Android CLI Specialist
Orchestrates Android development tasks including project creation, deployment, SDK management, and environment diagnostics using the `android` command-line tool.
Authorized Android/iOS application reverse engineering and security testing: APK/IPA analysis, runtime instrumentation (Frida/Objection), SSL-pinning and jailbreak/root-detection bypass, per OWASP MASTG.
---
name: mobile-reverse
description: "Authorized Android/iOS application reverse engineering and security testing: APK/IPA analysis, runtime instrumentation (Frida/Objection), SSL-pinning and jailbreak/root-detection bypass, per OWASP MASTG."
risk: offensive
source: "https://github.com/zhaoxuya520/reverse-skill"
source_repo: "zhaoxuya520/reverse-skill"
source_type: community
date_added: "2026-08-25"
license: "MIT"
license_source: "https://github.com/zhaoxuya520/reverse-skill/blob/main/LICENSE"
---
> **⚠️ AUTHORIZED USE ONLY**
> This skill is for educational purposes or authorized security assessments only.
> You must have explicit, written permission from the system owner before using this tool.
> Misuse of this tool is illegal and strictly prohibited.
> **Mandatory confirmation gate**
> Before running any command that probes, exploits, changes, persists on, extracts data from, or attempts credential access against a target:
> 1. Ask the user to state the exact target URL, IP, account, or resource.
> 2. Ask the user to confirm written authorization and the permitted scope.
> 3. Show the exact command(s) and explain their expected effect.
> 4. Wait for explicit confirmation in the current conversation.
>
> Without that confirmation, remain read-only and provide defensive guidance only. Prefer a sandbox, disposable VM, or controlled lab.
# Mobile Reverse Engineering
## When to Use
- Assessing a mobile app's security posture within an approved scope.
- Instrumenting runtime behavior or bypassing transport protections in tests.
## 适用场景
- Android APK 逆向与安全测试
- iOS IPA 逆向与安全测试
- 移动应用运行时动态插桩
- SSL Pinning / Root 检测 / 越狱检测绕过
- 移动端加密算法提取(AES/RSA/HMAC 密钥)
- 移动应用渗透测试(OWASP MASTG)
- 非 Root/越狱环境下的应用测试
## 四阶段工作流
### Phase 1: 信息收集
```text
Android:
□ APK 获取(Google Play / APKMirror / adb pull)
□ Manifest 分析: 权限、导出组件、Intent Filter、backup 标志
□ androguard: androguard analyze APK → 组件/权限/签名
□ APKLeaks: 硬编码 API Key / Token / Secret 扫描
□ 加固检测: 是否加壳(360/腾讯/梆梆/爱加密)
iOS:
□ IPA 获取(App Store / ipatool / Apple Configurator)
□ 解密 App Store 二进制: frida-ios-dump / Clutch
□ Info.plist 分析: ATS 配置、URL Scheme、Queries Schemes
□ class-dump: 导出 ObjC 类结构
□ 加固检测: 是否使用 Swift/ObjC 混淆
```
### Phase 2: 静态分析
```text
跨平台:
□ JADX-GUI: APK → Java 源码(Android)
□ Ghidra / Hopper: .so / Mach-O 反编译
□ radare2 / Cutter: CLI 快速侦察
Android 专项:
□ apktool d app.apk → smali 代码 + 资源
□ dex2jar: DEX → JAR → JD-GUI
□ smali/baksmali: Dalvik 字节码修改
iOS 专项:
□ class-dump: 导出 ObjC 头文件
□ Swift 符号恢复: swift-demangle
□ dsymutil: 调试符号提取
□ otool -L: 查看动态库依赖
□ jtool2: Mach-O 分析
```
### Phase 3: 动态分析
```text
Frida — 通用动态插桩:
□ frida-ps -U: 列出设备进程
