Use Atera browser connect to automate browser-based tasks

Atera browser connect lets Robin operate a real Chrome browser on an endpoint to complete IT tasks that have no API, such as internal portals, HR systems, printer admin pages, and SaaS consoles. Robin navigates, reads pages, fills forms, and clicks through flows the same way a technician would, then reports a structured result back to Atera.

Sessions run inside the endpoint's actual Chrome (or Edge) profile, using the signed-in user's real, authenticated session. Nothing is simulated or run headlessly. Robin works in the same browser session the user sees.

Why use Atera browser connect?

Closes the API gap: Resolves goals inside web apps that have no API, so tasks that used to need a human's clicks can complete on their own.
Runs on real, authenticated sessions: Operates in the endpoint's own Chrome profile, so it can act on internal portals and tools that require real credentials.
Grounded in live page state: Every action is based on a fresh read of the current page, so Robin never acts on stale or assumed state.
Visible and controllable: An on-page overlay shows what Robin is doing in real time, and the end user can stop a session at any point.
Bounded by design: Sessions run within a hard step limit and timeout watchdogs, so there's no risk of a runaway session.

Before you begin

Take note of the following:

Atera browser connect requires the Robin module to be enabled on your account.
Playbooks aren’t required to connect to the integration. However, browser sessions can only be triggered by a playbook. You'll need to configure at least one playbook before Robin can run a session.
Supported browsers are Chrome and Edge (Chromium-based). Other browsers aren't supported.
Supported endpoint operating systems are Windows, and macOS.
Sessions run in the end user's real, visible browser. Headless or sessions aren't supported.
The extension needs to be installed on the endpoint before a session can start. If it isn't already installed, the Chrome Web Store page opens on the endpoint so the user can install and enable it.
On managed fleets that block external extensions via policy, force-install may not go through. Check your Chrome/Edge management policy if the extension doesn't appear. If you have the Google workspace integration enabled in Atera, you can update this policy from the browser connect integration's manage dialog.

Connect Atera browser connect

Connect the integration once per account (ITD) or once per customer (MSP). After it's connected, Robin can request browser sessions on any covered endpoint.

To connect Atera browser connect:

Go to AI center > Integrations.
Find Atera browser connect and click Connect.
Choose a Deployment option:

On demand: When a session is first requested, the Chrome Web Store page opens on the endpoint so the user can install the extension. It isn't installed silently.
Google workspace: Available if you have the Google workspace integration enabled. Selecting it lets you choose the organizational units (OUs) to apply the force-install policy to.

Review the AI capabilities Robin will have once connected (see What Robin can do in your browser below), then click Connect.

Note: With On demand selected, you'll need to install Atera browser connect on the endpoint from the Chrome Web Store before it can connect to the integration.

Atera browser connect now shows as Connected. Robin can request browser sessions on your endpoints.

What Robin can do in your browser

Robin has around 40 action types available. The most common ones are:

Navigate and read: Opens URLs, follows links, and reads text, tables, and fields from any page.
Fill and submit forms: Types into fields, selects options, and submits forms on any web portal.
Click and interact: Clicks buttons, opens menus, and handles pop-ups and confirmation dialogs.
Extract structured data: Pulls tables, lists, and key-value pairs from a page for use in tickets.

How a browser session works

Once Robin identifies a goal it needs a browser to complete, it requests a session on the relevant endpoint. From there:

Chrome (or Edge) opens on the endpoint, and the extension connects to Robin.
Robin reads the current page (a summary of interactive elements plus a screenshot) and decides the next action.
The extension executes that action in the live tab, such as clicking a button or filling a field, and returns the result.
Robin reads the result and decides the next step. This repeats until the goal is reached, Robin determines it's blocked, or a session limit is hit (see Session limits below).
Robin reports the outcome back to Atera as a structured result.

Note: Because each step re-reads the live page rather than assuming an action worked, Robin can catch and adjust to things like a slow-loading page or an unexpected pop-up mid-task.

