웹, 모바일, 데스크톱 애플리케이션을 감사하여 리버스 엔지니어링을 허용하는 취약점을 식별하고 설정 수정 방안을 제안합니다.
다른 이름 GuardBee/guardbee-mcp, @guardbee/mcp-reverse-shield
- 스타
- 2
- Tools
- 5
- 다운로드, 30일
- 2,992
- 최신 버전
- 0.8.0
- 마지막 커밋
- 2026년 11월 12일
- 라이선스
- MIT
reverse-shield MCP란?
reverse-shield은(는) MCP server입니다. 웹, 모바일, 데스크톱 애플리케이션을 감사하여 리버스 엔지니어링을 허용하는 취약점을 식별하고 설정 수정 방안을 제안합니다. 5개 도구를 제공합니다. 예: scan_artifact, suggest_fixes, scan_url.
프로젝트 README에서 생성된 요약
reverse-shield MCP server 정보
보안 엔지니어와 애플리케이션 개발자는 reverse-shield를 사용하여 배포된 소프트웨어에 민감한 정보가 얼마나 노출되어 있는지 평가할 수 있습니다. APK, IPA, Electron 폴더 또는 웹 번들과 같은 빌드 아티팩트를 서버로 지정하면 리버스 엔지니어링을 가능하게 하는 취약점을 식별합니다. 임베디드 소스 맵, 압축되지 않은 코드, Flutter 또는 .NET의 난독화 누락, 바이너리 또는 리소스 파일 내의 하드코딩된 비밀 정보 등의 문제를 탐지합니다.
서버는 파일 스캔을 위해 로컬 및 오프라인 상태로 작동하지만, scan_url을 통해 원격 점검도 수행할 수 있습니다. DEX, ELF, Mach-O, ASAR와 같은 형식을 위해 특화된 내부 파서를 사용하므로 apktool과 같은 외부 도구가 필요하지 않습니다. 사용자는 AI 클라이언트를 통해 서버와 상호작용하여 보안 태세에 대한 A–F 등급을 받을 수 있습니다. 취약점 식별을 넘어 suggest_fixes 도구는 Vite, Gradle 또는 React Native 프로젝트 설정을 분석하여 수정을 위한 통합 diff를 제공합니다.
활용 사례
- 1모바일 애플리케이션 보안 감사Android APK 또는 iOS IPA를 분석하여 디버그 가능 설정 및 난독화 누락 여부를 확인합니다.
- 2유출된 자격 증명 식별바이너리, 설정 파일 및 리소스 풀을 검색하여 노출된 개인 키와 비밀 정보를 찾습니다.
- 3빌드 설정 수정 자동화웹 또는 모바일 빌드 설정의 보안 격차를 해결하기 위한 코드 패치를 생성합니다.
- 4웹 번들 노출 평가라이브 웹사이트 또는 로컬 JS 번들에서 노출된 소스 맵과 압축되지 않은 코드를 확인합니다.
프로젝트의 README를 바탕으로 작성되었습니다. 모든 활용 사례는 해당 내용의 인용문과 대조하여 확인합니다.
