Teach Your Agent to Prove It's Done

A hands-on workshop with Nick Nisi at AI Engineer New York on building verification loops that let coding agents check their own work. We covered giving agents goals they can check, closing the feedback loop with CLI health checks and browser snapshots, making every part of an app feedback-loopable, and pairing deterministic checks with LLM judges and eval baselines, so each failure feeds the agent's next fix. Built around a live demo app, Midnight Snack Club, with a repo to follow along.