mcp-remote-call-ping-pong

This MCP server implements a Ping-Pong interaction via MCP calls using a FastAPI backend. It exposes two ways to interact: a standard API endpoint and a Server-Sent Events (SSE) transport. The API lets clients retrieve prompts and invoke MCP tool commands such as ping, pong, and count through HTTP requests. The SSE variant provides a streaming, event-driven channel suitable for interactive sessions or long-running conversations. Use the UI file mcp-api-client.html to interact with the API in a browser, or run the SSE client script to exchange commands over SSE. The server runs by default at http://localhost:8080, and uses the Python scripts mcp-api-server.py and mcp-sse-server.py to handle requests and sessions.

Prerequisites:

• Python 3.8+ installed on your system
• Poetry (optional but recommended for dependency management)

Install and run (API server):

# Using Poetry (recommended)
poetry install
poetry run python mcp-api-server.py

# Or without Poetry
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python mcp-api-server.py

Access the API at http://localhost:8080. The UI file mcp-api-client.html can be opened in a browser to interact with the API.

Install and run (SSE server):

# Using Poetry
poetry install
poetry run python mcp-sse-server.py

# Or without Poetry
python mcp-sse-server.py

This will start the SSE server for interactive command sessions.

Tips and notes:

• The API provides endpoints like GET /ping-pong?prompt_name=<prompt_name> and POST /ping-pong to execute commands such as ping, pong, and count.
• SSE mode offers a real-time, streaming interface for MCP interactions; you can run mcp-sse-client.py to connect and issue commands.
• The default port is 8080 unless otherwise configured in the server scripts.
• Ensure that dependencies are installed (e.g., FastAPI, uvicorn) when running the API server. If using Poetry, use poetry install to fetch dependencies.
• If you modify prompts or endpoints, verify that the UI or client scripts point to the correct server URL.
• Session management is described as thread-safe in the features; for production, consider appropriate deployment considerations for Python web services.
• The repository references external MCP guides and related projects for inspiration on MCP transport and UI integration.