Skip to main content
Once you’ve started a simulation, monitoring its progress is crucial to ensure your agent learns effectively. This guide covers how to track simulation status, respond to agent questions, and ensure successful exploration.

Understanding Simulation Status

Status Indicators

Your simulations will show different status indicators: Pending 🟡:
  • Simulation is queued and waiting to start
  • Agent is preparing for exploration
  • Usually brief, lasting a few seconds
Running 🟢:
  • Agent is actively exploring your application
  • You’ll see real-time progress updates
  • This is when you should monitor most closely
Completed ✅:
  • Simulation finished successfully
  • Agent has completed its exploration
  • Results are available for review
Failed ❌:
  • Simulation encountered an error
  • Agent couldn’t complete the exploration
  • Check error details for troubleshooting
Paused ⏸️:
  • Agent is waiting for your response to a question
  • Simulation will resume after you provide an answer
  • Respond promptly to keep exploration moving

Real-Time Monitoring

Progress Updates

As your agent explores, you’ll see detailed status messages: Navigation Updates:
  • “Navigating to the dashboard”
  • “Exploring the user settings page”
  • “Accessing the checkout process”
  • “Reviewing the help documentation”
Discovery Updates:
  • “Found the account management section”
  • “Located the project creation workflow”
  • “Identified the payment processing flow”
  • “Discovered the user onboarding steps”
Learning Updates:
  • “Understanding how users create accounts”
  • “Learning the product browsing interface”
  • “Mapping the complete user journey”
  • “Recording interaction patterns”

Activity Timeline

The simulation dashboard shows:
  • Timestamp of each activity
  • Detailed descriptions of what the agent is doing
  • Progress milestones and major discoveries
  • Time elapsed since simulation started

Responding to Agent Questions

Types of Questions

Your agent may ask different types of questions during exploration: Clarification Requests:
  • “Should I focus on the mobile or desktop version?”
  • “Which user role should I explore as?”
  • “Should I test the free or paid features?”
Permission Requests:
  • “Can I create a test account to explore the signup process?”
  • “Should I proceed with the checkout process using test data?”
  • “Is it okay to explore the admin panel features?”
Context Questions:
  • “What should I do if I encounter an error during exploration?”
  • “How should I handle two-factor authentication prompts?”
  • “Should I explore the help documentation as well?”
Confirmation Requests:
  • “I’ve completed exploring the dashboard. Should I move to user settings?”
  • “I found the main features. Should I also explore the mobile app?”
  • “I’ve mapped the basic workflow. Should I test error scenarios?”

How to Respond

Be Clear and Specific:
  • Provide direct answers to questions
  • Give context when helpful
  • Be concise but complete
  • Avoid ambiguous responses
Examples of Good Responses:
  • “Yes, focus on the desktop version for now”
  • “Create a test account using the email test@example.com
  • “If you encounter errors, try to understand what went wrong and how to recover”
  • “Yes, move to user settings after completing the dashboard”
Examples of Poor Responses:
  • “Maybe” (too vague)
  • “I don’t know” (doesn’t help the agent)
  • “Whatever you think” (no guidance)
  • “Try your best” (not specific enough)

Response Best Practices

Respond Promptly:
  • Agent questions pause the simulation
  • Quick responses keep exploration moving
  • Delayed responses can cause timeouts
  • Set aside time to monitor active simulations
Provide Context:
  • Explain why you’re giving specific instructions
  • Share relevant business context when helpful
  • Help the agent understand your goals
  • Guide the agent toward valuable discoveries
Be Patient:
  • Agents may ask many questions initially
  • Questions help ensure accurate exploration
  • Better questions lead to better learning
  • Quality exploration takes time

Monitoring Best Practices

Active Monitoring

Check Status Regularly:
  • Monitor simulations every 10-15 minutes
  • Respond to questions within 30 minutes
  • Watch for error messages or stuck states
  • Ensure exploration is progressing as expected
Review Progress Updates:
  • Read status messages to understand what’s happening
  • Look for signs of effective exploration
  • Identify areas where agent might be struggling
  • Note interesting discoveries or insights