reverse-shield MCP 설치 방법
형식 확인: 2026년 11월 10일Cursor에 추가Goose에 추가Claude Code에 reverse-shield MCP 설치
$ claude mcp add reverse-shield -- npx -y @guardbee/mcp-reverse-shieldCursor에 reverse-shield MCP 설치
~/.cursor/mcp.json
{
"mcpServers": {
"reverse-shield": {
"command": "npx",
"args": [
"-y",
"@guardbee/mcp-reverse-shield"
]
}
}
}VS Code에 reverse-shield MCP 설치
.vscode/mcp.json
{
"servers": {
"reverse-shield": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@guardbee/mcp-reverse-shield"
]
}
}
}VS Code CLI에 reverse-shield MCP 설치
$ code --add-mcp '{"name":"reverse-shield","type":"stdio","command":"npx","args":["-y","@guardbee/mcp-reverse-shield"]}'Claude Desktop에 reverse-shield MCP 설치
claude_desktop_config.json
{
"mcpServers": {
"reverse-shield": {
"command": "npx",
"args": [
"-y",
"@guardbee/mcp-reverse-shield"
]
}
}
}Codex에 reverse-shield MCP 설치
$ codex mcp add reverse-shield -- npx -y @guardbee/mcp-reverse-shieldCodex config.toml에 reverse-shield MCP 설치
~/.codex/config.toml
[mcp_servers.reverse-shield] command = "npx" args = ["-y", "@guardbee/mcp-reverse-shield"]
Gemini CLI에 reverse-shield MCP 설치
$ gemini mcp add -s user reverse-shield npx -- -y @guardbee/mcp-reverse-shieldGemini settings.json에 reverse-shield MCP 설치
~/.gemini/settings.json
{
"mcpServers": {
"reverse-shield": {
"command": "npx",
"args": [
"-y",
"@guardbee/mcp-reverse-shield"
]
}
}
}reverse-shield MCP 도구 5
README에서 감지됨scan_artifactsuggest_fixesscan_urllist_checksbundle_secret
| scan_artifact | — | |
| suggest_fixes | — | |
| scan_url | — | |
| list_checks | — | |
| bundle_secret | — |
자동 검사
2026년 11월 13일에 실행됨전체 보고서 →자동 검사를 통과했습니다. 이는 보안 감사가 아닙니다. 코드를 직접 읽고 서버에 필요한 최소한의 권한만 부여하십시오. 검사 방식
- 소스 저장소 연결 및 접근 가능통과
- 오픈 소스 라이선스 (OSI)통과
- 최근 1년 내 커밋 또는 릴리스통과
- 패키지 이름이 인기 패키지와 유사하지 않음통과
- 원격 엔드포인트가 HTTPS 사용검사되지 않음
- 파괴적인 도구에 라벨 표시됨통과
- 모델에 대한 숨겨진 지침 없음통과
- 패키지 및 해당 종속성에 중요한 보안 권고 없음통과
README
GitHub에서 보기@guardbee/mcp-reverse-shield
🇬🇧 English | 🇹🇷 Türkçe | 🇨🇳 中文
An MCP server that tells you how much your shipped app gives a reverse engineer for free.
Reverse-engineering MCPs now let an agent take an app it has no source for — a web bundle, an Electron app, an APK, an iOS app, a game, a native or .NET binary — and explain how a feature works, down to the binary. This package is the other side of that: point it at your own build artifact and it reports what such a tool (or a person with jadx, ILSpy, Il2CppDumper, Hopper or the browser's devtools) would get without any effort, with the fix for each — and suggest_fixes turns those fixes into a diff against your build config.
Telemetry is off by default. With
GUARDBEE_TELEMETRY=1it sends the server, tool, success, duration and short parameter values such as a file path; longer strings and fields likecontent,tokenorpasswordare replaced with a placeholder. See@guardbee/mcp-telemetry.
Claude ──► reverse-shield ──► your build artifact / live site
│
├─ web source maps (shipped, inline, publicly reachable), unminified / dev bundles, keys, internal hosts, WebAssembly debug info and names
├─ electron fuses (RunAsNode, --inspect, NODE_OPTIONS, ASAR integrity), unpacked app code, code inside app.asar
├─ android APK and AAB: debuggable, R8/ProGuard missing, signing scheme, backup, plain React Native bundle, unstripped .so
├─ ios IPA and .app: Debug-configuration builds, development profiles, plain React Native bundle, .dSYM in the bundle, the executable
├─ engines Unity (IL2CPP metadata, Mono assemblies), PyInstaller, Java/Spring Boot JARs, Flutter (Dart obfuscation)
├─ native debug info, symbols, get-task-allow, hardened runtime, signing, PIE/NX/RELRO/canary (ELF, Mach-O, PE)
├─ secrets keys in everything above: dex, resources.arsc, plists, config files, Hermes bytecode, .so and executables
└─ .NET obfuscation missing, embedded PDB, leaked build paths
Everything runs locally and offline (except scan_url), with its own parsers for ASAR, the Electron fuse wire, binary AndroidManifest (AXML), DEX, ELF, Mach-O (including universal binaries and code signatures) and PE/.NET metadata — no apktool, jadx or binutils needed.
