Onboarding onto a large LMS codebase, I used Cursor not to write features faster, but to build a mental model: where legacy PHP meets newer REST layers, how events propagate, and where permission checks actually live.

This post walks through two real explorations (user impersonation across stacks and a permissions trace), with anonymized prompts, what the tool got right and wrong, and a small playbook you can reuse on your own brownfield project.

You bump a version, run install, and the project catches fire. That’s a React upgrade.

Half the bugs you find were already there — the rest is types, tests, and Claude Code keeping the investigation from eating you alive.

A young engineer asked me a deceptively simple question. Years in the problem, and the answer came out as a fumble.

Every short definition slips. Architecture is not the diagrams, not a phase, not one person’s role. It is the practice of deciding which tensions to accept. Organizations and systems co-evolve. The architect’s value lies in building that capacity in teams, not owning every decision.

This is the answer that took almost four months to write: why the question resists a tidy definition, and why that resistance is the most honest signal we have about where architecture actually lives.

I generated six AI-powered learning roadmaps for Go in a year. Each one was well-structured, personalized, and genuinely good. None survived past day three.

This isn’t about AI being bad for learning. It’s about a specific trap: when everything accelerates, starting over becomes frictionless, and depth never has a chance.

Modern B2B sales has evolved from annoying persistence into systematic manipulation, using tactics that exploit identity, manufacture intimacy, and weaponize professional boundaries.

These approaches persist because the conversion rate justifies burning bridges with the majority of prospects.

This article dissects the playbooks from an engineer’s perspective, revealing why they work, who they target, and how we can collectively demand better.

Most AI coding workflows assume you’re building something new from scratch. But what if you’re building something that spans multiple existing codebases, each with years of accumulated quirks?

This is how I stopped fighting context loss and built a system that actually works with existing codebases.