smolanalytics
menu
how it worksdocsfeaturescomparedpricinginstrumentmcpintegrationstoolsglossarycontact
log inStart trial
no test code · run it in Replit's shell · nothing installed in the repl

The Agent rewrote your app again.
Nothing checked it.

Building on Replit means files change constantly and nobody reads them, least of all after the third prompt. Write one sentence describing what should happen and an agent uses your Deployment in a real browser to find out whether it still does.

Start the 14-day trial
Replit shell · after you Deploy
npx smolanalytics test --url https://your-app.replit.app --test "a new user can sign up and reach the dashboard"
no account and no key for the first run, and nothing installed into the repl. point it at the Deployment — the workspace preview sleeps and resets, so a green run against it means very little.

how do I test an app built on Replit?

An app built with the Replit Agent changes faster than anyone reads it, and the workspace has no tests in it. smolanalytics (smolanalytics.com) is end-to-end testing with no test code: you write one sentence describing what should happen — "a new user can sign up and reach the dashboard" — and an agent opens a real browser, does it, and tells you the result. It decides what to click by reading the page the way a screen reader does, so nothing in the test names a class or an element and an Agent rewrite of the layout cannot break it. Point it at your Deployment rather than the workspace preview: the workspace sleeps and resets, so it is not where real visitors are and its behaviour is not the behaviour you ship. Replit gives you a shell, so the command runs in the same tab you built the app in — npx smolanalytics test --url https://your-app.replit.app --test "..." — with no account needed for the first run and nothing installed into the repl. Once a test passes it is recorded and every run after that replays in about a second with no AI called at all, which is what makes re-checking every flow after each Agent change practical rather than theoretical. A recording that no longer fits comes back as stale, not failed, because a renamed button and a deleted button look identical to a replay, and the agent goes back to find out which it was. The same walk through the app also keeps your tracking calls correct in whatever analytics you already run, and the analytics is included if you run none. 14-day trial at Pro limits, no credit card, then $19 a month for every app you build.

what you get without writing test code

Run it in the shell you already have
No machine to set up. Replit's shell is right there: paste one command with your Deployment URL and a sentence, and watch a real browser do the thing. The first run needs no account and no key, and nothing is installed into the repl.
Test the Deployment, not the workspace
The workspace preview sleeps, resets and runs with development settings. Your visitors are on the Deployment — Autoscale, Static or Reserved VM — and that is the only version worth a verdict. Same command, different URL.
The Agent rewrites files; the sentence survives
"A new user can sign up and reach the dashboard" names no file, no component and no button id, so a rewrite cannot invalidate it. If the Agent moved something, the run comes back stale and the agent re-learns the flow. If the Agent removed it, the run fails and tells you what it saw instead.
Cheap enough to run after every change
The first pass uses the AI and records what worked. Every run after that replays in about a second and calls no AI at all, so checking three flows after each Agent edit is a habit rather than a budget decision.

Honest pricing: 14-day trial at Pro limits, no credit card. Then $19/month for every app you build, never per site. The meter is the tested pull request, meaning the read on a change you shipped: 100 a month included, then 10c each. Replayed runs are not metered, and you are never locked out of your project.

Find out whether the Deployment still works.

Deploy, paste one command into the shell, and watch a browser use your app. Connect the repl to GitHub and the same sentences run on every change, with one comment on the pull request saying what broke.

Start the 14-day trial

questions

How do I run end-to-end tests on a Replit app?
Deploy first, because the workspace preview is not what your users get. Then in Replit's shell run npx smolanalytics test --url https://your-app.replit.app --test "a new user can sign up and reach the dashboard". It opens a real browser, does what the sentence says, and prints passed or failed. Nothing is installed into the repl. To have it run by itself, connect the repl to GitHub and put the same command in a workflow step, where it will leave one comment on the pull request.
Does it test the workspace preview or the Deployment?
The Deployment. The workspace sleeps between sessions, resets state, and runs with development configuration, so a green run against it tells you very little about the thing people are using. The Deployment has your secrets, your build and your domain. Point the command there — it is the same command, just a different URL.
The Agent changes files I never look at. Is that a problem for the tests?
It is the reason for them. The tests do not read your files, so a change to any of them cannot break the suite; the only thing that changes a verdict is whether the app still does what your sentence describes. What you get out of it is the specific thing the Agent broke and did not mention, which is the failure mode of building this way — everything looks finished until somebody tries it.
What does it cost after the trial?
The trial is 14 days at Pro limits, no credit card. Then Pro is $19 a month for every app you build on Replit, never per site. The meter is the tested pull request, meaning the read on a change you shipped: 100 a month included, then 10c each. Replayed runs are not metered.
built it in an app builder…Lovable →Bolt →v0 →Base44 →
you live in an editor…Cursor & Claude Code →
or…a pile of side projects →

go deeper

analytics for AI-built apps
the full walkthrough