ClawHub
Back to skill

Security audit

Xborder Ecom Check - 跨境电商合规检查

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed cloud-scored compliance checker; its network use, API key handling, and local trial ID persistence are visible and aligned with its purpose.

Install only if you are comfortable sending the check responses to compliancehub.cn for scored runs. Use --non-interactive or --non-interactive-json when you want the bundled offline preview only, and avoid entering sensitive business details beyond the yes/no/na compliance statuses requested.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Scope Creep

Medium
Confidence
96% confidence
Finding
The manifest permits writes only to ~/.config/compliancehub, but the skill documentation also states it persists a separate anon_id file under that directory. If the platform expects explicit declaration of each persisted artifact or enforces more granular path review, this understates local state creation and can bypass user understanding of what data is retained between runs.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The docstring states that network usage is limited to scoring runs with a valid API key, but the code later submits user responses to the cloud even when no key is present by using anonymous trial mode. This is a transparency and consent issue: users may believe preview/offline expectations apply more broadly and may disclose sensitive compliance information that is transmitted to a third party unexpectedly.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.