gstack-open-gstack-browser

To launch a visible Chromium browser for UI automation debugging, trigger the session by asking the AI to open the browser. This opens an on-screen instance so you can directly observe navigation, clicks, and sidebar activity in real time.

The activity sidebar helps with observing automated browser actions by displaying a live feed and chat alongside the browser. This shows what the session is doing and why, making it easier to debug UI flows instead of guessing what actions occurred.

Yes, you can use stealth measures to reduce bot detection during automated browsing. The browser session includes anti-bot stealth support intended to minimize detection problems while navigating and interacting with web pages.

The visible browser session works for UI walkthroughs and browsing tasks where the model needs to navigate, interact, and present ongoing status. It requires running the browser-launch workflow for trigger phrases like launch chromium or control my browser.

A limitation of running a GStack-controlled browser session is that it uses only approved tools for plan-mode safety and user prompting. It is specifically designed for tasks requiring real-time visual observation rather than fully background automated processes.