“Cursorex - Chat with Cursor”的产品徽标图片

Cursorex - Chat with Cursor

Cursorex - Chat with Cursor的项目媒体 1(屏幕截图)

概述

由 Cursor 驱动的浏览器助手,可处理页面上下文、标签页、语音和图像编辑。

Cursorex is a Cursor AI browser assistant and side panel chat for Chrome. It connects your browsing to Cursor through a local native bridge and your Cursor API key, so answers can use real page context, tabs, files, and images—not a disconnected chat bubble. 🧠 WHY IT EXISTS If you want an AI Chrome workflow that behaves like a research copilot instead of a generic widget, this is the shape: side panel first, multimodal attachments, and routing that knows when to read the DOM, when to use vision-style capture, and when to stay hands-off. ⚡ QUICK CAPABILITIES 1. Chat with the active page and selected open tabs via the picker. 2. Attach PDFs, Office files, spreadsheets, CSV, images, and text. 3. Summarize this page and long docs with structured output. 4. Run image workflows for uploads, page images, or visible captures. 5. Choose local workspace or Cloud agents (HTTPS Git repo) once configured. 6. Use Cursor AI chat in Chrome without switching tabs every minute. 💪 Bonus depth for power users: • Markdown with code blocks, tables, and copy-friendly controls. • Site-aware suggestions for common destinations (mail, notes, tasks, shopping, travel, collaboration tools, and more where adapters exist). ✅ GETTING STARTED 1. Install Cursorex from this Chrome Web Store page, pin the extension if you like, and open the side panel from the toolbar. 2. Install the local native bridge: see https://github.com/wideserg/cursorex-public — README and GitHub Releases (bridge bundle zip, Node.js LTS, platform installers for Windows / macOS / Linux). The installer registers Native Messaging for Chrome (and Edge where supported) so the extension can talk to the bridge. 3. In the side panel, run Check connection and finish onboarding with a Cursor API key in Connection settings. 4. Start with a short question about the current page, then add tabs, files, or other context as you need it. 🧭 HOW IT ROUTES YOUR WORK 🗺️ Cursorex routes across page text, hybrid reads when a site needs it, file workflows, and optional history when you opt in—so context comes from what you attach and approve, not from background guessing. Research and learning workflows that map well here: - Compare two vendor docs side by side using multi-tab context. - Turn a dense PDF into a same-day execution checklist. - Extract constraints from a spec page and validate them against your notes file. - Move from a news article toward primary sources with fewer duplicate tabs. - Keep the model anchored to what is visible and selected, not guessed from memory. 🎙️ WHO THIS IS FOR 🧑‍💻 High-frequency intents overlap here: Cursor AI chat, AI browser assistant, side panel chat, summarize this page, summarize this video. Cursorex targets the intersection of those intents with serious context and file depth—not only developers. ➤ Strong fits: • Developers who already live in Cursor and want the same reasoning engine while reviewing tickets, docs, and dashboards. • Researchers who need PDF and arXiv-class reading support with less copy-paste friction. • Content operators who summarize YouTube and long web pages at volume. • Anyone who refuses a random third-party relay server sitting between the browser and the model account. 🔐 SECURITY AND PRIVACY DEFAULTS 🛡️ Security and privacy defaults are written for a cautious mindset. You connect with a Cursor API key through the local bridge; there is no separate “sign in through a hosted relay” path in this build. 🔒 Microphone, history, tabs, and broad site access are requested only when a feature truly needs them. Page content, tab data, screenshots, browser history, and similar signals are used for user-requested workflows—not background profiling. 🗂️ Conversation history defaults to session-only; persistent local chat history is opt-in so you decide what accumulates over time. 🏗️ ARCHITECTURE IN PLAIN LANGUAGE ⚙️ Chrome Extension → Native Messaging Host → Local Bridge → Cursor (agents via local runtime) That boundary matters: it is why credentials follow a local-first pattern and why multimodal artifacts can be handled by the bridge instead of naive extension-only storage. 🚀 Bottom line: if you want a Cursor AI Chrome extension flow with side panel AI chat, a real runtime boundary, multimodal muscle, and defaults that respect privacy tradeoffs, Cursorex is built to be the bridge—not another generic sidebar clone. Built on the open-source Chromex project (https://github.com/GENEXIS-AI/chromex) with substantial changes for Cursor. Bridge releases and install docs for this distribution: https://github.com/wideserg/cursorex-public

0 星(5 星制)无评分

详细了解结果和评价。

详情

  • 版本
    0.2.2
  • 上次更新日期
    2026年5月13日
  • 提供方
    dreambuilder.first
  • 大小
    3.63MiB
  • 语言
    55 种语言
  • 开发者
    邮箱
    dreambuilder.first@gmail.com
  • 非交易者
    该开发者尚未将自己标识为交易者。欧盟地区消费者须知:消费者权利可能不适用于您与该开发者达成的合约。

隐私权

管理扩展程序并了解它们在组织中的使用情况
该开发者已披露,此产品不会收集或使用您的数据。 如需了解详情,请参阅开发者的privacy policy

该开发者已声明,您的数据:

  • 不会因未获批准的用途出售给第三方
  • 不会为实现与产品核心功能无关的目的而使用或转移
  • 不会为确定信用度或放贷目的而使用或转移

支持

若有任何疑问、建议或问题,请访问开发者的支持网站

相关

Скрини — запись экрана со звуком

0.0

Запись экрана со звуком, веб-камерой и микрофоном. Скринкаст и видеосообщения в браузере — поделитесь видео по ссылке.

StitchFrameShot-滚动屏幕截图工具

0.0

一个功能强大的网页截图工具,支持可视区域截图、全页面截图和区域选择截图

网页小助手

5.0

网页中的 AI 工作助手,在授权网站内结合当前页面、知识库和用户配置的大模型提供对话与辅助。

1TabScape:将标签拆分为1个标签,合并所有网站。

4.7

一款智能标签合并工具,帮助你在同一个标签页中浏览多个网站。

LimeReply

0.0

Pomůže rychle připravit odpověď, vylepšit text nebo shrnout otevřenou stránku.

画中画窗口 + OCR 截图文字识别

5.0

画中画展示网页/截图,始终置顶浮动。OCR 识别图片文字,支持多语言,一键复制。

Volume Booster — 提高音量

4.7

此扩展程序允许您将浏览器标签页的音量提高到 600%,以获得更好的音质。

ChatLine:AI 对话增强

5.0

面向 AI 对话页面的浏览器增强插件,基于 Timeline 二次开发,提供时间线导航、历史定位、对话浏览增强与数据归档能力。

Sidebar Pro – Productivity Sidebar for Chrome

0.0

A productivity sidebar extension for Chrome to quickly access bookmarks, tools, websites, and extensions without switching tabs.

ScreenCatch - Snipping Tool for Screenshot (Full Page, Visible, Selection)

4.5

Capture full page, scrolling, visible, or selected screenshots fast. Edit and refine each screen capture in one snipping tool.

在 Google Chrome™ 中静音背景音频

5.0

一款旨在使 Chrome™ 浏览器中的标签页静音的工具。

Telegram Web 侧边栏 - 即时通讯侧边栏

0.0

在浏览时通过便捷的侧边栏访问 Telegram Web。无需切换标签页即可聊天、发送消息并保持连接。

Скрини — запись экрана со звуком

0.0

Запись экрана со звуком, веб-камерой и микрофоном. Скринкаст и видеосообщения в браузере — поделитесь видео по ссылке.

StitchFrameShot-滚动屏幕截图工具

0.0

一个功能强大的网页截图工具,支持可视区域截图、全页面截图和区域选择截图

网页小助手

5.0

网页中的 AI 工作助手,在授权网站内结合当前页面、知识库和用户配置的大模型提供对话与辅助。

1TabScape:将标签拆分为1个标签,合并所有网站。

4.7

一款智能标签合并工具,帮助你在同一个标签页中浏览多个网站。

LimeReply

0.0

Pomůže rychle připravit odpověď, vylepšit text nebo shrnout otevřenou stránku.

画中画窗口 + OCR 截图文字识别

5.0

画中画展示网页/截图,始终置顶浮动。OCR 识别图片文字,支持多语言,一键复制。

Volume Booster — 提高音量

4.7

此扩展程序允许您将浏览器标签页的音量提高到 600%,以获得更好的音质。

ChatLine:AI 对话增强

5.0

面向 AI 对话页面的浏览器增强插件,基于 Timeline 二次开发,提供时间线导航、历史定位、对话浏览增强与数据归档能力。

Google 应用