ClawHub
Back to skill

Security audit

流感数据周报

Security checks across malware telemetry and agentic risk

Overview

This skill is a focused China CDC report finder/extractor with scoped official-site access and disclosed local artifact writes.

Before installing, expect the skill to open China CDC pages, run included Python helper scripts, download verified PDFs when requested, and write report artifacts locally. It does not ask for credentials or browser profiles, but administrators who require manifest-level permission declarations may want that metadata made explicit.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
96% confidence
Finding
The skill instructs the agent to use shell commands, network access, and file read/write operations, including downloading remote PDFs and writing artifacts, but it declares no explicit permissions. That mismatch is a real security issue because it can bypass user/admin expectations and allow broader execution than the skill manifest transparently communicates, especially when handling untrusted remote content.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.