911Proxy Universal Web Scraper MCP Server - supports HTML extraction and screenshots
npm install 911-scraper-mcp~/Library/Application Support/Claude/claude_desktop_config.json
%APPDATA%/Claude/claude_desktop_config.json
json
{
"mcpServers": {
"911-scraper": {
"command": "npx",
"args": ["-y", "911-scraper-mcp"],
"env": {
"SCRAPER_USERNAME": "your-api-username",
"SCRAPER_PASSWORD": "your-api-password"
}
}
}
}
`
#### Cherry Studio
In Settings ā MCP Servers, add the same configuration as above.
---
$3
If you use NVM (Node Version Manager), MCP clients may not be able to find npx automatically. You'll need to specify the full path to npx in your configuration.
#### Step 1: Find Your Node and NPX Paths
Run the following commands to locate your NVM installation:
Windows (Command Prompt or PowerShell):
`cmd
where node
where npx
`
macOS:
`bash
which node
which npx
`
Linux:
`bash
which node
which npx
`
#### Step 2: Note the Output
You should see output similar to:
Windows:
`
C:\Users\Administrator\AppData\Roaming\nvm\v22.21.1\node.exe
C:\Users\Administrator\AppData\Roaming\nvm\v22.21.1\npx.cmd
`
macOS:
`
/Users/yourname/.nvm/versions/node/v22.21.1/bin/node
/Users/yourname/.nvm/versions/node/v22.21.1/bin/npx
`
Linux:
`
/home/yourname/.nvm/versions/node/v22.21.1/bin/node
/home/yourname/.nvm/versions/node/v22.21.1/bin/npx
`
#### Step 3: Update Your MCP Configuration
Use the full path to npx from Step 2 in your configuration file:
Windows Configuration:
`json
{
"mcpServers": {
"911-scraper": {
"command": "C:\\Users\\Administrator\\AppData\\Roaming\\nvm\\v22.21.1\\npx.cmd",
"args": ["-y", "911-scraper-mcp"],
"env": {
"SCRAPER_USERNAME": "your-api-username",
"SCRAPER_PASSWORD": "your-api-password",
"PATH": "C:\\Users\\Administrator\\AppData\\Roaming\\nvm\\v22.21.1;%PATH%"
}
}
}
}
`
macOS Configuration:
`json
{
"mcpServers": {
"911-scraper": {
"command": "/Users/yourname/.nvm/versions/node/v22.21.1/bin/npx",
"args": ["-y", "911-scraper-mcp"],
"env": {
"SCRAPER_USERNAME": "your-api-username",
"SCRAPER_PASSWORD": "your-api-password",
"PATH": "/Users/yourname/.nvm/versions/node/v22.21.1/bin:${PATH}"
}
}
}
}
`
Linux Configuration:
`json
{
"mcpServers": {
"911-scraper": {
"command": "/home/yourname/.nvm/versions/node/v22.21.1/bin/npx",
"args": ["-y", "911-scraper-mcp"],
"env": {
"SCRAPER_USERNAME": "your-api-username",
"SCRAPER_PASSWORD": "your-api-password",
"PATH": "/home/yourname/.nvm/versions/node/v22.21.1/bin:${PATH}"
}
}
}
}
`
Important Notes:
- Replace yourname with your actual username
- Replace v22.21.1 with your actual Node.js version from Step 2
- On Windows, use double backslashes (\\) in paths
- On macOS/Linux, use forward slashes (/) in paths
- Don't forget to replace your-api-username and your-api-password with your actual 911Proxy credentials
$3
After configuration, restart Claude Desktop or Cherry Studio.
š¬ Usage Examples
Once configured, simply send instructions to your AI:
$3
`
Help me scrape https://www.example.com and summarize the content
`
$3
`
Take a screenshot of https://www.google.com homepage
`
$3
`
Access https://www.google.com using Japan proxy and take a screenshot
`
$3
`
Scrape https://www.sojson.com with JavaScript rendering enabled
`
$3
`
Help me batch scrape these websites:
1. https://www.example1.com
2. https://www.example2.com
3. https://www.example3.com
`
š¼ļø Screenshot Location
Screenshots are automatically saved to:
- Windows: C:\Users\YourName\Downloads\scraper-screenshots\
- macOS: /Users/YourName/Downloads/scraper-screenshots/
- Linux: /home/yourname/Downloads/scraper-screenshots/
AI can help you open the screenshots folder:
`
Open screenshots folder
`
āļø Advanced Configuration
$3
Add to env in config file:
`json
{
"env": {
"SCRAPER_USERNAME": "your-username",
"SCRAPER_PASSWORD": "your-password",
"SCREENSHOTS_DIR": "/custom/path/screenshots"
}
}
`
$3
Common codes: US (United States), HK (Hong Kong), JP (Japan), GB (United Kingdom), DE (Germany), FR (France), SG (Singapore)
Note: CN (Mainland China) will be automatically mapped to HK (Hong Kong)
$3
Common codes: en-US (English), zh-CN (Chinese Simplified), ja-JP (Japanese), ko-KR (Korean), es-ES (Spanish)
ā FAQ
$3
A: 911Proxy API can only return one format per request. AI will automatically make two separate requests: first for HTML, then for screenshot.
$3
A: Screenshots are saved in the scraper-screenshots subfolder in your Downloads folder. Tell AI "open screenshots folder" to view them.
$3
A: You need to set both region and language. For example:
`
Access Google using Japan proxy and Japanese language
`
$3
A:
- Recommend scraping no more than 3 websites at once
- Or make requests in batches
- Avoid using batch scraping with JavaScript rendering
$3
A: Send to AI:
`
Help me test the web scraping feature
`
If the tool list includes scrape_url, configuration is successful.
š§ Troubleshooting
$3
1. Check JSON format is correct (use JSON validator)
2. Verify environment variable names are spelled correctly
3. Restart AI client
4. Check client logs
$3
Ensure:
- Username and password are correct
- API sub-account type is Scraper API (product_type: 18)
- Account has sufficient balance
$3
Tell AI:
`
Where are screenshots saved?
``