What is Mcp Sse Proxy

mcp-sse-proxy is a script that serves as a bridge between an MCP server using standard input/output (STDIO) transport and a client utilizing Server-Sent Events (SSE) transport. It converts MCP STDIO communication to the SSE protocol for real-time data streaming over web connections.

Use cases

Use cases for mcp-sse-proxy include real-time notifications in web applications, live data updates in dashboards, and streaming data from IoT devices to web interfaces.

How to use

To use mcp-sse-proxy, execute the ‘mcp_sse_proxy.py’ script with the required ‘–sse-url’ parameter, specifying the URL of your SSE server endpoint. You can also set this using the ‘ST_PROXY_SSE_URL’ environment variable. Install the package using pip and ensure all dependencies are met.

Key features

Key features of mcp-sse-proxy include the ability to convert MCP STDIO communication to SSE, support for real-time data streaming, and configurable command-line arguments for debugging and ping intervals.

Where to use

mcp-sse-proxy can be used in applications that require real-time data streaming from an MCP server to web clients, such as web applications, dashboards, and IoT systems.

MCP SSE Proxy

Description

This script acts as a proxy between an MCP (Model Context Protocol) server using STDIO (standard input/output) transport and a client using SSE (Server-Sent Events) transport. It bridges the communication gap by converting MCP STDIO communication to the SSE protocol, enabling real-time data streaming over web connections.

Communication Diagram:

[App](https://raw.githubusercontent.com/getsimpletool/mcp-sse-proxy/master/STDIO) ----<>---- (STDIO)[PROXY](https://raw.githubusercontent.com/getsimpletool/mcp-sse-proxy/master/SSE) ----<>---- (SSE)[SSE-Server]

flowchart LR
    App -- STDIO --> PROXY
    PROXY -- SSE --> SSEServer
    PROXY -- STDIO --> App
    SSEServer -- SSE --> PROXY
    SSEServer["SSE-Server"]
    App["APP"] 

    App@{ icon: "gcp:tools-for-powershell", form: "square", pos: "b"}
    SSEServer@{ icon: "aws:res-toolkit", form: "square", pos: "b"}

Usage

To run the MCP SSE Proxy, execute the mcp_sse_proxy.py script with the required --sse-url or ST_PROXY_SSE_URL environment variable, which specifies the URL of your SSE server endpoint. This is the only required mandatory parameter.

Installation

pip3 install mcp-sse-proxy
python3 -m mcp_sse_proxy -h

The package installer (pip) creates an executable script in the bin directory of your virtual environment (or Python installation). If you are in a virtual environment, you can run the script using directly as a command.

mcp-sse-proxy -h

Before you start

If you do installation by copy-past, remember to install dependencies:

pip3 install -r requirements.txt

Configuration

Command-line Arguments

• --sse-url <url>: (Required) The URL of the external SSE endpoint to connect to. You can also set this using the ST_PROXY_SSE_URL environment variable.
• --debug-enabled: Enable debug logging. You can also set this using the ST_PROXY_DEBUG_ENABLED environment variable (set to true to enable).
• --ping-interval <seconds>: Set the interval (in seconds) for sending ping messages to maintain the SSE connection. Default is 10 seconds. You can also set this using the ST_PROXY_PING_INTERVAL environment variable.

Environment Variables

The proxy can be configured using the following environment variables:

• ST_PROXY_SSE_URL: URL of the external SSE endpoint.
• ST_PROXY_DEBUG_ENABLED: Enable debug logging (set to true to enable).
• ST_PROXY_DEBUG_FILENAME: Filename for the debug log file. Defaults to mcp_sse_proxy.log.
• ST_PROXY_PING_INTERVAL: Ping interval in seconds.
• ST_PROXY_ENV_PREFIXES: Comma-separated list of environment variable prefixes to forward to the proxied application. This allows you to pass specific environment variables through the proxy. For example: ANTHROPIC_,OPENAI_,GEMINI_.

Claude AI configuration example

Windows Config location: C:\Users<USERNAME>\AppData\Roaming\Claude\claude_desktop_config
[or open via clinent menu: File->Settings then Developer->Edit Config]
Config example for Windows:

{
  "globalShortcut": "",
  "mcpServers": {
    "mcp-sse-proxy": {
      "command": "C:\\Users\\<USERNAME>\\AppData\\Local\\Programs\\Python\\Python312\\python.exe",
      "args": [
        "C:\\mcp-sse-proxy-claude\\src\\mcp_sse_proxy.py",
        "--base-url",
        "http://<IP_ADDRESS>:8000",
        "--debug-enabled"
      ],
      "env": {
        "OPENAI_API_KEY": "sk-xxx",
        "BRAVE_API_KEY": "BSAxxxx",
        "ST_PROXY_ENV_PREFIXES": "OPENAI_,BRAVE_"
      }
    }
  }
}

Debug Logging

Debug logging can be enabled using the --debug-enabled command-line argument or by setting the ST_PROXY_DEBUG_ENABLED environment variable to true. When enabled, detailed logs are written to the file specified by the ST_PROXY_DEBUG_FILENAME environment variable (default: mcp_sse_proxy.log).

Ping Interval

To ensure the SSE connection remains active, the proxy sends ping messages to the SSE server at regular intervals. The default interval is 10 seconds, configurable via the --ping-interval argument or the ST_PROXY_PING_INTERVAL environment variable.

Before you start debuging

• Note that by design MCP SSE Proxy will not print any messages after running in CLI mode. As by design, MCP will not log anything to STDIO, as this will interfere with protocol operation.
• Log file is stored in mcp_sse_proxy.log by default in same folder where mcp_sse_proxy.py script is located.
• Provides detailed information about SSE connections, tool calls, and system events (in debug file, you will see both side of connection messages - MCP and SSE)

⚠️ Warning: Enabling debug logging may expose sensitive information in logs. ⚠️

License

This project is likely licensed under the MIT License. See the LICENSE file for details.