MCP Server SSH Client

You can run MCP tools hosted on remote machines through a secure SSH tunnel by using the MCP Server SSH Client. It acts as a local proxy that forwards MCP requests over SSH to a remote MCP server, letting you access tools and resources as if they were on your local network while keeping communications protected.

How to use

Use the MCP Server SSH Client to connect to a remote MCP server over SSH. You will run a local proxy command (via your MCP client or editor) that establishes an SSH session to the remote host and forwards JSONRPC MCP requests to the remote server.

How to install

Prerequisites: you need Python tooling and the uv runtime available in your environment.

# Install the MCP SSH client in a Python project using uvx tooling
uv add m2m-mcp-server-ssh-client

# Or install via pip in a standard Python project
pip install m2m-mcp-server-ssh-client

# If you want to run directly from source
git clone https://github.com/Machine-To-Machine/m2m-mcp-server-ssh-client.git
cd m2m-mcp-server-ssh-client
pip install -e .

# Run the server inside your project
uv run m2m-mcp-server-ssh-client

Additional configuration and usage notes

You can connect to remote MCP servers by specifying host, port, and authentication options. Common workflows include using a key server for automatic key exchange or relying on your existing SSH keys. The client supports both remote HTTP-based key management and local key handling via an SSH tunnel.

Standard usage patterns include configuring your MCP client or editor to point at the local proxy started by the MCP Server SSH Client, which then forwards requests to the remote MCP server over SSH.

Example configurations

{
  "mcpServers": {
    "remote-mcp-tools": {
      "command": "uvx",
      "args": [
        "m2m-mcp-server-ssh-client",
        "--host", "mcp-beta.machinetomachine.ai",
        "--port", "8022",
        "--use-key-server"
      ]
    }
  }
}

Security notes

Always verify the host key in production deployments. Use passphrase-protected SSH keys for stronger security and run under a dedicated, limited-privilege user on the remote server.

Troubleshooting

If you encounter connection issues, check that the SSH service is reachable on the specified host/port and that the key-server workflow (if used) is correctly configured. For authentication issues, verify your private key path and permissions.

Notes on usage with AI platforms

You can configure AI-assisted environments to use the local MCP Server SSH Client as the remote MCP tool provider by pointing the tool to the local proxy endpoint created by the client.