Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计提醒

helpmetest-visual-checkhelpmetest 目视检查

Agent Skill

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

总安装

559

周安装

24

GitHub Stars

公开资料未说明

下载量

196
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:helpmetest-visual-check(helpmetest 目视检查)
来源仓库:https://github.com/help-me-test/skills
仓库路径:skills/helpmetest-visual-check
安装命令:
npx skills add https://github.com/help-me-test/skills --skill helpmetest-visual-check
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/help-me-test/skills --skill helpmetest-visual-check

简介

helpmetest-visual-check 用于辅助界面设计、视觉规范和交互体验优化。

  • 适用于需要生成 UI 方案、检查视觉一致性或改进组件层级的开发场景。
  • 使用时需结合现有品牌、设计系统和用户任务,避免仅堆砌装饰元素。
  • 涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Who you are: If .helpmetest/SOUL.md exists in this project, read it before starting — it defines your character and shapes how you work.
No MCP? The CLI has full feature parity — use helpmetest <command> instead of MCP tools. See the CLI reference.

Visual Check

Quick visual verification using screenshots. Tests how things look without creating formal test files.

When to Use This Skill

Use when user asks casual verification questions like:

  • "Check if the login button is blue"
  • "Does the header look right on mobile?"
  • "See if the pricing table is centered"
  • "Is the footer visible?"
  • "Verify the hero image loaded"
  • "Check that modal is showing correctly"

NOT for:

  • Creating formal tests (use /helpmetest-test-generator)
  • Comprehensive site testing (use /helpmetest)
  • Debugging failing tests (use /helpmetest-debugger)

Process

Step 1: Understand What to Check

Extract from user's request:

  • URL - Where to look (if not provided, ask)
  • Element/Feature - What to verify ("login button", "header", "pricing table")
  • Expected state - What should it look like? ("blue", "centered", "visible")

Step 2: Navigate and Capture

Use helpmetest_run_interactive_command with navigation + screenshot:

Go To  <url>

Tool returns:

  • Screenshot (automatically included)
  • Page state
  • Any console errors

Step 3: Describe What You See

Based on screenshot, describe:

  1. Is the element there? - Present or missing
  2. How does it look? - Color, size, position, styling
  3. Matches expectation? - Yes/No with specifics

Example response:

✓ Login button found in top-right navigation
✓ Color: Blue (#0066FF) - matches expected blue
✓ Size: Standard button size (~120px wide)
✓ Position: Aligned right, 20px from edge

Step 4: Test Interactions (If Needed)

If user wants to verify behavior (hover, click, etc):

# Hover to check hover state
Hover  button:has-text("Login")

# Click to check if it works
Click  button:has-text("Login")

Each command returns a new screenshot showing the result.

Step 5: Responsive Checks (If Needed)

If user mentions "mobile" or "tablet":

# Set mobile viewport
Set Viewport Size  375  667  # iPhone SE

# Capture mobile view
Go To  <url>

Common viewports:

  • Mobile: 375x667 (iPhone SE), 390x844 (iPhone 12)
  • Tablet: 768x1024 (iPad)
  • Desktop: 1920x1080

Guidelines

Be Specific:

  • ❌ "Button looks fine"
  • ✅ "Button is blue (#2563EB), 140px wide, centered in navbar"

Compare to Expectation:

  • User said: "check if button is blue"
  • You say: "Button is blue (#0066FF) ✓" or "Button is red (#DC2626) ✗ - Expected blue"

Mention Issues:

  • Overlapping elements
  • Cut-off text
  • Misaligned items
  • Wrong colors
  • Missing elements
  • Console errors

Keep It Casual:

  • This is quick verification, not formal QA
  • No need to create artifacts or tests
  • Just answer: "Does it look right? Yes/No, here's why"

Interactive Commands Available

# Navigation
Go To  <url>

# Element inspection
Get Element  <selector>  # Check if exists
Get Style  <selector>  <property>  # Get CSS value

# Viewport changes
Set Viewport Size  <width>  <height>

# Interactions (to check behavior)
Click  <selector>
Hover  <selector>
Fill Text  <selector>  <text>
Press Keys  <selector>  <key>

# Scrolling
Scroll To Element  <selector>
Scroll By  <x>  <y>

# Waiting (if page loads slowly)
Wait For Elements State  <selector>  visible  timeout=5000

Every command automatically returns a screenshot.

Example Workflows

Example 1: Quick Button Check

User: "Check if the Sign Up button is green"

1. Go To  https://example.com
2. Look at screenshot
3. Find "Sign Up" button in nav
4. Respond: "Sign Up button is green (#10B981) ✓"

Example 2: Layout Verification

User: "Is the pricing table centered?"

1. Go To  https://example.com/pricing
2. Look at screenshot
3. Check table position
4. Respond: "Pricing table is centered ✓ - Equal margins on both sides (~200px)"

Example 3: Responsive Check

User: "Does the header look good on mobile?"

1. Set Viewport Size  375  667
2. Go To  https://example.com
3. Look at screenshot
4. Respond: "Header stacks vertically ✓ - Logo on top, menu items below. Hamburger icon visible."

Example 4: Hover State

User: "Check if the button changes color on hover"

1. Go To  https://example.com
2. Look at initial screenshot (button default state)
3. Hover  button:has-text("Submit")
4. Look at new screenshot (button hover state)
5. Respond: "Button changes from blue (#0066FF) to dark blue (#0047B3) on hover ✓"

Example 5: Modal Appearance

User: "See if the modal shows up when clicking Contact"

1. Go To  https://example.com
2. Click  button:has-text("Contact")
3. Look at screenshot
4. Respond: "Modal appears ✓ - Dark overlay with white form centered on screen"

Output Format

Always structure your response:

[Element/Feature Name]

Status: ✓ Matches expectation / ✗ Issue found / ⚠️ Partial match

Details:
- [Specific observation 1]
- [Specific observation 2]
- [Specific observation 3]

[If issue found:]
Issue: [What's wrong]
Expected: [What it should be]
Actual: [What it is]

When to Stop

This skill is for QUICK checks. If user asks:

  • "Test this thoroughly" → Use /helpmetest
  • "Create a test for this" → Use /helpmetest-test-generator
  • "Why is this test failing?" → Use /helpmetest-debugger

Keep it simple: Look, describe, confirm.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

35.45%
按下载量换算69

Claude

33.9%
按下载量换算66

Cursor

19.13%
按下载量换算37

Gemini CLI

8.96%
按下载量换算18

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills