AI agents can do much more today than generate code. Used well, they can read code, inspect HTTP responses, analyze build logs, evaluate deployment metadata, and help systematically find performance, cost, and infrastructure issues.
In this 3-hour workshop, we build a practical agent workflow for web applications. We start with an external audit of a running app, then analyse common issues in a Next.js codebase and connect the findings with infrastructure data such as logs, deployments, caching behaviour, and metrics. The Vercel CLI is used as a concrete example, but the concepts also apply to other platforms with a CLI or API, such as Cloudflare, Netlify, or AWS.
Participants will learn how to create repeatable agent skills for audits and debugging, how to build safe read-only workflows, and how to structure results so they lead to concrete technical actions instead of vague recommendations.
This workshop has been presented at React Advanced 2026, check out the latest edition of this React Conference.
















