Custom Context MCP

🚀 Custom Context MCP Server

This Model Context Protocol (MCP) server provides tools for structuring text and extracting data based on JSON templates. It addresses the challenge of transforming unstructured text into structured JSON data, which is valuable for downstream applications.

✨ Features

Text-to-JSON Conversion

• Group and structure text according to a JSON template containing placeholders.
• Extract information from AI-generated text and convert it into a structured JSON format.
• Support arbitrarily complex JSON structures, including nested placeholders.
• Intelligently extract key-value pairs from text.
• Process AI outputs to generate structured data for downstream applications.

🚀 Quick Start

📦 Installation

npm install

Running the Server

npm start

For development with hot reloading:

npm run dev:watch

💻 Usage Examples

Basic Usage

This MCP server offers two main tools:

1. Group Text by JSON (group-text-by-json)

This tool takes a JSON template with placeholders and generates prompts to guide AI in grouping text according to the template structure.

{
    "template": "{ \"type\": \"<type>\", \"text\": \"<text>\" }"
}

This tool analyzes the template, extracts placeholder keys, and returns a prompt to guide the AI in extracting information in key-value format.

2. Text to JSON (text-to-json)

This tool takes the grouped text generated in the previous step and converts it into a structured JSON object based on the original template.

{
    "template": "{ \"type\": \"<type>\", \"text\": \"<text>\" }",
    "text": "type: pen\ntext: This is a blue pen"
}

It extracts key-value pairs from the text and structures them according to the template.

Advanced Usage

Example Workflow

1. Define a JSON template with placeholders:

{
    "item": {
        "name": "<name>",
        "price": "<price>",
        "description": "<description>"
    }
}

2. Generate an AI prompt using group-text-by-json:

• The tool identifies the placeholder keys: name, price, description.
• It generates a prompt to guide the AI in extracting information by these keys.

3. Send the prompt to an AI model and receive grouped text:

name: Blue Pen
price: $2.99
description: A smooth-writing ballpoint pen with blue ink

4. Convert the grouped text to JSON using text-to-json:

{
    "item": {
        "name": "Blue Pen",
        "price": "$2.99",
        "description": "A smooth-writing ballpoint pen with blue ink"
    }
}

📚 Documentation

Template Format

Templates can contain placeholders anywhere within a valid JSON structure:

• Define placeholders using angle brackets: <name>, <type>, <price>, etc.
• The template must be a valid JSON string.
• Placeholders can appear at any nested level.
• Support complex nested structures.

Example template with nested placeholders:

{
    "product": {
        "details": {
            "name": "<name>",
            "category": "<category>"
        },
        "pricing": {
            "amount": "<price>",
            "currency": "USD"
        }
    },
    "metadata": {
        "timestamp": "2023-09-01T12:00:00"
    }
}

Installation of Dependencies

npm install mcp-server mcp-inspector uuid express

Starting the Server

To start the MCP server, run:

node index.js

By default, the server will run on port 3000.

Example Request

Use curl or any HTTP client to send the following request to http://localhost:3000/process:

Request Body

{
    "template": {
        "product": {
            "details": {
                "name": "<name>",
                "category": "<category>"
            },
            "pricing": {
                "amount": "<price>",
                "currency": "USD"
            }
        }
    },
    "text": "The product is a new electronic gadget priced at $49.99, category: electronics"
}

Response

{
    "product": {
        "details": {
            "name": "The product",
            "category": "electronics"
        },
        "pricing": {
            "amount": "49.99",
            "currency": "USD"
        }
    }
}

🔧 Technical Details

Development Considerations

• Ensure all dependencies are installed.
• Make sure the server is running on the correct port.
• Consider performance optimization when handling large amounts of data.