🌐 Browser Automation Tools
Browser tools are available exclusively to the qa-kitten agent for automated browser testing and web scraping.
🐱 QA Kitten Required
Switch to the QA Kitten agent to use browser tools: /agent qa-kitten
Browser Control
browser_initialize(headless=True)
Start a browser session. Set headless=False to see the browser.
browser_close()
Close the browser and clean up resources.
browser_status()
Check if browser is running and get current URL.
Navigation
browser_navigate(url)
Navigate to a URL. Waits for page load.
browser_go_back()
Go back in browser history.
browser_go_forward()
Go forward in browser history.
browser_reload()
Reload the current page.
Interactions
browser_click(selector)
Click an element. Supports CSS selectors and semantic locators.
browser_set_text(selector, text)
Type text into an input field. Clears existing text first.
browser_get_text(selector)
Get text content from an element.
browser_select(selector, value)
Select an option from a dropdown.
browser_check(selector, checked=True)
Check or uncheck a checkbox.
Semantic Element Location
Find elements using semantic attributes (better for testing):
browser_find_by_role(role, name=None)
Find by ARIA role: button, link, textbox, heading, etc.
browser_find_by_text(text, exact=True)
Find by visible text content.
browser_find_by_label(label)
Find form elements by their label text.
browser_find_by_placeholder(placeholder)
Find inputs by placeholder text.
browser_find_by_test_id(test_id)
Find by data-testid attribute.
Visual Analysis
browser_screenshot_analyze(question=None)
Take a screenshot and optionally analyze it with vision AI.
JavaScript Execution
browser_execute_script(script)
Execute JavaScript in the page context.
Workflow Management
browser_save_workflow(name)
Save current browser actions as a reusable workflow.
browser_load_workflow(name)
Load and execute a saved workflow.
Example Session
/agent qa-kitten
> Initialize a browser and go to example.com
🐱 Starting browser...
[browser_initialize(headless=False)]
[browser_navigate("https://example.com")]
> Click the "More information" link
[browser_find_by_text("More information")]
[browser_click(...)]
> Take a screenshot and describe what you see
[browser_screenshot_analyze("Describe the current page")]
> Close the browser
[browser_close()]Complete Tool List
| Category | Tools |
|---|---|
| Control | initialize, close, status |
| Navigation | navigate, go_back, go_forward, reload |
| Interactions | click, set_text, get_text, select, check |
| Locators | find_by_role, find_by_text, find_by_label, find_by_placeholder, find_by_test_id |
| Visual | screenshot_analyze |
| Scripts | execute_script |
| Workflows | save_workflow, load_workflow |