Browser tasks
Use Agent's separate browser workspace.
Learn where browser tasks run, how to handle sign-in, and how to continue after taking over briefly.
On this page
Work in Agent's separate browser
Browser tasks run in a browser workspace owned by Agent. It stays separate from the tabs in your everyday browser and does not require a browser extension. The pages needed for the task stay together in that Agent workspace.
Use the Agent browser for the sites needed by the request. When a task moves between research, a document editor, and another web page, Agent keeps those pages together in that workspace.
Agent uses its own virtual cursor inside the browser. It does not move your system pointer, so your mouse stays where you left it while Agent works.
Take over for sign-in
- 1
Let Agent open the account page
Agent navigates to the sign-in page inside its browser workspace.
- 2
Enter private details yourself
Type the email, password, security key response, or other private account information directly in that browser.
- 3
Return to Agent
After the account page finishes signing in, ask Agent to continue the original task.
Keep the handoff in the same window. You do not need to move the task into your personal browser. Whether a site keeps the session signed in later depends on that site's account rules.
Keep the browser task available
Closing the main LowLighter dashboard does not quit Agent. The Agent overlay and its browser task can remain available while you work elsewhere. Reopen the dashboard from the system tray on Windows or the menu bar on macOS.
Choosing Quit stops LowLighter and its background features. Open LowLighter again before starting or continuing another Agent task.