Fibonacci MCP Server

A secure Model Context Protocol (MCP) server that provides tools for calculating Fibonacci numbers. This server is designed to meet OpenAI's MCP specifications and includes comprehensive security measures for production deployment.

Features

• Fibonacci Calculator: Calculate the nth Fibonacci number with input validation

• BMI Calculator: Calculate BMI from weight and height with realistic bounds

• Code Review Prompt: Generate prompts for code review with size limits

• Web Interface: Beautiful home page explaining the server capabilities

• Session Management: Stateful MCP server with automatic session cleanup

• Security Features: Rate limiting, CORS, input validation, and error handling

• Production Ready: Health checks, logging, and environment-specific configurations

Security & Compliance Features

🔒 Security Measures

• Rate Limiting: 1000 requests per 15-minute window per IP

• Input Validation: Comprehensive validation using Zod schemas

• Request Size Limits: 10MB limit on request payloads

• CORS Configuration: Proper CORS headers for MCP client compatibility

• Session Timeout: Automatic cleanup of sessions older than 1 hour

• Error Handling: Proper error responses following MCP protocol standards

🛡️ Production Readiness

• Health Check Endpoint: /health for monitoring and load balancers

• Structured Logging: Comprehensive error logging and monitoring

• Environment Detection: Automatic configuration for production vs development

• Memory Management: Automatic cleanup of rate limit and session data

• Graceful Error Handling: Proper HTTP status codes and JSON-RPC error responses

📋 MCP Protocol Compliance

• Proper Error Codes: JSON-RPC 2.0 compliant error responses

• Server Metadata: Complete server information and capabilities

• Session Management: Proper MCP session lifecycle management

• Resource Validation: Input validation for dynamic resources

Local Development

Prerequisites

• Node.js 18.0.0 or higher

• npm

Installation

1. Clone the repository:

2. Install dependencies:

3. Start the server:

The server will be available at:

• Home page: http://localhost:7171/

• MCP endpoint: http://localhost:7171/mcp

• Health check: http://localhost:7171/health

Deployment on Render

Option 1: Using render.yaml (Recommended)

1. Push your code to a Git repository (GitHub, GitLab, etc.)

2. Connect your repository to Render

3. Render will automatically detect the render.yaml file and deploy your service

Option 2: Manual Deployment

1. Create a new Web Service on Render

2. Connect your Git repository

3. Configure the service:

   • Environment: Node

   • Build Command: npm install

   • Start Command: npm start

   • Environment Variables:

     • NODE_ENV: production

Environment Variables

• PORT: Port number (automatically set by Render)

• NODE_ENV: Set to production for production deployments

MCP Tools

fibonacci

Calculate the nth Fibonacci number with input validation.

Parameters:

• n (number): The position in the Fibonacci sequence (0-indexed, max 1000)

Example:

Response:

calculate-bmi

Calculate BMI from weight and height with realistic bounds.

Parameters:

• weightKg (number): Weight in kilograms (0-1000)

• heightM (number): Height in meters (0.1-3.0)

review-code

Generate a prompt for code review with size limits.

Parameters:

• code (string): The code to review (max 10,000 characters)

MCP Resources

• config://app: Static application configuration

• users://{userId}/profile: Dynamic user profile data (with input validation)

API Endpoints

• GET /: Home page with server information

• GET /health: Health check endpoint for monitoring

• POST /mcp: MCP protocol endpoint

• GET /mcp: Server-sent events (SSE) for notifications

• DELETE /mcp: Session termination

Security Headers

The server automatically includes the following security headers:

• Access-Control-Allow-Origin: *

• Access-Control-Allow-Methods: GET, POST, DELETE, OPTIONS

• Access-Control-Allow-Headers: Content-Type, mcp-session-id, Authorization

Rate Limiting

• Limit: 1000 requests per 15-minute window per IP address

• Response: HTTP 429 with JSON-RPC error when limit exceeded

• Storage: In-memory with automatic cleanup

Error Handling

The server returns proper JSON-RPC 2.0 error responses:

Common error codes:

• -32000: Bad Request

• -32029: Rate limit exceeded

• -32601: Method not found

• -32603: Internal server error

Monitoring & Health Checks

Health Check Response

Logging

The server logs:

• Session creation and cleanup

• Rate limit violations

• Error conditions

• Server startup information

Technical Details

• Server Name: fibonacci-mcp-server

• Version: 1.0.0

• Protocol: Model Context Protocol (MCP)

• Transport: HTTP/Express with session management

• Node.js Version: >=18.0.0

• Security: Rate limiting, CORS, input validation

• Compliance: JSON-RPC 2.0, MCP protocol standards

OpenAI MCP Compliance

This server is designed to meet OpenAI's MCP specifications:

✅ Authentication & Security: Rate limiting, input validation, CORS
✅ Error Handling: Proper JSON-RPC error responses
✅ Session Management: Secure session lifecycle
✅ Production Readiness: Health checks, logging, monitoring
✅ Protocol Compliance: MCP and JSON-RPC 2.0 standards
✅ Resource Management: Memory cleanup and timeout handling

License

ISC