Scientific Paper Analyzer (Gemini)

🚀 Basic MCP Application

A simple application demonstrating how to implement the Model Context Protocol (MCP) using FastAPI and Gradio (because I can't stand the headache caused by Streamlit).

🚀 Quick Start

Prerequisites

• Python 3.11 or higher (sorry, dinosaurs still using Python 2).
• pip package manager (or its cooler, faster cousin uv).
• Understanding and patience (optional but recommended, just like RCB. Google it).

Setup Steps

1. Clone the repository:

   git clone https://github.com/your-repository.git
   cd basic-mcp-app
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Start the backend server:

   python backend/main.py
   

4. Start the Gradio frontend interface:

   python frontend/app.py
   

✨ Features

• Search for scientific papers using Semantic Scholar (because Google Scholar is too mainstream).
• Paper analysis tool (works 60% of the time, every time).
• Simple chat interface (simple for users, a nightmare for developers).
• Easy installation process (if you've ever climbed Mount Everest, this will feel like a stroll in the park).

📦 Installation

This project is perfect to use with uv, an ultra-fast Python package installation tool! Instead of waiting for pip to take an eternity to finish the task, you can use uv to install dependencies in seconds. I highly recommend it for a smoother experience (and to get back a few hours of your life from global dependencies).

📚 Documentation

Project Overview

This project demonstrates a basic MCP server with a Gradio frontend interface (Streamlit is a headache, and life is too short to endure unnecessary pain). Users can chat with the AI model through the "simple interface" (marketers call it a "simple interface", developers know it's "doing the best they can, after all, time and patience are limited 🥲").

Technology Stack

• Backend: FastAPI + MCP Python SDK (heaven, unlike pineapple on pizza).
• Frontend: Gradio (because pretty buttons release dopamine quickly).
• AI Integration: Google Gemini API (not an astrological symbol, but something like Sundar Pichai's AI AI AI AI AI AI).

Project Structure

Basic MCP App
├── backend/
│   └── main.py          # Backend server code, containing MCP (where the real magic happens)
└── frontend/
    └── app.py           # Gradio frontend interface (where all the pretty buttons are)

🔧 Technical Details

Known Issues

⚠️ Important Note

• The citation tool is currently not working properly. Errors may occur when trying to analyze paper citations or use certain advanced search functions. I'm working on fixing this issue. When will it be fixed? Who knows ¯_(ツ)_/¯. Maybe when AGI (Artificial General Intelligence) is achieved.
• The Semantic Scholar API has rate limits, which may cause the search function to sometimes return the error message "I can't directly search and provide papers for you". This is the result of a free API meeting enthusiastic users - we love them too much. Take a break and try again (or have a cup of coffee while waiting for the rate limit to reset).

📄 License

MIT (because I'm generous and don't want to read long license agreements).

Acknowledgments

• Anthropic for MCP — https://www.anthropic.com/news/model-context-protocol
• https://modelcontextprotocol.io/introduction
• Claude adjusted part of the DEMO. Although it's not fully adjusted, still thanks 🤏