Cybersecurity

You can run multiple MCP servers for cybersecurity tasks in parallel, letting you query, manage, and transform data through modular, self-contained endpoints. This enables you to experiment with SQL injection testing, network space asset searching, and Markdown-to-Word processing in a unified environment.

How to use

You interact with each MCP server through a client that supports the MCP protocol. Start the servers locally, then connect your client to the named MCP endpoints to perform actions such as running scans, searching assets, or converting content. Each server exposes its own set of capabilities, so you can mix and match the ones you need for a given workflow.

Typical usage patterns include enabling a server, performing a task (for example, initiating a SQL injection test or searching for network assets), and then saving or reviewing notes and results. You can also run the servers in development mode to automatically rebuild on code changes, which accelerates iteration when you are configuring or extending MCP functionality.

How to install

Prerequisites: you need Node.js installed on your system. You also need a compatible package manager, such as npm. Ensure you have a modern JavaScript runtime environment to run the MCP servers.

Install dependencies for all implementations, then build and run in development mode if you want automatic rebuilds.

npm install
npm run build
npm run watch

Configuring MCP servers for Claude Desktop

To use these MCP servers from Claude Desktop, add entries to the MCP configuration. You can point each server to either a local build or a remote endpoint. The example below shows three MCP servers configured as local stdio servers. Adapt the paths to where you built each server on your machine.

{
  "mcpServers": {
    "sqlmap_mcp": {
      "command": "/path/to/sqlmap-mcp/build/index.js",
      "args": [],
      "disabled": false,
      "alwaysAllow": []
    },
    "quake_mcp": {
      "command": "node",
      "args": [
        "/path/to/quake-server/build/index.js"
      ],
      "env": {
        "QUAKE_API_KEY": "YOUR_API_KEY"
      },
      "disabled": false,
      "alwaysAllow": []
    },
    "docproc_mcp": {
      "command": "node",
      "args": [
        "/path/to/doc-processor/build/index.js"
      ],
      "disabled": false,
      "alwaysAllow": []
    }
  }
}

Notes on environment variables

If a server requires an API key or other credentials, place them in the server's env block. For example, the quake server uses QUAKE_API_KEY in the configuration. Keep credentials out of source control and use environment-specific values in your deployment.

Troubleshooting

If a server does not start, check that the build output exists at the expected path and that your runtime command and arguments exactly match what is shown in your configuration. Ensure you have network access if you run servers remotely, and verify that you provided any required environment variables.

Security considerations

Run MCP servers with least-privilege permissions and isolate them from sensitive data when possible. Use separate environments for development and production, and rotate API keys or credentials periodically. When testing SQL injection capabilities, constrain scans to authorized targets and use safe, controlled environments.