Browser Agent
Last updated
Last updated
Browser Agent is an AI-powered node that automates web browser interactions. It can navigate websites, click through processes, and complete tasks just like a human user would, then return the results to your workflow.
⚠️ Premium Feature: Browser Agent is only available on Pro and Enterprise plans. This feature is not included in free accounts.
AI-powered web browser automation
Real-time interaction with websites
Human-like navigation and clicking
Task completion with intelligent problem-solving
Recording playback of browser sessions
Session viewing capabilities
Open the Add Action menu using one of three methods:
Click the "Add Action" button in the top left corner
Click on a connection circle on an existing action card
Click and drag from one action to create a connection
Navigate to the Tools tab in the action menu
Find the AI subcategory under Tools
Click on Browser Agent to add it to your canvas
💡 Tip: Browser Agent actions will be automatically numbered if you add multiple instances (Browser Agent 1, Browser Agent 2, etc.)
Locate the Browser Agent card on your canvas which includes:
Header with the action name (which can be renamed)
Variable chip in the top left corner for referencing outputs
Controls for expanding/collapsing, running the agent, and additional options
Connection points for inputs and outputs
Configure the Task input field:
Describe what you want the browser agent to accomplish
Be specific about the website and actions you want performed
Use clear, step-by-step language for complex tasks
The AI will interpret your task and execute the browser interactions automatically
Be specific about the target website:
Example: "Go to leverage.ai and describe what Lleverage does"
Example: "Navigate to Amazon.com and search for wireless headphones"
Describe the desired outcome clearly:
What information you want extracted
What actions should be completed
What format you want the results in
Use action-oriented language:
"Go to [website]"
"Click on [element]"
"Search for [term]"
"Extract [information]"
"Fill out [form]"
Combine multiple steps in one task:
Example: "Go to LinkedIn, search for companies in the tech industry, and list the top 5 results"
Click the play button on the Browser Agent action card to start execution
Be patient during execution:
Browser Agent tasks take longer than typical AI actions
The agent works in real-time, simulating human interactions
Processing time varies based on task complexity
Monitor the progress:
The agent will work through each step systematically
It can handle unexpected situations and adapt as needed
Some complex interactions may require additional processing time
⚠️ Warning: Browser Agent tasks can take significantly longer than other workflow actions due to real-time web interaction requirements.
View the Output:
The agent will return the requested information or confirmation of completed actions
Results appear in the standard output format for use in subsequent workflow steps
Access Session Recording:
After execution, two new buttons appear on the Browser Agent card:
View Recording - Opens a popup showing the step-by-step browser interactions
Open Session - Opens the browser session in a new window for review
Analyze the Process:
Use the recording to understand how the agent navigated the task
Review tab structure and interaction patterns
Identify any areas for task optimization
Start with simple tasks to understand how the agent interprets instructions
Be explicit about required information to ensure accurate extraction
Test tasks thoroughly before deploying in production workflows
Use clear, unambiguous language in task descriptions
Consider timeout implications for time-sensitive workflows
Review recordings to optimize future task descriptions
If the agent gets stuck: Review your task description for clarity and specificity
For failed executions: Check the session recording to identify where issues occurred
For slow performance: Consider breaking complex tasks into smaller, focused steps
For unexpected results: Refine your task instructions with more specific requirements
The Browser Agent action outputs:
Extracted information or data as requested in the task
Confirmation of completed actions
Structured results ready for use in subsequent workflow steps
Access to session recordings for process review
💡 Tip: Browser Agent results can be used as variables in other workflow actions, making it perfect for automated data collection, form filling, and web-based research tasks.