mayanks0369/mcp-server

MCP Server - Model Context Protocol API

The MCP Server is a FastAPI-based HTTP server that exposes three powerful NLP endpoints:

• 🔹 /v1/summarize: Generate a concise summary of input text.
• 🔹 /v1/keywords: Extract key phrases from text.
• 🔹 /v1/sentiment: Perform sentiment analysis on input text.

It includes request logging, API key authorization, OpenAPI schema, test coverage, and a minimal web frontend.

🚀 Features

• ✅ 3 NLP Tasks via Hugging Face Transformers
• ✅ API Key Authentication (middleware-based)
• ✅ Structured Error Handling with unique requestId
• ✅ Request Logging (path, status, duration)
• ✅ OpenAPI Docs available at /docs
• ✅ Minimal Frontend UI to test endpoints visually (index.html)
• ✅ Test Suite using pytest + httpx

📦 Project Structure

MCP_Server/
├── app/
│   ├── main.py
│   ├── routes/
│   │   ├── summarize.py
│   │   ├── keywords.py
│   │   └── sentiment.py
│   └── utils/
│       └── logger.py
├── static/
│   └── index.html
├── tests/
│   ├── conftest.py
│   ├── test_keywords.py
│   ├── test_sentiment.py
│   └── test_summarize.py
├── openapi.json
├── requirements.txt
└── README.md

🔐 API Key Protection

All routes are protected with API key authentication, except:

• /
• /docs
• /openapi.json
• /static/*

Use this test key in your headers:

X-Api-Key: my-secret-api-key

📋 Example Usage

✅ 1. Health Check

curl http://localhost:8000/health

✅ 2. Sentiment Analysis

curl -X POST http://localhost:8000/v1/sentiment \
     -H "Content-Type: application/json" \
     -H "X-Api-Key: my-secret-api-key" \
     -d '{ "text": "I love working on AI projects!" }'

✅ 3. Summarize

curl -X POST http://localhost:8000/v1/summarize \
     -H "Content-Type: application/json" \
     -H "X-Api-Key: my-secret-api-key" \
     -d '{ "text": "Artificial Intelligence is transforming industries by enabling machines to learn from data..." }'

✅ 4. Keyword Extraction

curl -X POST http://localhost:8000/v1/keywords \
     -H "Content-Type: application/json" \
     -H "X-Api-Key: my-secret-api-key" \
     -d '{ "text": "Transformers like BERT and GPT have revolutionized NLP." }'

🧪 Running Tests

To run all unit tests:

python -m pytest -v tests/

🖥️ Frontend UI

A minimal client interface is available at:

📍 http://localhost:8000/

You can:

• Paste input text
• Choose task (Summarize, Keywords, Sentiment)
• View output directly in the browser
• API key is auto-attached via frontend JS

📄 OpenAPI Spec

The server is OpenAPI 3.0 compliant.

Download the OpenAPI spec with:

curl http://localhost:8000/openapi.json -H "X-Api-Key: my-secret-api-key" -o openapi.json

Swagger UI available at:

📍 http://localhost:8000/docs

✅ Setup Instructions

# Create virtual environment
python -m venv venv

# Activate
# Linux/macOS
source venv/bin/activate
# Windows
venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the server
uvicorn app.main:app --reload

🧠 Models Used

✨ Future Enhancements

• 🔐 Switch to token-based auth (OAuth2 / JWT)
• 💻 Frontend improvements using React/Vue
• ⚡ Add caching for repeated results

🙋 Author

Mayank Singh

AI/ML Engineer – Take-home assignment

Built with ❤️ using FastAPI + HuggingFace + HTML