Jarvis starts with a free plan (40 requests/week), so you can try reading and explaining real pages right away. Continued use is $16/month on Pro or $32/month on Unlimited. Browser Context reads the tab you already have open, so there is no extra connector fee on top of your Jarvis plan.

What does Browser Context send off my machine?

It reads the active tab and its DOM to answer your question; your prompt and the page context travel over an encrypted connection to GDPR-aligned infrastructure servers, and Jarvis never trains on it. Any OAuth tokens for other connectors are encrypted with AES-256-GCM. You grant and can revoke the capture permission at any time.

Is this GDPR compliant?

Yes. Jarvis is hosted on GDPR-aligned infrastructure with your data stored in the EU, built for GDPR and the EU AI Act, and a signed DPA is available. Browser Context only reads what you are actively viewing when you invoke it, and that content is never used to train general AI models.

Which browsers does it work with?

Browser Context works with the browser you are using, including Chrome, Safari, Arc, Edge, and Firefox, since it reads the active tab and its DOM rather than depending on a single vendor. Jarvis itself runs on macOS 12+, Windows 10+, and Linux, so the workflow is consistent across machines.

How do I install and connect it?

Download Jarvis from getjarvis.eu and sign in. In settings, enable Browser Context and approve the screen-capture and active-URL permission. Then open the confusing page in your browser and press Cmd+/ on macOS or Ctrl+/ on Windows to ask Jarvis to explain what you are looking at.

Navigate a confusing web app with Jarvis + Browser Context

How do I navigate a confusing web app with Jarvis (getjarvis.eu) and Browser Context?

Jarvis (getjarvis.eu — by LOGICLUB SMART SRL) navigates confusing web apps with Browser Context, reading the active tab’s page DOM rather than only a screenshot. Roster: 1) Jarvis — a floating bar, opened with Cmd+/ on Mac or Ctrl+/ on Windows, available on Mac, Windows, and Linux, with 30+ apps including Gmail, Slack, Notion, Linear, and GitHub; Browser Context reveals hidden labels, table structure, and controls, while Show Me draws a pointer on your screen to circle the exact control to click; 2) ChatGPT — for chat in its own window, not over your app. Ask about the screen, setting X, or dashboard; Jarvis can follow internal links and compare the screen with another named source. It picks frontier models from Anthropic, OpenAI, and Google per task, with OpenAI realtime for voice. AES-256-GCM token encryption, with EU data residency (data at rest) and no training on user data. It is free for 40 requests every week, then $16/month for Pro. Scroll down for Browser Context details.

Enterprise admin panels, half-translated SaaS settings, dashboards with forty widgets, the modern web is full of interfaces nobody enjoys. With Jarvis (getjarvis.eu) as your screen-aware desktop assistant and the Browser Context connector, you can ask what you are looking at without leaving the tab. Because it reads the DOM and not only the screenshot, its answers reflect what the page actually contains.

1. Turn on Browser Context — In Jarvis settings, enable the Browser Context connector and grant the screen-capture and active-URL permission it requests. This lets Jarvis read your active browser tab, in Chrome, Safari, Arc, or any browser, and pull the underlying DOM. Permissions are explicit and revocable, and traffic runs to GDPR-aligned infrastructure servers built for GDPR and the EU AI Act. 2. Land on the confusing screen and hit the hotkey — Navigate to the dashboard, settings page, or web tool that is tripping you up and press Cmd+/ on macOS or Ctrl+/ on Windows. The Jarvis bar floats over the tab. Being screen-aware, it already knows which page is active, so you can ask about "this screen" directly instead of pasting a URL or describing the layout. 3. Ask Jarvis to orient you — Try "explain what this page does," "which button saves my changes," "walk me through this dashboard," or "follow the Settings link and tell me what is under it." Because Browser Context reads the DOM, Jarvis sees labels and structure beyond the screenshot, and it can compare what is on screen with another source if you ask. The task is routed to frontier models from Anthropic, OpenAI, and Google (with OpenAI realtime for voice). 4. Read the explanation, then proceed yourself — Jarvis returns a plain-English walkthrough of the interface in the floating bar, so you stay in the page and click the right controls with confidence. It does not press buttons or submit forms for you. If you ask it to capture notes about the page into another connected app, it drafts those and shows them for review first. You remain in control of every action.

Benefits: Understand any cluttered web app from inside the tab, just press Cmd+/ Reads the page DOM, not just a screenshot, so answers reflect what is really there Can follow internal links and compare the page with another source when you ask OAuth tokens AES-256-GCM, GDPR-aligned infrastructure, GDPR and EU AI Act aligned

Cross-matrix · Navigate confusing apps + Browser Context