Skip to main content
Glama
nikita-pakhno

parking-mcp-server

by nikita-pakhno
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

Parking MCP Server

MCP (Model Context Protocol) server that exposes a tool for writing confirmed parking reservations to a file. Called by the admin agent once an administrator approves a reservation.

Tool

write_reservation(
    name: str,
    car_number: str,
    reservation_period: str,   # "start - end"
    approval_time: str         # ISO timestamp
) -> str   # confirmation or error

Appends one line per reservation to data/confirmed_reservations.txt:

Name | Car Number | Reservation Period | Approval Time

Related MCP server: MCP Security Framework

Security

  • Bearer token auth on every request (Authorization: Bearer <token>).

  • Input validation — names, plate numbers, timestamps are validated and rejected if they contain pipe characters or newlines (anti-injection for the delimited file format).

  • Append-only, atomic writes — file is opened in append mode with a brief file lock so concurrent writes don't interleave.

  • No arbitrary file paths — the output file is fixed by the server config, not by tool arguments.

Setup

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env
# edit .env — change MCP_AUTH_TOKEN

python -m mcp_server.run
# MCP endpoint at http://localhost:8002/mcp

Calling the tool from a Python MCP client

import asyncio
from mcp import ClientSession
from mcp.client.streamable_http import streamablehttp_client

async def main():
    async with streamablehttp_client(
        "http://localhost:8002/mcp",
        headers={"Authorization": "Bearer change-me-parking-mcp-secret"},
    ) as (read, write, _):
        async with ClientSession(read, write) as session:
            await session.initialize()
            result = await session.call_tool(
                "write_reservation",
                {"name": "Nikita", "car_number": "BC1234AB",
                 "reservation_period": "2025-12-01 10:00 - 2025-12-01 12:00",
                 "approval_time": "2026-08-11T12:00:00Z"},
            )
            print(result.content[0].text)

asyncio.run(main())

Project layout

mcp_server/
  server.py        # FastMCP server + tool definition
  storage.py       # atomic append-only writer with input validation
  auth.py          # Bearer token middleware
  config.py
tests/
  test_storage.py
  test_server.py
  test_auth.py

Tests

pytest -q

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivitySlowing
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers