Apps built with AI, checked and fixed before they take money, or once they already do.
I check and fix apps people build with Lovable, Bolt, Cursor or Claude Code, with a focus on the money: payments, refunds, totals and reports that should agree. A fix comes with a test.
this one you can open yourself
From a report
Finding 27 of 28CRITICAL
Any signed-in account can open any guest's order by its number
- AS RECEIVED
- The order page opens: name, street, city, email, phone, what was bought.
- AFTER THE FIX
- "We couldn't open this order."FIXED
- TEST
- A stranger's account asks for a guest's order. Failed before the fix, passes after.
Review log
All cases →
28problems in a shop built with AI24 fixed, live
29problems in a client portal built with AI on Firebaseall high fixed, live
26findings worth fixing in a real, live app on Supabase and Stripe, 1 critical and 9 high; 63 in allsent, not applied yet
13findings worth fixing in a real subscription app on Supabase and Stripe, 2 high; 39 in allsent, not applied yet
54findings worth fixing in a real app on Supabase and Stripe, 13 high; 102 in allowners contacted
7findings worth fixing in a real app on Supabase and Stripe, 2 high; 17 in allowners contacted
The shop and the portal are sample apps with test data. The real apps stay unnamed while the findings are open. Worth fixing means critical, high or medium.
What the report looks like
Real pages from the report on the sample shop, 22 pages in all: what happens, how to repeat it, the fix and the test. The shop is fictional and holds test data.
Open the report (PDF)
Have an app like these?
Tell me what it does, or send a link to the code. I answer by email, no call needed.