Web Browser MCP Server

@MCP-Mirror

リポジトリ

基本情報

カテゴリ

ブラウザ自動化

トランスポート

stdio

作者

MCP-Mirror

Web Browser MCP Server について

Mirror of

#web-browser#mcp-server #ai-integration

Web Browser MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "blazickjp_web-browser-mcp-server": {
      "command": "uv",
      "args": [
        "pip",
        "install",
        "web-browser-mcp-server"
      ]
    }
  }
}

Web Browser MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is 🌐 Web Browser MCP Server?

This MCP server gives AI applications the ability to browse the web, read pages, and extract structured content using CSS selectors. It is built for developers who want their AI tools to understand live web pages.

How to use 🌐 Web Browser MCP Server?

Install with pip install web-browser-mcp-server or uv pip install web-browser-mcp-server. Then integrate with Claude Desktop by adding the provided JSON configuration to your MCP settings. To browse a page, call the browse_webpage tool with a URL and optional CSS selectors.

Key features of 🌐 Web Browser MCP Server

Smart content extraction with CSS selectors
Async processing for fast page loading
Rich metadata including titles, links, and structure
Built-in error handling and timeout management
Cross‑platform (runs anywhere Python does)

Use cases of 🌐 Web Browser MCP Server

Fetching a full webpage for analysis
Extracting specific elements like headlines or navigation links
Gathering structured data from news or documentation sites
Letting an AI assistant read and reference live web content

よくある質問

What does this MCP server do?

It enables AI agents to browse web pages, extract content, and retrieve structured data using CSS selectors, making web information accessible to the AI.

How do I install it?

Use `pip install web-browser-mcp-server` or the recommended `uv pip install web-browser-mcp-server`.

What environment variables can I configure?

You can set `REQUEST_TIMEOUT` (default 30s), `USER_AGENT` (default modern Chrome UA), `LOG_LEVEL` (default "info"), and `MAX_RETRIES` (default 3).

What are the prerequisites?

Python 3.8 or higher. No other runtime or platform restrictions.

How do I use it with Claude Desktop?

Add the provided JSON configuration block to your Claude Desktop config file, pointing the `args` to the installed package directory. Then the tool becomes available in your Claude conversations.