FeaturesTest Execution

Test Execution

Execute your test cases on real browsers using BugBrain’s AI-powered autonomous agents. Watch tests run, capture screenshots, and analyze results.

How Test Execution Works

When you run a test, BugBrain:

  1. Launches a browser (Chrome, Firefox, or Safari)
  2. Executes each step using an AI agent
  3. Captures screenshots at each step
  4. Records detailed logs of all actions
  5. Returns results with pass/fail status

Execution Providers

BugBrain supports three execution providers:

Browser-Use Local

What it is: Local Python service with Browser-Use library and Anthropic AI

Best for:

  • Standard test execution
  • Most test scenarios
  • Cost-effective testing

Features:

  • Claude AI for intelligent interactions
  • Self-healing selectors
  • Screenshot capture
  • Detailed logging

Hybrid Executor

What it is: Playwright + Gemini AI with self-healing capabilities

Best for:

  • Complex applications
  • Dynamic selectors
  • Self-healing tests

Features:

  • AI-powered script generation
  • Automatic selector healing on failures
  • Playwright stability
  • Fast execution

Browser-Use Cloud

What it is: Cloud-hosted Browser-Use service

Best for:

  • No local setup needed
  • Scalable test execution
  • Distributed teams

Features:

  • Managed infrastructure
  • Automatic updates
  • No maintenance required
  • Pay-per-use pricing

Your organization admin configures which execution provider to use. All providers support the same test case format.

Execution Options

Browser Selection

Choose which browser to use:

  • Chrome (default) - Most compatible
  • Firefox - Good for cross-browser testing
  • Safari - Mac/iOS testing

Headless Mode

Run browsers without a visible window:

  • Headless: On - Faster, uses less resources (default)
  • Headless: Off - See the browser window, useful for debugging

Screenshot Capture

Control when screenshots are captured:

  • Every Step - Maximum visibility (recommended)
  • On Failure - Only when tests fail
  • On Assertions - Only at verification points
  • Disabled - No screenshots (faster)

Timeouts

Set maximum execution time:

  • Default: 300 seconds (5 minutes)
  • Range: 30-600 seconds
  • Per Test: Can override in test settings
⚠️

Important: Increase timeout for slow-loading applications or complex workflows. Tests will fail if they exceed the timeout.

Execution Flow

1. Test Queued

When you click “Run Test”:

  • Test is added to the execution queue
  • Status shows as “Queued”
  • Wait time depends on current load

2. Browser Launches

Once execution starts:

  • Browser opens (or starts headless)
  • Initial page loads
  • Status updates to “Running”

3. Steps Execute

For each test step:

  • AI agent reads the step description
  • Locates the target element
  • Performs the action
  • Captures screenshot (if enabled)
  • Logs the result

4. Results Generated

After all steps complete:

  • Overall status determined (Pass/Fail)
  • Screenshots compiled
  • Logs finalized
  • Notifications sent (if configured)

Understanding Results

Pass Status ✅

Test passed if:

  • All steps completed successfully
  • No errors occurred
  • All assertions passed

Fail Status ❌

Test failed if:

  • A step couldn’t complete
  • Element not found
  • Assertion failed
  • Timeout exceeded
  • Browser error occurred

Warning Status ⚠️

Test passed but with warnings:

  • Slow page loads
  • Deprecated selectors
  • Minor issues that didn’t fail the test

Real-Time Monitoring

While tests run, you can:

Watch Progress:

  • See which step is executing
  • View live screenshots
  • Monitor execution time

View Browser:

  • If headless mode is off, watch the browser
  • See exactly what the AI agent does

Cancel Execution:

  • Click “Cancel” to stop a running test
  • Partial results are saved

Execution History

Every test execution is saved with:

  • Timestamp - When it ran
  • Duration - How long it took
  • Status - Pass/Fail/Warning
  • Screenshots - Visual record
  • Logs - Detailed output
  • Environment - Browser, provider used

Access history:

  1. Go to Test Cases
  2. Click on a test
  3. View “Execution History” tab

Concurrent Execution

Run multiple tests at the same time:

Limits by Plan:

  • Starter: 2 concurrent executions
  • Growth: 5 concurrent executions
  • Pro: 10 concurrent executions

Benefits:

  • Faster test suite completion
  • Parallel test plans
  • Efficient resource usage

Best Practices

💡

1. Enable Screenshots Always capture screenshots - they’re invaluable for debugging failures.

2. Set Appropriate Timeouts Allow enough time for pages to load, but not so much that stuck tests run forever.

3. Use Descriptive Selectors Text-based selectors (e.g., “Button with text ‘Submit’”) are more maintainable than CSS selectors.

4. Add Wait Steps Insert waits before assertions to ensure pages are fully loaded.

5. Monitor Execution Time If tests are slow, investigate and optimize page load times or selectors.

6. Review Logs Check execution logs for warnings even when tests pass.

Troubleshooting

Test times out?

  • Increase timeout in project settings
  • Add explicit wait steps
  • Check if page is actually loading

Element not found?

  • Verify selector is correct
  • Check if element is hidden or in iframe
  • Add wait before interacting with element

Flaky tests (pass/fail randomly)?

  • Add more wait time
  • Use more specific selectors
  • Check for race conditions in application

Browser crashes?

  • Reduce concurrent executions
  • Check for memory-heavy pages
  • Contact support if persistent

Performance Tips

Speed up test execution:

  • Disable unnecessary screenshots
  • Use headless mode
  • Reduce wait times where safe
  • Run tests in parallel
  • Use faster execution provider

Reduce costs:

  • Delete old execution results
  • Run only necessary tests
  • Use test plans for batch execution
  • Schedule tests during off-peak hours

Next Steps

AI Test GenerationDiscovery