Click-test my app and list what breaks
Minda tests your app like a new user, clicks through the flows you name and lists every bug it finds, sorted by severity. It is a ready-made task for the SI agent in Minda Super Intelligence. It runs when you start it. You fill in [app link], [test login, or 'none'], [flows] and [tracker].
In Minda: Tasks → New → Browse tasks → Click-test my app and list what breaks. Minda Super Intelligence for Windows is coming soon.
- Engineering
- The web
- Go through [flows]
- On each step, click every main button and link,…
- For each bug
- Finish with the bug list sorted by severity and…
What you get
Tests like a new user It goes through the flows you name, clicks every main button and link, and tries empty and wrong inputs.
Each bug written up Every bug comes with steps to reproduce, the expected result against what happened, a screenshot, the console error and a severity.
Test account, dummy data It uses only the test account you give it and dummy data like 'Test User'.
Stops before anything real It stops before a real payment, an email to real people or deleting data, and it does not file tickets until you say so.
Why use this?
Minda goes through your app's main flows the way a new user would, trying empty and wrong inputs along the way. You get a bug list sorted by severity, and it asks before filing anything in your tracker.
What it does
- When
- When you start it
- Works with
- The web
- For
- Engineers, Founders
- You fill in
- [app link], [test login, or 'none'], [flows], [tracker]
How it works
- WhenWhen you start itSave it as a workflow to run it daily at 9am instead
- Each runMinda does the taskI want you to test my app at [app link] like a new user, using only the test account [test login, or 'none'] and dummy data like 'Test User'.
- ThenTasks shows what it didOpen it there to read the result and ask for changes
Go through [flows]
On each step, click every main button and link, try empty and wrong inputs, and read the console errors
For each bug: steps to reproduce, expected vs what happened, a screenshot, the console error and a severity
Finish with the bug list sorted by severity and ask: file these in [tracker]?
Rules it keeps
The agent follows these on every run.
Stop before any real payment, an email to real people or deleting data
If a step needs a login or code I didn't give, ask once
Don't file tickets yet, list them first
The full prompt
This is what the SI agent is asked to do. The parts in brackets are what you fill in.
I want you to test my app at [app link] like a new user, using only the test account [test login, or 'none'] and dummy data like 'Test User'. Go through [flows, e.g. sign up, onboarding, search, settings, checkout up to payment]. On each step, click every main button and link, try empty and wrong inputs, and read the console errors. For each bug: steps to reproduce, expected vs what happened, a screenshot, the console error and a severity. Rules: stop before any real payment, an email to real people or deleting data; if a step needs a login or code I didn't give, ask once. Don't file tickets yet, list them first. Finish with the bug list sorted by severity and ask: file these in [tracker]?
Who is it for?
Engineers who ship an app You want a new user's view of the app, with every bug written up so you can reproduce it.
Founders testing their own app You want to see what breaks in flows like sign up, onboarding or checkout.
Frequently asked questions
What does it click on?
Every main button and link on each step, and it tries empty and wrong inputs. It also reads the console errors as it goes.
Which flows does it test?
The ones you list, such as sign up, onboarding, search, settings, or checkout up to payment.
Will it make a real payment or send email?
No. It stops before any real payment, an email to real people or deleting data, and uses only the test account and dummy data.
What if a step needs a login I did not give?
It asks you once. It uses only the test account you gave it, and dummy data for everything else.
Does it file the bugs for me?
It lists them first, sorted by severity, and then asks whether to file them in your tracker. It files nothing until you answer.
What do I fill in?
Your app link, the test login or none, the flows to test, and the tracker to file the bugs in.
Run this task in the Minda app
In Minda: Tasks → New → Browse tasks → Click-test my app and list what breaks. Minda Super Intelligence for Windows is coming soon.
Last updated