The AI Prompt That Replaced My Senior Engineer Code Reviews

I used to dread code reviews.
Not because of the feedback — because of the wait. You open a PR, ping your team, and sometimes sit idle for hours or days waiting for a review that takes 10 minutes to read.
Then I found a prompt that changed how I work.
The Prompt
Review this code as a senior engineer at a top tech company. Evaluate:

1. Correctness — does it do what it's supposed to?
2. Security vulnerabilities — SQL injection, XSS, exposed secrets, improper auth
3. Performance — bottlenecks, memory leaks, unnecessary loops
4. Readability — naming, structure, complexity
5. Test coverage gaps — what's untested?

Prioritize every issue you find as Critical / Medium / Low.

Code: [PASTE YOUR CODE HERE]
Why It Actually Works
Most people prompt AI with "review my code" and get a generic response. The difference here is two things:
Role assignment. Telling the AI it's a senior engineer at a top tech company consistently produces more thorough, production-minded feedback. It stops treating your code like a homework assignment and starts treating it like something that will run in production.
Priority tiers. Without them, AI gives you a wall of feedback with no sense of urgency. Critical/Medium/Low forces it to think about impact — and forces you to act on what actually matters instead of fixing a variable name when there's an auth vulnerability.
Real Example
I ran this on a Node.js API endpoint last week. It caught:

A missing input validation that could cause a crash (Critical)
An N+1 database query I'd completely missed (Medium)
Three variable names that were genuinely confusing (Low)

The whole review took 30 seconds. Not bad.
How To Use It

Copy the prompt above
Replace [PASTE YOUR CODE HERE] with your actual code
Paste into ChatGPT, Claude, or any LLM
Work through the Critical issues first, ignore Low until you have time

One More Thing
This is one of 40 prompts I've collected specifically for developers. If you want the full set — covering debugging, testing, documentation, system design, DevOps, and more — I packaged them into a clean PDF here: [b4m.gumroad.com/l/wehfa]
What do you use AI for most in your dev workflow? Drop it in the comments — always looking for what to cover next.