MCP ChatGPT - AI Collaboration Server

Enable seamless collaboration between Claude and ChatGPT through the Model Context Protocol

🚀 Features

• Multi-Conversation Management: Create and manage multiple ChatGPT conversations
• Model Selection: Choose between GPT-4, GPT-4 Turbo, and GPT-3.5 Turbo
• Specialized Analysis: Code review, security analysis, performance optimization
• Approach Comparison: Compare different solutions side-by-side
• Conversation History: Full access to conversation history and context
• System Prompts: Set custom system prompts for specialized contexts

📋 Prerequisites

• Node.js 18 or higher
• OpenAI API Key - Get yours at platform.openai.com

🛠️ Installation

Quick Install for Claude Code

# Install globally for all projects
claude mcp add chatgpt -s user -e OPENAI_API_KEY=your_api_key -- npx @chinchillaenterprises/mcp-chatgpt

# Or install for specific project
claude mcp add chatgpt -s project -e OPENAI_API_KEY=your_api_key -- npx @chinchillaenterprises/mcp-chatgpt

Manual Installation

1. Install the package:

   npm install -g @chinchillaenterprises/mcp-chatgpt

2. Configure Claude Code (~/.claude.json):

   {
     "mcpServers": {
       "chatgpt": {
         "type": "stdio",
         "command": "npx",
         "args": ["-y", "@chinchillaenterprises/mcp-chatgpt"],
         "env": {
           "OPENAI_API_KEY": "your_openai_api_key_here"
         }
       }
     }
   }

🎯 Quick Start

1. Start a conversation:

   Use chatgpt_start_conversation with title "Code Review Session" and model "gpt-4"

2. Send a message:

   Use chatgpt_send_message with "Can you help me optimize this Python function?"

3. Analyze code:

   Use chatgpt_analyze_code with the code and focus on "performance"

4. Compare approaches:

   Use chatgpt_compare_approaches to compare two different implementations

🔧 Available Tools

Conversation Management

| Tool | Description | Parameters | |------|-------------|------------| | chatgpt_start_conversation | Start a new ChatGPT conversation | title: string, model?: string, system_prompt?: string | | chatgpt_send_message | Send a message in active conversation | message: string, system_prompt?: string | | chatgpt_list_conversations | List all conversations | None | | chatgpt_get_conversation | Get full conversation history | conversation_id: string | | chatgpt_switch_conversation | Switch to different conversation | conversation_id: string | | chatgpt_delete_conversation | Delete a conversation | conversation_id: string | | chatgpt_clear_conversation | Clear messages but keep conversation | None |

Specialized Tools

| Tool | Description | Parameters | |------|-------------|------------| | chatgpt_analyze_code | Analyze code with specific focus | code: string, focus?: enum, language?: string | | chatgpt_compare_approaches | Compare two different approaches | problem: string, approach1: string, approach2: string, criteria?: string[] | | chatgpt_get_models | Get available OpenAI models | None |

📊 Usage Examples

Basic Conversation

// Start a new conversation
await chatgpt_start_conversation({ 
  title: "Python Development", 
  model: "gpt-4",
  system_prompt: "You are a Python expert"
})

// Send a message
await chatgpt_send_message({ 
  message: "How do I implement async generators in Python?" 
})

Code Analysis

// Analyze code for security issues
await chatgpt_analyze_code({
  code: `
def login(username, password):
    query = f"SELECT * FROM users WHERE username='{username}' AND password='{password}'"
    return db.execute(query)
  `,
  focus: "security",
  language: "python"
})

Compare Approaches

// Compare two sorting algorithms
await chatgpt_compare_approaches({
  problem: "Sort a large list of integers",
  approach1: "Use Python's built-in sorted() function",
  approach2: "Implement quicksort from scratch",
  criteria: ["performance", "maintainability", "memory usage"]
})

🎨 Code Analysis Focus Options

• performance - Optimization, complexity analysis, bottlenecks
• security - Vulnerabilities, input validation, authentication
• best practices - Style, patterns, maintainability
• bugs - Logic errors, edge cases, error handling
• general - Comprehensive analysis of all aspects

🔍 Example Outputs

Conversation List

[
  {
    "id": "conv_1234567890_abc123",
    "title": "Python Development",
    "model": "gpt-4",
    "messageCount": 5,
    "createdAt": "2024-01-15T10:30:00Z",
    "updatedAt": "2024-01-15T10:45:00Z",
    "isActive": true
  }
]

Code Analysis Response

Based on my security analysis of your Python login function:

**Critical Security Issues:**
1. **SQL Injection Vulnerability**: The code uses string formatting to build SQL queries, making it vulnerable to SQL injection attacks.
2. **Plain Text Passwords**: The code appears to compare passwords directly, suggesting passwords are stored in plain text.

**Recommendations:**
1. Use parameterized queries or prepared statements
2. Implement password hashing with bcrypt or similar
3. Add input validation and sanitization
...

⚙️ Configuration

Environment Variables

| Variable | Description | Required | |----------|-------------|----------| | OPENAI_API_KEY | Your OpenAI API key | ✅ Yes |

Available Models

• gpt-4 - Most capable, best for complex tasks
• gpt-4-turbo-preview - Faster GPT-4 with 128k context
• gpt-3.5-turbo - Fast and cost-effective

🚨 Error Handling

The server handles common errors gracefully:

• Missing API Key: Clear error message with setup instructions
• API Rate Limits: Passes through OpenAI error messages
• Invalid Model: Lists available models in error
• Network Issues: Descriptive connection errors

🧪 Development

Local Development Setup

# Clone the repository
git clone https://github.com/ChinchillaEnterprises/ChillMCP.git
cd ChillMCP/mcp-chatgpt

# Install dependencies
npm install

# Build the project
npm run build

# Test locally
claude mcp add chatgpt-local -s user -e OPENAI_API_KEY=your_key -- node $(pwd)/dist/index.js

Testing

# Run unit tests
npm test

# Run with coverage
npm run test:coverage

# Watch mode for development
npm run test:watch

🤝 Contributing

1. Fork the repository
2. Create your feature branch: git checkout -b feature/amazing-feature
3. Make your changes and add tests
4. Commit your changes: git commit -m 'Add amazing feature'
5. Push to the branch: git push origin feature/amazing-feature
6. Open a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🔗 Links

• NPM Package: @chinchillaenterprises/mcp-chatgpt
• GitHub Repository: ChillMCP
• OpenAI Documentation: platform.openai.com/docs
• Model Context Protocol: modelcontextprotocol.io

🆘 Support

• Issues: GitHub Issues
• Documentation: Claude Code Docs

Built with ❤️ by Chinchilla Enterprises for AI collaboration