REA : reverse engineer anything
A new open-source project called REA gets an AI assistant to work out how an app is built.
Ever used an app and wondered how one of its features works? You can’t open it up and read the code. All you have is the finished product.
Working it out anyway is called reverse engineering, a bit like tasting a dish and guessing what went into it. It takes time and experience, and a new open-source project called REA (“Reverse Engineer Anything”) wants an AI agent to do part of the work. I haven’t used it, so everything below is what the project says about itself.
This is written for people new to tech. You don’t need to code to follow it.
Actual Working :Simple
Two pieces of background first. An AI agent is an AI assistant that can do things on your computer instead of only chatting. And Ghidra is one of several free and paid tools that open up a finished app so you can study it. Its output is rough and hard to read if you’re new.
Get Nirajthakare’s stories in your inbox
Join Medium for free to get updates from this writer.
REA sits between those two. You ask your agent a question about an app in plain English. REA passes the question to the existing tools, they look inside the app, and the agent reports back with what turned up and the evidence for it.
You ask the agent. The agent asks REA, REA uses the existing tools, and the findings come back up to you.
According to its README, it works with agents like Claude Code, Codex, Cursor and Gemini CLI, and the analysis runs on your own machine. It’s free under the MIT license, and the repo is morluto/rea on GitHub.
The evidence part caught my eye. AI can sound completely sure and still be wrong, so an answer you can check is worth more than a confident one. Whether REA delivers on that, I can’t say yet.