What it flags
| Check | Severity | Target | What it gives away |
|---|---|---|---|
sourcemap_embeds_source |
critical | web | Full original source via sourcesContent |
sourcemap_shipped / sourcemap_inline / sourcemap_publicly_reachable |
high | web | Original file paths and structure |
eval_devtool_bundle |
high | web | Dev bundle with each module's path |
bundle_secret |
per key type | all | A working credential in client code, an app package or a binary |
wasm_debug_info |
high | web | WebAssembly with DWARF: source files, lines, types |
unminified_bundle / dev_build_framework / wasm_name_section |
medium | web | Readable code / framework dev runtime / named WebAssembly functions |
sourcemap_reference / internal_endpoint |
low | web | Where the map lives / localhost, private IPs, staging hosts |
electron_fuse_run_as_node / electron_fuse_inspect_args |
high | electron | Node runtime or debugger inside the signed app |
electron_app_unpacked |
high | electron | App code as an editable folder |
electron_fuse_node_options / _asar_integrity_off / _load_outside_asar |
medium | electron | Code injection and ASAR tampering |
apk_debuggable |
critical | android | Debugger attach and run-as into app data |
apk_no_obfuscation / apk_js_bundle_plaintext / apk_unsigned |
high | android | jadx output with real names / readable RN bundle / no tamper evidence |
apk_allow_backup / apk_v1_only_signature |
medium | android | Data pulled via backup / Janus-style repackaging |
native_debug_info / native_debuggable_entitlement |
high | native | Source-level debug data / lldb attach |
native_symbols_not_stripped / native_no_hardened_runtime / native_injection_entitlement / native_unsigned |
medium | native | Named functions / dylib injection / patched copies run |
native_no_pie / native_no_nx |
medium | native | Exploit-friendly fixed addresses / executable stack |
native_no_relro / native_no_stack_canary / native_adhoc_signed / native_build_path_leak |
low | native | Hardening gaps and build-host details |
ios_debug_build / ios_js_bundle_plaintext / ios_dsym_shipped |
high | ios | Unoptimised code with full symbols and preview stubs / readable RN bundle / full DWARF symbols in the bundle |
ios_development_profile |
medium | ios | Debugger attach on registered devices |
unity_il2cpp_metadata / unity_mono_assemblies |
high | engines | Every class, method and string of the game / C# scripts via dnSpy |
flutter_not_obfuscated |
high | engines | Dart class and method names for blutter / reFlutter |
hermes_names_not_minified |
medium | engines | React Native function names in Hermes bytecode for hermes-dec / hbctool |
python_pyinstaller_bundle / java_not_obfuscated |
high | engines | Decompilable .pyc files / Java classes with original names |
dotnet_not_obfuscated / dotnet_embedded_pdb |
high | .NET | ILSpy/dnSpy output close to the original source |
list_checks returns the full catalog with fixes. Every result carries an A–F grade: F when anything critical (or several high findings) is exposed, A when nothing is.
How the trickier checks decide
- R8/ProGuard: reads class descriptors from every
classes*.dexand looks at classes under your application id (and under the packages of components your manifest names, for.debug-style ids). Dozens of descriptive class names with almost no one- or two-letter names means R8 did not run. - .NET obfuscation: reads the metadata
#Stringsheap. Known obfuscator markers (ConfuserEx, Dotfuscator, SmartAssembly, …) pass immediately; otherwise an assembly passes when at least 10% of its identifiers are one- or two-letter names, bare numbers or non-ASCII junk. Checked against an Obfuscar build (21%, passes) and 400 unobfuscated NuGet assemblies (383 flagged; the other 17 are resource-only assemblies with too few names to judge). - Private keys: only reported when base64 key material follows the PEM header — crypto libraries bundled into apps carry the header as a string constant.
- Credentials outside JavaScript: Android dex strings, the
resources.arsc/resources.pbstring pool (wherestrings.xmland google-services values end up),assets/andres/raw/; iOS plists (GoogleService-Info.plist) and the executable; JSON/plist/.env/properties/YAML files in an ASAR, a JAR (Spring Bootapplication.properties) or a build folder; Hermes bytecode,.solibraries and native/.NET binaries by their printable strings. Only fixed-format keys are matched there, as in bundles. A Google/Firebase API key in an Android or iOS package is reported as medium: those ship by design and must be restricted to the app's package name or bundle id. - Repeats: one key in 30 chunks is one finding (“also in 29 other files”), and long lists of unminified files collapse into one.
