Smartsheet Server

🚀 Intelligent Server Development Documentation

The Intelligent Server is a highly modular system that supports development in multiple programming languages (including TypeScript and Python) and communicates with the MCP (Machine Learning Control Protocol) through a unified protocol. This server is designed to handle complex data operations and business logic.

✨ Features

1. Multi - language Support
   • Supports TypeScript and Python code.
   • Provides a cross - language data conversion module.
2. Extension Mechanism
   • Plugin - based architecture for easy function extension.
   • High flexibility and strong customizability.
3. Protocol Compatibility
   • Fully compatible with the MCP protocol.
   • Supports multiple communication modes (including long - connections and polling).
4. Performance Optimization
   • Built - in data caching mechanism.
   • Efficient task processing capabilities.
5. Error Handling
   • Hierarchical error handling architecture.
   • Detailed error log recording.

🚀 Quick Start

📦 Installation

npm install mcp-server --save
pip install mcp-server

💻 Usage Examples

Basic Usage

1. Create a new TypeScript module:

// Calculator.ts
export class Calculator {
    add(a: number, b: number): number {
        return a + b;
    }

    multiply(a: number, b: number): number {
        return a * b;
    }
}

2. Create the corresponding MCP configuration file:

// calculator.json
{
  "name": "Calculator",
  "description": "Basic arithmetic operations.",
  "operations": [
    {
      "name": "add",
      "parameters": [
        { "type": "number", "name": "a" },
        { "type": "number", "name": "b" }
      ],
      "returns": "number"
    },
    {
      "name": "multiply",
      "parameters": [
        { "type": "number", "name": "a" },
        { "type": "number", "name": "b" }
      ],
      "returns": "number"
    }
  ]
}

3. Start the service:

mcp-server calculator.json

Advanced Usage

Batch Processing

The server supports batch processing through the batchSize parameter, with a default value of 500. For example:

const result = await mcpClient.processBatch(batchData, {
  batchSize: 100,
  lenientMode: true
});

Multi - thread Support

The Python version supports multi - thread processing. You can control the number of threads by configuring the threadCount parameter:

from mcp_server import Server

server = Server(thread_count=5)
server.start()

📚 Documentation

Example: Calculator Service

1. Create a TypeScript implementation:

class Calculator {
    add(a: number, b: number): number {
        return a + b;
    }

    multiply(a: number, b: number): number {
        return a * b;
    }
}

2. Define the MCP configuration file:

{
  "name": "Calculator",
  "description": "Basic arithmetic operations.",
  "operations": [
    {
      "name": "add",
      "parameters": [
        { "type": "number", "name": "a" },
        { "type": "number", "name": "b" }
      ],
      "returns": "number"
    },
    {
      "name": "multiply",
      "parameters": [
        { "type": "number", "name": "a" },
        { "type": "number", "name": "b" }
      ],
      "returns": "number"
    }
  ]
}

3. Start the service:

mcp-server calculator.json

Development

Automatically Reconstruct the Development Environment:

npm run watch

Debugging

Since the MCP server communicates through standard input and output, debugging can be challenging. The server implements comprehensive error log recording and provides detailed error information through the MCP protocol.

Key debugging features include:

• Recording error information to the standard error stream.
• Detailed error information in MCP responses.
• Multi - level type verification.
• Comprehensive operation result reports.
• Dependency analysis for column operations.
• Formula reference tracking.

Error Handling

The server adopts a multi - layer error handling approach:

1. MCP Layer
   • Validate tool parameters.
   • Handle protocol - level errors.
   • Provide formatted error responses.
   • Manage timeouts and retries.
2. CLI Layer
   • Validate input parameters.
   • Handle command - line errors.
   • Generate user - friendly error messages.
3. Business Logic Layer
   • Customize error handling.
   • Perform data validation.
   • Implement an error propagation mechanism.

Example: Data Conversion Service

1. Create a TypeScript module:

class DataConverter {
    toUpperCase(str: string): string {
        return str.toUpperCase();
    }

    reverseString(str: string): string {
        return str.split('').reverse().join('');
    }
}

2. Define the MCP configuration file:

{
  "name": "DataConverter",
  "description": "String manipulation operations.",
  "operations": [
    {
      "name": "toUpperCase",
      // The rest of the content in the original JSON block should be continued here if there is more
    }
  ]
}