MCP Server - Model Context Protocol API

MCP Server is a FastAPI-based implementation of the Model Context Protocol (MCP) that provides a standardized interface for interaction between LLM models and applications.

Peculiarities

🚀 High-performance API based on FastAPI and asynchronous operations
🔄 Full MCP support with resources, instruments, prompts and sampling
📊 Monitoring and metrics via Prometheus and Grafana
🧩 Extensibility through simple interfaces to add new tools
📝 GraphQL API for flexible work with data
💬 WebSocket support for real-time interaction
🔍 Semantic search via integration with Elasticsearch
🗃️ Caching via Redis for improved performance
📦 Manage dependencies via Poetry for reliable package management

Getting Started

Installation

Clone repository:
git clone https://github.com/yourusername/myaiserv.git cd myaiserv
Install Poetry (if not already installed):
curl -sSL https://install.python-poetry.org | python3 -
Install dependencies via Poetry:
poetry install

Starting the server

poetry run uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

Or via the just utility:

just run

After launch, the API is available at: http://localhost:8000

API Documentation

Swagger UI: http://localhost:8000/docs
ReDoc: http://localhost:8000/redoc
GraphQL Playground: http://localhost:8000/graphql

Project structure

myaiserv/
├── app/
│   ├── core/             # Базовые компоненты MCP
│   │   ├── base_mcp.py   # Абстрактные классы MCP
│   │   └── base_sampling.py  # Базовые классы для сэмплирования
│   ├── models/           # Pydantic модели
│   │   ├── mcp.py        # Модели данных MCP
│   │   └── graphql.py    # GraphQL схема
│   ├── services/         # Бизнес-логика
│   │   └── mcp_service.py # Сервис MCP
│   ├── storage/          # Хранилище данных
│   ├── tools/            # Инструменты MCP
│   │   ├── example_tool.py   # Примеры инструментов
│   │   └── text_processor.py # Инструмент обработки текста
│   ├── utils/            # Утилиты
│   └── main.py           # Точка входа FastAPI
├── app/tests/            # Тесты
├── docs/                 # Документация
│   └── MCP_API.md        # Описание API
├── pyproject.toml        # Конфигурация Poetry и инструментов
└── .justfile             # Задачи для утилиты just

Available tools

File System Tool

A file system tool that supports reading, writing, deleting and listing files.

curl -X POST "http://localhost:8000/tools/file_operations" \
     -H "Content-Type: application/json" \
     -d '{"operation": "list", "path": "."}'

Weather Tool

A tool for obtaining weather data by coordinates.

curl -X POST "http://localhost:8000/tools/weather" \
     -H "Content-Type: application/json" \
     -d '{"latitude": 37.7749, "longitude": -122.4194}'

Text Analysis Tool

A tool for text analysis, including sentiment detection and summarization.

curl -X POST "http://localhost:8000/tools/text_analysis" \
     -H "Content-Type: application/json" \
     -d '{"text": "Example text for analysis", "analysis_type": "sentiment"}'

Text Processor Tool

A tool for text processing, including formatting, statistics calculation, entity extraction.

curl -X POST "http://localhost:8000/tools/text_processor" \
     -H "Content-Type: application/json" \
     -d '{"operation": "statistics", "text": "Example text", "stat_options": ["chars", "words"]}'

Image Processing Tool

An image processing tool that supports resizing, cropping and applying filters.

curl -X POST "http://localhost:8000/tools/image_processing" \
     -H "Content-Type: application/json" \
     -d '{"operation": "resize", "image_data": "base64...", "params": {"width": 800, "height": 600}}'

WebSocket API

To connect to the WebSocket API:

const socket = new WebSocket("ws://localhost:8000/ws");

socket.onopen = () => {
  socket.send(JSON.stringify({
    type: "initialize",
    id: "my-request-id"
  }));
};

socket.onmessage = (event) => {
  const data = JSON.parse(event.data);
  console.log("Received:", data);
};

GraphQL API

Examples of queries via GraphQL:

# Получение списка всех инструментов
query {
  getTools {
    name
    description
  }
}

# Выполнение инструмента
mutation {
  executeTool(input: {
    name: "text_processor",
    parameters: {
      operation: "statistics",
      text: "Example text for analysis"
    }
  }) {
    content {
      type
      text
    }
    is_error
  }
}

Running tests

To run tests, use Poetry:

poetry run pytest

Or via the just utility:

just test

Docker

Building and running via Docker Compose

docker compose up -d

To launch individual services:

docker compose up -d web redis elasticsearch

Integration with LLM

MCP Server provides a standardized interface for integration with LLM models from various vendors:

import httpx

async def query_mcp_with_llm(prompt: str):
    async with httpx.AsyncClient() as client:
        # Запрос к MCP для получения контекста и инструментов
        tools_response = await client.get("http://localhost:8000/tools")
        tools = tools_response.json()["tools"]

        # Отправка запроса к LLM с включением MCP контекста
        llm_response = await client.post(
            "https://api.example-llm.com/v1/chat",
            json={
                "messages": [
                    {"role": "system", "content": "You have access to the following tools:"},
                    {"role": "user", "content": prompt}
                ],
                "tools": tools,
                "tool_choice": "auto"
            }
        )

        return llm_response.json()

Metrics and monitoring

MCP Server provides metrics in Prometheus format via the /metrics endpoint. Metrics include:

Number of requests to each tool
Query execution time
Errors and exceptions

Development

To format code and check it with linters:

just fmt
just lint

License

MIT License

MyAIServ MCP Server

Integrations