sec-mcp 0.1.3

Python toolkit providing security checks for domains, URLs, IPs, and more.

sec-mcp: Security Checking Toolkit

A Python toolkit providing security checks for domains, URLs, IPs, and more. Integrate easily into any Python application, use via terminal CLI, or run as an MCP server to enrich LLM context with real-time threat insights.

Features

• Comprehensive security checks for domains, URLs, IP addresses, and more against multiple blacklist feeds
• On-demand updates from OpenPhish, PhishStats, URLhaus and custom sources
• High-performance, thread-safe SQLite storage with in-memory caching for fast lookups
• Python API via SecMCP class for easy integration into your applications
• Intuitive Click-based CLI for interactive single or batch scans
• Built-in MCP server support for LLM/AI integrations over JSON/STDIO

Installation

pip install sec-mcp

Usage via CLI

1. Install the package:

   pip install sec-mcp
   

2. Check a single URL/domain/IP:

   sec-mcp check https://example.com
   

3. Batch check from a file:

   sec-mcp batch urls.txt
   

4. View blacklist status:

   sec-mcp status
   

5. Manually trigger an update:

   sec-mcp update
   

Usage via API (Python)

1. Install in your project:

   pip install sec-mcp
   

2. Import and initialize:

   from sec_mcp import SecMCP
   
   client = SecMCP()
   

3. Single check:

   result = client.check("https://example.com")
   print(result.to_json())
   

4. Batch check:

   urls = ["https://example.com", "https://test.com"]
   results = client.check_batch(urls)
   for r in results:
       print(r.to_json())
   

5. Get status and update:

   status = client.get_status()
   print(status.to_json())
   
   client.update()
   

Usage via MCP Client

To run sec-mcp as an MCP server for AI-driven clients (e.g., Claude):

1. Install in editable mode (for development):

   pip install -e .
   

2. Start the MCP server:

   sec-mcp-server
   

3. Configure your MCP client (e.g., Claude) to point at the command:

   {
     "mcpServers": {
       "sec-mcp": {
         "command": ".venv/bin/python3",
         "args": ["-m", "sec_mcp.start_server"]
       }
     }
   }
   

Clients will then use the built-in check_blacklist tool over JSON/STDIO for real-time security checks.

Configuration

The client can be configured via config.json:

• blacklist_sources: URLs for blacklist feeds
• update_time: Daily update schedule (default: "00:00")
• cache_size: In-memory cache size (default: 10000)
• log_level: Logging verbosity (default: "INFO")

Configuring sec-mcp with Claude (MCP Client)

To use your MCP Server for security checking (sec-mcp) with an MCP client such as Claude, add it to your Claude configuration as follows:

{
  "mcpServers": {
    "sec-mcp": {
      "command": ".venv/bin/python3",
      "args": ["-m", "sec_mcp.start_server"]
    }
  }
}

• Ensure you have installed all dependencies in your virtual environment (.venv).
• The command should point to your Python executable inside .venv for best isolation.
• The args array should launch your MCP server using the provided script.
• You can add other MCP servers in the same configuration if needed.

This setup allows Claude (or any compatible MCP client) to connect to your sec-mcp server and use its check_blacklist tool for real-time security checks on URLs, domains, or IP addresses.

For more details and advanced configuration, see the Model Context Protocol examples.

Development

Clone the repository and install in development mode:

git clone <repository-url>
cd sec-mcp
pip install -e .

License

MIT