morluto/rea

What it solves

REA (Reverse Engineer Anything) simplifies the complex process of reverse engineering software by allowing AI agents to handle the tedious work of tool orchestration, evidence gathering, and code analysis. Instead of a human operator manually switching between decompilers and debuggers, REA provides a unified interface for agents to investigate how an application's features work—even without source code—and then help the user recreate those features in their own project.

How it works

REA acts as a bridge between AI agents and professional reverse engineering tools. It integrates with tools like Hopper and Ghidra to perform deep native binary analysis, recover readable code, and extract function dossiers. It also supports passive observation of websites, Electron pages, and Node/Electron V8 Inspector for JavaScript-based apps. The system uses a Model Context Protocol (MCP) registration to give agents specific skills and commands to decompile, understand, and recreate software behavior.

Who it’s for

It is designed for developers and security researchers who want to understand the inner workings of existing software to learn from it or implement similar functionality in their own applications.

Highlights

• Agent-First Design: Allows agents to perform the investigation and explain the results rather than requiring the user to manually operate the tools.
• Multi-Tool Support: Integrates with Hopper and Ghidra for native binary analysis across macOS, Linux, and Windows (experimental).
• Local Analysis: All analysis runs on the user's local host to ensure privacy and security; no apps are uploaded to a hosted service.
• Comprehensive Tooling: Includes capabilities for process capture, JavaScript/source-map reconstruction, and versioned domain graphs to connect application layers.