Help us improve
Share bugs, ideas, or general feedback.
How this command is triggered — by the user, by Claude, or both
Slash command
/mint:browseThe summary Claude sees in its command listing — used to decide when to auto-load this command
# /browse Command Navigate to a URL and perform browser automation tasks. ## Usage ## Arguments | Argument | Required | Description | |----------|----------|-------------| | `url` | Yes | URL to navigate to | | `task description` | No | What to do on the page (e.g., "fill in the form", "check if the button renders") | ## Examples ## Process 1. **Parse arguments:** - Extract URL (first argument) - Remaining text is the task description - If no URL provided: show usage error 2. **Normalize URL:** - If URL has no protocol, prepend `http://` - If URL is just a port (e...
/agent-browserAutomates browser for web interaction and testing: navigate to URLs, take screenshots, fill forms, click elements, or run tests using Playwright CLI.
/browseLaunches headless browser via Playwright to navigate URL, click elements, fill forms, take screenshots. Outputs screenshot path, page title, and URL for visual verification and e2e testing.
/exploreExplores web pages or navigation goals using browser automation. Analyzes with accessibility snapshots and reports elements, navigation options, content found.
/testLaunches Chrome for manual UI testing at given URL, monitors console errors during interaction, generates report with screenshots and logs on completion.
/exploreStarts a browser session for the given URL to explore a website and discover API endpoints. Reports site ID, session details, and page summary.
Share bugs, ideas, or general feedback.
Navigate to a URL and perform browser automation tasks.
/browse <url> [task description]
| Argument | Required | Description |
|---|---|---|
url | Yes | URL to navigate to |
task description | No | What to do on the page (e.g., "fill in the form", "check if the button renders") |
# Navigate and explore
/browse https://myapp.com/login
# Navigate and perform a task
/browse https://myapp.com/login "fill in the form and submit"
# Check local dev server
/browse localhost:3000 "check if the new button renders"
# Scrape information
/browse https://docs.example.com "find the API endpoint for user creation"
Parse arguments:
Normalize URL:
http://:3000), prepend http://localhostLoad browser config from .mint/config.json:
browser key missing: suggest running mint initbrowser.enabled is false: inform user browser features are disabledInvoke browser-runner agent with:
Return result from browser-runner agent
Error: URL is required.
Usage: /browse <url> [task description]
Examples:
/browse https://myapp.com/login
/browse localhost:3000 "check the new button"
Error: Browser plugin not configured.
Add browser config to .mint/config.json:
{
"browser": {
"enabled": true,
"baseUrl": "http://localhost:9867"
}
}
Or run: mint init (to auto-configure)
Warning: PinchTab not running at http://localhost:9867
Start PinchTab:
pinchtab &
Or install:
curl -fsSL https://pinchtab.com/install.sh | sh
http:// prepended automaticallybrowser.token from config for authenticated PinchTab instances