CV Digital Twin MCP Server

An MCP server that takes your CV PDF as input and behaves like your digital twin, answering questions about you based on your CV content.

Features

• Single Tool Interface: One simple chat_with_me tool handles everything
• Automatic CV Loading: CV is loaded automatically on first use
• AI-Powered Digital Twin: Uses OpenAI GPT models to answer questions naturally and conversationally
• Intelligent Q&A: Ask about work experience, skills, education, contact info, and more

Getting Started

Installation

1. Install dependencies:

pip install -r requirements.txt

1. Set up your OpenAI API key:

export OPENAI_API_KEY='your-api-key-here'

Or create a .env file:

OPENAI_API_KEY=your-api-key-here

Usage

As an MCP Server

The server provides a single tool:

chat_with_me: Chat with your digital twin based on your CV

• Parameters:
  • message (required): Your message or question
  • cv_path (optional): Path to your CV PDF file (only needed on first call)
• Examples:
  • First call: chat_with_me("Tell me about yourself", cv_path="path/to/my_cv.pdf")
  • Subsequent calls: chat_with_me("What is your work experience?")
  • chat_with_me("What are your skills?")
  • chat_with_me("How can I contact you?")

Local Testing

You can also test the server locally:

python cv_digital_twin.py path/to/your_cv.pdf

Example Questions

The digital twin can answer questions like:

• "What is your work experience?"
• "What are your skills?"
• "Tell me about your education"
• "What technologies do you know?"
• "How can I contact you?"
• "Tell me about yourself"

Files

• cv_digital_twin.py - Main CV Digital Twin MCP server
• echo.py - Original echo server example (for reference)
• requirements.txt - Python dependencies

Dependencies

• fastmcp - FastMCP framework
• openai - OpenAI API client for AI-powered responses
• pdfplumber - PDF text extraction (preferred)
• PyPDF2 - Alternative PDF library (fallback)

Configuration

The server uses OpenAI's GPT models to generate responses.

Default Model

• Default: gpt-4o-mini (cost-efficient)
• Can be changed via environment variable: export OPENAI_MODEL="gpt-4o"

Available Models

• gpt-4o-mini - Fast and cost-efficient (default)
• gpt-4o - More capable, higher cost
• gpt-4 - Legacy GPT-4
• gpt-3.5-turbo - Fast and economical
• gpt-4-turbo - Enhanced GPT-4

Note: You'll need an OpenAI API key. Get one at https://platform.openai.com/api-keys

Deployment

This repository is ready to be deployed!

• Create a new FastMCP Cloud account
• Connect your GitHub account
• Select Clone our template and a deployment will be created for you!

Learn More

• FastMCP Documentation
• MCP Protocol

This repository was created from the FastMCP quickstart template and extended with CV Digital Twin functionality.