ClawHub
Back to skill

Security audit

AI网站一键发布

Security checks across malware telemetry and agentic risk

Overview

The skill is a disclosed static-site publishing workflow that uploads a vetted static ZIP to a fixed deployment service only when the user asks to publish.

Install this only if you are comfortable using the fixed coze-js-api.devtool.uk deployment service for public static-site publishing. Before deploying, check that your build output contains no secrets or private data, and note that the artifact does not explain remote retention or undeploy behavior.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill instructs the agent to read local project files and perform network operations against hard-coded external endpoints, yet it declares no permissions or equivalent user-facing capability disclosure. This creates a transparency and consent problem: a user may believe the skill is local-only or minimally scoped when it can exfiltrate deployment artifacts and metadata to a remote service.

Tp4

High
Category
MCP Tool Poisoning
Confidence
84% confidence
Finding
The skill markets itself as one-click static site publishing, but its operative instructions are dominated by hard-coded remote endpoint checks, strict private deployment routing, and artifact hashing/report generation. That mismatch can mislead users about what third-party infrastructure is involved and what data leaves the environment, increasing the chance that sensitive project contents are sent to an unexpected service.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill directs the agent to upload deployment ZIPs to an external endpoint and even instructs it not to ask the user again for confirmation. Even if intended for deployment, this suppresses an important consent checkpoint for data transmission and can result in source-derived assets, embedded secrets, or proprietary content being transferred to a third-party service without adequate privacy disclosure.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.