跳到正文
arXiv cs.AI· Chenyang Ma, Clyde Fare, Matthew Wilson, Dave Braines·· 4 小时前AI 评分38

EmbeWebAgent:将 Web 智能体嵌入企业自定义 UI

EmbeWebAgent: Embedding Web Agents into Custom Enterprise UIs

AI 导读

EmbeWebAgent 提出一种将智能体直接嵌入企业现有 UI 的框架,通过轻量前端钩子(精选 ARIA 与基于 URL 的观测、经 WebSocket 暴露的每页函数注册表)和可复用后端工作流完成推理与操作。

正文 · 原文

View PDF HTML (experimental)

Abstract:Most web agents operate at the human interface level, observing screenshots or raw DOM trees without application-level access, which limits robustness and action expressiveness. In enterprise settings, however, explicit control of both the frontend and backend is available. We present EmbeWebAgent, a framework for embedding agents directly into existing UIs using lightweight frontend hooks (curated ARIA and URL-based observations, and a per-page function registry exposed via a WebSocket) and a reusable backend workflow that performs reasoning and takes actions. EmbeWebAgent is stack-agnostic (e.g., React or Angular), supports mixed-granularity actions ranging from GUI primitives to higher-level composites, and orchestrates navigation, manipulation, and domain-specific analytics via MCP tools. Our demo shows minimal retrofitting effort and multi-step behaviors grounded in a live UI setting. Demo Video: this https URL
Comments: Technical Report; Demo Video: this https URL
Subjects: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
Cite as: arXiv:2602.14865 [cs.AI]
  (or arXiv:2602.14865v2 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2602.14865

arXiv-issued DOI via DataCite

Submission history

From: Chenyang Ma [view email]
[v1] Mon, 16 Feb 2026 15:59:56 UTC (3,323 KB)
[v2] Tue, 17 Nov 2026 18:20:19 UTC (3,326 KB)

来源:arXiv cs.AI · arxiv.org