Skip to content
04TEST AUTOMATION

Point, click, assert. Watch it run for real.

Record a flow on any site or compose it from 90+ blocks, then watch it run live across real browsers, with assertions, visual regression, and a report for every run.

checkout-smoke / live runlive · watching
alpinesupply.co/checkoutChrome
Pay $91.80
142/142 · 12.1s
alpinesupply.co/checkoutFirefox
Pay $91.80
apply code · ALPINE10
alpinesupply.co/checkoutSafari
Pay $91.80
fill #email · dana@…
no code · no Seleniumreal browsers, headedvideo · trace · screenshots3 of 6 browsers · streaming in-app
90+
no-code blocks
25+
built-in assertions
6
browsers, watched live
0
lines of code
4.0START WITH DATA

Your data is the fixture. Test on real rows.

Bind a saved query from the Data Console and fan one test out across every row, one browser tile per row, in parallel or sequence, up to five hundred at once. The same Alpine orders the storefront writes and the dashboard charts become the cases your test runs against.

data source / alpine_supply
one tile per row
query · recent_ordersData Console
ordercustomertotal
#4821Dana Whitfield$612.50
#4822Marcus Lee$88.20
#4823Maya Chen$204.00
bind → {{ row.total }}
fan out · 3 rows × 1 browser
1expect total = $612.50pass
2expect total = $88.20run
3expect total = $204.00run
parallel or sequential · up to 500 tiles
4.1RECORD OR BUILD

Point, click, assert. No code, no Selenium.

Record a journey on any site with the browser extension and it wires itself into steps. Or point at a single element and capture it, ranked across ten accessibility-first locators, straight into a reusable element library that self-heals as your UI moves.

checkout-smoke / build
90+ blocks
record · browser extensionREC 0:42▮▮ ■
1navigate → /checkout
2fill #email · dana@whitfield.co
3click · Apply code
4expect · order total visible
captured clicks → wired blocks, automatically
capture an element⌘⇧P
Apply code
hovering · /checkout
role=button · "Apply code"strongest
testId · promo-apply
css · .cart-form > button
save to element library · heal when the UI moves
records clicks · fills · drags · uploads10 locator strategies · accessibility-first
4.2THE CANVAS

Every test is a flow. Branch when it breaks.

A test isn't a script. It's a flow on the same block canvas as workflows. Compose from 90+ blocks across a six-tab palette, give every step its own timeout and failure branch, wrap the fragile parts in Try/Catch, and build it together, live cursors and all.

checkout-smoke / canvas
SAVED
Start
fill #email
click · Apply code
expect total
screenshot + notify
Add block ⌘KBrowserAssertionsStateLogicNetworkAI
error path per blockTry / Catchtimeout per stepsame canvas as workflows
4.3ASSERT EVERYTHING

Catch the bug. See the pixel that moved.

25+ built-in assertions check elements, pages and API responses (equals, contains, starts, ends, regex), and every one auto-retries until it passes or times out. Visual regression captures a baseline on the first run, then pixel-diffs each run after with a tolerance you set and dynamic regions you mask.

assertions / 25+ built in
all green
expect visiblePlace order
expect textOrder #4821
expect value$612.50
expect URL/order/confirmed
=containsstartsendsregexnot
visual regression0.00% diff
baseline
current
auto-baseline · tolerance + region masking
4.4BROWSER AND API

Test the page. Test the endpoint. One canvas.

Author browser tests and API tests side by side. A Postman-style HTTP request block covers seven methods with status, JSON-path, response-time and header assertions, and the same API blocks drop into a browser test to seed data and clean up after, so setup never lives in a second tool.

orders-api / API test
200 · 84ms
POST/api/orders
GETPOSTPUTPATCHDELETEOPTIONSHEAD
assert status200
assert JSON path$.order.id = 4821
assert response time< 500ms
extract JSON path→ {{ orderId }}
shared blocks ·drop the same API steps into a browser test for setup & teardown
4.5RUN EVERYWHERE

One click. Every browser, every condition.

Run across Chrome, Edge, Brave, Firefox and Safari in parallel, and watch each one live. Throttle the network from slow-2G to WiFi, set locale, timezone and dark mode, then snapshot the storage state to reuse a login and skip the sign-in on every run after.

run config / one click
6 browsers
Browsers
ChromeEdgeBraveFirefoxSafariChromium
Real-world conditions
slow-3gfast-3gslow-4gfast-4gwifioffline
viewport · locale · timezone · dark mode · geo
Reuse the session
snapshot storage state
restore login · skip the sign-in
parallel tiles · run again with last config
4.6SEE THE TREND

Every run reported. Every flake surfaced.

Workspace, suite and per-test dashboards track pass-rate, duration and the top-failing, slowest and flakiest tests over time. Auto-retry surfaces the flaky ones before your users do. Export any run as a PDF or HTML report with secrets redacted, or trigger the whole thing from your CI/CD pipeline.

dashboard / pass rate
96.4% · 30d
Runs · last 7 days
1,284
Total runs
96.4%
Pass rate
31.8s
Avg
3
Flaky
report / export & ship
Flaky detection
auto-retry 0–3 attempts
flags tests that pass only on retry
top-failing · slowest · flakiest
From your pipeline
POST /api/v1/ci/run
JUnit XMLreport.json
secrets redacted in every report

The result: checkout-smoke, green across the board.

The same Alpine order the storefront writes, the dashboard charts and the workflow processes, now guarded by a test on every browser.

checkout-smoke / chromium · firefox · webkit
running
open /checkout1.2s
fill #email · dana@whitfield.co0.3s
apply code · ALPINE100.8s
expect {{block.cart.total}} = $91.800.1s
pay · card ending 42422.4s
video ✓console ✓network ✓1280×720 · en-US

Common questions

No. Record a flow on any site with our browser extension and it wires itself into steps automatically. The same extension captures single elements off your page, ranked across 10 accessibility-first locators (Test ID, role and name, label, text and beyond), into a reusable element library that heals as your UI changes. Or compose a test from 90+ blocks, including interactions, 25+ assertions, waits, network mocks and more.