Scene · competitor feature teardown

Agent reverse engineering of competitor & app features

See a feature you like — understand how it works, down to the binary. REA (Reverse Engineer Anything, open-source morluto/rea) is the hot MCP toolkit that wires Hopper / Ghidra / IDA, JS/Electron, .NET, and website analysis into your coding agent — local, evidence-based, ready to implement.

MCP toolkit MIT npx rea-agents setup rea.tools

GitHub · morluto/rea Official site rea.tools Guides

REA: one MCP for reverse engineering across binaries, apps, and runtime behavior.

Who this is for

Product and engineering teams who want to understand how a competitor — or their own app — implements a feature, without hand-walking every disassembly from scratch.

Drive local analysis from a coding agent (Claude Code, Cursor, Codex, and more)
Need evidence (pseudocode, module graphs, IPC boundaries), not vibes
Lawful scope: software you own or are authorized to examine; research, education, interoperability

How it works

Your agent calls REA over MCP to inspect and trace a local target. REA returns findings with evidence. The agent then explains, follows up, or writes and tests an implementation. The same workflows are available from the CLI.

Ask about a feature and point at a local target
REA uses Hopper, Ghidra, or IDA for native work — or JS/Electron / browser tools as needed
Results include evidence and stated limits so you can verify and ship

REA with Hopper analysis UI screenshot — Native analysis can use Hopper, Ghidra, or IDA; static JavaScript analysis needs no native engine.

Quick start

With Node.js and npm installed, register REA's MCP server and matching workflow instructions (Claude Code, Codex, Cursor, Gemini CLI, Grok Build, and others):

npx rea-agents setup

Review the proposed changes, approve, and restart your agent. Then ask something like: “Understand how search works in the Notes app, show me the evidence, and build a similar feature for my project.”

Or install the global CLI:

npm install --global rea-agents
rea --help

What you can analyze

Native binaries

Pseudocode, assembly, strings, symbols, and references. Needs Hopper, Ghidra, or IDA.

JavaScript / Electron

Modules, imports, source maps, routes, IPC, and native add-on relationships. Node only.

.NET assemblies

Metadata, CIL, declared native deps, and build comparisons (static).

Websites & more

Page structure, scripts, and network observations — plus Android, WASM, firmware, captures, and more.

Full matrix and prerequisites: rea.tools/guides and the GitHub README.

Authoritative links

github.com/morluto/rea — source & README
rea.tools — website, guides, showcases
Guides · Showcases
Discord — community Q&A
npm: rea-agents

FAQ

Is it free? What license?

REA-authored code is MIT. Commercial engines (Hopper, IDA, JEB, …) require your own lawful licenses; REA grants none.

Legal / ethics?

Analyze only software you are authorized to examine. REA is for lawful research, education, interoperability, and preservation. Analysis runs locally. See the project's DISCLAIMER.

Vs classic RE tools?

Hopper / Ghidra / IDA remain the native engines. REA connects them (plus JS/.NET/browser and more) to agent workflows, emphasizing MCP, evidence, and repeatable steps — not replacing those tools.

Node requirements?

Node.js 22.x (≥22.19), 24.x (≥24.11), or 26+, plus npm. Some targets need extra engines or hosts.

Does REA upload my app?

REA analyzes targets locally; your agent receives the results. Model providers have their own data policies.

Unofficial scene overview. This page summarizes the public README and rea.tools so you can scan the “agent reverse-engineering of competitor features” scene. Not affiliated with morluto / REA. Authoritative sources: GitHub and rea.tools.