On-page overlay

While a session runs, an overlay on the page shows the end user:

The agent's current mode: Thinking, Executing, Waiting, Error, Done, or Inactive.
The current step count.
A list of recent actions.

The end user can stop the session from the overlay at any time.

Session limits

Sessions are automatically bounded so they can't run indefinitely:

Step budget: Session caps to a maximum of 50 steps , whether or not the goal was reached.
Watchdog timeouts: A session is ended if a command goes unacknowledged or the session sits inactive too long.
Manual stop: The end user can end a session at any time from the overlay.

When a session ends for any reason, Robin receives a structured result explaining the outcome.

Disconnect Atera browser connect

Disconnecting removes the integration from your account. Any playbooks that depend on browser sessions will stop working until you reconnect.

To disconnect Atera browser connect:

Go to AI center > Integrations.
Find Atera browser connect and click Disconnect. If you used the Google workspace integration to apply a force-install policy to OUs, remove it here first. This screen lets you both apply and remove the policy.
Click Disconnect again in the confirmation window.

Atera browser connect is now disconnected. You'll need to reconnect it to use the integration again.

Privacy and security

No long-term storage of page content: Page state and screenshots are used to drive the session and are retained only briefly afterward for security-audit purposes, then discarded. Login credentials are never extracted, transmitted, or stored.
Credentials never leave the endpoint: Atera browser connect operates inside the endpoint's own browser profile. Robin never extracts, transmits, or stores login credentials.
Harmless without Robin: A user can install the extension from the Chrome Web Store, but it does nothing on its own. It can only act on instructions from Robin, so an installed-but-unassigned extension is effectively inert.
User visibility and control: The on-page overlay makes every session visible to the person at the machine, who can stop it at any time.
Scoped access: The extension stays dormant by default, even during a Robin session. It only becomes active after the user approves it in the chat and Robin activates it. It has no access to browsing activity outside assigned tasks.

Use cases

Vendor portal account resets: A technician needs to reset a user's password in a vendor portal that has no API. Robin logs in with the endpoint's real session, navigates to the account, and completes the reset, then reports back to the ticket.

Printer configuration: A network printer only exposes settings through its web-based admin UI. Robin navigates to the admin page, updates the setting, and confirms the change.

HR system updates: An HR platform requires manual clicks to update an employee record. Robin fills out the relevant fields and submits the form on the technician's behalf.

Data extraction for tickets: A SaaS console shows account details only in the UI. Robin extracts the relevant table or key-value data and attaches it to the ticket, no manual copy-paste required.

FAQ

Q: Which browsers are supported?

A: Chrome and Edge (Chromium-based). Other browsers aren't supported for this integration.

Q: Does this work with headless?

A: No. Sessions always run in the end user's real, visible browser session.

Q: What happens if the extension can't be force-installed on a managed fleet?

A: Some fleet policies (such as a block on external extensions) can prevent force-install. Check your Chrome/Edge management policy. If you use the Google workspace deployment option, you can adjust the force-install policy for the relevant OUs from the browser connect manage dialog, or switch to On demand and install from the Chrome Web Store instead.

Q: Can the end user see what Robin is doing?

A: Yes. An on-page overlay shows the current mode, step count, and recent actions for the duration of the session, and the end user can stop it at any time by clicking the End session button.

Q: What happens when a session hits the step limit?

A: The session ends automatically at 50 steps, and Robin reports back whatever result it reached at that point.

Q: Does Robin store my credentials or page content?

A: No. Atera browser connect runs in the endpoint's own browser profile and never extracts or stores credentials. Page content and screenshots are used to drive the active session and are retained only briefly afterward for security-audit purposes, then discarded.

Q: What happens if I disconnect the integration?

A: Any playbooks or automations that rely on browser sessions stop working immediately. You'll need to reconnect Atera browser connect to use them again.

How can we help?

Common topics: