test automationBeta

Tests that write themselves once, then run free forever.

Point CodeReviewer at part of your app. It reads the code to learn what the feature should do, clicks through it like a user, and records real end-to-end tests. Authoring uses AI once; every replay after that is free.

Start free trial Talk to sales

Invoices suite · v3

12 tests · written by AI · replayed free

nightly 02:00on demand

14 replays · 13 green · 1 flaky test benched

Flaky? Quarantined and retried on its own. Your suite stays green.
App changed? The broken test is repaired and re-verified.
screenshots · video · step-by-step trace on every replay

$0to replay a suite, however often

Screenshots · video · tracekept for every replay

the problem

End-to-end tests are the ones nobody writes.

Writing browser tests by hand takes longer than building the feature, so coverage never catches up.

Every UI change breaks a handful of selectors, and the suite goes red for reasons that aren't bugs.

One flaky test teaches the whole team to ignore failures.

how it works

From trigger to result.

Pick a module

Pick the part of your app to cover. Invoices, checkout, onboarding.

Review the test plan

CodeReviewer proposes positive and negative cases from your code. Curate the list before anything is written.

Watch it author

It signs in and drives your app in a real browser while you watch live, recording each test.

Replay forever

Run the suite on demand or on a schedule, as often as nightly. Replays never use AI, so they're free.

capabilities

What you get.

Author once, replay free

AI is used to write a test, never to replay it. Scheduled and on-demand replays are unlimited.

Self-healing tests

When your app changes and a test breaks, it's repaired and re-verified before it counts again.

Flaky-test quarantine

A flaky test is benched and retried on its own instead of failing the suite, until it earns its way back.

Works behind login

Tests run as a signed-in user using a saved session or stored test credentials, kept encrypted and never shown to the AI at replay.

Rich run reports

Every replay keeps screenshots, video and a step-by-step trace, with a one-click repro for failures.

Excel export

Download the suite's test cases as an .xlsx catalogue for manual QA, auditors or stakeholders.

Why “author once” changes the economics

Most AI testing tools charge for thinking every time a test runs. CodeReviewer uses AI to write a test: to understand the feature, explore it and record the steps. Then it replays the recording without AI. That’s why replays are free and unlimited. A nightly suite costs the same in its hundredth week as in its first.

A suite you can trust

Coverage only helps if people believe the results. Before authoring, you review the plan of cases so the suite tests what matters. After authoring, self-healing keeps tests aligned with a changing UI, and quarantine keeps a single flaky test from turning the dashboard red. Authoring keeps going until the plan you approved is covered.

faq

Questions, answered.

What does "replays are free" mean?

Authoring a test uses AI and counts against your plan's allowance. Replaying a test that's already written doesn't use AI, so replays are unlimited on every plan.

Can it test pages behind a login?

Yes. Give it a saved session or test credentials. Credentials are stored encrypted and are never shown to the AI during replays.

What happens when our UI changes?

Tests that break because the app changed are repaired and re-verified automatically. Tests that fail for real reasons stay failed and tell you why.

How do we run suites automatically?

Schedule replays (nightly, hourly, before a release) or run them on demand from the dashboard.

Why is it marked Beta?

Test automation is fully usable today and improving quickly. The Beta label means we're still widening what it covers. We'd love your feedback.

Give every PR the review it deserves.

Start your 15-day free trial. We onboard a few teams every week.