Effective Question Handling

Prepare for Common Questions:
  • Have test credentials ready
  • Know your application’s key features
  • Understand your business context
  • Be ready to provide specific guidance
Guide Agent Focus:
  • Direct agent attention to important areas
  • Clarify priorities when agent asks
  • Help agent understand user goals
  • Ensure comprehensive coverage

Troubleshooting During Simulation

Agent Seems Stuck:
  • Check if agent is waiting for a response
  • Look for error messages in status updates
  • Provide more specific instructions
  • Consider stopping and restarting with clearer guidance
Exploration Going Off-Track:
  • Answer questions to redirect focus
  • Provide additional context about priorities
  • Clarify what you want the agent to learn
  • Guide agent back to important areas
Technical Issues:
  • Check connection status and credentials
  • Verify application is accessible
  • Look for network or authentication errors
  • Contact support if issues persist

Simulation Dashboard Features

Overview Panel

Current Status:
  • Active simulation count
  • Recent simulation history
  • Quick access to running simulations
  • Status summary for each simulation
Quick Actions:
  • Start new simulation
  • View simulation details
  • Access simulation history
  • Manage simulation settings

Detailed View

Simulation Information:
  • Connection and agent details
  • Instructions provided
  • Start time and duration
  • Current status and progress
Activity Log:
  • Chronological list of all activities
  • Detailed status messages
  • Question and answer history
  • Error messages and warnings
Question Interface:
  • Current questions from agent
  • Response input field
  • Send response button
  • Question history

Managing Multiple Simulations

Running Concurrent Simulations

When to Run Multiple:
  • Different agents exploring different areas
  • Separate simulations for different features
  • Testing different instruction approaches
  • Exploring different user roles or scenarios
Best Practices:
  • Monitor each simulation individually
  • Respond to questions promptly for all simulations
  • Avoid overwhelming yourself with too many at once
  • Focus on quality over quantity

Simulation Prioritization

High Priority:
  • Core user workflows
  • New feature exploration
  • Error scenario testing
  • User onboarding processes
Medium Priority:
  • Secondary features
  • Advanced functionality
  • Integration testing
  • Edge case exploration
Low Priority:
  • Nice-to-have features
  • Experimental functionality
  • Rare use cases
  • Optional workflows

Common Monitoring Scenarios

New Feature Exploration

What to Monitor:
  • Agent’s understanding of new functionality
  • Discovery of key features and workflows
  • Identification of user interaction patterns
  • Recognition of potential user issues
Questions to Expect:
  • “Should I explore this new feature in detail?”
  • “How does this relate to existing functionality?”
  • “What should I focus on for this feature?”

User Journey Mapping

What to Monitor:
  • Complete workflow coverage
  • Logical progression through steps
  • Understanding of user goals
  • Recognition of decision points
Questions to Expect:
  • “Should I continue with the next step?”
  • “What should I do if I encounter this scenario?”
  • “Is this the complete user journey?”

Error Scenario Testing

What to Monitor:
  • Agent’s ability to trigger errors
  • Understanding of error messages
  • Knowledge of recovery procedures
  • Recognition of when to escalate
Questions to Expect:
  • “Should I intentionally trigger this error?”
  • “What should I do when I see this error message?”
  • “How should I test error recovery?”

Next Steps

After monitoring your simulation:
  1. Review simulation results - See what your agent learned
  2. Test agent performance - Verify improved capabilities
  3. Create additional simulations - Explore more areas
  4. Update agent knowledge - Integrate new learning
Effective monitoring ensures your simulations produce valuable learning for your agents. By staying engaged and providing helpful guidance, you’ll create agents that truly understand your product.

Getting Help

If you need assistance monitoring simulations:
  • Check the troubleshooting guide for common issues
  • Review monitoring best practices for effective oversight
  • Test with simple scenarios to build experience
  • Contact support for additional help
You’re now ready to monitor simulations effectively and guide your agents to success! 🚀
I