You commented KIT, so here is the whole thing: the repo link, a 5 minute setup and 5 prompts you can paste straight into your AI. You do not need to know anything about reverse engineering to follow along. We go step by step.

What it is

Reverse Engineer Anything (REA on GitHub) is a free, MIT licensed tool that plugs into the coding agent you already use. Once it is connected, your agent can look inside a program, find out how a feature works, show you the evidence, and help you build your own version.

It is one of the fastest growing repos on GitHub right now: 177K stars in under six months.

rea.tools home page: Find out how software works

The REA home page at rea.tools.

Reverse engineering, in plain words: working backward from a finished app to understand how it was built. It used to need specialist tools and years of practice. REA hands those tools to your AI, so you just ask in plain English.

It works on a lot more than you would expect: websites, JavaScript and Electron desktop apps, Android apps, .NET programs, native binaries, firmware and more. The README lists 19 kinds of targets.

Before you start

  1. A Mac, Windows or Linux computer.

  2. Node.js and npm installed. REA needs Node 22 (22.19 or newer), Node 24 (24.11 or newer) or Node 26+. Grab it from nodejs.org if you do not have it.

  3. A coding agent. Setup supports Claude Code, Codex, Cursor, Gemini CLI, Grok Build and others.

  4. About 5 minutes.

Step 1: install it with one command

The easiest way: paste this into your coding agent and let it do the work.

Install REA and connect it to this coding agent using npx rea-agents@latest setup. Show me the setup plan for approval, then verify the installation.

Prefer the terminal? Run this instead:

npx rea-agents@latest setup

Pick your agent, read the setup plan, approve it, then restart your agent. That restart is the step people forget. Setup backs up your existing config before it changes anything.

Set up REA: the agent prompt or the terminal command

Both setup options, straight from rea.tools.

Step 2: ask your first question

Start small so you see how it thinks. Paste this into your agent:

Use REA to inspect this dinosaur game. Why does it get faster? Show the speed rule, then build a small version with an adjustable speed.

Your agent reads the script that is actually running, finds the speed rule, explains it and builds a tiny version you can play with. That is the whole loop: ask, see the evidence, build your own.

The dinosaur game example: the prompt and the recovered speed rule

The rule REA pulled out of the game: start at 6, add 0.001 each update, stop at 13. Example from rea.tools.

The kit: 5 prompts to paste

Swap the bits in brackets for your own target. Each one asks for evidence, so you can check the answer instead of trusting it blindly.

1. The warm up

Use REA to inspect Windows Calculator. Why does 200 + 10% give 220?

2. Understand a feature you love

Use REA to understand how [feature] works in [app]. Show the evidence for each conclusion and list anything you could not confirm.

3. Build your own version

Use REA to understand how [feature] works in [app], show the evidence, and build a similar feature for my project. Write tests that compare it with the behavior you found.

4. Take apart a website

Use REA to inspect [website URL]. Map its page structure, scripts and network requests, then explain how [feature] works.

5. Look inside a desktop app

Use REA to analyze the Electron app at [absolute path to the app folder]. Map its modules, routes and IPC, then explain how [feature] works with the evidence.

What people have already done with it

The project's own case studies: tracing how Notion copies text to your clipboard, and rebuilding a sound calculation from an old Windows game that matches the original byte for byte.

REA showcases: DX-Ball, Notion and TH04

Three showcases from the REA repo. Full write ups at rea.tools/showcase.

Going deeper: native apps

Websites, JavaScript and .NET apps work right away. To dig into native apps down to the assembly, REA uses a disassembler: Hopper, Ghidra or IDA. Setup can install Hopper for you after you approve it. IDA and Hopper are commercial tools, so you need a legit license or their demo mode.

REA running inside Hopper while inspecting a native binary

REA running inside Hopper on a native binary. Screenshot from the repo README.

If something goes wrong

  1. Your agent does not see REA? Restart it. Setup only takes effect after a restart.

  2. Hit a bug? REA ships fixes fast. Run npx rea-agents@latest setup again to update, restart your agent and retry.

  3. Hopper opens a dialog on Mac? Choose demo mode or activate your license once, then run your question again.

The honest part (read this)

REA is built for software you own or are allowed to inspect, plus research, education and interoperability. Cracking license checks, cheating in games or copying someone's protected code is off the table, and the project says so in its disclaimer.

The analysis runs on your own computer, but what it finds is sent to your AI agent's model provider, so keep that in mind with private apps. And AI answers can be incomplete or wrong: ask for the evidence and test anything it builds.

Who this is for

Builders who keep asking "how did they do that?" If you see a feature you love and want your own version of it, install it tonight and ask your first question.

Website and guides: https://rea.tools

The AI Leverage is how a few thousand builders get sharper every week: the tools that quietly change the game, and the exact way to run them, so AI saves you real time and real money instead of just adding another tab. Come get better with us. by bestapps.ai on instagram.