Lepuroshi - Scheduling Service with MCP Integration
A scheduling service built with React and .NET, featuring an MCP server that lets AI agents manage appointments through natural language.
Overview
Lepuroshi is a scheduling platform where businesses manage appointments and clients book services online. The interesting part is the MCP (Model Context Protocol) server underneath it — AI assistants like Claude or ChatGPT can create, reschedule, and cancel appointments through natural language, with no custom integration needed per model.
Key Features
- Clients browse availability and book time slots through a clean booking interface
- Real-time sync across devices via SignalR, so availability is always current
- Email reminders sent automatically before upcoming appointments
- Conflict detection that prevents double-bookings at the data layer
- MCP server exposing the full scheduling lifecycle to AI agents over JSON-RPC
The MCP Server
The MCP server treats appointments, availability windows, and client records as first-class tools that any AI agent can invoke. In practice, this means an AI can:
- Check what time slots are open on a given day
- Create a new appointment for a client
- Reschedule or cancel an existing booking
- Trigger a confirmation message
No bespoke integration per AI model — any client that speaks the Model Context Protocol works out of the box.
Technical Stack
- Frontend: React 18, TypeScript
- Backend: ASP.NET Core (.NET 8), Entity Framework Core
- Real-time: SignalR
- Database: SQL Server
- AI Integration: Custom MCP server, OpenAI API for natural language understanding
- Deployment: Docker, cloud hosting
My Role
Full-stack — I designed the database schema, built the REST API and MCP server from scratch, and created the React frontend. The MCP implementation was the most novel part; there weren't many reference examples at the time, so a lot of it was working directly from the protocol spec.