ClawHub
Back to skill

Security audit

北京聚英电子｜聚英云设备控制

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed IoT device-control skill that uses a user-provided API token to read and control Juying Cloud devices, with no hidden code or persistence found.

Install this only if you intend to let the agent control real Juying Cloud hardware. Use a token limited to the devices and operations you need, review target device and channel names carefully before control commands, and prefer read-only status checks when unsure.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The file is structured around fixed language sections for Simplified Chinese, Traditional Chinese, and English, but does not state that the user may choose their preferred language or that responses should follow user preference. Under the policy for natural-language constraints, hardwiring language behavior without opt-in or documented justification can be a locale/language policy issue.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill documents physical control actions that can open or close device channels, but it does not warn users or require any confirmation before potentially affecting real-world equipment. In an IoT device-control context, ambiguous or accidental invocation could cause unintended operation of connected systems, making this a genuine safety and security concern even if the API itself is authenticated.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The examples instruct the agent to open and close physical device channels without any cautionary guidance, confirmation step, or acknowledgement of possible real-world consequences. In an IoT control context, these actions may affect powered equipment, doors, relays, pumps, or other actuators, so omission of safety warnings increases the chance of accidental or unsafe actuation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill explicitly supports active device-control operations such as opening and closing channels, yet the overview does not clearly warn users that using the skill can change real-world device state. Because the same API token is sent on every request and likely authorizes both read and write actions, a user may invoke the skill without understanding it can perform potentially unsafe or unintended physical actions.

Content

No source excerpt is available for this finding.