noname-mcp-server by braveenth - MCP Server

NoName MCP Server

Overview

This project implements a Model-Context-Protocol (MCP) server that provides an intelligent, conversational interface for searching product data. The server leverages a Large Language Model (LLM) to understand natural language queries, interact with a product database stored in a CSV file, and return structured, relevant information. It's built with FastAPI and LangGraph, creating a powerful agent-based system for product discovery.

Features

Natural Language Product Search: Ask for products in plain English (e.g., "find me some milk").
Similar Product Recommendations: Discover other products in the same aisle (e.g., "what's similar to butter?").
Recipe Generation: Get recipe ideas based on a list of ingredients.
Product Link Retrieval: Instantly get a direct link to the product page.
Nutritional Information: Access mock nutritional data for products.
Extensible Toolset: The server's capabilities can be easily expanded by adding new tools.

Architecture

The server uses a LangGraph-based agent to process incoming requests. The agent decides which tool to use based on the user's query, executes the tool, and then formulates a final response.

Data Sources

noname_products.csv: This file contains the core product data, and it was obtained from Kaggle. It includes columns such as Name, aisle, brand, price, and Code (the product identifier for generating links). The data is based on products from a Canadian supermarket.
OpenAI API: The get_recipe tool utilizes the gpt-4 model from OpenAI to dynamically generate recipes from a given list of ingredients.

Available Tools

Tool	Description	Input Parameter(s)
`get_products`	Searches for products by name.	`query` (string)
`get_similar_products`	Finds products in the same aisle as a given product.	`product_name` (string)
`get_recipe`	Generates a recipe from a list of ingredients.	`ingredients` (list of strings)
`get_nutritional_info`	Returns (mock) nutritional information for a product.	`product_name` (string)
`get_products_with_links`	Searches for products and includes a web link for each.	`query` (string)
`get_similar_products_with_links`	Finds similar products and includes a web link for each.	`product_name` (string)

API Endpoints

List Available Tools

GET /tools/list
Description: Retrieves a list of all available tools, including their names, descriptions, and input schemas.
Response: A JSON array of ToolDefinition objects.

Run the Agent

POST /mcp
Description: Sends a query to the MCP agent for processing.

Request Body:

{
  "query": "your natural language query here"
}

Response: An MCPResponse object containing the agent's structured SearchResponse.

Setup & Running the Server

Clone the repository:

git clone <repository-url>
cd mcp_server

Install dependencies:
```
pip install -r requirements.txt
```
Set up environment variables: Create a .env file in the project root and add your OpenAI API key:
```
OPENAI_API_KEY="your-api-key-here"
```
Run the server:
```
uvicorn app:app --host 0.0.0.0 --port 8000
```
The server will be accessible at http://localhost:8000.

Usage Examples

Get a list of tools

curl -X GET http://localhost:8000/tools/list

Search for a product

curl -X POST http://localhost:8000/mcp \
-H "Content-Type: application/json" \
-d '{
  "query": "find me some eggs"
}'

Find similar products with links

curl -X POST http://localhost:8000/mcp \
-H "Content-Type: application/json" \
-d '{
  "query": "find products similar to milk and give me links"
}'

Project Structure

.
├── CSVs/
│   └── noname_products.csv   # Product data
├── product_search/
│   ├── __init__.py
│   ├── agent.py              # Defines the LangGraph agent
│   ├── graph.py              # Wires up the tools for the agent
│   ├── schema.py             # Pydantic models for data structures
│   └── tools.py              # Core tool implementations
├── .env                      # Environment variables (needs to be created)
├── app.py                    # FastAPI application and endpoints
├── Dockerfile                # Containerization configuration
└── requirements.txt          # Python dependencies

Contributing

Contributions are welcome! If you have suggestions for improvements or find any issues, please feel free to:

Open an issue: Describe the issue or feature proposal.
Submit a pull request: Fork the repository, make your changes, and submit a PR with a clear description of your work.