
Agents all the way down
When agents write, run, and “review” code for us, the real risk isn’t bugs—it’s what happens when we stop verifying and start trusting outputs by default.

When agents write, run, and “review” code for us, the real risk isn’t bugs—it’s what happens when we stop verifying and start trusting outputs by default.

A weekend build: a data-driven resume with a digital assistant, streaming Q&A, and a job-fit analyzer.

Six more months, we have better tools. Agentic workflows are here: but code review, tests, and budgets still matter.

A frank, detailed review of Vibe Coding: hype, metaphors, contradictions, and real-world AI coding lessons.

Ten years, three articles, and one more trip through MobileNativeJavaScript land.