跳到正文
cccskills
免费在 GitHub 公开

mobile-insecure-storage

Detects sensitive data stored insecurely on mobile devices (Android/iOS). Trigger on: SharedPreferences, NSUserDefaults, SQLite, Room DB, DataStore, Core Data, Keychain misconfiguration, external storage, backup exposure, plaintext files, unencrypted databases, adb backup, iCloud backup, NSFileProtection, EncryptedSharedPreferences, SQLCipher, allowBackup, FLAG_SECURE, keyboard cache, sensitive logs. Covers MASVS-STORAGE-1 (local storage) and MASVS-STORAGE-2 (exposure to unauthorized actors).

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md7.1 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Mobile Insecure Data Storage

What Is Broken and Why

Mobile apps often store sensitive data (credentials, tokens, PII, keys) in locations accessible to other apps, backups, or physical device extraction. Android's SharedPreferences and iOS's NSUserDefaults are plaintext XML/plist files readable with root/jailbreak. External storage is world-readable. Backups (ADB/iCloud) can expose the entire app sandbox unless explicitly excluded. Logging APIs persist sensitive data in system logs readable by other apps. The attacker gains access to credentials or session tokens without ever touching the backend.

Key Signals

  • allowBackup="true" in AndroidManifest.xml without fullBackupContent exclusion rules
  • SharedPreferences files in /data/data/<pkg>/shared_prefs/ containing tokens, passwords, or keys
  • SQLite databases in the app sandbox without SQLCipher encryption
  • Files in /sdcard/ or getExternalStorageDirectory() containing sensitive content
  • iOS files lacking NSFileProtectionComplete data protection class
  • Keychain items with kSecAttrAccessibleAlways or no accessibility constraints
  • Log statements (Log.d, NSLog, print) containing session tokens or user data
  • Input fields without inputType="textPassword" or secureTextEntry=true
  • App switcher screenshots capturing password or payment screens

Methodology

Android:

  1. Decompile APK: apktool d app.apk — review AndroidManifest.xml for allowBackup, fullBackupContent
  2. Pull sandbox via ADB: adb backup -f backup.ab -noapk <pkg> → extract with android-backup-extractor
  3. Inspect /data/data/<pkg>/ (rooted): shared_prefs/, databases/, files/
  4. Check SharedPreferences XML for plaintext credentials or tokens
  5. Open SQLite databases: sqlite3 app.db .dump — look for unencrypted sensitive tables
  6. Search for external storage usage: grep -r "getExternalStorage" in decompiled source
  7. Check for sensitive log output: adb logcat | grep -i "password\|token\|secret\|key"
  8. Use objection: android hooking list activities → android clipboard monitor

iOS:

  1. Pull IPA and extract app bundle; examine Info.plist for NSAllowsArbitraryLoads
  2. Use iMazing or ideviceinstaller to pull app sandbox data
  3. Check Data Protection class on files: objection --gadget TARGET run ios filesystem list
  4. Inspect NSUserDefaults: <uuid>.plist in Library/Preferences
  5. Check Keychain: objection run ios keychain dump
  6. Dynamic: Frida hook NSFileManager writeToFile to observe Data Protection class used
  7. Background the app — screenshot the app switcher for sensitive data capture

Payloads & Tools

# Android — pull and extract backup
adb backup -f backup.ab -noapk TARGET_PKG
java -jar abe.jar unpack backup.ab backup.tar
tar xvf backup.tar

# Android — inspect SharedPreferences (rooted)
adb shell "cat /data/data/TARGET_PKG/shared_prefs/*.xml"

# Android — check logcat for leaks
adb logcat | grep -iE "password|token|secret|api.?key|session|auth"

# iOS — Frida: dump NSUserDefaults
frida -U TARGET -e "ObjC.classes.NSUserDefaults.standardUserDefaults().dictionaryRepresentation()"

# iOS — objection: keychain dump
objection --gadget TARGET run ios keychain dump

# iOS — check Data Protection class
frida -U TARGET -l data_protection_check.js

Bypass Techniques

  • Backup extraction without root — ADB backup works on non-rooted devices if allowBackup=true
  • Keychain extraction on jailbroken device — Keychain items without kSecAttrAccessibleWhenPasscodeSetThisDeviceOnly survive device-to-device migration
  • Memory scraping — even encrypted storage decrypts into memory; dump process memory with gcore or Frida memory scanner
  • Log persistence — adb logcat -d captures prior log buffer; sensitive data logged before crash is preserved

Exploitation Scenarios

Scenario 1 — ADB Backup Token Theft Setup: App stores JWT in SharedPreferences, allowBackup=true. → Trigger: Attacker with USB access runs adb backup. → Impact: Extracts valid session token, authenticates to backend as victim.

