Understand the app.
Find your next step.
Turn an unfamiliar Electron app into a useful starting point. Explore its files, find the entry point, and take the evidence into REA.
1 {
2 "name": "fieldnotes",
3 "version": "1.2.0",
4 "main": "src/main.js",
5 "dependencies": {
6 "electron-store": "^10.0.0",
7 "papaparse": "^5.4.1"
8 }
9 }Setup, a published walkthrough, and the Electron example are here. The longer notes stay linked underneath.
Eric Tech published this walkthrough on October 10, 2026. It explains what REA is, how the agent connection works, runs the setup command, and then tries to rebuild one voice-to-text app. This workbench has not repeated that rebuild.
Watch on YouTubeUse Node 22.19 or later within 22.x, Node 24.11 or later within 24.x, or Node 26+. Then review the plan before anything is written.
npx rea-agents@latest setupChoose your coding agent, read the proposed configuration, and approve it. Restart the agent so the new tools show up in that session.
Read the full setup guideName a single action and the evidence you want. An entire application is a weak first question.
Use REA to inspect my local Electron application.
Trace what happens when I export a note as CSV.
Cite the entry point, IPC channel, and file-writing code.
Separate static findings from verified runtime behavior.
List unknowns and suggest one small verification step.Open app.asar, read the declared entry point, and follow one feature from the renderer through preload into the main process.
The video above installs REA with Hopper and tries to rebuild a voice-to-text desktop app, including a hotkey and local transcription.
Deeper native analysis needs Ghidra, Hopper, or IDA on your machine. JavaScript inspection can run without those engines. This site does not decompile binaries.
You don’t need a full reverse engineering setup
to find your first useful clue.
Open an Electron archive without installing anything. Browse files, inspect declared dependencies, locate the main entry, and export a summary for your REA investigation.
See application metadata, source map candidates, and native modules. Every clue points back to a file you can inspect.
Inspection happens in your browser. No archive uploads, no API keys, and no account between you and your first finding.
REA stands for Reverse Engineer Anything. It connects coding agents to tools that inspect software, so you can investigate how an application actually works.
Through MCP or the command line, REA can examine JavaScript and Electron apps, native binaries, and other supported targets. The agent follows the evidence, asks another question, and helps you explain or reconstruct a specific behavior.
This independent workbench makes that first step easier. Inspect an ASAR here, choose a feature worth understanding, then continue locally with REA. Each tool has a clear scope: an archive overview is a starting point, not proof of what executes at runtime.
One useful distinction: REA is the investigation toolkit. Your coding agent decides how to use its findings.
Find the official Reverse Engineer Anything repository,
catch up on releases, and choose the right companion tool.
The source behind Reverse Engineer Anything. Explore how REA connects coding agents to application analysis, native tools, and runtime evidence.
Three updates for your Electron investigation.
Improved recognition of renamed APIs, packaged ASAR pages, and unpacked companions.
Configurable memory and time limits, with completed findings retained when analysis stops early.
Summaries and selected views help agents explore large results without loading every payload.
Choose an app.asar from an application you can examine. Or open our original example to explore the inspector without finding a file first.
Check package.json, the declared entry point, and relevant source files. A dependency declaration or source map can suggest where to look next.
Copy a focused investigation task into your connected agent. Ask REA to trace one behavior, cite file evidence, and explain what still needs verification.
Understand the skill, the MCP connection, and the analysis engine. Then run your first focused investigation.
Go from an ASAR archive to an application entry point, an IPC boundary, and a question you can actually test.
What Reverse Engineer Anything does,
where this workbench fits,
and what to expect.
REA, short for Reverse Engineer Anything, is an open-source toolkit that connects coding agents to software analysis tools through MCP and a command-line interface. It helps investigate application behavior, JavaScript packages, and native binaries using evidence from the target.
No. REA Workbench is an independent companion with free browser tools and practical guides. The official project is morluto/rea on GitHub, and its website is rea.tools. Our ASAR inspector works on its own, without installing REA.
This workbench currently inspects Electron ASAR archives. It reveals files, declared application metadata, and useful investigation clues. Native decompilation and runtime observation need additional local tools. A browser inspection cannot recover a remote backend or guarantee the original source code.
No. Archive parsing and text previews run in your browser. File bytes, filenames, and generated reports are not sent to our servers. Downloaded reports stay on your device unless you choose to share them.
Not for this inspector. REA can also inspect JavaScript applications without a native analysis engine. Deeper native analysis requires a supported provider, such as Ghidra, Hopper, or IDA. Your coding agent and any commercial tools have their own requirements and costs.
Start with a small behavior: an export button, a settings value, or a clipboard action. Use the declared entry point and file paths to frame a focused REA question. Ask for evidence and unresolved details, then verify the explanation against the original app.
Start your Reverse Engineer Anything workflow with one file and one question.
