Browser Automation

Pando includes a full browser automation suite using chromedp, with support for multiple browser types including the lightweight Lightpanda browser.

Browser Types

Available Tools

Navigation & Content

• browser_navigate - Navigate to URL
• browser_get_content - Extract HTML, text, or title
• browser_screenshot - Capture page or element screenshot

Interaction

• browser_click - Click on elements
• browser_fill - Fill form inputs
• browser_scroll - Scroll the page
• browser_evaluate - Execute JavaScript

Analysis

• browser_console_logs - Get JavaScript console messages
• browser_network - Get network requests
• browser_pdf - Generate PDF of page

Lightpanda Support

Lightpanda is a lightweight, headless browser designed for automation:

[InternalTools]
BrowserType = 'lightpanda'

Unlike local browsers, Lightpanda is launched as a CDP server process and connected via WebSocket. No user profile or headless flags apply.

Configuration

[InternalTools]
BrowserEnabled = true
BrowserType = 'chrome'
BrowserExecutable = ''         # Auto-detect
BrowserHeadless = false
BrowserTimeout = 30            # seconds
BrowserUserDataDir = ''
BrowserMaxSessions = 3

Fetch Tool Integration

The fetch tool supports browser backends for JavaScript-heavy pages:

{
  "url": "https://example.com",
  "format": "markdown",
  "browser": "auto"
}

Browser backends: auto, chrome, firefox, curl, http