□ frida-trace -U -i "open*" com.app: 追踪函数调用
□ 自定义 Hook 脚本: 修改参数/返回值、调用私有方法
Objection — Frida 增强层(无需写脚本):
□ objection -g "com.app" explore
□ android root disable / ios jailbreak disable
□ android sslpinning disable / ios sslpinning disable
□ android keystore list / ios keychain dump
□ env / ls / sqlite connect
Frida Gadget(免 Root/越狱):
□ 注入 frida-gadget.so / FridaGadget.dylib 到 APK/IPA
□ 重新签名并满足设备/开发签名、调试及安装条件后测试;不代表绕过系统权限或所有平台完整性校验
□ objection patchapk --source app.apk(全自动)
```
### Phase 4: 网络分析
```text
□ Burp Suite: 拦截 HTTP/HTTPS,修改请求/响应
□ mitmproxy: 脚本化代理(Python API)
□ Wireshark: PCAP 抓包分析
□ 证书安装: Android 用户证书 → 系统证书(Magisk + MoveCert)
□ SSL Pinning 绕过: Frida/Objection/Xposed/SSL Kill Switch 2
□ WebSocket / gRPC 流量分析
```
## 常见绕过速查
### SSL Pinning
```bash
# Objection(最简)
objection -g "com.app" explore
android sslpinning disable
# Frida 通用脚本
frida -U -l ssl_pinning_bypass.js -f com.app
# Xposed(Android)
TrustMeAlready 模块 → 全局禁用证书校验
```
### Root / 越狱检测
```bash
# Objection
android root disable
ios jailbreak disable
# Frida 自定义(多层检测)
Java.perform(function() {
var RootBeer = Java.use("com.scottyab.rootbeer.RootBeer");
RootBeer.isRooted.implementation = function() { return false; };
// 额外绕过: Magisk su 检测、frida-server 检测、/proc/self/maps 检测
});
```
### 反调试
```bash
# Android
frida -U -l anti_debug_bypass.js -f com.app
# 绕过: ptrace(TracerPid)、/proc/self/status、isDebuggerConnected()
# iOS
# 绕过: PT_DENY_ATTACH、sysctl CTL_KERN/KERN_PROC/KERN_PROC_PID
frida -U -l ios_anti_debug.js -f com.app
```
## 移动端加密提取
```javascript
// Android — Hook Cipher.getInstance 获取密钥+算法
Java.perform(function() {
var Cipher = Java.use("javax.crypto.Cipher");
Cipher.getInstance.overload('java.lang.String').implementation = function(algo) {
console.log("[Cipher] Algorithm: " + algo);
return this.getInstance(algo);
};
Cipher.init.overload('int', 'java.security.Key').implementation = function(mode, key) {
var encoded = key.getEncoded();
console.log("[Cipher] encoded key available: " + (encoded !== null)); // do not log real key bytes
return this.init(mode, key);
};
});
// iOS — Hook CCCrypt
Interceptor.attach(Module.findExportByName("libcommonCrypto.dylib", "CCCrypt"), {
onEnter: function(args) {
console.log("CCCrypt op: " + args[0] + " alg: " + args[1]);
console.log("Key length: " + args[4].toInt32()); // capture only approved synthetic-key metadata
}
});
```
## 工具链
| 工具 | 平台 | 用途 |
|------|:--:|------|
| JADX-GUI | A | Java 反编译 |
| apktool | A | APK 解包/重建 |
| Ghidra | A+I | 多架构反编译 |
| Hopper | I | iOS 专用反汇编 |
| Frida | A+I | 动态插桩 |
| Objection | A+I | Frida REPL 增强 |
| MobSF | A+I | 自动化 SAST+DAST |
| class-dump | I | ObjC 类导出 |
| frida-ios-dump | I | IPA 解密 |
| jtool2 | I | Mach-O 分析 |
| Burp Suite | A+I | HTTP 拦截 |
| mitmproxy | A+I | 脚本化代理 |
> A=Android, I=iOS
## 参考
- `references/frida-objection-deep.md` — Frida + Objection 深度用法
- `references/ios-reverse-guide.md` — iOS 逆向专项
- `references/anti-detection-bypass.md` — Root/越狱/反调试/SSL Pinning 绕过
## 任务完成自检(声称完成前 MUST 通过)
- [ ] 我是否执行了工作流中的每一步(而不是只阅读)?
