Before you buy
Test the work you need your software to do.
Build a practical test plan, try it in your shortlisted tools, and compare what actually worked.
No account needed. Your plan stays in this page until you download it.
- Six tests for a real appA small request tracker, built the same way in every builder you try.
- Essential firstPick up to three things that must work. One failure there is never averaged away.
- Your own evidenceRecord what you saw per product. Print it, download it, or copy questions for the supplier.
Your plan lives only in this page. Reloading or closing the page clears it: download your results to continue later.
- 1 Shortlist
- 2 Requirements
- 3 Test plan
- 4 Decision sheet
How the test drive works
App builders are easy to demo and hard to judge: the first screen always looks good. This test drive asks each builder to make the same small request tracker, then checks the things that decide whether you can live with it: whether saved data survives a reload, whether you can change the data model later, whether bad input is refused, and whether you can leave with your code and data.
Choose one or two builders and the plan you will try. Mark up to three requirements as essential. The plan then shows six test cards, essential ones first. You run each step in the builder yourself and record what you saw: passed, partly passed, failed, blocked or not tested.
The decision sheet gives each builder its own verdict. If an essential test failed, it says so at the top, however well the other tests went.
The six tests for an AI app builder
Find out whether an app builder can make, change and restore a small working app, and what you can take with you. Test version 1.0, test data version 1.0.
Test 1: Create and retain a record
Check that the app keeps what you save.
- Ask the builder for a request list, a detail view and a form, with the fields ID, title, status and owner.
- Enter REQ-001, REQ-002 and REQ-003.
- Add REQ-004 through the form.
- Reload the page and open REQ-004 again.
Passes when: REQ-004 is still there after the reload, with the values you entered. The three sample requests are unchanged.
Test 2: Change the data model
Check that you can add a field without breaking the records you already have.
- Ask for a new field, priority, with the values low, normal and high.
- Ask for normal as the value of every existing request.
- Create a new request with priority high.
- Open every request again.
Passes when: The existing requests keep their values and now show priority normal. The new request is saved with priority high.
Test 3: Reject invalid input
Check that the app refuses data that should never be stored.
- Try to save a request with an empty title.
- Try to save a request with the status archived.
- Reload the list and look for either request.
Passes when: A clear error appears both times. Neither request appears in the list after the reload.
Test 4: Check access separation
Check that someone with read-only rights cannot change data.
- Sign in as the read-only account and try to change REQ-001 through the form.
- Open the copied edit link as the read-only account and try to save.
Passes when: The read-only account cannot change REQ-001 by either route.
Test 5: Recover a change
Check that you can go back to an earlier state, and what that costs you.
- Create a restore point.
- Change the title of REQ-002.
- Restore the restore point.
- Open REQ-002 and the other requests.
Passes when: REQ-002 shows its original title again. You checked the other requests for lost or changed data.
Test 6: Check your exit
Find out what you can take with you if you leave.
- Export the code and the data with whatever the tool offers.
- Write down separately what you got: code, data, settings, and anything that depends on an outside service.
- If you can, start the exported app somewhere else.
Passes when: You know which of code, data, settings and outside services you can take with you. You call the app portable only if you started it somewhere else.
Test data: three fictional requests
Every builder gets the same three requests. Add REQ-004 yourself during test 1. All of it is fictional.
| id | title | status | owner |
|---|---|---|---|
| REQ-001 | Create a sample dashboard | open | Demo A |
| REQ-002 | Update a test form | in_progress | Demo B |
| REQ-003 | Archive a sample record | done | Demo A |
App builders we are researching for this test drive
Replit
Product-specific steps: not ready. A source check on 28 September 2026 left open questions about plan limits and documented steps, so the general tests above apply to Replit as to any other product.
What the test drive does not do
- It does not operate your software. You carry out every step yourself, in your own test set-up.
- It does not rate or rank products, and it adds no score. Each product keeps its own results.
- It does not send anything. The sample contacts use example.invalid addresses, which cannot receive email.
- It keeps your plan only in this page. Download your results to continue later; nothing is stored elsewhere.
- A passed test drive is not a verdict on reliability, security, scale or legal compliance.
Questions about the test drive
Does a builder that passes all six tests suit every project?
No. It passed these six tests in your set-up, on the plan you tried. Larger apps, more users or other integrations raise questions these tests do not cover.
Why is a code export not enough for test 6?
Because an app is more than its code. Data, settings, secrets and outside services may stay behind. Call it portable only when you have started the exported app somewhere else.
Do I need a paid plan?
Not to begin. Many builders have a free plan or trial, but limits differ. Record the plan you used, and treat a trial with every feature switched on as a hint, not proof, of what a cheaper plan does.
Tests written and maintained by Daniel Haket. Version 1.0 of 2026-09-29: First version, from the build plan of 28 September 2026.