Security audit

电商价格比较

Security checks across malware telemetry and agentic risk

Overview

The skill is a mostly coherent price-comparison scraper, but its documentation encourages anti-bot evasion and possible logged-in scraping without clear consent or compliance boundaries.

Install only if you are comfortable with a scraper that may contact shopping platforms and whose docs discuss anti-bot workarounds. Prefer anonymous, low-rate, user-approved checks; do not provide cookies, passwords, browser sessions, proxies, or CAPTCHA-solving workflows unless you have explicit authorization and have reviewed the platform terms.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (4)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 89% confidence
Finding: The skill explicitly describes network-enabled scraping of multiple e-commerce platforms, but no permissions are declared. Undeclared network capability weakens transparency and policy enforcement, making it harder for users or a hosting system to understand and constrain outbound access.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 94% confidence
Finding: The documented behavior promises broad multi-platform comparison, price history, and recommendation features, while the actual implementation reportedly only supports limited JD-focused functionality and includes extra behaviors not clearly disclosed. This mismatch can mislead operators into over-trusting outputs, and hidden or undocumented behavior increases the risk of unauthorized data collection or unsafe downstream decisions.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The document explicitly recommends anti-scraping evasion measures such as rotating proxy IPs, changing user agents, and handling CAPTCHAs, but does not include any legal, contractual, or operational safeguards. In the context of an e-commerce price comparison skill, this materially increases the likelihood that the skill will be used to bypass platform protections, violate terms of service, trigger account/IP bans, or facilitate unauthorized data collection.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The README explicitly advises anti-bot evasion tactics such as frequent IP rotation and handling login state for scraping protected platforms. While this is documentation rather than executable code, it encourages behavior that can violate platform terms, trigger account bans, or expose operators to legal and operational risk, especially because it lacks warnings, compliance guidance, or rate-limiting/consent boundaries.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.