- iOS Debug builds: Xcode 16 puts a Debug build's code in
<App>.debug.dylibnext to__preview.dylib; either one in a bundle means the Debug configuration was shipped. - Unity: IL2CPP's
global-metadata.datis flagged only when its standard header is intact (protected builds encrypt it); a MonoAssembly-CSharp.dllgoes through the .NET name check. - PyInstaller: the CArchive cookie is searched across the whole file (on macOS the archive sits in its own section before the code signature); PyArmor and Nuitka builds are noted, not flagged.
- Flutter: reads the identifiers in the AOT snapshot (
libapp.so,App.framework/App).--obfuscaterenames everything the program defines to two- or three-letter names while the Dart VM's own names stay, so a snapshot passes when at least 10% of its identifiers are that short. Checked against one program built both ways with the Dart SDK'sgen_snapshot: 1% without the flag, 31% (unstripped) and 37% (stripped) with it. Fewer than 200 identifiers is a note, not a verdict. - WebAssembly: custom sections only —
.debug_*(DWARF),name(function names) andsourceMappingURL; the code itself is not judged. - Live URLs: only the site's own scripts are read. Platform-served code (HubSpot
/hs/hsstatic/, WordPress/wp-includes/) and open-source libraries (jQuery, Bootstrap, polyfills, vendor bundles) are skipped, and a reachable map whose sources are all dependencies is not reported. Internal hosts count only when they point into the site (private IPs,.internal/.corp/.lan, its own staging hosts); thelocalhostURLs frameworks ship are ignored. A Google browser key is reported as medium (restrict it to the site's referrers), and a Supabaseanonkey is not reported, since the browser is meant to hold it (aservice_rolekey still is). - React Native with Hermes: reads the bytecode's string table and function headers. Metro's minifier renames functions to one- or two-letter names; without it hermes-dec and hbctool print the source's names back. A bundle passes when at least 15% of its named functions have names that short. Checked against one bundle (lodash + zod + app code) compiled with hermesc
-Oboth ways: 1% unminified, 57% minified. Function names are read for bytecode versions 84–100 that share the v98 12-byte small-header layout; other versions, 16-byte headers, and bundles with fewer than 50 named functions are a note, not a verdict.
Quick start
{
"mcpServers": {
"guardbee-reverse-shield": {
"command": "npx",
"args": ["-y", "@guardbee/mcp-reverse-shield"]
}
}
}
Then ask your agent: “How easy is it to reverse engineer ./release/MyApp.app? Fix what you can in the build config.”
npx @guardbee/mcp-reverse-shield scan ./dist --fail-on=high
npx @guardbee/mcp-reverse-shield scan ./app/build/outputs/apk/release/app-release.apk
npx @guardbee/mcp-reverse-shield scan "/Applications/My App.app" --format=json
npx @guardbee/mcp-reverse-shield scan https://app.example.com --format=sarif > reverse-shield.sarif
npx @guardbee/mcp-reverse-shield scan ./build/MyApp.ipa
npx @guardbee/mcp-reverse-shield suggest . --format=patch | git apply # review first: suggest . prints the same as text
Tools
| Tool | Purpose |
|---|---|
scan_artifact |
Scan a file or directory: APK, AAB, IPA, iOS or macOS .app, Electron folder, app.asar, JAR, ELF/Mach-O/PE/.NET binary, JS bundle, source map, or a build output folder (Unity and Python builds included). Optional ignore and baseline |
suggest_fixes |
Read the project's build config (Vite, webpack, Next.js, Angular, CRA, Android Gradle and manifest, React Native, electron-builder / Forge, .NET project files, Flutter) and return the changes as unified diffs. Never writes files |
scan_url |
Fetch a live page, check its same-site bundles, follow sourceMappingURL and probe <bundle>.map (third-party scripts are not fetched) |
list_checks |
The check catalog with what each finding gives away and how to fix it |
Directories are walked including dist/, build/ and out/ (that is the point); node_modules, VCS and cache folders are skipped, and a macOS .app or Electron folder found inside is scanned as one unit — its fuses, its ASAR and its own executable, not the bundled third-party frameworks.
