Back to skill

Security audit

Android Control

Security checks for vulnerabilities and agentic risk

Overview

This is a clearly disclosed Android-control skill with broad but purpose-aligned device automation guidance and visible safety limits.

Install this only if you want an agent to control an Android device through ADB, Termux, intents, screenshots, and UI automation. Keep device debugging authorized only for devices you trust, and require clear confirmation before message sending, app install or uninstall, data deletion, security changes, or screenshots of sensitive screens.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The usage guidance says only to use the skill when 'the conditions in the description are met,' which is circular and too vague to safely constrain activation. In a skill that can control Android devices via ADB, intents, UI automation, screenshots, and logging, ambiguous triggering can cause the agent to invoke powerful capabilities in unintended contexts, increasing the chance of unsafe actions or privacy-impacting behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

Most of the operational instructions and identity directives are written as mandatory Indonesian-language guidance, with no indication that the skill can adapt to the user's preferred language or locale. This creates a natural-language policy concern because the skill effectively imposes a specific language without user opt-in or documented justification.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill encourages screenshot and screen inspection for understanding state and verifying UI actions, but it does not clearly warn that screenshots may capture messages, contacts, tokens, photos, or other sensitive on-screen data. Because this skill targets Android control and observability, the omission increases privacy risk and may normalize collecting sensitive visual data without adequate notice or minimization.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill ends with broad trigger phrases such as 'Jalankan tugas Android ini' and positions itself to infer app, method, tool, and execution details automatically. In a high-privilege Android control skill, overly generic invocation language can cause accidental activation for loosely related requests and may lead the agent to take device-control actions the user did not explicitly intend.

Content

No source excerpt is available for this finding.