home / mcp / claude mcp gateway mcp server

Claude MCP Gateway MCP Server

Bridges Claude Desktop with MCP servers by translating stdio to HTTP SSE, enabling resource discovery and tool use.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "boilingdata-mcp-server-and-gw": {
      "command": "npx",
      "args": [
        "mcp-server-and-gw",
        "http://localhost:8808/"
      ],
      "env": {
        "MCP_HOST": "localhost",
        "MCP_PORT": "8808"
      }
    }
  }
}

You can bridge Claude Desktop with MCP-enabled servers by running a gateway that translates between the standard input/output flow and the MCP HTTP SSE transport. This lets Claude discover resources and use tools provided by remote MCP servers while remaining flexible in how you deploy and test the integration.

How to use

Install the gateway support, start a local bridge, and connect Claude Desktop so it can discover resources, tools, and actions exposed by your MCP server. The bridge runs as a small command that you execute in the terminal; Claude then communicates through the bridge to your MCP server, enabling actions like listing tools and running queries.

How to install

Prerequisites: you need Node.js and npm installed on your system. You also need a shell environment to run commands.

Step by step commands you can run exactly as shown:

# 1. install
npm install -g mcp-server-and-gw
# 2. Or run directly with npx
npx mcp-server-and-gw http://localhost:8808/
# ...you can use environment variables too
MCP_HOST=localhost MCP_PORT=8808 npx mcp-server-and-gw

Additional notes

The gateway example demonstrates how to bridge a local MCP server to Claude by using a simple stdio to HTTP SSE bridge. You can also configure Claude to discover the gateway by adding a configuration entry that points to the bridge command and its destination URL.

Available tools

query

Run a read-only SQL query on a DuckDB database

visualise

Visualise SQL query results as an Apache ECharts chart. Provide the SQL clause that produces the data for the visualisation and a chart configuration