The AI Code Review Gauntlet
AI code fails in predictable, well-documented ways. Here's the review gauntlet — five failure classes and the fast tells that catch the most damage per minute of attention.
Read articleRunning a Team of Agents Without the Chaos
You've gone from conductor to orchestrator. The bottleneck moved from the agent's capability to coordination — here's how to run parallel agents without quietly corrupting your codebase.
Read articleLoop Engineering
The people who build the most-used coding agents stopped hand-prompting. The unit of work is no longer a prompt — it's a loop that prompts, checks, remembers, and re-runs the agent for you.
Read articleStop Shipping AI-Looking Websites
Every AI-generated site has the same tell. The fix isn't a better tool — it's directing the model with a real brief, a design plan, and your boldness spent in exactly one place.
Read articleYour AI Coding Agent Doesn't Need a Better Prompt
The honeymoon with “vibe coding” ends the moment you point an agent at a real, messy codebase. What replaces it is a discipline, not a trick.
Read articleA Bigger Context Window Won't Save You
The fix for an AI agent that keeps getting your codebase wrong is rarely a smarter model or a longer context window. It's deciding, deliberately, what the model gets to see.
Read articleFive Ways to Keep Your AI Agent From Shipping Insecure Code
AI agents write code fast, and they write insecure code with exactly the same confidence as secure code. Here are five practices that raise the floor.
Read article