- [ ] 我是否基于 `tool-index` 使用了真实工具路径?
- [ ] 我是否产出了可复现证据(命令/脚本/截图/报告)?
- [ ] 我是否完成并回写了 RULES 要求的 Checklist 项?
## Limitations
- iOS instrumentation requires a jailbroken device or patched build.
- Bypass techniques break with app-shield vendor updates.
> Adapted from [zhaoxuya520/reverse-skill](https://github.com/zhaoxuya520/reverse-skill) (MIT).
## Instrumentation and evidence limits
Use approved app builds and dedicated test devices/accounts; obtaining an APK/IPA from a mirror does not establish provenance, license or permission. Store original hashes/signatures and label patched/re-signed builds. Re-signing changes entitlements, keychain/access groups, integrity checks and update behavior; results do not automatically reproduce in the production build. Pinning/root-detection bypass demonstrates a local instrumentation capability, not a server authorization failure or a universal Play Integrity bypass. Restore trust stores/hooks after the authorized test; never globally disable TLS validation on a daily-use device.
The scripts and tool-index/RULES ecosystem named in the guide are not bundled. Frida APIs/Java/ObjC bridges and Objection commands vary by version; verify the installed implementation, target ABI, module/class/method availability and overload before hooking. Several examples remain pseudocode (empty callbacks, missing bytesToHex helper and app-specific symbols). Hardware-backed keys may have getEncoded() return null. dsymutil cannot recreate stripped debug information; demangling does not undo semantic obfuscation. Avoid real key/keychain/cookie dumps and validate with canary material, keeping artifacts private. No app, device, hook, proxy or bypass ran during review.
---
name: mobile-reverse
description: "Authorized Android/iOS application reverse engineering and security testing: APK/IPA analysis, runtime instrumentation (Frida/Objection), SSL-pinning and jailbreak/root-detection bypass, per OWASP MASTG."
risk: offensive
source: "https://github.com/zhaoxuya520/reverse-skill"
source_repo: "zhaoxuya520/reverse-skill"
source_type: community
date_added: "2026-08-25"
license: "MIT"
license_source: "https://github.com/zhaoxuya520/reverse-skill/blob/main/LICENSE"
---
> **⚠️ AUTHORIZED USE ONLY**
> This skill is for educational purposes or authorized security assessments only.
> You must have explicit, written permission from the system owner before using this tool.
> Misuse of this tool is illegal and strictly prohibited.
> **Mandatory confirmation gate**
> Before running any command that probes, exploits, changes, persists on, extracts data from, or attempts credential access against a target:
> 1. Ask the user to state the exact target URL, IP, account, or resource.
> 2. Ask the user to confirm written authorization and the permitted scope.
> 3. Show the exact command(s) and explain their expected effect.
> 4. Wait for explicit confirmation in the current conversation.
>
> Without that confirmation, remain read-only and provide defensive guidance only. Prefer a sandbox, disposable VM, or controlled lab.
# Mobile Reverse Engineering
## When to Use
- Assessing a mobile app's security posture within an approved scope.
- Instrumenting runtime behavior or bypassing transport protections in tests.
