Jul 10, 2026 / AI / 5 min read Claude's Working Memory Is Smaller Than You Think Anthropic's J-space research shows Claude's internal reasoning workspace is small. Here's what that changes about how I sequence prompts when building a feature.
Jun 28, 2026 / AI / 6 min read CLAUDE.md is RAM, not disk A simple memory model for structuring docs in a Claude Code project, plus a Laravel starter repo you can copy.
May 22, 2026 / AI / 8 min read Stop Typing, Keep Going Claude Code and Codex now have a /goal command that turns your coding agent into an autonomous loop. Here's when it helps and when it doesn't.
Feb 13, 2026 / AI / 4 min read When Your AI Writes Confident Nonsense AI coding agents are powerful, but they hallucinate. Here are practical tips to spot it before it reaches production.