Browser Automation
Control a Chromium browser with MyOctopus. Navigate pages, fill forms, take screenshots, and read content. Login sessions persist across app restarts.
How It Works
The Browser Action automation type executes a sequence of browser steps:
| Step | Description |
|---|---|
| Navigate | Go to a URL |
| Click | Click an element on the page |
| Fill | Type text into an input field |
| Read Text | Get text content of a specific element |
| Read Page | Get all visible text on the page |
| Screenshot | Capture the page as an image |
| Wait For | Wait for an element to appear |
| Wait | Pause for N seconds |
| Evaluate | Run JavaScript on the page |
| Close | Close the browser |
Persistent Sessions
The browser remembers your login sessions between uses. This means:
- WhatsApp Web stays logged in after QR scan
- Social media sessions persist
- No need to re-authenticate on every use
Screenshot + AI Analysis
Combine screenshots with AI to let Claude see web pages:
"screenshot WhatsApp and list unread messages"
This navigates to WhatsApp Web, takes a screenshot, and sends it to Claude for analysis.
Examples
| Command | What happens |
|---|---|
| "open google.com in the browser" | Launches Chromium, navigates to Google |
| "fill the search box and click search" | Types into input, clicks button |
| "read this webpage and summarize it" | Reads page text, sends to Claude |
| "take a screenshot of the page" | Captures visible viewport |
First Use
On first use, MyOctopus installs Chromium automatically. The browser opens visibly (not headless) so you can see what's happening.