sqlew

LLMs Get Confused by "Everything at Once": Good Prompts Still Matter in 2026

Have you ever noticed that the more detail you pack into a prompt, the less reliable the AI's output becomes? You carefully elaborate specifications, list every…

108,000 Lines of Code, 26,000 Lines of Docs: The Maintenance Reality of AI Agent "External Memory"

In February 2026, independent researcher Aristidis Vasilopoulos published "Codified Context: Infrastructure for AI Agents in a Complex Codebase." AI c…

What Is Context Engineering?

If you've ever used AI coding tools and felt that no matter how you refine your prompts, the AI still doesn't quite do what you want — you're not alone. Writing…

Is Loading Everything into CLAUDE.md Really the Best Practice?

Project rules, architecture overviews, style guides, tool configurations — it's tempting to dump everything into CLAUDE.md, thinking "the AI will read it a…

Why AI Ignores Your "Use TDD" Instruction: The Compliance Gap Between Instruction and Rationale

You wrote "Use TDD" in CLAUDE.md, but the AI doesn't write tests. Or it writes them, but they're clearly going through the motions. If you've tried pr…

Recording Design Intent for AI Efficiency: What Measured Data Shows About the Power of "Why"

Write a function that returns a string. Give this instruction to an AI coding tool, and the implementation you get back varies wildly depending on context. In C…