Remote Clinic MCP Server

A stateless Streamable HTTP MCP server for the remote-server requirement of CC3067 Project 1. It provides a laboratory preparation catalog, checks preparation requirements, and returns the remote server time.

Academic demonstration project for CC3067 Redes using synthetic preparation policies.

Tools

The MCP endpoint is POST /mcp; GET /health is available for platform checks. The server is stateless and returns JSON responses over Streamable HTTP.

Related MCP server: Veridexa MCP Gateway

Local installation

uv sync
uv run remote-clinic-mcp

Then connect an MCP Inspector or client to http://localhost:8080/mcp.

Run automated tests with:

uv run pytest

Container

docker build -t remote-clinic-mcp .
docker run --rm -p 8080:8080 remote-clinic-mcp

Google Cloud Run deployment

The following commands require the Google Cloud CLI, an authenticated account, and a project with Cloud Run enabled:

gcloud auth login
gcloud config set project YOUR_PROJECT_ID
gcloud run deploy remote-clinic-mcp `
  --source . `
  --region us-central1 `
  --allow-unauthenticated `
  --port 8080

Cloud Run prints the service URL. The MCP endpoint is that URL followed by /mcp. Verify it with:

Invoke-RestMethod https://YOUR-SERVICE-URL/health

Set REMOTE_MCP_URL in the chatbot .env, for example:

REMOTE_MCP_URL=https://remote-clinic-mcp-xxxxx-uc.a.run.app/mcp

For this academic demonstration the service allows unauthenticated requests for testing.

Wireshark preparation

The deployed connection uses TLS over TCP port 443. Capture traffic while starting the chatbot and invoking get_remote_server_time. Wireshark can show the DNS, TCP, TLS, and HTTPS exchange, but the JSON-RPC body remains encrypted unless TLS key logging is configured.