ClawHub
Back to skill

Security audit

Youzan

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Youzan connector wrapper with limited tool scope and no hidden persistence or destructive behavior in the artifact.

Before installing, understand that this skill can access Youzan business data through your connected OOMOL account. Review the oo CLI installer if setup is needed, and confirm any operation that could change store or order data even though this artifact only lists read/list actions.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The skill labels `get_order` and `get_order_logistics` as `[write]` even though their descriptions indicate read-only retrieval. Misclassifying safe read actions as state-changing can distort downstream safety logic, cause unnecessary confirmations, and more importantly undermine operator trust in the action taxonomy used to separate read vs write behavior.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The instruction to use this skill for "ANY Youzan request" is overly broad and can cause the agent to route all Youzan-related tasks through a Bash-capable connector without sufficient task-level scrutiny. Broad invocation criteria increase the chance of unintended execution in contexts where a narrower, purpose-scoped skill or direct user clarification would be safer.

Static analysis

No suspicious patterns detected.