NabilRFG12/Reverse-Engineer-anything

Reverse engineer anything with agents, from app behavior down to native binaries.

REA (Reverse Engineer Anything) – Quick Overview

What it is – REA is a Node‑based platform that lets large‑language‑model agents (Claude, Gemini, Grok, etc.) automatically investigate software they don’t have source code for. It talks to the agent through a local MCP (Modular Capability Provider) server, runs a suite of analysis tools (Hopper, Ghidra, IDA, pwntools, JADX, Binwalk, browsers, etc.) and returns evidence – pseudocode, assembly, module graphs, network captures, etc. – that the LLM can then explain, modify, or re‑implement for you.

Why it matters – Traditional reverse‑engineering is manual, requires deep expertise and a handful of heavyweight tools. REA abstracts those tools behind a simple CLI (rea …) and an MCP contract, letting an AI assistant do the heavy lifting: ask a question like “How does the clipboard work in Notion?” and get a step‑by‑step trace, code snippets, and a ready‑to‑use implementation.

Key capabilities

Target type What REA extracts Required tooling
Native binaries (PE/ELF/Mach‑O) Pseudocode, assembly, strings, symbols, call graphs Hopper, Ghidra or IDA (installed or auto‑installed)
JavaScript/Electron apps Module list, imports, source‑maps, IPC bridges Node.js/npm only
.NET assemblies Metadata, CIL, native dependencies Static inspection
Android APKs Manifest, decompiled classes Head‑less JADX + JDK
Firmware images Regions, extracted files Binwalk/Unblob
Web sites & network captures DOM structure, scripts, HAR data, screenshots Chrome‑family browser, mitmdump
EVM bytecode, recorded Linux crashes, etc. Dispatch tables, registers, signals pwntools, optional GDB

How you use it

  1. Install – npx rea-agents setup registers the MCP server and installs a “skill” for your chosen LLM agent.
  2. Ask the agent – type a natural‑language request; the agent calls REA, which runs the appropriate analysis tool and returns structured evidence.
  3. CLI fallback – you can run the same commands directly, e.g. rea analyze-javascript-application /path/to/app --json.
  4. Update – rea update or repeat the setup command to get the latest bug‑fixes.

Showcase examples – The README links to three detailed case studies:

  • Reconstructing a sound‑pan calculation in the classic DX‑Ball game (x86 → C).
  • Tracing the clipboard bridge in the Electron‑based Notion app.
  • Recovering a DOS‑era bullet‑ring calculation from a PC‑98 game. Each includes a public GitHub repo with the reconstructed code.

Community & support – A Discord server, extensive docs (website, guides, MCP contract catalog), and a growing ecosystem of agents that can plug into REA.

Legal note – REA is intended for lawful reverse‑engineering research; users must ensure they have permission to analyze the target software.

TL;DR

REA is a Node.js‑driven, AI‑augmented reverse‑engineering framework that connects LLM agents to traditional binary‑analysis tools via a local MCP server, delivering evidence that the model can turn into explanations or new code. It supports many artifact types (native binaries, JavaScript/Electron, .NET, Android, firmware, web captures, etc.) and works with popular agents like Claude, Gemini, and Grok.

Related

  • Project
    jin-zi-xuan/kaobuddy-pwa

    KaoBuddy is a PWA that lets you upload PDFs, docs, handwritten images, and B‑site video subtitles, then uses an LLM (DeepSeek, Kimi, OpenAI, etc.) to extract exam‑relevant knowledge points, generate explanations, flash‑cards, practice questions and full mock exams. It builds a daily study plan, stores everything locally in IndexedDB, and runs via a FastAPI backend + React‑TypeScript frontend. Windows users can use a portable zip; macOS and other platforms run from source or Docker. MIT‑licensed.

  • Project
    Pancakes-Labs/astrbot_plugin_proactive_chat

    A plug‑in for the open‑source AstrBot framework that lets the bot automatically start a conversation after a period of silence. It supports private and group chats, configurable timing, context‑aware prompts, TTS, message splitting, persistence, and includes its own WebUI for monitoring and configuration.

  • Project
    waynesutton/builder-skills

    builder‑skills provides a set of concise, markdown‑based “skills” that teach AI coding assistants how to follow Convex best‑practice rules and disciplined project‑workflow habits (PRDs, task tracking, safe git). Installable via skills.sh, Claude plugins, or npm, it drops starter docs and lets agents load the appropriate SKILL.md on demand, keeping code, docs, and safety checks in sync.

  • Project
    marcusquinn/aidevops

    aidevops is an open‑source AI‑DevOps framework (CLI + OpenCode plugin) that lets you describe a task in natural language and have specialised AI agents automatically plan, implement, verify and merge the change in a Git repository. It includes model‑routing, secret handling, background “Pulse” workers, and a large library of domain‑specific agents.