
REA
Lets AI agents reverse engineer apps, JavaScript/Electron code, .NET, and native binaries, with evidence.
- Stars
- 421k
- Forks
- 105k
- Licence
- MIT
- Language
- TypeScript
- Last update
- 4 days ago
- Type
- AI model / agent
What it is
REA is an MCP server that connects an AI agent to tools for inspecting native binaries, JavaScript and Electron apps, .NET assemblies, and websites. You can ask your agent to figure out how a feature in an app works, and it returns the code it found along with the evidence and limitations behind each conclusion. The same tools can also be run from a terminal, and the README says analysis runs locally.
- Ask an agent to explain how a feature in a local app works and show the supporting evidence
- Run analyze-javascript-application on an extracted Electron app directory or ASAR file from the terminal
- Use an agent to build a similar feature for your own project based on investigated behavior
- Configure a native analysis provider and run native commands through the rea CLI
Great for
- Developers studying how a feature in an existing app works without its source code
- Security researchers who want an agent to help inspect binaries and Electron or JavaScript apps
- People already using MCP-compatible agents such as Claude Code, Codex, or Cursor
- Users who want to inspect a JavaScript/Electron app directory or ASAR from the terminal
Not ideal for
- Users who need native binary analysis without installing or configuring an external engine such as Hopper, Ghidra, IDA, or Kuna, since the README says native analysis needs one of these
- Users on Linux who lack a procps-compatible ps command, since the README says Ghidra, JADX, firmware launches and static JavaScript analysis require it
- Anyone who needs a fully packaged, binary-downloadable tool, because the latest release has no binaries
- Readers who need a plain-language guide to reverse engineering itself, since this is a tool rather than a tutorial
How to get it
- 1
Install with a package manager
You'll need: Node.js 22.x (>=22.23.4), 24.x (>=24.22.1), or 26.12.1+ · npm
Set up your agent with REA (run from a trusted directory outside the target project)
npx rea-agents setupAnalyze an extracted JavaScript/Electron app directory or ASAR from the terminal
npx -y rea-agents@latest analyze-javascript-application /absolute/path/to/app --jsonInstall the rea command globally
npm install --global rea-agents - 2
Run it
- Run npx rea-agents setup and approve the proposed changes for your agent
- Restart your agent after setup finishes
- Ask your agent to investigate a feature in a target app and show the evidence
- Or run analyze-javascript-application on an extracted app directory from the terminal
- 3
Get the most out of it
- Set REA_MCP_INPUT_SCHEMA_PROFILE=full in your MCP registration when you need full field guidance and nested validation details
- Set REA_MCP_OUTPUT_SCHEMA_PROFILE=compact or off if your client limits the size of tools/list responses
- Use rea update for global CLI installs to keep REA current, then re-run setup to refresh agent registrations
- Read the CLI and Evidence guide (docs/cli.md) for native commands, provider selection, snapshots, and scripting
Downloads
Strengths
- Connects agents to reverse engineering tools through MCP, so the same workflow works across several supported agents
- Returns findings with evidence and stated limitations rather than unsupported conclusions
- Static JavaScript analysis needs no native analysis engine
Trade-offs
- Native analysis depends on an existing external tool installation (Hopper, Ghidra, IDA, or Kuna) before it works
- Setup and recovery involve several configuration details, including Node.js version requirements and package-runner cache handling, so it is not a one-click install
- The README describes the project as changing quickly, so behavior and commands may shift between releases
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| rea (this) | 421k | AI model / agent | intermediate | 4 days ago | Developers studying how a feature in an existing app works without its source code |
| agentic-awesome-skills | 49k | Dev tool | intermediate | 4 days ago | Developers using Codex or Claude who want agent-chosen instructions for a project |
| agents | 42.1k | Dev tool | beginner | 4 days ago | Developers who already use one of the listed AI coding tools and want ready-made specialist agents |
| ponytail | 182k | AI model / agent | beginner | 8 days ago | Developers who use Claude Code or similar AI agents and find the generated code too long |
| taste-skill | 108k | AI model / agent | beginner | 6 days ago | Developers using Claude Code or Codex who want less generic frontend output |
| career-ops | 79.4k | AI model / agent | intermediate | 4 days ago | Job seekers who apply to many listings and want to filter them before spending time on each one |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.