End-to-End Browser Testing Using Natural Language
Write tests in plain English. They run in real browsers — Chromium, Firefox, or WebKit — no selectors, no brittle scripts, no test framework to learn.
How It Works
Leverage AI for enhanced automation, or save on costs with a pure statement-driven approach. Either way, you write the same plain English steps.
Statement-Driven
The resilience framework interprets and executes your steps directly. Fast, predictable, and cost-effective — ideal when you know exactly what to test.
AI-Enhanced
AI drives the browser on the first run and figures out the interactions for you. Subsequent runs replay without AI — ideal when you want to describe intent and let AI handle the details.
Record-and-Replay Architecture
AI Natural Language monitors use a record-and-replay approach that keeps costs predictable while adapting to UI changes automatically.
First Run — AI Records
AI executes your instructions in a real browser, interpreting your intent and recording every interaction it performs.
Replay — No AI Tokens
Subsequent runs replay the recorded steps through the resilience framework. No AI involvement means consistent, low-cost execution.
Self-Healing — AI Re-Engages
If a replay fails because the UI changed, AI re-engages to figure out the new interactions and re-records for future runs.
The Resilience Framework
Both monitor types execute through the same resilience framework — a smart element-finding engine that adapts to page changes instead of relying on brittle selectors.
The framework handles a wide range of browser interactions: clicking, typing, form interaction, hovering over menus, drag-and-drop, and managing multiple tabs. When the page layout changes, the framework adapts rather than breaking.
You can also extend the framework by injecting additional commands into test steps, making it flexible enough for complex testing scenarios.
Advanced Features
Encrypted Secrets
Securely inject credentials and API keys into test runtime. Organization-level, encrypted at rest.
Dynamic Test Data
Generate unique values per run — randomized strings, numbers, emails. No stale test data.
Screenshots
Capture at every step, only on error, or disable entirely. Visual evidence of test execution.
Console Capture
Record browser console errors, warnings, and uncaught exceptions during test runs.
Browser Selection
Run tests in Chromium, Firefox, or WebKit to validate cross-browser compatibility.
Viewport Control
Set custom viewport dimensions to test responsive layouts across device sizes.
Run From Anywhere
Browser automation isn't locked inside a dashboard. Trigger tests wherever they fit into your workflow.
Scheduled Runs
Set monitors to run daily so critical workflows are validated continuously without manual effort.
CLI Tool
Trigger tests from your terminal or CI/CD pipeline. Integrate browser testing into your build and deploy process.
MCP Server
Connect to AI coding tools like Claude Code, Cursor, and Windsurf. Run monitors, check results, and react to failures without leaving your editor.
Ideal Use Cases
User Registration & Signup
Test complete signup flows from form submission through email confirmation to account activation.
Login & Authentication
Verify login flows including multi-factor authentication, password resets, and session handling.
Checkout & Payments
Ensure e-commerce checkout flows complete successfully from cart through payment confirmation.
Onboarding Flows
Validate that multi-step onboarding experiences guide new users through setup correctly.
Why Browser Automation?
Detect Issues Uptime Can't
A website can respond with 200 OK but still have broken user workflows. Browser tests catch these issues before your users do.
Maintainable & Expressive
Natural language tests are easier to write, read, and maintain than fragile selector-based frameworks.
Continuous Validation
Run tests on a schedule to continuously validate that critical user paths keep working after every deployment.
Visual Example
Test Your Critical Workflows
Write tests in plain English. Run them on a schedule or on demand — including from your CI/CD pipeline via our CLI and MCP Server.