Browser Automation
Browser Automation Tutorial: Let Your AI Employee Operate Web Pages
Let your OpenMax AI Employee log into websites, take screenshots, fill forms, and scrape data — anything a browser can do, your AI can do for you.
Workflow: Install browser component → (Optional) Set up VNC remote monitoring → Give tasks in natural language → AI Employee automates the browser and reports results
Prerequisites
OpenMax AI Employee
No additional accounts needed — the component works out of the box
What Can Browser Automation Do?
Once the browser component is installed, you can ask your AI Employee to complete these tasks with a single message:
Log into accounts: Remotely log into Gmail, company dashboards, or any website
Take screenshots: Capture any web page and send it to your chat
Fill out forms: Automatically fill in and submit forms
Scrape data: Extract structured data from web pages and organize it into tables
Scheduled monitoring: Periodically check a page for price changes, status updates, or content changes
Automated workflows: Multi-step web operations like placing orders, approvals, or publishing content
Step 1: Install the Browser Component
Simply tell your OpenMax AI Employee:
纯文本
Install the browser componentOr send the following command:
纯文本
zylos add browserThe AI Employee will automatically download and install it, usually within 1–2 minutes.
Verify Installation
After installation, ask your AI Employee:
纯文本
Is the browser component installed?The AI Employee will confirm the installation status. Once installed, you can start giving browser tasks right away.
No restart needed — ready to use immediately after installation.
Step 2: (Optional) Set Up VNC Remote Monitoring
VNC lets you see your AI Employee's browser screen in real time. This is especially useful when:
You want to watch the AI's operations live
You need to manually complete a CAPTCHA or SMS verification
You want to log into an account yourself, then hand off to the AI
Enable VNC
Tell your AI Employee:
纯文本
Enable VNC remote desktop accessThe AI Employee will provide you with a web link and a connection password.
Connect via VNC
The browser component provides remote desktop access through noVNC web interface — no client software needed:
Open the VNC web link provided by the AI Employee in your browser
Enter the connection password
You'll see the AI Employee's live desktop right in your browser
Typical Usage: Log In via VNC, Let AI Take Over
Connect to the AI Employee's desktop via VNC
Manually log into your account in the browser (e.g., Google, company systems)
Disconnect from VNC after logging in
Tell the AI Employee: "I've logged in, now help me with..."
The AI Employee reuses the logged-in browser session and continues the remaining tasks
The AI Employee preserves your login state — no need to re-login for future related tasks.
Step 3: Give Browser Tasks
After installation, simply tell your AI Employee what you need in natural language.
Screenshots
纯文本
Open https://openmax.com and send me a screenshot纯文本
Take a screenshot of my Gmail inbox to check for new emailsLogin & Operations
纯文本
Log into my Gmail, check if there are any emails from support@xxx.com today, and send me the subject lines and timestamps纯文本
Open our company's CRM dashboard and screenshot the latest 10 customer recordsFill Out Forms
纯文本
Open this link and fill out the form: name "John Smith", email "john@example.com", then submitData Scraping
纯文本
Open https://example.com/pricing and organize all plan names and prices into a table for me纯文本
Every morning at 9 AM, check the competitor's pricing page and notify me if anything changesDaily Usage Flow
1. Give a Task
Describe what you want the AI Employee to do in natural language — the more specific, the better:
纯文本
Open Gmail, find emails with "invoice" in the subject, download the attachments and send them to me2. AI Executes and Reports
The AI Employee will:
Automatically navigate to the target web page
Understand the page structure and locate interactive elements
Perform clicks, typing, scrolling, and other operations
Take screenshots of key steps
Report the results
3. Handle Exceptions
If the AI encounters login verification, CAPTCHA, or other situations requiring human intervention, it will proactively notify you:
纯文本
I encountered an SMS verification code. Please connect via VNC to handle it, or tell me the code.Once you've handled it, the AI Employee continues with the remaining operations.
常见问题
Q: Can the AI log into accounts with two-factor authentication (2FA)? A: Yes, but you'll need to help with the verification step. We recommend connecting via VNC to manually complete 2FA, then letting the AI take over. Alternatively, you can send the verification code directly to your AI Employee.
Q: Is the AI's browser the same as my local browser? A: No. The AI Employee runs an independent browser on the server, completely separate from your local browser. VNC lets you remotely view the AI's browser.
Q: Can the AI still use the account I logged into via VNC? A: Yes. The AI Employee preserves the login session and can use it for future tasks without re-logging in, unless the session expires or is manually cleared.
Q: Can the AI run browser tasks on a schedule? A: Yes. Just tell the AI Employee: "Check ... for me every morning at 8 AM," and it will automatically set up a scheduled task.
Q: Will the AI trigger anti-bot detection? A: The browser component is built on a real browser engine (Playwright), so its behavior closely resembles a normal user. The risk of triggering anti-bot measures is low, but some platforms have strict detection — we recommend the AI maintain reasonable intervals between actions.
Q: Which browsers does the component support? A: It's built on Chromium (Chrome engine) and is compatible with the vast majority of modern websites.
Q: What if installation fails? A: Tell your AI Employee "The browser component installation failed, help me troubleshoot," and it will automatically diagnose and fix the issue.
Learn more: openmax.comContact us: support@openmax.com
