Skip to main content
Get quick answers to the most common questions about Keystone. Can’t find what you’re looking for? Contact our support team.

Getting Started

No. Keystone is completely cloud-based. All test execution happens in our managed browser infrastructure. The only thing you add to your repository is a single GitHub Action step or API call.Your CI system just triggers test execution and receives results - no complex setup required. You can install a CLI tool to create tests from local branches that will be recorded.
Yes. Keystone supports exporting tests to popular testing frameworks:
  • Playwright: Export as TypeScript or JavaScript test files
  • Cypress: Generate Cypress test specs with proper selectors
  • Raw JSON: Export test steps as structured data for custom integrations
Exported tests include all recorded interactions, assertions, and variable handling. This gives you flexibility to run tests locally or migrate to other platforms if needed.
About 5 minutes for a basic setup:
  1. Sign up with GitHub OAuth (30 seconds)
  2. Add your staging environment URL (1 minute)
  3. Record your first test flow (2-3 minutes)
  4. Add GitHub Action to your workflow (30 seconds)
Your first test will run on the next pull request.
Keystone supports multiple authentication methods:
  • Basic authentication (username/password)
  • OAuth 2.0 flows
  • API key headers
  • Custom session cookies
  • SAML/SSO (for environments behind corporate SSO)
Configure authentication in your Environment settings - no code changes needed.
Absolutely. Keystone only needs access to a staging or preview environment. We never interact with production systems unless you explicitly configure a production environment.Start with a free account and test against your staging environment risk-free.

Test Creation & Management

Record browser interactions directly:
  1. Click “Record New Test” in the dashboard
  2. A cloud browser opens with your staging environment
  3. Click through your user journey naturally
  4. Keystone captures every interaction automatically
  5. Save and name your test
No selectors, no scripting - just point and click.
AI healing automatically handles most dynamic content. For more control:
  • Use data attributes that don’t change (data-testid="submit-button")
  • Configure healing to ignore timestamp elements
  • Set up test data with consistent IDs
  • Use environment-specific test accounts

Integration & CI/CD

Any CI system that can make HTTP requests:
  • Native support: GitHub Actions (recommended)
  • REST API: CircleCI, GitLab CI, Jenkins, Azure DevOps
  • Docker: Official runner image for any container-based CI
  • Custom: Direct API integration for proprietary systems
See our CI/CD guide for specific examples.
Yes. Keystone automatically distributes tests across multiple browsers and intelligently coordinates the tests.
Network stubbing is on our roadmap:
  1. We’re developing a Network Stubs panel for test recording
  2. This will allow capturing real API responses as HAR files
  3. Keystone will replay these responses during test runs
  4. No code changes will be needed in your application
Planned support for REST APIs, GraphQL, WebSocket connections, and file uploads.Contact us for current workarounds or to join the beta program.

Data & Security

Test execution artifacts:
  • Video recordings of browser sessions
  • Screenshots at key test steps
  • Network traffic (HAR files)
  • DOM snapshots for debugging
  • Browser console logs
No source code or sensitive credentials are stored. Configure retention periods from 1-180 days.
Default regions:
  • Primary: US West 1 (N. California)
  • Secondary: US West 2 (Oregon)

Migration & Importing (Beta)

We are developing custom import capabilities (in beta) for popular frameworks:
  • Cypress
  • Playwright
  • Selenium
Import process handles common patterns:
  • Page Object Model → Reusable Keystone test components
  • Custom commands → Equivalent recorded actions
  • Test fixtures → Environment-specific configuration
Manual steps may be needed for:
  • Complex custom assertions
  • Framework-specific APIs
  • Unusual wait strategies
No. Your test logic and user journeys are preserved:
  • Same test scenarios, more reliable execution
  • Keep existing CI integration (just change the runner)
  • Gradual migration (run both frameworks in parallel)
  • Team knowledge transfers directly
Think of it as upgrading your test infrastructure, not replacing your tests.

Troubleshooting

Common causes:
  • Different environment URLs (check Environment config)
  • Missing authentication (add auth to Environment)
  • Feature flags disabled in staging
  • Timing differences (AI healing usually fixes this)
Enable debug logging and check test artifacts for detailed diagnosis.
Rich debugging artifacts:
  1. Video recording: See exactly what the browser did
  2. Screenshots: Visual state at each step
  3. HAR files: Network requests and responses
  4. DOM snapshots: Page HTML at failure point
  5. Console logs: JavaScript errors and warnings
Download from the test results page or via API.

Advanced Features

Yes. We have native integrations with Slack, email, and Pagerduty

Still Have Questions?

Contact Support

Email us at [email protected]Response time: <2 hours during business days

Schedule Demo