Fixed price. Written report.

Built it with AI? I will check it before your customers do.

Lovable, Bolt, Replit, Cursor, and v0 make it fast to ship. They also make it easy to ship a leaked API key, an open database, or a login anyone can walk past.

No rewrite unless it truly needs one. Nothing changes until you approve it.

Eight problems, almost every time

The same issues show up in nearly every AI-built app I open.

Keys and secrets

API keys sitting in front-end code or git history, where anyone with browser tools can copy them and run up your bill.

Database rules

Supabase and Firebase projects with row-level security off or rules left open, so any logged-in user can read every customer record.

Logins and permissions

Can one customer see another's data by changing a number in the URL? Can a regular user reach admin pages? I check every route.

Payments

Prices set in the browser, payment webhooks that are never verified, and orders marked paid when they were not.

AI and API costs

No limits on the calls that cost you money. One bad actor or one runaway loop can turn a small monthly bill into a large one overnight.

Forms and uploads

Inputs and file uploads that trust whatever they are given, the classic path to injected code and stolen data.

Hosting and backups

Environment settings, backups, error logging, and what happens when the server restarts at 2 a.m.

Handover

Whether another developer could pick this up next year, or whether it only works as long as nobody touches it.

What you get.

  • 1A written report. Every issue ranked by severity, explained in plain language, with the fix for each.
  • 2A call to walk through it, so you know what matters now and what can wait.
  • 3A list of what is fine, so you can stop worrying about it.
  • 4Your choice on fixes. I make them at $149 an hour, or your own developer works from the report.

How it works.

  • Tell me about the app. What it does, who uses it, which tool built it.
  • Give me read access. A GitHub invite or a zip of the code. Nothing gets changed.
  • I review it. The code, the database rules, and the live app from the outside.
  • We go through the report. Then you decide what gets fixed.

Questions people ask.

Do I need to understand code?

No. The report is written for the person who owns the app. The technical detail sits in its own section for whoever does the fixing.

Will you tell me to rewrite everything?

Rarely. Most AI-built apps need targeted fixes, not a rebuild. If yours truly needs one, I will explain why and what it would take.

Which tools do you review?

Any of them. Most AI coding tools produce JavaScript or TypeScript on Node with a Postgres database such as Supabase, the same stack my own platform runs on.

Will you change my app during the review?

No. I work from read-only access. Nothing changes until you approve a fix.

Ask for a review price.

Tell me what the app does and which tool built it. I will reply with a fixed price.