JMRMEDEV/amazon-q-gpt-mcp-server

GPT Agent MCP Server

An MCP server that provides a ChatGPT agent specialized in software architecture, full-stack development, and code debugging.

Features

• Specialized Agent: Focused on software development, architecture, and debugging
• Real Web Search: DuckDuckGo integration for live search results
• Conversation Memory: Maintains context across interactions
• Attempt Tracking: Monitors failed solutions to trigger web search automatically
• GPT-4o-mini: Cost-efficient model optimized for development tasks
• Three Core Tools: Chat, reset conversation, and status checking

Setup

1. Install dependencies:

   npm install
   

2. Set up environment variables:

   cp .env.example .env
   # Edit .env and add your OPENAI_API_KEY
   

3. Make executable:

   chmod +x server.js
   

Usage

Available Tools

chat_with_agent

Chat with the software development agent.

Parameters:

• message (required): Your development question or request
• context (optional): Additional context about your project or environment

Example:

{
  "message": "How should I structure a Node.js microservices architecture?",
  "context": "Building an e-commerce platform with high traffic requirements"
}

reset_conversation

Clear conversation history and attempt tracking.

get_agent_status

Get current agent status and conversation information.

Web Search Integration

The agent includes real web search functionality powered by DuckDuckGo that triggers:

• Version Checks: When asking about current versions, latest releases, or updates
• API Documentation: When requesting API docs, references, or endpoints
• Failed Attempts: After 3 unsuccessful solution attempts on the same topic

The search provides live results including titles, URLs, and snippets from DuckDuckGo, giving the agent access to current information for better responses.

Benefits of DuckDuckGo Integration:

• No API keys required (completely free)
• No rate limiting or bot detection issues
• Privacy-focused search without tracking
• Reliable results for development topics
• Automatic fallback to simulated results if search fails

Agent Capabilities

The agent specializes in:

• Software architecture and design patterns
• Full-stack development (frontend, backend, databases)
• Code debugging and optimization
• Technology stack recommendations
• Best practices and code review

Amazon Q Integration

To use with Amazon Q, add this server to your MCP configuration:

{
  "mcpServers": {
    "gpt": {
      "command": "node",
      "args": ["/home/jmrmedev/mcp-servers/gpt/server.js"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key-here"
      }
    }
  }
}

Configuration

Environment variables:

• OPENAI_API_KEY: Your OpenAI API key (required)
• OPENAI_MODEL: Model to use (default: gpt-4o-mini)
• DEBUG: Enable debug logging (default: false)