Hyperbrowser MCP
Hyperbrowser MCP is a server for scraping, extracting structured data, and browsing webpages using AI agents. It supports tools like scrape_webpage, crawl_webpages, and various browser automation agents.
How to Install and Use Hyperbrowser MCP
Hyperbrowser MCP is a great tool that helps you scrape, extract, and crawl web content easily. It also works with different browser agents like OpenAI and Anthropic Claude. Here’s a simple guide to get it installed and running with some clear code examples.
Manual Installation
First, you can install Hyperbrowser MCP on your computer manually. To do this, you just need to run one simple command in your terminal. This will install the MCP server and link it with your Hyperbrowser API key.
Here is the command you use:
npx hyperbrowser-mcp <YOUR-HYPERBROWSER-API-KEY>
Replace <YOUR-HYPERBROWSER-API-KEY> with your actual API key from Hyperbrowser. This command will start the MCP server using your key.
Running on Cursor
If you use Cursor, the tool supports easy setup by adding a configuration file. You need to add the MCP server details to a file called ~/.cursor/mcp.json. This file should look like this:
{
"mcpServers": {
"hyperbrowser": {
"command": "npx",
"args": ["-y", "hyperbrowser-mcp"],
"env": {
"HYPERBROWSER_API_KEY": "YOUR-API-KEY"
}
}
}
}
Make sure to replace "YOUR-API-KEY" with your actual Hyperbrowser API key. This configuration will let Cursor use the Hyperbrowser MCP automatically.
Running on Windsurf
Similarly, if you use Windsurf, you can run Hyperbrowser MCP by adding it inside your Windsurf model configuration. You add it to the file ./codeium/windsurf/model_config.json like this:
{
"mcpServers": {
"hyperbrowser": {
"command": "npx",
"args": ["-y", "hyperbrowser-mcp"],
"env": {
"HYPERBROWSER_API_KEY": "YOUR-API-KEY"
}
}
}
}
This makes sure Windsurf can start and use the MCP server with your API key.
Development Setup
If you want to change or develop Hyperbrowser MCP yourself, you can run it from the source code. Here’s what to do step by step:
- Clone the MCP GitHub repository:
git clone [email protected]:hyperbrowserai/mcp.git hyperbrowser-mcp
cd hyperbrowser-mcp
- Install the required software packages:
npm install
npm run build
- Start the server by running:
node dist/server.js
This lets you work directly with the MCP code and test your changes.
Example Setup for Claude Desktop App
If you want to use Hyperbrowser MCP with the Claude Desktop client, here’s an example JSON configuration for it:
{
"mcpServers": {
"hyperbrowser": {
"command": "npx",
"args": ["--yes", "hyperbrowser-mcp"],
"env": {
"HYPERBROWSER_API_KEY": "your-api-key"
}
}
}
}
Just replace "your-api-key" with your real API key to get started.
With these easy instructions and code examples, you can install and run Hyperbrowser MCP quickly on your system or inside different applications. Just remember to add your API key where needed, and you will be ready to scrape and crawl web pages in no time!