otx-mcp

A Model Context Protocol (MCP) server for AlienVault OTX (Open Threat Exchange) that provides a comprehensive interface to the OTX API.

This MCP server allows you to interact with the AlienVault OTX platform through the Model Context Protocol. The Model Context Protocol (MCP) is an innovative standard that enables applications to provide context and functionality to Large Language Models (LLMs) in a secure, standardized way. Think of it like a web API specifically designed for LLM interactions.

MCP servers can:

• Expose data through Resources (used to load information into the LLM's context)
• Provide functionality through Tools (used to execute code or produce side effects)
• Define interaction patterns through Prompts (reusable templates for LLM interactions)

This server implements the Tools functionality of MCP, providing a comprehensive set of tools for interacting with the OTX platform. It allows AI systems like Claude to retrieve and analyze threat indicators and Pulses in real-time.

• Indicator Search and Analysis: Search for indicators, get detailed information about specific indicators, and validate indicators
• Pulse Management: Create, edit, and manage threat intelligence pulses
• User Interaction: Follow users, subscribe to pulses, and manage your OTX network
• URL Analysis: Submit URLs for analysis to identify potential threats
• Event Monitoring: Track recent events and activities on OTX

1. Export your OTX API key as an environment variable:

   export OTX_API_KEY=your_api_key_here
   

2. Authenticate with GitHub Container Registry:

   # Create a GitHub Personal Access Token (PAT) with at least 'read:packages' scope
   # Go to GitHub → Settings → Developer settings → Personal access tokens → Tokens (classic)
   # Generate a new token with 'read:packages' scope
   
   # Login to GitHub Container Registry
   docker login ghcr.io -u YOUR_GITHUB_USERNAME
   # When prompted, enter your Personal Access Token as the password
   

3. Pull the Docker image from GitHub Container Registry:

   docker pull ghcr.io/mrwadams/otx-mcp:main
   

1. Clone this repository
2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Export your OTX API key as an environment variable:

   export OTX_API_KEY=your_api_key_here
   

   Or create a .env file with:

   OTX_API_KEY=your_api_key_here
   

To use this MCP server with Claude Desktop, add the following to your Claude Desktop config file (claude_desktop_config.json):

"mcpServers": {
  "otx": {
    "command": "docker",
    "args": [
      "run",
      "-i",
      "--rm",
      "-e",
      "OTX_API_KEY",
      "ghcr.io/mrwadams/otx-mcp:main"
    ],
    "env": {
      "OTX_API_KEY": "your_api_key_here"
    }
  }
}

Make sure you have:

1. Exported your OTX API key as an environment variable before starting Claude Desktop
2. Authenticated with GitHub Container Registry using a Personal Access Token as described in the installation section

This MCP server is designed to be used with any MCP-compatible client. The server listens for MCP protocol messages on stdin/stdout, making it compatible with various MCP clients that can execute Docker containers.

The MCP server provides the following tools:

• search_indicators: Search OTX for pulses matching a keyword (supports pagination via page and limit arguments).
• get_indicator_details: Get detailed information about a specific indicator
• get_indicator_details_full: Get all available details about a specific indicator
• validate_indicator: Validate an indicator before adding it to a pulse

• get_pulse: Get full details of a Pulse using its ID
• extract_indicators_from_pulse: Extract a paginated list of indicators from a given Pulse ID (supports page and limit arguments).
• create_pulse: Create a new pulse with threat intelligence information
• get_my_pulses: Get pulses created by the authenticated user
• get_subscribed_pulses: Get pulses the user is subscribed to

• search_users: Search for users in OTX
• get_user: Get information about a specific user
• get_user_pulses: Get pulses created by a specific user
• follow_user: Follow a user to receive notifications about their activities
• unfollow_user: Unfollow a user to stop receiving notifications

• subscribe_to_pulse: Subscribe to a pulse to receive updates
• unsubscribe_from_pulse: Unsubscribe from a pulse to stop receiving updates

• submit_url: Submit a URL for analysis
• submit_urls: Submit multiple URLs for analysis
• get_recent_events: Get recent events/activities from OTX

Here are some example queries you can run using the MCP server with an LLM like Claude:

Can you search OTX for any information about recent ransomware attacks?

I need to find threat intelligence about CVE-2023-1234. Can you search OTX for me?

OTX: Check indicators for google.com

Is 8.8.8.8 a malicious IP address? Can you check its reputation in OTX?

I found a suspicious domain called example.com. Can you get all the information about it from OTX?

I have a pulse ID 5f7c8e9a1b2c3d4e5f6a7b8c9d0e1f2. Can you get the details for me?

Can you extract all the indicators from pulse 5f7c8e9a1b2c3d4e5f6a7b8c9d0e1f2?

I need to create a new pulse about a ransomware campaign targeting healthcare organizations. The indicators include malicious-domain.com, 192.168.1.1, and https://malicious-domain.com/payload.exe. Can you help me create this pulse?

Can you search for users with "AlienVault" in their name or username?

I want to follow the AlienVault user to get notifications about their activities.

Can you subscribe me to pulse 5f7c8e9a1b2c3d4e5f6a7b8c9d0e1f2?

I found a suspicious website at https://suspicious-website.com. Can you submit it for analysis?

What are the 10 most recent events from OTX?

These natural language queries demonstrate how an LLM can understand your intent and use the appropriate MCP tools to fulfill your request, making it much easier to interact with the OTX platform without needing to know the specific API calls or parameters.

This project is licensed under the MIT License - see the LICENSE file for details.

• AlienVault OTX
• Model Context Protocol