Reverse Engineer Anything with REA: The Complete Guide
REA (Reverse Engineer Anything) connects AI coding agents to Ghidra, Hopper and IDA. What it is, how to install it, and how to run your first AI-assisted analysis.
Last updated: 2026-10-15
What is REA (Reverse Engineer Anything)?
REA is an open-source tool that lets AI coding agents — Claude Code, Codex, Cursor, Gemini CLI and others — reverse engineer software for you. Ask your agent how a feature works, and REA gives it the tools to inspect the target: native binaries, JavaScript and Electron apps, .NET assemblies, Android APKs, firmware and more.
Technically, REA is an MCP server (Model Context Protocol — the open standard that lets AI assistants call external tools). It doesn't replace your disassembler. It orchestrates the engines you already have:
- Ghidra, Hopper or IDA for native binaries and decompilation
- JADX for Android APKs, Apktool for Android resources
- Built-in static analysis for JavaScript, Electron and .NET — no extra engine needed
In October 2026 the project went viral on GitHub, becoming one of the fastest-growing repositories of the year — likely the "viral reverse engineering repo" you saw on YouTube or Reddit.
Why it's a big deal
Reverse engineering has always been gated by two costs: expensive tools (an IDA Pro license runs into the thousands of dollars) and a steep learning curve (reading raw disassembly takes years to master). AI agents short-circuit both:
- Plain-English questions. Ask "how does this app sync my notes?" instead of hunting for the function that handles the sync protocol.
- Evidence, not guesses. REA's results include the evidence and the limitations behind every conclusion, and your agent can ask follow-up questions in the same session.
- Local and private. Analysis runs on your machine. Nothing is uploaded to a cloud service.
- Free. REA is MIT licensed — you only pay for the AI agent you choose to use with it.
What you can analyze
- Native binaries — pseudocode, assembly, strings, symbols, calls and references, via Hopper, Ghidra or IDA
- JavaScript and Electron apps — modules, imports, source maps, routes, IPC bridges and native add-on relationships
- .NET assemblies — metadata, CIL instructions, declared native dependencies and build comparisons
- Android APKs — manifest declarations, classes, decompiled methods and references
- Go and Rust binaries — embedded compiler versions, modules, crate names and dependency edges
- Firmware — regions, extraction results and handoffs to native analysis
- Websites and network captures — page structure, scripts, requests, exposed payloads and source locations
- WASM and EVM bytecode — validation, import/export forms, dispatch selectors and inferred arguments
That is not the full list — recorded Linux crashes, packet captures, logic signals, SQLite databases and process behavior are all on the menu.
Quick start: your first analysis
1. Check the prerequisites. REA needs Node.js 22.19+ (or 24.11+, or 26+) and npm. For native binary analysis you'll want Ghidra, Hopper or IDA installed; static JavaScript and Electron analysis needs no engine at all.
2. Run the setup.
npx rea-agents setup
Choose your agents (Claude Code, Codex, Cursor, Gemini CLI and more), review the proposed changes, and approve them. Setup registers REA's MCP server and installs matching workflow instructions — with backups of your existing configuration. Restart your agent afterwards.
3. Ask your first question. Inside your agent session, try:
Understand how search works in my target app, show me the evidence, and build a similar feature for my project.
4. Or work from the terminal. The same tooling is exposed as CLI commands:
npx -y rea-agents@latest analyze-javascript-application /absolute/path/to/app --json
How it works under the hood
Your agent calls REA through MCP. REA inspects the target and traces the relevant code, then returns findings with their evidence. The agent uses those findings to explain behavior, answer follow-ups, or write and test an implementation. The same workflows are available from the command line, so you can script them or wire them into CI.
Is this legal?
Reverse engineering is legal in many contexts — interoperability research, security auditing, malware analysis — and restricted in others, depending on your jurisdiction and the software's license. REA is explicitly built for authorized targets: software you own, software you're licensed to inspect, or research you have permission to perform. The project ships a disclaimer, and you should take it seriously. This guide teaches methodology, not piracy.
Keep exploring
- REA on GitHub — source code, docs and showcases, including a Notion clipboard-bridge teardown and a DX-Ball sound-engine reconstruction
- rea.tools — the official site with illustrated guides and case studies
- MCP setup guide — go deeper on wiring agents to reversing engines
- Tools directory — every major AI reversing tool in one place