REA is an open-source, MIT-licensed project written in TypeScript and published on npm as rea-agents. It connects an AI agent to reverse-engineering tools through the Model Context Protocol (MCP), so the agent can inspect native binaries, JavaScript and Electron apps, .NET assemblies and websites without access to their source code, and then explain how a feature works, show the evidence and write an implementation. The same workflows are available from a terminal through the rea command. Setup is a single command, npx rea-agents setup, which registers REA's MCP server and installs matching workflow instructions for supported agents including Claude Code, Codex, Cursor, Gemini CLI and Grok Build, keeping backups of existing configuration. The README lists a broad set of targets: native binaries (pseudocode, assembly, strings, symbols, calls and references), JavaScript and Electron apps (modules, imports, source maps, routes and IPC), websites, saved network captures, .NET assemblies, Android APKs and devices, firmware, and process behavior. Requirements differ by target. Deep native analysis uses an existing Hopper, Ghidra or IDA installation, with Hopper optionally installed by setup after approval, while static JavaScript and .NET inspection need no native analysis engine. Node.js 22.x, 24.x or 26+ and npm are required, and several targets are limited to particular hosts such as Linux x64. REA states that it analyzes targets locally and that results include the evidence and limitations behind each conclusion; the agent's model provider has its own data policy. The README showcases case studies such as reconstructing a DX-Ball sound-pan calculation and tracing Notion's Electron clipboard bridge, and includes a disclaimer that users are responsible for authorization and legal compliance.