​
Overview

Starts a browser operator session that can autonomously interact with web pages and perform complex actions based on the provided prompt.

​
Use Cases

  • Web Research: Gather information from various websites automatically.
  • Data Extraction: Collect structured data from web pages without manual interaction.
  • Website Testing: Verify functionality and user flows across web applications.
  • Automated Workflows: Perform routine web-based tasks within larger automated processes.

​
Configuration Fields

​
Prompt

  • Description: Enter a question or statement to guide the browser operator on what tasks to perform.
  • Example: “Research the current price of Bitcoin on three different cryptocurrency exchanges” or “Find contact information for tech companies in San Francisco with over 100 employees”
  • Required: Yes

​
Output Variable Name

  • Description: Assign a variable name to store the initialized browser operator session details for later reference.
  • Example: “browser_operator_session” or “web_research_session”
  • Validation: Only letters, numbers, and underscores (_) are allowed in variable names.
  • Required: Yes

​
How It Works

When this action runs, it creates a new browser operator session that:

  1. Analyzes the provided prompt to determine required web navigation steps
  2. Opens a managed browser instance in the background
  3. Autonomously navigates to relevant websites
  4. Interacts with web elements as needed (filling forms, clicking buttons, etc.)
  5. Collects information according to the prompt’s requirements
  6. Returns session details including:
    • session_id: Unique identifier for the session
    • uuid: Universal unique identifier for tracking
    • live_url: URL to watch the browser operator in real-time
    • ws_endpoint: WebSocket endpoint for live updates
    • task: The original prompt submitted

​
Beta Feature

This action is currently in beta. While fully functional, it may undergo changes based on user feedback.

​
Usage Notes

  • For optimal results, be specific in your prompts about what information you need
  • The browser operator can handle complex multi-step tasks but may take longer to complete
  • Use the “Browser Operator Results” action to retrieve the session results once completed