Scenario 2 — External Storage Credential Exposure Setup: App writes exported reports to getExternalStorageDirectory(). → Trigger: Malicious app with READ_EXTERNAL_STORAGE reads the files. → Impact: PII and embedded tokens from reports leaked to third-party app.

Scenario 3 — iOS Keychain Accessible After Reboot Setup: App stores password in Keychain with kSecAttrAccessibleAlways. → Trigger: Attacker with physical device access extracts Keychain via jailbreak. → Impact: Password retrieved even when device is locked/rebooted.

False Positives

  • SharedPreferences containing non-sensitive config (theme, language preference)
  • SQLite databases storing only public content (cached news articles, offline maps)
  • Keychain items scoped to the correct access control group — confirm accessibility attribute
  • Log output in debug builds only — verify the production build strips debug logs

Fix Patterns

// Android — encrypted SharedPreferences
val masterKey = MasterKey.Builder(context).setKeyScheme(MasterKey.KeyScheme.AES256_GCM).build()
val prefs = EncryptedSharedPreferences.create(context, "secure_prefs", masterKey,
    EncryptedSharedPreferences.PrefKeyEncryptionScheme.AES256_SIV,
    EncryptedSharedPreferences.PrefValueEncryptionScheme.AES256_GCM)

// Android — exclude from backup
// res/xml/backup_rules.xml
// <exclude domain="sharedpref" path="." />
// iOS — Keychain with strict access control
let access = SecAccessControlCreateWithFlags(nil, kSecAttrAccessibleWhenPasscodeSetThisDeviceOnly,
    .userPresence, nil)
// iOS — exclude file from backup
var url = URL(fileURLWithPath: sensitiveFilePath)
try url.setResourceValues({ v in v.isExcludedFromBackup = true }())

Related Skills

[[mobile-weak-crypto]] is the direct mitigation path: data that must be stored locally should be protected using properly implemented cryptography with keys in the Android Keystore or iOS Secure Enclave. [[mobile-auth-bypass]] overlaps when insecure storage holds auth tokens — a stolen JWT from SharedPreferences enables direct backend access without triggering biometric checks. [[mobile-resilience]] controls (root/jailbreak detection) are the last line of defense if storage protections fail, since extraction typically requires root or jailbreak access.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Exploit pervasive logical flaws in Multi-Factor Authentication (MFA/2FA) implementations to bypass the secondary authentication challenge entirely. Techniques include response manipulation, referal spoofing, token reuse, and predictable backup codes.

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

401/403 bypass playbook. Use when encountering access-denied responses on admin panels, API endpoints, or restricted paths. Covers path manipulation, HTTP method tampering, header injection, protocol downgrade, and automated bypass tools.

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

Complete PortSwigger deep-dive with exact payloads for every lab variant including zero-day techniques

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

Active Directory ACL abuse playbook. Use when exploiting misconfigured AD permissions including GenericAll, WriteDACL, DCSync rights, shadow credentials, LAPS reading, GPO abuse, and BloodHound-guided attack paths.

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

Execute AS-REP Roasting to extract and crack the NTLM hashes of Active Directory user accounts that have the "Do not require Kerberos preauthentication" flag explicitly enabled. This attack generates a recoverable Ticket Granting Ticket (TGT) without requiring the attacker to authenticate first.

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

AD Certificate Services attack playbook. Use when targeting misconfigured AD CS for privilege escalation via ESC1-ESC13 template abuse, NTLM relay to enrollment, CA officer abuse, and certificate-based persistence.

中文简介正在准备中。当前显示原文说明。

ShulkwiSEC/bb-huge232026年7月11日 更新

查看 ShulkwiSEC 的全部技能

报告此技能的问题