Getting Started
Get the NAICS MCP Server up and running in minutes.
Prerequisites
- Python 3.11 or higher
- An MCP-compatible client (e.g., Claude Desktop, Cursor)
Installation
Option 1: Docker (Recommended)
docker pull ghcr.io/mfbaig35r/naics-mcp-server:latest
docker run -p 9090:9090 ghcr.io/mfbaig35r/naics-mcp-server:latestOption 2: pip
pip install naics-mcp-server
# Initialize the database
naics-mcp init
# Generate embeddings (takes a few minutes)
naics-mcp embeddings
# Run the server
naics-mcp serveOption 3: From Source
git clone https://github.com/mfbaig35r/naics-mcp-server.git
cd naics-mcp-server
pip install -e .
naics-mcp init
naics-mcp embeddings
naics-mcp serveConfigure Your MCP Client
Claude Desktop
Add to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):
{
"mcpServers": {
"naics": {
"command": "naics-mcp",
"args": ["serve"]
}
}
}Or with Docker:
{
"mcpServers": {
"naics": {
"command": "docker",
"args": ["run", "-i", "--rm", "ghcr.io/mfbaig35r/naics-mcp-server:latest"]
}
}
}Cursor
Add to your Cursor MCP settings:
{
"naics": {
"command": "naics-mcp",
"args": ["serve"]
}
}Verify Installation
Once configured, you can verify the server is working by asking your AI assistant:
“Check the NAICS server health”
The assistant should use the get_server_health tool and return component status.
Next Steps
- Human Guide - Learn how to classify businesses
- LLM Guide - System prompts and best practices for AI usage
- API Reference - Full tool documentation