Why I'm Building AI-Native Developer Tools
A short note on why the next wave of developer tools should be built around agents, evidence, and human judgment.
AI-native developer tools are interesting because they change the shape of the work, not just the speed of it.
The useful version is not a chatbot bolted onto an existing workflow. It is a system that can gather evidence, form a hypothesis, make a change, run the loop again, and then hand a human a crisp decision instead of a pile of vibes.
That is the thread I keep pulling on: tools that make software engineering more observable, more inspectable, and more honest about uncertainty.
The Bar
The bar for agentic tooling should be higher than "it wrote code." Code generation is cheap. Engineering judgment is not.
The better question is whether the system can explain what changed, why it changed, what evidence supports it, and where the edge cases still are.
Where I Am Pointing
I am especially interested in systems that sit close to runtime truth: browsers, tests, logs, diffs, screenshots, DOM state, review comments, and production behavior.
That is where agents can become useful partners instead of confident autocomplete with a bigger costume.