Fixing the PR Bottleneck in OroCommerce: What I Learned from Matt Pocock's Talk
AI agents now write code faster than humans can review it.
InfoMukesh Blog
Build better with practical developer notes.
Practical developer notes for PHP, Laravel, AI, commerce platforms, and useful web apps from InfoMukesh.
Recent posts
Latest from the notebook
What a CCTV Business Card Taught Me About Listening Before Designing
A client gave me "full freedom" to design his CCTV business's business card — and I learned the hard way why that's actually the riskiest brief you can get. Here's the three-question process I now use before touching any design tool.
Stop Leaking Secrets: How to Catch Security Flaws Before Your Pull Request
Finding security vulnerabilities and exposed API keys during pull request reviews or CI/CD pipelines causes unnecessary rework and security risks. This guide explains how to implement "Shift-Left Security" by running static application security testing (DevSkim) alongside dedicated secret scanning (Gitleaks) directly on your local workstation using Git pre-commit hooks. By catching insecure coding patterns and credentials the moment you commit, you eliminate awkward PR reviews and prevent leaks before they ever enter Git history.
From Lost Paper Receipts to One-Tap Profit: How I Built a Custom App for a Cab Driver
Independent taxi drivers juggling multiple aggregators like Uber, Ola, Rapido, and offline private bookings face daily accounting chaos, often relying on paper slips that get misplaced. This case study documents the end-to-end development of SD Travels (Driver Portal)—a lightweight, driver-centric mobile app engineered to deliver instant net-profit visibility, quick single-tap entries for fares, CNG refills, and maintenance, and reliable offline-first local data persistence. Within its first week of real-world use, the app completely eliminated month-end bookkeeping guesswork and provided effortless, real-time daily profit tracking
The AI Code Review Bottleneck Nobody Warned Us About
AI coding tools have made writing an app faster than ever, but that speed didn't remove the bottleneck — it just moved it downstream to code review. This post breaks down why AI-generated code tends to skip architecture and testing by default, how that shows up as cascading production failures, and why the actual engineering work now happens at review time instead of at the keyboard.
Why Your AI-Generated Code Keeps Turning Into a Mess
Why Your AI-Generated Code Keeps Turning Into a Mess (the Pipeline That Actually Fixes It)
AI Models vs AI Tools 2026 — Which One to Use, and When
Developer's Practical Guide
Paytm Dynamic QR Payment Integration in a Custom POS: Auto-Detecting Payments
In A Custom POS Auto-Detecting Payments
Building Scalable AI Infrastructure with Stateless MCP 2026
Death to the Sticky Session: Why MCP Stateless is the Production Paradigm You’ve Been Waiting For
Product connection
Need the actual tools?
Use the live InfoMukesh product for QR generation and web utilities, then come back here for the build notes behind them.