# Upgrade .NET, React, and Next.js apps to latest versions with multiple AI Agents > Source: > Published: 2026-08-31 03:42:39+00:00 **Teaching an AI agent to upgrade .NET, React, and Next.js apps for real — not just talk about it** Every engineering team has that repo. The one running a framework version from three or four years ago. Everyone knows it needs an upgrade. Nobody wants to be the one who breaks production doing it. That's the problem UpgradePilot — an open-source, multi-agent upgrade pipeline — is built to solve. And this week we shipped the piece that made it stack-agnostic: real, working upgrade automation for .NET, React, and Next.js, including repos that mix a .NET backend with a React or Next.js frontend in the same codebase. Here's what that actually means, because "AI upgrades your code" is a claim that's earned a lot of well-deserved skepticism. **The design principle: shell out to the real tool, never fake it** The easy version of this feature is an LLM that reads your package.json, guesses at new version numbers, and writes some plausible-looking code changes. That's not what we built. Every step in UpgradePilot's pipeline calls the actual toolchain: **.NET** — real dotnet restore, dotnet build, dotnet list package --outdated, dotnet ef migrations add. Package version bumps are verified by an actual restore, not assumed to work. **React / Next.js** — real npm install, npm run build, npm outdated. Codemods run through the actual react-codemod and [@next](https://dev.to/next)/codemod CLIs — we pulled the real transform names directly from those projects' GitHub repos rather than guessing, because a fabricated transform name just fails at runtime. **Target versions aren't invented.** PackageTargetVersions come from dotnet list package --outdated and npm outdated — the same commands you'd run yourself. **Codemod selection isn't invented either.** UpgradePilot pulls React's and Next.js's own GitHub release notes, classifies breaking changes, and matches them against a verified catalog of real codemod transforms. If a step can't do something for real, it says so — with a confidence score and an explanation — instead of pretending. **What's real today, stack by stack** **.NET** **React** **Next.js** **Mixed repos (.NET + React/Next.js in one codebase)** **Why this matters more than it sounds like it should** A huge amount of AI coding tooling right now optimizes for demo-ability: something that looks impressive on a green-path example. The failure mode that actually costs teams trust is quieter — a tool that claims success on a real repo but didn't actually verify anything. Every one of these steps is designed around one question: **does this survive contact with a real, imperfect codebase?** That's why package bumps are verified by a real restore, why codemods run through the actual official tools instead of hand-rolled regex, and why every gap that isn't solved yet — renaming a method based on a release note, auto-coordinating a genuinely ambiguous mixed repo — is documented in the code as an open limitation, not hidden behind a confident-sounding summary. **It's open source** UpgradePilot is Apache 2.0 licensed, and the agents that run in the free CLI are the same agents that power the hosted product — nothing is held back to make a paid tier look better. If you're maintaining a repo that's overdue for a .NET, React, or Next.js upgrade, the code is here: github.com/qmmughal/upgradepilot-core I'd genuinely welcome issues, PRs, and "this codemod catalog is missing X" reports — especially from teams running these upgrades for real. That's exactly the feedback that makes the difference between a tool that looks good in a demo and one you'd actually trust on your own repo.