## 适用场景
- Android APK 逆向与安全测试
- iOS IPA 逆向与安全测试
- 移动应用运行时动态插桩
- SSL Pinning / Root 检测 / 越狱检测绕过
- 移动端加密算法提取(AES/RSA/HMAC 密钥)
- 移动应用渗透测试(OWASP MASTG)
- 非 Root/越狱环境下的应用测试
## 四阶段工作流
### Phase 1: 信息收集
```text
Android:
□ APK 获取(Google Play / APKMirror / adb pull)
□ Manifest 分析: 权限、导出组件、Intent Filter、backup 标志
□ androguard: androguard analyze APK → 组件/权限/签名
□ APKLeaks: 硬编码 API Key / Token / Secret 扫描
□ 加固检测: 是否加壳(360/腾讯/梆梆/爱加密)
iOS:
□ IPA 获取(App Store / ipatool / Apple Configurator)
□ 解密 App Store 二进制: frida-ios-dump / Clutch
□ Info.plist 分析: ATS 配置、URL Scheme、Queries Schemes
□ class-dump: 导出 ObjC 类结构
□ 加固检测: 是否使用 Swift/ObjC 混淆
```
### Phase 2: 静态分析
```text
跨平台:
□ JADX-GUI: APK → Java 源码(Android)
□ Ghidra / Hopper: .so / Mach-O 反编译
□ radare2 / Cutter: CLI 快速侦察
Android 专项:
□ apktool d app.apk → smali 代码 + 资源
□ dex2jar: DEX → JAR → JD-GUI
□ smali/baksmali: Dalvik 字节码修改
iOS 专项:
□ class-dump: 导出 ObjC 头文件
□ Swift 符号恢复: swift-demangle
□ dsymutil: 调试符号提取
□ otool -L: 查看动态库依赖
□ jtool2: Mach-O 分析
```
### Phase 3: 动态分析
```text
Frida — 通用动态插桩:
□ frida-ps -U: 列出设备进程
□ frida-trace -U -i "open*" com.app: 追踪函数调用
□ 自定义 Hook 脚本: 修改参数/返回值、调用私有方法
Objection — Frida 增强层(无需写脚本):
□ objection -g "com.app" explore
□ android root disable / ios jailbreak disable
□ android sslpinning disable / ios sslpinning disable
□ android keystore list / ios keychain dump
□ env / ls / sqlite connect
Frida Gadget(免 Root/越狱):
□ 注入 frida-gadget.so / FridaGadget.dylib 到 APK/IPA
□ 重新签名并满足设备/开发签名、调试及安装条件后测试;不代表绕过系统权限或所有平台完整性校验
□ objection patchapk --source app.apk(全自动)
```
### Phase 4: 网络分析
```text
□ Burp Suite: 拦截 HTTP/HTTPS,修改请求/响应
□ mitmproxy: 脚本化代理(Python API)
□ Wireshark: PCAP 抓包分析
□ 证书安装: Android 用户证书 → 系统证书(Magisk + MoveCert)
□ SSL Pinning 绕过: Frida/Objection/Xposed/SSL Kill Switch 2
□ WebSocket / gRPC 流量分析
```
## 常见绕过速查
### SSL Pinning
```bash
# Objection(最简)
objection -g "com.app" explore
android sslpinning disable
# Frida 通用脚本
frida -U -l ssl_pinning_bypass.js -f com.app
# Xposed(Android)
TrustMeAlready 模块 → 全局禁用证书校验
```
### Root / 越狱检测
```bash
# Objection
android root disable
ios jailbreak disable
# Frida 自定义(多层检测)
Java.perform(function() {
var RootBeer = Java.use("com.scottyab.rootbeer.RootBeer");
RootBeer.isRooted.implementation = function() { return false; };
// 额外绕过: Magisk su 检测、frida-server 检测、/proc/self/maps 检测
});
```
### 反调试
```bash
# Android
frida -U -l anti_debug_bypass.js -f com.app
# 绕过: ptrace(TracerPid)、/proc/self/status、isDebuggerConnected()
# iOS
# 绕过: PT_DENY_ATTACH、sysctl CTL_KERN/KERN_PROC/KERN_PROC_PID
frida -U -l ios_anti_debug.js -f com.app
```
## 移动端加密提取
```javascript
// Android — Hook Cipher.getInstance 获取密钥+算法
Java.perform(function() {
var Cipher = Java.use("javax.crypto.Cipher");
Cipher.getInstance.overload('java.lang.String').implementation = function(algo) {
console.log("[Cipher] Algorithm: " + algo);
return this.getInstance(algo);
};
Cipher.init.overload('int', 'java.security.Key').implementation = function(mode, key) {
var encoded = key.getEncoded();