guardbee.yml:
reverse-shield:
fail-on: high
max-files: 5000
baseline: .reverse-shield-baseline.json
exclude:
- "vendor/"
ignore: # accepted exposures: a rule id, or rule:path-substring
- bundle_secret:firebase-config.js
CI and baselines
Record what the current release exposes, then let CI fail only on what a change adds:
npx @guardbee/mcp-reverse-shield scan ./dist --update-baseline # writes .reverse-shield-baseline.json
npx @guardbee/mcp-reverse-shield scan ./dist --baseline=.reverse-shield-baseline.json --fail-on=high
# Compared to baseline (grade C → D): 1 new, 2 fixed, 4 unchanged
Findings are matched by check and location, with bundler content hashes (index-BvL4J5f8.js, 7321-dd7cb90a7de65945.js) normalised, so a rebuild does not make old findings look new. SARIF results carry baselineState (new / unchanged).
The repository ships a composite action that scans, uploads SARIF, writes a job summary and exposes grade / new-findings outputs:
- uses: GuardBee/guardbee-mcp/.github/actions/reverse-shield@main
with:
target: dist
fail-on: high
baseline: .reverse-shield-baseline.json
Library use
@guardbee/mcp-reverse-shield/scan exports the scanners (scanUrl, scanPath, analyzeJavaScript, …) without the MCP server, for tools that embed reverse-shield:
import { scanUrl } from "@guardbee/mcp-reverse-shield/scan";
const result = await scanUrl("https://example.com/", { maxScripts: 15, timeoutMs: 8_000 });
console.log(result.grade, result.findings);
Limits
- Obfuscation is judged from names, not control flow: a build that only renames passes, and string encryption or control-flow flattening is not measured.
- Web checks look at what is shipped; they cannot tell whether a key is meant to be public (some, like Firebase web keys, are restricted server-side). Rotate or restrict it, then accept it with an
ignoreentry (bundle_secret:firebase-config.js). - Strings are read from dex and resource tables by their structure, from other binaries as printable runs; a key split across string constants, encrypted or assembled at runtime is not found (that is the point of doing so, but it only slows a reverse engineer down).
- Native hardening checks (PIE/NX/RELRO/canary) are exploit-mitigation signals rather than reverse-engineering ones; they are reported low/medium for that reason.
- App Store IPAs are FairPlay-encrypted on distribution; the scanner reads the IPA you export (development, ad hoc or App Store export before Apple encrypts it), which is what a jailbroken device would also yield.
suggest_fixesedits the common shapes of each config; unusual layouts (computed webpack configs, Gradle convention plugins) get no diff, and Forge and Flutter get instructions rather than a patch. Always review before applying.
License
MIT
reverse-shield MCP 질문
reverse-shield MCP를 Claude Code에 추가하려면 어떻게 하나요?
터미널에서 claude mcp add reverse-shield -- npx -y @guardbee/mcp-reverse-shield를 실행합니다. 모든 프로젝트에서 사용할 수 있게 하려면 --scope user를 추가합니다.
reverse-shield MCP를 Cursor에 추가하려면 어떻게 하나요?
~/.cursor/mcp.json를 열고 mcpServers 아래에 서버를 추가합니다. 위 설치 섹션에 표시된 Cursor 구성을 사용합니다.
reverse-shield MCP는 어떤 도구를 제공하나요?
5개 도구를 제공합니다. 예: scan_artifact, suggest_fixes, scan_url, list_checks, bundle_secret.
어떤 유형의 애플리케이션을 스캔할 수 있나요?
웹 번들, Electron 앱, APK, iOS 앱, 게임, 네이티브 바이너리, .NET 바이너리, Unity, PyInstaller, Java/Spring Boot JAR를 감사할 수 있습니다.
jadx나 apktool 같은 외부 리버스 엔지니어링 도구가 필요한가요?
아니요, ASAR, DEX, ELF, Mach-O, PE를 포함한 다양한 파일 형식을 처리하기 위해 자체 내부 파서를 사용합니다.