Gpt5 MCP Server

GPT-5 MCP服务器

 

GPT-5交互和问答工作流的模型上下文协议服务器

简单的MCP服务器，提供提问、继续对话、分析回复和生成GPT-5跟进的工具。

特性

• 5核心工具 -提问、继续对话、分析回答
• 响应分析 -质量评分和反馈
• 后续生成 -自动生成相关问题
• 响应比较 -比较多个响应
• 错误处理 -优雅的API错误处理与测试模式

快速开始

# Clone and install
git clone https://github.com/yourusername/gpt5-mcp-server
cd gpt5-mcp-server
npm install

# Set up environment
cp .env.example .env
# Add your OpenAI API key to .env

# Build and run
npm run build
npm start

安装

1. 克隆存储库：

   git clone https://github.com/yourusername/gpt5-mcp-server
   cd gpt5-mcp-server
   npm install

1. 设置环境：

   cp .env.example .env
   # Edit .env and add your OpenAI API key

1. 构建和测试：

   npm run build
   npm run test-server  # Test mode (no API key required)

核心工具

1. ask_gpt5 -问GPT-5问题

{
  "question": "What is machine learning?",
  "context": "I'm a beginner learning about AI",
  "temperature": 0.7,
  "max_tokens": 500
}

2. 继续对话 -多回合对话

{
  "messages": [
    {"role": "user", "content": "What is Python?"},
    {"role": "assistant", "content": "Python is a programming language..."},
    {"role": "user", "content": "How do I install it?"}
  ]
}

3. 分析响应 -分析响应质量

{
  "response": "Your GPT-5 response here...",
  "criteria": ["Accuracy", "Clarity", "Completeness"]
}

4. 生成后续问题 -生成跟进

{
  "original_question": "What is Python?",
  "response": "Python is a programming language...",
  "count": 3
}

5. 比较响应 -比较多个响应

{
  "question": "What is the best programming language?",
  "responses": [
    {"id": "response_1", "content": "Python is the best because..."},
    {"id": "response_2", "content": "JavaScript is better because..."}
  ]
}

MCP客户端配置

克劳德桌面

增添 ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "gpt5-qa": {
      "command": "node",
      "args": ["/path/to/gpt5-mcp-server/build/index.js"],
      "env": {
        "OPENAI_API_KEY": "your-api-key-here"
      }
    }
  }
}

测试模式（不需要API密钥）

{
  "mcpServers": {
    "gpt5-qa-test": {
      "command": "tsx",
      "args": ["./src/test-server.ts"],
      "cwd": "/path/to/gpt5-mcp-server"
    }
  }
}

使用示例

连接到MCP客户端后：

1. 问一个基本问题：

   Use tool: ask_gpt5
   Arguments: {"question": "What is Python?", "temperature": 0.7}

1. 继续对话：

   Use tool: continue_conversation
   Arguments: {
     "messages": [
       {"role": "user", "content": "What is Python?"},
       {"role": "assistant", "content": "Python is a programming language..."},
       {"role": "user", "content": "How do I install it?"}
     ]
   }

1. 分析响应：

   Use tool: analyze_response
   Arguments: {"response": "Your GPT response here", "criteria": ["Accuracy", "Clarity"]}

发展

以开发模式运行

npm run dev  # Hot reload with tsx

测试模式（无API调用）

npm run test-server  # Uses mock responses

为生产而建

npm run build
npm start

错误处理

服务器包括优雅的错误处理：

• 超出配额：显示带有解决步骤的有用错误消息
• API错误：详细的错误信息和故障排除
• 测试模式：超过配额时无需API调用的完整功能

许可证

MIT许可证-请参阅 许可证 了解详情。

贡献

1. 分叉存储库
2. 创建要素分支
3. 进行更改
4. 测试您的更改
5. 提交拉取请求