Back to skill

Security audit

Wot UI Plus

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly normal component documentation, but some copy-ready examples could send user-selected files or personal data to places developers may not expect.

Review upload and mini-program authorization examples before use. Replace all demo upload URLs with an endpoint you control, consider disabling automatic upload until the user confirms, and avoid logging raw profile, phone, file, or upload event payloads in production.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
references/upload.md:153
Finding

Copy-ready upload examples transmit user-selected files to a fixed third-party endpoint

Content
View full analysis
``` The custom upload example explicitly sends the selected local file to the same service: ```ts const action = 'https://mockapi.eolink.com/zhTuw2P8c29bc981a741931bdd86eb04dc1e8fd64865cb5/upload' const customUpload: UploadMethod = (file, formData, options) => { const uploadTask = uni.uploadFile({ url: action, header: options.header, name: options.name, fileName: options.name, fileType: options.fileType, formData, filePath: file.url, success(res) { if (res.statusCode === options.statusCode) { options.onSuccess(res, file, formData) } else { options.onError({ ...res, errMsg: res.errMsg || '' }, file, formData) } }, fail(err) { options.onError(err, file, formData) } }) uploadTask.onProgressUpdate((res) => { options.onProgress(res, file) }) } ``` The image-cropper example also uploads the generated image to that endpoint: ```ts await startUpload(file, { action: 'https://mockapi.eolink.com/zhTuw2P8c29bc981a741931bdd86eb04dc1e8fd64865cb5/upload', onSuccess() { imgSrcUpload.value = tempFilePath showToast({ msg: t('shang-chuan-cheng-gong') }) } ...[truncated 2632 chars]
Remediation
View remediation
``` Require an explicit confirmation before transmitting selected content. 4. Move the destination into application configuration rather than embedding it in reusable source code: ```ts const action = import.meta.env.VITE_UPLOAD_ENDPOINT ``` 5. Require production upload endpoints to implement: - Authentication and authorization - HTTPS with certificate validation - File size and type allowlists - Server-side content validation - Malware scanning where appropriate - Randomized server-side file names - Storage access controls and retention limits - Rate limiting and audit logging 6. Avoid logging file contents, sensitive form data, authentication headers, or private upload responses. 7. Update the Skill instructions to require safe placeholders and privacy warnings whenever generating examples that transmit user-selected files. 8. Replace the endpoint in both `references/upload.md` and `references/imgCropper.md`, then search the entire reference set to ensure no remaining examples send user data to uncontrolled demonstration services. ]]>
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (105)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 70)May include surrounding context.

md
- `dropMenu.md`

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/avatar.md (reported line 146)May include surrounding context.

vue
<template>
  <view class="avatar-row">
    <!-- 图标头像 -->
    <wd-avatar icon="logo-wechat" fontSize="22" shape="circle"></wd-avatar>
    <wd-avatar icon="logo-douyin-o" fontSize="22" shape="square"></wd-avatar>

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/avatar.md (reported line 146)May include surrounding context.

vue
<template>
  <view class="avatar-row">
    <!-- 图标头像 -->
    <wd-avatar icon="logo-wechat" fontSize="22" shape="circle"></wd-avatar>
    <wd-avatar icon="logo-douyin-o" fontSize="22" shape="square"></wd-avatar>

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/button.md (reported line 133)May include surrounding context.

vue
<template>
  <view>
    <!-- 默认圆角按钮 -->
    <wd-button>主要按钮</wd-button>
    <wd-button type="success">成功按钮</wd-button>
    <wd-button type="info">信息按钮</wd-button>

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/button.md (reported line 370)May include surrounding context.

md
获取手机号
    </wd-button>

    <!-- 打开客服 -->
    <wd-button
      open-type="contact"
      session-from="weapp"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/calendar.md (reported line 176)May include surrounding context.

vue
<template>
  <wd-cell-group border>
    <!-- 单个日期选择 -->
    <wd-calendar
      label="单个日期选择"
      v-model="value1"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/calendar.md (reported line 281)May include surrounding context.

md
:formatter="formatter"
    />

    <!-- 快捷选项:快速选择近N天 -->
    <wd-calendar
      label="快捷选项"
      type="daterange"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/calendar.md (reported line 281)May include surrounding context.

md
:formatter="formatter"
    />

    <!-- 快捷选项:快速选择近N天 -->
    <wd-calendar
      label="快捷选项"
      type="daterange"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 111)May include surrounding context.

