--- title: "Control Chrome from Bearly" description: "Let a Bearly desktop chat read, navigate, and inspect your regular Chrome tabs, including sites you are already signed in to." canonical: "https://bearly.ai/docs/features/chrome-control/" markdown: "https://bearly.ai/docs/features/chrome-control.md" --- # Control Chrome from Bearly HTML: [https://bearly.ai/docs/features/chrome-control/](https://bearly.ai/docs/features/chrome-control/) > Let a Bearly desktop chat read, navigate, and inspect your regular Chrome tabs, including sites you are already signed in to. With the Bearly extension installed, a chat in the Bearly desktop app can work in the Chrome you already use: open and switch tabs, read pages, click and type, and inspect console and network activity. It uses your normal Chrome profile, so sites you are signed in to stay signed in. > **Note — Desktop app and Chrome** > This works in the Bearly desktop app for macOS, Windows, and Linux, with Chrome. It is not available on the web, on mobile, or in another browser. A team admin can turn it off for managed accounts. Bearly can work in a browser three ways: - **Chrome control** uses your everyday Chrome and the sites you are already signed in to. - [Bearly Browser](https://bearly.ai/docs/features/browser.md) is a separate browser beside the chat, with its own profile and its own sign-ins. - [The extension side panel](https://bearly.ai/docs/features/browser-extension.md) answers questions about the page you are reading, without the desktop app. ## Connect Chrome 1. Install **Bearly** from the [Chrome Web Store](https://chrome.google.com/webstore/detail/bearlyai/lkennebljmnnpimcgdjlocbncpamgipp). 2. Open the Bearly desktop app and go to **App Settings** → **Desktop** → **Chrome Extension**. 3. Keep Chrome and the desktop app open. The status updates on its own and reads **Connected to Chrome** when Bearly is ready. Nothing else to configure. Chrome and Bearly connect locally on your machine, and the connection survives restarting either one. You do not need to sign the extension in to Bearly for this — that sign-in is only for the extension's own side panel. ## Ask for a browser task Say what you want and name Chrome. Bearly picks the tab and the tools. ```text copy In Chrome, open the invoice list in my billing tab, find every invoice still unpaid after 30 days, and give me a table with vendor, amount, and due date. Don't pay or cancel anything. ``` Bearly can list your open tabs, so you can also point at one you already have: *"read the Chrome tab with the pull request and tell me what changed"*. ## Approve what Bearly can do Bearly asks in the chat before it touches Chrome, at one of three levels: | Prompt | What it allows | |---|---| | **Read your Chrome tabs** | Tab list, page text, screenshots, console output, and page events | | **Control Chrome** | Everything above, plus opening and closing tabs, navigating, clicking, typing, scrolling, and key presses | | **Use Chrome DevTools** | Everything above, plus network requests and responses, storage, running JavaScript, debugging, profiling, and tracing | Select **Allow** for the single action, or **Allow for…** to keep the level for **This response** or **For now**. **Deny** stops the action and lets the chat continue without it. Approving a higher level covers the lower ones for as long as the grant lasts. While Bearly is attached to a tab, Chrome shows its own debugging banner. Closing the chat, stopping the response, or quitting either app detaches Bearly from every tab. > **Caution — This is your real browser** > Bearly acts in your signed-in Chrome profile and can see whatever those tabs can see, including account pages and, at the DevTools level, request contents and stored session data. Page text, console output, and network data are also untrusted input: a page can contain text written to mislead an assistant. State limits in your request, keep an eye on the tab, and use [Bearly Browser](https://bearly.ai/docs/features/browser.md) instead when a task does not need your own sign-ins. ## Turn it off - Remove the grant for the current work by finishing or stopping the response. A **For now** grant ends when the chat session does. - Quit the Bearly desktop app to disconnect Chrome entirely. Chrome control is unavailable until it is running again. - Disable or remove the Bearly extension in Chrome to end it from the browser side. The extension side panel goes away with it. ## If Chrome control does not work | What you see | What to do | |---|---| | **Chrome extension not connected** | Install or enable the Bearly extension in Chrome, then leave Chrome and the desktop app open until the status changes. | | **Finishing secure connection…** stays on screen | Give it a moment. If it does not finish, restart Chrome, then the desktop app. | | **Update Bearly** | Install the current desktop release and reopen it. | | **Disabled by your workspace** | Ask a team admin whether your assigned policy allows Chrome control. | | Bearly never offers to use Chrome | Confirm you are in the desktop app, not the web app, and that the status reads **Connected to Chrome**. | | **Chrome extension control could not complete that action.** | The tab may have closed, navigated, or lost the attachment. Ask Bearly to list tabs and try again. | | Bearly is stuck on a sign-in wall or CAPTCHA | Finish that step yourself in the tab, then ask Bearly to continue. | Managed-team administrators control availability from **Teams** → team name → **Policies** → policy name → **Code & Automation** → **Chrome extension browser**. On means members can use it. ## Continue through the documentation - [Documentation index](https://bearly.ai/docs.md): Browse every Bearly guide. - [Previous: Use Bearly on a web page](https://bearly.ai/docs/features/browser-extension.md) - [Next: Find and check domain names](https://bearly.ai/docs/features/domain-research.md) - [Related: Browse beside a Bearly chat](https://bearly.ai/docs/features/browser.md)