Browser Use
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
作者提供的说明(英文原文)
MCP server for Safari DevTools — browser debugging and automation for AI coding agents
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"safari-devtools": {
"command": "npx",
"args": [
"-y",
"safari-devtools-mcp"
]
}
}
}在终端中运行这条命令
claude mcp add safari-devtools -- npx -y safari-devtools-mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"safari-devtools": {
"command": "npx",
"args": [
"-y",
"safari-devtools-mcp"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"safari-devtools": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"safari-devtools-mcp"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 HayoDev 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(1.8.1)于7个月前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。
同一分类中有:Browser Use, Chrome DevTools MCP, MCP Server Browser
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Browser agents and stealth cloud browsers for AI: run web tasks, fill forms, extract data.
MCP server for Chrome DevTools
MCP server for browser use access
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
safari-devtools-mcp lets your coding agent (such as Claude, Cursor, Copilot or Gemini) control and inspect a live Safari browser on macOS. It acts as a Model-Context-Protocol (MCP) server, giving your AI coding assistant access to Safari DevTools for debugging, automation, and testing.
Chrome developers get powerful AI debugging through chrome-devtools-mcp by Google. Safari developers should have something equivalent — this project brings that same experience to Safari, with matching tool names and parameter schemas so you can swap between browsers with minimal friction.
Note: This server exposes browser content (page data, console logs, network traffic) to MCP clients. Avoid browsing sensitive websites or entering credentials while a session is active.
This project uses WebDriver for capabilities that scripting alone cannot provide:
sudo safaridriver --enable
Standard MCP config:
{
"mcpServers": { "safari-devtools": { "command": "npx", "args": ["-y", "safari-devtools-mcp@latest"] } } }
Claude Code
claude mcp add safari-devtools -- npx -y safari-devtools-mcp@latest
Or add to your project's .mcp.json using the standard config above.
Claude Desktop
Add the standard config to ~/Library/Application Support/Claude/claude_desktop_config.json.
Cursor
Add the standard config to your Cursor MCP settings.
VS Code
Add the standard config to .vscode/mcp.json.
Copilot CLI
copilot mcp add safari-devtools -- npx -y safari-devtools-mcp@latest
Gemini CLI
gemini mcp add safari-devtools -- npx -y safari-devtools-mcp@latest
Or add the standard config to your ~/.gemini/settings.json.
Gemini Code Assist
Add the standard config to your .gemini/settings.json in the project root.
JetBrains AI Assistant & Junie
Add the standard config to your .junie/mcp.json in the project root.
Raycast
Open "Install MCP Server" in Raycast and fill in:
npx-y safari-devtools-mcp@latestOr copy the standard config JSON above before opening the command — Raycast will auto-fill the form.
Visual Studio
Add the standard config to your .vs/mcp.json in the solution root.
Warp
Add the standard config to your Warp MCP settings file at ~/.warp/mcp.json.
From source
git clone https://github.com/HayoDev/safari-devtools-mcp.git
cd safari-devtools-mcp npm install && npm run build
Then point your MCP client to the built entry point:
{
"mcpServers": { "safari-devtools": { "command": "node", "args": [ "/path/to/safari-devtools-mcp/build/src/bin/safari-devtools-mcp.js" ] } } }
Navigate to https://example.com, take a snapshot, and list any console errors.
Use --slim to reduce token usage with shorter tool descriptions. Useful when working with context-limited models or long conversations:
{
"mcpServers": { "safari-devtools": { "command": "npx", "args": ["-y", "safari-devtools-mcp@latest", "--slim"] } } }
The server exposes guided debugging workflows as MCP prompts. Clients that support prompts (e.g. Claude Desktop, Claude Code) can invoke them by name:
| Prompt | Description |
|---|---|
safari-devtools | Getting started — tool overview, recommended workflow, and Safari-specific notes |
a11y-debugging | Accessibility audit — a11y tree inspection, axe-core injection, WCAG checks |
safari-specific-debugging | Debug WebKit quirks — CSS prefixes, JS feature gaps, ITP/CORS issues |
performance-debugging | Performance analysis — Navigation Timing, Core Web Vitals, resource waterfall |
| Tool | Description |
|---|---|
list_console_messages | List console messages with filtering by level (log, warn, error) |
get_console_message | Get a detailed message including stack trace and arguments |
clear_console | Clear all captured console messages |
list_network_requests | Monitor network requests — Fetch, XHR, and resource loads |
get_network_request | Get full request/response details with headers and body |
clear_network | Clear all captured network requests |
evaluate_script | Execute JavaScript in the browser context and return results |
take_screenshot | Capture a PNG screenshot of the page or a specific element |
take_snapshot | Accessibility-tree snapshot of the DOM with stable UIDs for element targeting |
| Tool | Description |
|---|---|
get_page_content | Get the page title, URL, and visible text content |
get_html_source | Get the full HTML source of the page |
extract_links | Extract all links with their text, href, and rel attributes |
extract_meta | Extract meta tags (og:, twitter:, description, etc.) |
| Tool | Description |
|---|---|
list_pages | List all open Safari tabs across windows |
select_page | Switch to a specific tab |
new_page | Open a new tab and navigate to a URL |
close_page | Close a tab |
navigate_page | Navigate to a URL, go back, forward, or reload |
wait_for | Wait for specific text to appear on the page |
resize_page | Resize the browser window |
handle_dialog | Accept or dismiss browser dialogs (alert, confirm, prompt) |
| Tool | Description |
|---|---|
scroll | Scroll the page in any direction by a given amount |
scroll_to_element | Scroll an element into view by its UID |
| Tool | Description |
|---|---|
get_computed_style | Get computed CSS styles for any element by UID |
| Tool | Description |
|---|---|
get_cookies | Get browser cookies, optionally filtered by name or domain |
set_cookie | Set a cookie with name, value, and optional attributes |
delete_cookie | Delete a cookie by name, or delete all cookies |
get_storage | Read from localStorage or sessionStorage |
set_storage | Write a key-value pair to localStorage or sessionStorage |
delete_storage | Delete a key or clear all entries from localStorage/sessionStorage |
| Tool | Description |
|---|---|
click | Click an element by UID from a snapshot |
click_at | Click at specific x/y coordinates |
right_click | Right-click (context menu) on an element |
select_option | Select an option from a dropdown by value or label |
hover | Hover over an element |
fill | Type into an input field or select from a dropdown |
fill_form | Fill multiple form fields at once |
type_text | Type text into the currently focused element |
drag | Drag and drop between elements or coordinates |
press_key | Press a key or combination (e.g., Meta+A, Enter) |
upload_file | Upload a file through a file input |
| Tool | Description |
|---|---|
inspect_viewport_meta | Parse the viewport meta tag and validate against iOS best practices (width, zoom, viewport-fit) |
| get_safe_area_insets | Read CSS safe-area-inset values and check whether t
只显示了开头部分;其余内容请在仓库中阅读