BrowserGPT by CIVAI

What's New: MCP Browser Control for AI assistants, Selection toolbar, Image gen, Tabs, Voice Browsing

Introducing BrowserGPT by CIVAI: Your AI-Powered Browser Assistant BrowserGPT is a Chrome extension that helps you control your browser with voice or text, automate everyday tasks, and save time online. Speak naturally or type what you want. BrowserGPT runs in your own Chrome profile and works on the pages you already have open. 🆕 Latest update • Browser control via MCP: Connect your AI agents or assistants to your browser so they can list your tabs, switch tabs, navigate, capture page snapshots, click and type, and run small scripts on a tab you choose, in your real browser, signed in as you. • Live Page / Tab Agent: Start a persistent agent on a tab to watch for meaningful changes and run your prompt when the page updates. Useful for dashboards, job boards, and status pages where you would otherwise refresh and watch manually. • Automation sharing: Share visual summaries of automations as images or links so others can see what happened without installing the extension. • Image generation: Describe what you want and generate images in the browser without extra tools or credits. • Quick Task: From the context menu, describe an action on the page you're viewing; BrowserGPT runs it in seconds (close popups, fill forms, select checkboxes, extract visible details, and more). • Free text tools: Humanizer, AI detector, grammar fixer, summarizer, and readability tools are free to use from the selection toolbar. • Replays: Replay past automation runs to review, reuse, or troubleshoot. • Bug fixes and performance: Faster runs, smoother automation, and more reliable SmartSense suggestions. 🔗 Browser control via MCP BrowserGPT can act as the bridge between your Chrome and AI tools that speak MCP (Model Context Protocol). When you connect an agent from Nova by CIVAI, Claude, Cursor, Codex or another MCP client, that agent can work inside the browser you already use, with your logins and sessions intact. From a connected agent you can list open tabs, focus the right tab, open or refresh pages, take structured snapshots of what is on screen, click and type like a person would, and run short scripts on a specific tab when you need data from the page or a one-off in-page action. This is not a separate cloud browser. Actions run locally in your profile, on the tabs you choose, which makes it practical for email, admin panels, research, and everyday sites where you need an assistant to help without re-authenticating elsewhere. 🟢 Live Page / Tab Agent Start Live Page Agent from the context menu and give a prompt once. BrowserGPT monitors that tab for important changes and runs your instruction when something relevant updates. You do not need to keep refreshing the page. It helps with "wait, then act" workflows on dynamic sites. 🖼️ Image generation Describe what you want in plain language and BrowserGPT generates an image in your browser tab. Useful for design ideas, storytelling, and quick visuals without switching apps. ⚡ Quick Task Right-click, choose Quick Task, and describe what you want on the current page. Close dialogs, fill forms, bulk-select checkboxes, or pull visible contact details with minimal setup. 🧭 Tab management and TabInfo Manage tabs and tab groups with voice or text. TabInfo shows titles, URLs, and counts across open tabs. Group related tabs, close clutter, switch quickly, and back up or restore windows and tab groups when you change projects. 📱 Remote desktop control and SmartSense Use BrowserGPT Mobile to connect to your desktop browser, trigger SmartSense from your phone, and run tasks remotely when that fits your workflow. 🔍 AI detector Highlight text on a page and BrowserGPT estimates whether it reads like typical AI-generated or human-written content, directly in the browser, without copying text into a separate checker. Useful for reviewing essays, articles, and draft content before you publish or submit. 🧠 Humanizer Highlight stiff or robotic text and rewrite it in a more natural, human tone using the built-in humanizer, from the same selection toolbar, in one step. 🗂️ Add to memory Save page content to BrowserGPT memory with a command. Import memory from ChatGPT so saved context continues to work in the browser. 📧 AI email generator Highlight text or work in your email composer and draft polished, context-aware emails for outreach, replies, and follow-ups. 📥 Extract emails and usernames Use "Find And Select" from the context menu to scan the page and collect contact details it finds. Helpful for research and lead lists. ⚙️ More you can do • Voice control: Open sites, fill forms, and click controls with spoken commands. • SmartSense suggestions: Context-aware tips while you browse (jobs, forms, writing, and more). • Grammar and readability: Fix typos and improve clarity from highlighted text or voice. • Ask about the page: Summarize, explain, or draft responses based on what's on screen. • Teach mode: Record workflows once and replay them later. • Accessibility: Voice-first navigation, hands-free tab control, and support for users who prefer not to rely on mouse and keyboard alone. Works alongside common screen readers where you already use them. BrowserGPT helps you browse, write, automate, and create in one place. Start with voice or text: fewer clicks, less repetition, more done. 🚀