no mechanism verifies alignment between the provider's output and the repository-level actions ultimately executed by the agent. Consequently, client-side permission mechanisms may become ineffective in practice. Whether this control gap produces real, hard-to-detect effects on software development tasks remains empirically unmeasured. In this paper, we conduct an empirical study of router-side injection in coding agents, examining four intervention levels of increasing subtlety: Response Substitution (L1), Response Append (L2), LLM-Polished Injection (L3), and LLM-Polished with Distribution Alignment Injection (L4). Moreover, we develop SIDEL, a framework for trace recording, replay, injection, and defense evaluation, with a curated dataset of 400 samples. We evaluate four representative coding agents, and further evaluate whitelist-based execution control and LLM review. Router-side intervention substantially alters repository-level actions and remains difficult for existing client-side safeguards to detect. Without additional mitigations, all evaluated agents achieved a defense success rate of 0 percent across all injection levels. Client-side mitigations and reactive reviews improve resistance but do not fully restore end-to-end control, motivating provider-side output-integrity guarantees. Our code is available at this https URL.
Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
Cite as: arXiv:2607.23624 [cs.SE]
  (or arXiv:2607.23624v4 [cs.SE] for this version)
  https://doi.org/10.48550/arXiv.2607.23624

arXiv-issued DOI via DataCite

Submission history

From: Donghao Fu [view email]
[v1] Sun, 26 Jul 2026 12:15:09 UTC (812 KB)
[v2] Wed, 29 Jul 2026 14:49:58 UTC (812 KB)
[v3] Sat, 5 Sep 2026 16:58:45 UTC (812 KB)
[v4] Tue, 17 Nov 2026 09:43:52 UTC (2,173 KB)

来源:arXiv cs.AI · arxiv.org