Reverse Engineer Anything:
The Complete REA Tutorial
REA is an open-source MCP tool that lets AI coding agents (Claude Code, Cursor, Codex) analyze any app, binary, or website — without source code. This guide gets you from zero to your first reverse-engineered feature in 10 minutes.
What is REA?
REA (Reverse Engineer Anything) is an open-source MCP (Model Context Protocol) server that connects AI coding agents to reverse-engineering tools. Instead of manually loading a binary into Ghidra or decompiling JavaScript by hand, you simply ask your AI agent: "Understand how this app's search feature works, show me the evidence, and build something similar."
The tool was created by morluto and went viral on GitHub in October 2026. It supports:
- Native binaries — using Hopper, Ghidra, or IDA if installed
- JavaScript/Electron apps — static analysis, no decompiler needed
- .NET assemblies — decompilation with evidence
- Websites — runtime behavior analysis
Quick start (3 steps)
Step 1 — Install
Make sure Node.js and npm are installed, then run:
Choose your agent (Claude Code, Cursor, etc.), review the changes, and approve. Setup installs REA's MCP server and workflow instructions.
Step 2 — Ask your agent
Open your AI coding agent and type something like:
show me the evidence, and build a
similar feature for my project.
Step 3 — Review the results
REA returns the analysis with evidence, module maps, and confidence levels. You get conclusions and the proof behind them.
Who is REA for?
- Developers who want to understand how a feature works before building their own version
- Security researchers who need quick binary analysis without deep Ghidra expertise
- Product managers who want to understand competitor implementations
- Students learning how real-world software is structured