mcp-sanctions

What is mcp-sanctions?

mcp-sanctions is a Sanctions MCP Server designed to screen individuals and organizations against major global sanctions lists such as OFAC SDN, UN, OFSI, and more. It integrates with Claude through the Model Context Protocol, offering a robust solution for compliance and risk assessment.

How to Use mcp-sanctions

To use mcp-sanctions, you need Node.js v20 or higher, npm/yarn, and git. First, clone the repository, then install dependencies using npm install. Finally, configure your claude_desktop_config.json file by adding the mcpServers configuration, ensuring to update paths and add your OFAC API key. After configuration, restart Claude Desktop to enable the sanctions screening capability.

Key Features of mcp-sanctions

• Sanctions Screening: Screens against major global sanctions lists (OFAC SDN, UN, OFSI, etc.).
• Customizable Risk Threshold: Allows setting minimum risk scores for match relevance.
• Comprehensive Entity Support: Screens both individuals and organizations with detailed information.
• Multiple Identification Methods: Supports various identifiers like names, addresses, and documents.
• Source Selection: Enables choosing specific sanctions lists for compliance needs.

Use Cases of mcp-sanctions

• Screening an Individual: Ask Claude to check if a person (e.g., "John Smith born on 1980-01-15 from Syria") is on any sanctions lists.
• Screening an Organization: Ask Claude to check if an organization (e.g., "Tech Solutions Ltd based in Tehran, Iran") is a sanctioned entity.
• Custom Screening Parameters: Specify custom parameters for targeted screening, such as checking a person on the OFAC SDN list with a minimum match score of 90.

FAQ about mcp-sanctions

Q: What parameters does the Sanctions MCP Server process? A: The server processes screening requests with parameters such as cases (array of entities), type ("person" or "organization"), minScore (minimum score threshold, default: 95), and sources (sanctions lists to check against).

Q: What is the input schema for the server? A: The server accepts detailed entity information including Name, Date of birth, Gender, Citizenship/Nationality, Address details, Contact information, and Identification documents.

Q: What is the response format from the server? A: The server returns a JSON response with match details, match scores, source information, and entity profiles.

Sanctions MCP Server

A powerful sanctions screening API MCP Server, leveraging OFAC API to provide comprehensive risk assessments for individuals and organizations.

🎉 Ready for integration with Claude through the Model Context Protocol 🎉

https://github.com/user-attachments/assets/41ea6318-5b17-4da1-8d6d-8cb60b5342ea

Key Features

• Sanctions Screening: Screen individuals and organizations against major global sanctions lists including OFAC SDN, UN, OFSI, and more
• Customizable Risk Threshold: Set minimum risk scores to determine match relevance
• Comprehensive Entity Support: Screen both individuals and organizations with detailed information
• Multiple Identification Methods: Support for various identifiers including names, addresses, identification documents, and more
• Source Selection: Choose which sanctions lists to check against based on your compliance needs

How to Use

You can run the Sanctions MCP Server using Node.js directly.

Requirements

1. Node.js v20 or higher
2. npm or yarn for dependency management
3. git

Installation

1. Clone the repository

git clone https://github.com/yourusername/mcp-ofac
cd mcp-ofac

2. Install dependencies

npm install

3. Configure claude_desktop_config.json

Add the following configuration to your claude_desktop_config.json file. Make sure to:

• Update the paths to match your local Node.js installation and repository location
• Add your OFAC API key if required

{
  "mcpServers": {
    "sanctions": {
      "command": "/path/to/your/node",
      "args": ["/path/to/your/mcp-ofac/index.js"],
      "env": {
        "OFAC_API_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

Usage Examples

Screening an Individual

You can ask Claude to screen individuals against sanctions lists:

"Can you check if John Smith born on 1980-01-15 from Syria is on any sanctions lists?"

Screening an Organization

You can ask Claude to check organizations:

"Is there a sanctioned entity called Tech Solutions Ltd based in Tehran, Iran?"

Custom Screening Parameters

You can specify custom parameters for more targeted screening:

"Check if Maria Rodriguez with passport number AB123456 from Venezuela is on the OFAC SDN list with a minimum match score of 90."

How the Tool Works

The Sanctions MCP Server processes screening requests with the following parameters:

• cases: Array of entities to screen (individuals or organizations) with identifying information
• type: Entity type ("person" or "organization")
• minScore: Minimum score threshold for matches (0-100, default: 95)
• sources: Which sanctions lists to check against (SDN, NONSDN, DPL, UN, OFSI, etc.)

The tool returns detailed match information, including sanctioned entity details, match scores, and source information.

Data Format

Input Schema

The server accepts detailed entity information including:

• Name
• Date of birth (for individuals)
• Gender (for individuals)
• Citizenship/Nationality
• Address details
• Contact information
• Identification documents

Response Format

The server returns a JSON response with:

• Match details
• Match scores
• Source information
• Entity profiles

Restart Claude Desktop

After configuring your claude_desktop_config.json, restart Claude Desktop to apply the changes and enable the sanctions screening capability.