md
<template>
  <view>
    <wd-cell-group border>
      <!-- 基本用法:通过 columnChange 动态加载数据 -->
      <wd-col-picker
        label="选择地址"
        v-model="value1"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 111)May include surrounding context.

md
<template>
  <view>
    <wd-cell-group border>
      <!-- 基本用法:通过 columnChange 动态加载数据 -->
      <wd-col-picker
        label="选择地址"
        v-model="value1"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 203)May include surrounding context.

md
:column-change="asyncColumnChange"
      />

      <!-- 自动补全模式:columns 为空时自动触发 columnChange -->
      <wd-col-picker
        label="初始选项"
        v-model="value2"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 203)May include surrounding context.

md
:column-change="asyncColumnChange"
      />

      <!-- 自动补全模式:columns 为空时自动触发 columnChange -->
      <wd-col-picker
        label="初始选项"
        v-model="value2"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 305)May include surrounding context.

md
:column-change="columnChange"
      />

      <!-- 只读状态 -->
      <wd-col-picker
        label="只读"
        readonly

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/colPicker.md (reported line 411)May include surrounding context.

md
:display-format="displayFormat"
      />

      <!-- 确认前校验 -->
      <wd-col-picker
        label="before-confirm"
        v-model="value2"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/datetimePicker.md (reported line 182)May include surrounding context.

vue
<template>
  <wd-cell-group border>
    <!-- 完整日期时间(年月日时分) -->
    <wd-datetime-picker
      label="日期时间选择"
      v-model="datetimeValue"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/datetimePicker.md (reported line 182)May include surrounding context.

vue
<template>
  <wd-cell-group border>
    <!-- 完整日期时间(年月日时分) -->
    <wd-datetime-picker
      label="日期时间选择"
      v-model="datetimeValue"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/datetimePicker.md (reported line 286)May include surrounding context.

md
@confirm="handleConfirm"
    />

    <!-- 过滤分钟选项:仅显示5的倍数 -->
    <wd-datetime-picker
      label="过滤分钟(5分钟间隔)"
      v-model="filterValue"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/datetimePicker.md (reported line 286)May include surrounding context.

md
@confirm="handleConfirm"
    />

    <!-- 过滤分钟选项:仅显示5的倍数 -->
    <wd-datetime-picker
      label="过滤分钟(5分钟间隔)"
      v-model="filterValue"

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/sidebar.md (reported line 135)May include surrounding context.

md
<template>
  <view class="sidebar-container">
    <wd-sidebar v-model="active">
      <!-- 点状徽标 -->
      <wd-sidebar-item :value="0" label="新消息" is-dot />
      <!-- 数值徽标 -->
      <wd-sidebar-item :value="1" label="待处理" badge="5" />

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/steps.md (reported line 94)May include surrounding context.

vue
<template>
  <view>
    <!-- 基础步骤条:默认从第 1 步开始 -->
    <demo-block title="基本用法">
      <wd-steps :active="0">
        <wd-step />

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/steps.md (reported line 94)May include surrounding context.

vue
<template>
  <view>
    <!-- 基础步骤条:默认从第 1 步开始 -->
    <demo-block title="基本用法">
      <wd-steps :active="0">
        <wd-step />

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill description and all usage guidance in this README are written entirely in Chinese, which effectively imposes a language requirement on maintainers or users reading the documentation. The file does not mention that Chinese is optional, provide an alternative language, or justify a region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill sets a default requirement to answer in Chinese, which can override or conflict with a user's language preference if the runtime gives the skill strong influence over response behavior. This is primarily a policy/compliance issue rather than a direct security exploit, but it can reduce user control and cause incorrect handling in multilingual or accessibility-sensitive contexts.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The documentation explicitly promotes privacy-sensitive mini-program capabilities such as getUserInfo and getPhoneNumber, but provides no warning about consent, lawful basis, minimization, or secure handling of returned personal data. In a reusable UI skill, this omission can normalize unsafe implementation patterns and lead downstream developers to collect or log personal data without adequate privacy controls.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The example code demonstrates handlers for user info and phone number collection and logs returned event objects directly, while omitting any privacy or data-handling warning. Because sample code is commonly copied verbatim, this can propagate insecure handling of personally identifiable information and encourage overcollection or accidental exposure in logs.

Content

No source excerpt is available for this finding.