console.log("[Cipher] encoded key available: " + (encoded !== null)); // do not log real key bytes
return this.init(mode, key);
};
});
// iOS — Hook CCCrypt
Interceptor.attach(Module.findExportByName("libcommonCrypto.dylib", "CCCrypt"), {
onEnter: function(args) {
console.log("CCCrypt op: " + args[0] + " alg: " + args[1]);
console.log("Key length: " + args[4].toInt32()); // capture only approved synthetic-key metadata
}
});
```
## 工具链
| 工具 | 平台 | 用途 |
|------|:--:|------|
| JADX-GUI | A | Java 反编译 |
| apktool | A | APK 解包/重建 |
| Ghidra | A+I | 多架构反编译 |
| Hopper | I | iOS 专用反汇编 |
| Frida | A+I | 动态插桩 |
| Objection | A+I | Frida REPL 增强 |
| MobSF | A+I | 自动化 SAST+DAST |
| class-dump | I | ObjC 类导出 |
| frida-ios-dump | I | IPA 解密 |
| jtool2 | I | Mach-O 分析 |
| Burp Suite | A+I | HTTP 拦截 |
| mitmproxy | A+I | 脚本化代理 |
> A=Android, I=iOS
## 参考
- `references/frida-objection-deep.md` — Frida + Objection 深度用法
- `references/ios-reverse-guide.md` — iOS 逆向专项
- `references/anti-detection-bypass.md` — Root/越狱/反调试/SSL Pinning 绕过
## 任务完成自检(声称完成前 MUST 通过)
- [ ] 我是否执行了工作流中的每一步(而不是只阅读)?
- [ ] 我是否基于 `tool-index` 使用了真实工具路径?
- [ ] 我是否产出了可复现证据(命令/脚本/截图/报告)?
- [ ] 我是否完成并回写了 RULES 要求的 Checklist 项?
## Limitations
- iOS instrumentation requires a jailbroken device or patched build.
- Bypass techniques break with app-shield vendor updates.
> Adapted from [zhaoxuya520/reverse-skill](https://github.com/zhaoxuya520/reverse-skill) (MIT).
## Instrumentation and evidence limits
Use approved app builds and dedicated test devices/accounts; obtaining an APK/IPA from a mirror does not establish provenance, license or permission. Store original hashes/signatures and label patched/re-signed builds. Re-signing changes entitlements, keychain/access groups, integrity checks and update behavior; results do not automatically reproduce in the production build. Pinning/root-detection bypass demonstrates a local instrumentation capability, not a server authorization failure or a universal Play Integrity bypass. Restore trust stores/hooks after the authorized test; never globally disable TLS validation on a daily-use device.
The scripts and tool-index/RULES ecosystem named in the guide are not bundled. Frida APIs/Java/ObjC bridges and Objection commands vary by version; verify the installed implementation, target ABI, module/class/method availability and overload before hooking. Several examples remain pseudocode (empty callbacks, missing bytesToHex helper and app-specific symbols). Hardware-backed keys may have getEncoded() return null. dsymutil cannot recreate stripped debug information; demangling does not undo semantic obfuscation. Avoid real key/keychain/cookie dumps and validate with canary material, keeping artifacts private. No app, device, hook, proxy or bypass ran during review.
Imported SKILL.md format for Claude Code, Codex and Cursor. Review the complete instructions and required tools before use. Package integrity checked; individual agent behavior has not been tested.