Clone the official MCP Real-Time API server to your local machine. This downloads the server script, configuration templates, and dependency manifest.
git clone https://github.com/websitecategorization/mcp-realtime-api.git cd mcp-realtime-api
Install the required Python packages: mcp (Model Context Protocol SDK), httpx (async HTTP client), and python-dotenv (environment variable management).
pip install -r requirements.txt
mcp>=1.0.0, httpx>=0.27.0, and python-dotenv>=1.0.0. You can also install them individually: pip install mcp httpx python-dotenv
Create a .env file in the project directory with your API key. You can find your key on your profile page after signing up.
# Required: Your API key from websitecategorizationapi.com WCA_API_KEY=your-api-key-here
.env.example file as a starting point: cp .env.example .env and then edit it with your actual API key.Add the server configuration to your preferred MCP client. Select your client below and follow the instructions.
{
"mcpServers": {
"website-categorization-api": {
"command": "python",
"args": ["/full/path/to/mcp-realtime-api/wca_mcp_realtime.py"],
"env": {
"WCA_API_KEY": "your-api-key-here"
}
}
}
}
~/Library/Application Support/Claude/claude_desktop_config.json%APPDATA%\Claude\claude_desktop_config.json~/.config/Claude/claude_desktop_config.json
claude mcp add website-categorization-api \ -- python /full/path/to/mcp-realtime-api/wca_mcp_realtime.py
export WCA_API_KEY="your-api-key-here".env file in the cloned repository directory.
{
"mcpServers": {
"website-categorization-api": {
"command": "python",
"args": ["/full/path/to/mcp-realtime-api/wca_mcp_realtime.py"],
"env": {
"WCA_API_KEY": "your-api-key-here"
}
}
}
}
.cursor/mcp.json file in your project root directory. Cursor will automatically detect and load the MCP server when you open the project.
Restart your MCP client, then try these example prompts to verify everything is working.
"Categorize stripe.com and tell me what industry it belongs to"
"What is nytimes.com/politics about? Give me the IAB and IPTC categories"
"Batch categorize apple.com, shopify.com, and github.com"
"Check my remaining API credits"
"Categorize docs.google.com/spreadsheets with expanded categories and confidence scores"
The MCP server exposes four tools that your AI assistant can call automatically based on your natural language prompts.
Every URL classification returns categories across all seven supported taxonomy standards simultaneously.
| Taxonomy | Categories | Depth | Use Case |
|---|---|---|---|
| IAB v2 | 698 | 4 Tiers | Programmatic advertising, content classification |
| IAB v3 | 703 | 4 Tiers | Latest IAB standard for digital advertising |
| IPTC NewsCodes | 1,124 | 3 Tiers | News and media content classification |
| Google Shopping | 5,474 | Multi-level | Product categorization for e-commerce |
| Shopify | 10,560 | Multi-level | Shopify product taxonomy mapping |
| Amazon | 39,004 | Multi-level | Amazon product and browse node classification |
| Web Content Filtering | 44 | 1 Tier | Brand safety, content filtering, parental controls |
Beyond categorization, each API response includes rich metadata signals for deeper analysis.
Identify the target audience and buyer personas associated with the content.
Detect the overall sentiment (positive, negative, neutral) of the page content.
Extract the main topics and themes discussed on the page.
Identify the most relevant keywords and phrases from the content.
Extract named entities such as companies, people, locations, and products.
Detect competing brands, products, and services mentioned on the page.
Identify the technologies, frameworks, and platforms powering the website.
Flag URLs associated with malware distribution or phishing attempts.
Common issues and solutions when setting up the MCP Real-Time API server.
WCA_API_KEY environment variable. You can either:.env file in the mcp-realtime-api directory with WCA_API_KEY=your-key"env" section of your MCP client configurationexport WCA_API_KEY="your-key"
wca_mcp_realtime.py is an absolute path (starts with / on macOS/Linux or C:\ on Windows).python is in your system PATH. Try using python3 instead of python in the config.pip install -r requirements.txt in the mcp-realtime-api directory. If you have multiple Python versions, make sure you install to the same Python that your MCP client is using. Try pip3 install -r requirements.txt or python -m pip install -r requirements.txt.
.env file. If your credits are exhausted, you can purchase more at the pricing page.
batch_categorize tool has a limit of 10 URLs per request to prevent excessive API usage. Split larger batches into groups of 10 or fewer. Each URL in the batch counts as one API credit.
Our team is here to help you get started with the MCP Real-Time API server. Reach out through any of the channels below.