Skip to content

TINAA Best Practices

Follow these best practices to get the most out of TINAA's AI-powered browser automation and testing capabilities.

Test Execution Best Practices

1. Focused Test Requests

When using the HTTP API, be specific about what you want to test:

Good: 

{
  "action": "exploratory",
  "parameters": {
    "url": "https://example.com/login",
    "focus_area": "forms",
    "depth": 3
  }
}

Better: 

{
  "action": "exploratory",
  "parameters": {
    "url": "https://example.com/login",
    "focus_area": "authentication",
    "scenarios": ["valid_login", "invalid_credentials", "forgot_password"]
  }
}

2. Leverage WebSocket for Real-time Feedback

Monitor test progress in real-time:

const ws = new WebSocket('ws://localhost:8765/ws');

ws.onmessage = (event) => {
  const progress = JSON.parse(event.data);
  if (progress.type === 'error') {
    console.error('Test failed:', progress.details);
  }
};

3. Use Playbooks for Complex Scenarios

Create playbooks for multi-step test scenarios:

{
  "name": "checkout-flow",
  "actions": [
    {"type": "navigate", "url": "https://shop.example.com"},
    {"type": "exploratory", "focus": "product-selection"},
    {"type": "form_analysis", "target": "checkout"},
    {"type": "accessibility", "standard": "WCAG2AA"}
  ]
}

Performance Optimization

Parallel Test Execution

Run multiple test types simultaneously:

# Run in separate terminals or use a script
curl -X POST http://localhost:8765/test/exploratory -d '{"url": "page1"}'
curl -X POST http://localhost:8765/test/accessibility -d '{"url": "page2"}'

Resource Management

  • Close browser contexts after tests
  • Use headless mode for CI/CD environments
  • Implement proper timeouts for long-running tests

Integration Best Practices

1. CI/CD Integration

# GitHub Actions example
- name: Run TINAA Tests
  run: |
    docker-compose -f docker-compose.prod.yml up -d
    ./scripts/wait-for-tinaa.sh
    curl -X POST http://localhost:8765/test/accessibility \
      -d '{"url": "${{ env.STAGING_URL }}"}'

2. MCP Integration with Claude

When using with Claude Desktop: - Be specific in your requests - Ask for screenshots to verify actions - Request test summaries after execution

Example prompts: - "Navigate to example.com and check all form fields for accessibility" - "Take screenshots of each step during the checkout process" - "Analyze the login page for security best practices"

Test Result Analysis

1. Review Accessibility Reports

Pay attention to: - Critical violations (must fix) - Serious violations (should fix) - Moderate violations (consider fixing)

2. Analyze Exploratory Test Insights

Look for: - Broken links discovered - Form validation issues - JavaScript errors - Performance bottlenecks

Maintenance Best Practices

1. Regular Health Checks

# Add to your monitoring
curl http://localhost:8765/health

2. Log Analysis

Monitor TINAA logs for: - Failed test patterns - Performance degradation - Resource usage

3. Update Resources

Keep test patterns and heuristics updated: - Review /resources/test_patterns.json - Update accessibility rules as standards evolve - Refine exploratory heuristics based on findings

Common Pitfalls to Avoid

  1. Running tests without specific focus - Always specify what to test
  2. Ignoring WebSocket feedback - Use real-time updates for better insights
  3. Not reviewing test results - Analyze reports for actionable improvements
  4. Overloading the system - Limit concurrent tests based on resources

Security Considerations

  1. Don't expose TINAA publicly - Keep it on internal networks
  2. Validate input URLs - Ensure you're testing authorized domains
  3. Secure sensitive test data - Don't hardcode credentials