reverse-engineer-anything
- Use it when:
- A question depends on a shipped app or binary rather than its source: how a feature works, what changed between versions, or decompiled code.
- It does:
- Routes each target to the right REA tool, from Electron bundles, APKs and .NET assemblies to firmware, packet captures and native code in Ghidra or IDA.
- You get:
- Plain-language findings tied to evidence IDs, each kept apart as observed, inferred or unknown.
- Needs:
- The REA MCP server, or its rea-agents CLI for JavaScript apps; native code also needs Ghidra, IDA or Kuna, which it never installs.
- Also builds:
- AgentEnv bundles: a folder whose MCP tools stand in for a web app, answering from one recording of a live instance or a HAR file.
- Skips:
- Ordinary source repositories, where normal code tools do the job.
Needs the REA MCP server, or its rea-agents CLI for JavaScript apps; native analysis also needs an engine such as Ghidra or IDA, which it never installs.
TL;DR
REA, short for Reverse Engineer Anything, is one MCP server for reverse engineering. Installing the skill adds only instructions, except through REA's Claude Code or Codex plugin, which registers the server too. The skill tells the agent which tool fits the target: an ASAR or JavaScript tree, an archive or installer, an Android APK, a .NET assembly, firmware, WebAssembly, Go or Rust binaries, crash reports, model files, SQLite snapshots, saved HAR or packet captures, or a browser or Electron app already open. Native code goes through Ghidra, IDA or Kuna. It starts from summaries and pages into retained evidence instead of repeating an analysis. For broad questions it turns the request into a checklist, keeps a ledger of findings, and can split independent questions across parallel workers. Static analysis is never presented as observed execution; runtime tools, such as debuggers, Frida scripts or scripted browser clicks, run only the target and actions you asked for. When tools are missing or the connection drops, it diagnoses the actual blocker instead of re-running setup. It asks which app you mean rather than picking one, and ends by explaining the findings in plain language and closing its sessions.
Summarizes the skill at fdc4e39. Its SKILL.md has changed by 171 words since.
npx skills add https://github.com/morluto/rea --skill reverse-engineer-anything skills.sh pageSource on GitHubAll skills from morluto/rea
- Total Downloads
- 16,500 on skills.sh, Nov 9, 2026
- Repo stars
- 444,308 morluto/rea
- First seen
- Jul 18, 2026 on skills.sh
Weekly values
| Week to | Installs |
|---|---|
| Sep 21 | 4 |
| Sep 28 | 8 |
| Oct 5 | 4 |
| Oct 12 | 1,808 |
| Oct 19 | 5,004 |
| Oct 26 | 3,909 |
| Nov 2 | 3,265 |
| Nov 9 | 2,420 |
| Category |
Compliance, guardrails, skill vetting and browser fingerprinting. Find root causes, review code and test what you built. Everything for building software: planning, coding, debugging, design, frameworks, cloud, agents and skills. |
|---|---|
| Type |
Operates a CLI, API, app or bundled script to get something done. |
| Effect |
Its own steps create or change files, git state or local tools. |
| Needs |
Needs an MCP server connected. A specific command-line tool or desktop app must be installed, such as gh, agent-browser, LibreOffice, Obsidian or a vendor CLI. Drives a real Chromium or Chrome browser. |
| Publisher |
A verified publisher on skills.sh, or the company or maintainer behind the product the skill operates. |
| Topics | |
| Stack |
SKILL.md
No copy of its SKILL.md yet.