Security audit

get-net-pic

Security checks for vulnerabilities and agentic risk

Overview

The skill is a public-web image collection helper that writes downloaded images and a Word document locally, with some activation wording that users should treat carefully.

Install this only if you want an agent to fetch public website images and save local files. Avoid using it for login, account, or private-site tasks, and confirm the target site and save location before running it on large or unfamiliar pages.

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 (5)

Intent-Code Divergence

Medium

Confidence: 97% confidence
Finding: The skill metadata and examples say it triggers on phrases like '登录xxx官网', which implies authentication or account access, while the body later states it only handles publicly accessible pages and cannot process login-required sites. This mismatch can mislead the agent into activating in sensitive login contexts, increasing the chance of unintended navigation, credential-adjacent behavior, or user confusion around what the skill is authorized to do.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The trigger phrases are broad and overlap with ordinary browsing requests like opening a website or visiting an official page. In this skill, activation leads to fetching remote content, bulk-downloading images, writing files to the desktop, and generating documents, so accidental invocation can cause unexpected local side effects from a casual user request.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The applicability section lists broad scenarios but does not specify boundaries, exclusions, or consent requirements before downloading files and creating local artifacts. Because the skill performs network retrieval and local writes, unclear activation scope raises the risk of running destructive or privacy-impacting actions when the user only intended simple navigation or preview.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill instructs bulk downloading remote images to a desktop folder and creating a DOCX file, but it does not require an explicit warning or confirmation about local file creation, disk usage, or content provenance. This is risky because users may not expect persistent files to be written locally, and remote images from arbitrary sites may be undesirable, copyrighted, or misleading content.

Natural-Language Policy Violations

Low

Confidence: 88% confidence
Finding: 代码示例中固定使用 `toLocaleString('zh-CN')` 生成时间格式，意味着输出默认采用中文地区设置。文件中没有说明这是可配置项，也没有提供根据用户偏好选择语言/区域格式的机制。

Static analysis

No suspicious patterns detected.