# Test a Lovable, Bolt or Replit app by its address, no code needed

> Built your app in Lovable, Bolt, Replit or v0? Paste the published address, describe a user's journey in plain words, and get a verdict with a replay.

By Jason Nguyen, Founder, Testsy. Published 2026-11-21. https://testsy.dev/blog/test-a-lovable-bolt-or-replit-app-without-a-coding-agent/

## You do not need a coding agent to test

Most testing advice assumes a developer with a terminal. A large share of people building apps with AI never open one: they build in Lovable, Bolt, Replit or v0, press publish, and share the link.

That last step is where trouble starts. Agencies that fix these apps for a living keep listing the same failures after launch: a blank screen in production, sign-in that loops, data that does not save. One developer who shipped twelve such apps found [the same six things broke every time](https://dev.to/chalom_ellezam_5989bce65e/i-deployed-12-vibe-coded-apps-to-production-the-same-6-things-broke-every-single-time-341f), across different tools and founders, and that the code was fine: production exposed assumptions the model could not see.

You can catch most of them before a user does, without reading the code.

## Paste the address and describe the journey

In Testsy, **New test** takes your app's published address and a description. Write it the way you would brief a friend who is trying the app for you:

```text
Open https://my-app.lovable.app.
Sign up with this run's email address and a test password,
and confirm the account from the email.
Create a new habit called "Drink water" and mark it done for today.
Reload the page and check "Drink water" is still there and marked done.
Sign out and check the sign-in page is shown.
```

The assistant turns it into numbered steps and checks you can read and correct before anything runs. You do not write selectors, waits or code.

![New test: the address, the description in plain words, and the plan the assistant drafts from it.](https://testsy.dev/docs/img/new-test-light.webp)

## What you get back

A verdict, and the proof for it:

- **Each check** with the value it was decided on ("Drink water · done") and a crop of the screen it was read from.
- **A replay** of every step, with the click marked.
- **A video** of the whole run.
- For a failure: what was expected, what was seen, and the steps to reproduce it.

You can watch the replay without understanding any code. If something failed, **Copy fix prompt** gives you a message to paste back into your builder's chat that says exactly what broke. See [Turn a failed test into a fix prompt](https://testsy.dev/blog/copy-the-fix-prompt-into-your-coding-agent/).

## Five tests worth having before you share the link

1. Sign up and confirm the email. ([Why the link might point at localhost](https://testsy.dev/blog/sign-up-email-links-to-localhost/).)
2. Sign in, sign out, sign in again.
3. Create the main thing your app makes, reload, and check it is still there. ([Saved, then gone after a refresh](https://testsy.dev/blog/saved-then-gone-after-a-refresh/).)
4. Do the main thing on a phone-sized screen: in the test's options, choose **mobile** (412 × 914).
5. If you take payments, go as far as the checkout page in test mode.

## Limits to know

- The tester needs the **published** address. A preview that asks you to sign in to the builder cannot be opened by a stranger, and neither can the tester.
- It tests what a user can see and do. It does not check your database security rules. [What an AI tester cannot check for you](https://testsy.dev/blog/what-an-ai-tester-cannot-check-for-you/) covers that.

The free trial has 10 runs and needs no card.

This post is part of [Testing an app you built with AI](https://testsy.dev/blog/test-an-app-you-built-with-ai/).
