ClawHub

Android Armor Breaker

Security checks across malware telemetry and agentic risk

Overview

This is a transparent Android reverse-engineering skill, but it needs review because it can control rooted devices, bypass app protections, read app memory, and has unsafe package-name handling in ADB shell commands.

Install only for authorized Android security research on a disposable emulator or dedicated test device. Treat package names and APK paths as trusted input, review or patch package-name validation before use, expect local extraction of app code or memory, and avoid personal or production devices.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (47)

subprocess module call

Medium
Category
Dangerous Code Execution
Content
# 检查应用特征
        try:
            # 检查应用是否快速崩溃（强反调试特征）
            result = subprocess.run(
                ["adb", "shell", f"timeout 2 am start -n {package_name}/.MainActivity"],
                capture_output=True,
                text=True,
Confidence
96% confidence
Finding
result = subprocess.run( ["adb", "shell", f"timeout 2 am start -n {package_name}/.MainActivity"], capture_output=True, text=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Content
)
            
            # 检查是否有特定文件或特征
            result2 = subprocess.run(
                ["adb", "shell", f"pm path {package_name}"],
                capture_output=True,
                text=True,
Confidence
95% confidence
Finding
result2 = subprocess.run( ["adb", "shell", f"pm path {package_name}"], capture_output=True, text=True, timeout=5

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 1. 检查应用是否仍在运行
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
95% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
# 2. 检查应用稳定性（5秒后是否仍在运行）
                time.sleep(5)
                result2 = subprocess.run(
                    ["adb", "shell", f"pidof {package_name}"],
                    capture_output=True,
                    text=True,
Confidence
95% confidence
Finding
result2 = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, tim

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 先停止应用
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=5
Confidence
86% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
time.sleep(1)
            
            # 启动应用
            result = subprocess.run(
                ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"],
                capture_output=True,
                text=True,
Confidence
90% confidence
Finding
result = subprocess.run( ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"], capture_output=True, text=True

subprocess module call

Medium
Category
Dangerous Code Execution
Content
def get_package_pid(self, package_name: str) -> Optional[int]:
        """获取应用进程PID"""
        try:
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
87% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
self.log("getting_package_pid", package=package_name)
        
        try:
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
95% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
# pidof失败，尝试使用ps命令
            self.log("pidof_failed_trying_ps", "WARNING")
            ps_result = subprocess.run(
                ["adb", "shell", f"ps -A | grep {package_name}"],
                capture_output=True,
                text=True,
Confidence
99% confidence
Finding
ps_result = subprocess.run( ["adb", "shell", f"ps -A | grep {package_name}"], capture_output=True, text=True, timeout=5

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 先停止应用（确保干净启动）
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=5
Confidence
95% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
time.sleep(1)
            
            # 启动应用
            result = subprocess.run(
                ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"],
                capture_output=True,
                text=True,
Confidence
96% confidence
Finding
result = subprocess.run( ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"], capture_output=True, text=True

subprocess module call

Medium
Category
Dangerous Code Execution
Content
bypass = AntiDebugBypass(verbose=self.verbose, language=self.language)
            
            # 首先停止应用以确保干净状态
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=5
Confidence
95% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
self.log("starting_frida_injection", "INFO")
            
            # 尝试获取PID（如果应用已在运行）
            pid_result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
95% confidence
Finding
pid_result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=5

subprocess module call

Medium
Category
Dangerous Code Execution
Content
self.log("getting_package_pid", package=package_name)
        
        try:
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
98% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 先停止应用
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=5
Confidence
98% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=5 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
# 启动应用并监控
            start_time = time.time()
            result = subprocess.run(
                ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"],
                capture_output=True,
                text=True,
Confidence
98% confidence
Finding
result = subprocess.run( ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"], capture_output=True, text=True

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 策略1:使用pidof获取PID(最快)
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
98% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=10 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
# 策略2:pidof失败,使用ps -A | grep(兼容多进程/进程名隐藏)
            self.log("pidof_failed_trying_ps", "WARN")
            ps_result = subprocess.run(
                ["adb", "shell", f"ps -A | grep {package_name}"],
                capture_output=True,
                text=True,
Confidence
99% confidence
Finding
ps_result = subprocess.run( ["adb", "shell", f"ps -A | grep {package_name}"], capture_output=True, text=True, timeout=10

subprocess module call

Medium
Category
Dangerous Code Execution
Content
continue

            # 策略3:使用ps -A | grep -v grep(更精确)
            ps_result2 = subprocess.run(
                ["adb", "shell", f"ps -A | grep -v grep | grep {package_name}"],
                capture_output=True,
                text=True,
Confidence
99% confidence
Finding
ps_result2 = subprocess.run( ["adb", "shell", f"ps -A | grep -v grep | grep {package_name}"], capture_output=True, text=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 先停止应用(确保干净启动)
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=10
Confidence
98% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=10 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
time.sleep(2)

            # 启动应用
            result = subprocess.run(
                ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"],
                capture_output=True,
                text=True,
Confidence
98% confidence
Finding
result = subprocess.run( ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"], capture_output=True, text=True

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 使用pidof获取PID
            result = subprocess.run(
                ["adb", "shell", f"pidof {package_name}"],
                capture_output=True,
                text=True,
Confidence
96% confidence
Finding
result = subprocess.run( ["adb", "shell", f"pidof {package_name}"], capture_output=True, text=True, timeout=10 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
try:
            # 先停止应用（确保干净启动）
            subprocess.run(
                ["adb", "shell", f"am force-stop {package_name}"],
                capture_output=True,
                timeout=10
Confidence
96% confidence
Finding
subprocess.run( ["adb", "shell", f"am force-stop {package_name}"], capture_output=True, timeout=10 )

subprocess module call

Medium
Category
Dangerous Code Execution
Content
time.sleep(2)
            
            # 启动应用
            result = subprocess.run(
                ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"],
                capture_output=True,
                text=True,
Confidence
96% confidence
Finding
result = subprocess.run( ["adb", "shell", f"monkey -p {package_name} -c android.intent.category.LAUNCHER 1"], capture_output=True, text=True

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill advertises and documents powerful capabilities including shell execution, file read/write, environment inspection, ADB control, Frida injection, and root-based memory extraction, but does not declare explicit permissions. This creates a dangerous mismatch between what a reviewer or platform might expect and what the skill is capable of doing, reducing transparency and increasing the risk of misuse or unsafe execution in privileged environments.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal