VulnGPT MCP Server

Overview

VulnGPT MCP Server was built for the Puch AI Hackathon. It implements the Model Context Protocol (MCP) to provide vulnerability scanning capabilities that any MCP-compatible client can use.

What it does:

• MCP Integration — Works with any MCP-compatible AI client (Claude, etc.)
• Vulnerability Scanning — Scans code, configurations, and dependencies for vulnerabilities
• Context-Aware Analysis — Uses LLM context to understand code intent and find deeper issues
• Structured Output — Returns findings in MCP-compatible format

How It Works

┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  MCP Client     │────▶│  VulnGPT Server  │────▶│  Vulnerability  │
│  (Claude, etc.) │◀────│  (MCP Protocol)  │◀────│  Database       │
└─────────────────┘     └──────────────────┘     └─────────────────┘

Getting Started

# Clone the repository
git clone https://github.com/Anandhasasidharan/vulngpt-mcp-server.git
cd vulngpt-mcp-server

# Install dependencies
pip install -r requirements.txt

# Run the MCP server
python server.py

MCP Tools

The server exposes these tools via MCP:

• scan_code — Scan source code for vulnerabilities
• scan_dependencies — Check dependencies for known CVEs
• analyze_config — Analyze configuration files for misconfigurations
• get_report — Generate a vulnerability report

Hackathon Context

Built during the Puch AI Hackathon to demonstrate how MCP can extend AI capabilities with security tools. The server acts as a bridge between AI assistants and vulnerability databases, enabling natural language security analysis.