mcp-playwright-tools
Playwright browser automation tools for MCP enabling navigation, interaction, content extraction, and semantic locators.
README
mcp-playwright-tools
Playwright browser automation tools for MCP — navigation, interaction, content extraction, and semantic locators
Built on mcp-server-framework.
Installation
pip install mcp-playwright-tools
Usage
mcp-playwright-tools # stdio (default)
Claude Code / Claude Desktop
{
"mcpServers": {
"playwright": { "command": "mcp-playwright-tools" }
}
}
Tools
| Tool | Description |
|---|---|
navigate |
Navigate to a URL and wait for page load |
current_url |
Get the current page URL |
go_back |
Navigate back in browser history |
go_forward |
Navigate forward in browser history |
reload |
Reload the current page |
get_title |
Get the current page title |
get_text |
Get visible text of an element |
get_all_texts |
Get visible text of all matching elements |
get_page_content |
Get all visible text on the page |
get_html |
Get inner HTML of an element |
get_links |
Get all links on the page (text + href) |
get_attribute |
Get an attribute value from an element |
click |
Click an element |
fill |
Fill an input field with text |
type_text |
Type text character by character (for autocomplete etc.) |
press |
Press a keyboard key (Enter, Tab, Escape, etc.) |
select_option |
Select a dropdown option by value |
select_option_by_text |
Select a dropdown option by visible text |
check |
Check a checkbox |
uncheck |
Uncheck a checkbox |
hover |
Hover over an element |
focus |
Focus an element |
clear |
Clear an input field |
screenshot |
Take a screenshot of the page |
screenshot_element |
Take a screenshot of a specific element |
wait_for |
Wait for an element to appear |
wait_for_hidden |
Wait for an element to disappear |
wait_for_url |
Wait for the URL to match a pattern |
set_viewport |
Set the browser viewport size |
scroll_to |
Scroll an element into view |
scroll_page |
Scroll the page up or down |
find_by_role |
Find element by ARIA role (button, link, textbox, heading, etc.) |
find_by_text |
Find element by visible text |
find_by_label |
Find form element by its label text |
find_by_placeholder |
Find input element by placeholder text |
find_by_test_id |
Find element by data-testid attribute |
click_by_role |
Click element found by ARIA role |
click_by_text |
Click element found by visible text |
fill_by_label |
Fill input found by its label text |
describe_element |
Get ARIA role, text, and attributes of an element |
find_interactive_elements |
List all clickable and fillable elements on the page |
License
MIT — Part of uc-mcp-tools
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。