Smart MCP Server

🚀 Smart MCP Introduction

Smart MCP (Intelligent Control Panel) is a highly flexible and scalable platform that supports the integration of various tools and services. It can automatically select and execute the optimal combination of tools based on user needs and context information.

🚀 Quick Start

1. Start the Server

Run the main service:

npm run server

2. Test the Context-Aware Selector

Execute the test script:

npm run test

3. Run Example Scripts

Multiple example scripts are provided to demonstrate different functions:

# Test Gemini API integration
node examples/gemini-example.js

# Test the context-aware selector
node test-context-aware-selector.js

# Run the workflow example
node examples/test-workflow.js

✨ Features

1. Multithreading and Task Processing

// Example code: Multithreaded task dispatching
import { SmartMCP } from './SmartMCP.js';

const mcp = new SmartMCP();

// Define and register tool functions
const task1 = () => {
    return 'Task 1 completed!';
};

const task2 = () => {
    return 'Task 2 completed!';
};

mcp.registerTool('task1', task1);
mcp.registerTool('task2', task2);

// Execute tasks in parallel
const result = await mcp.executeParallel(['task1', 'task2']);
console.log(result); // Output: { task1: 'Task 1 completed!', task2: 'Task 2 completed!' }

2. Task Flow System

// Example code: Define and execute a workflow
const workflow = {
    id: "data_processing",
    description: "Data processing workflow",
    steps: [
        { 
            id: "Data cleaning", 
            type: "cleaner",
            parameters: { source: "db.csv" }
        },
        // Other steps...
    ]
};

const result = await mcp.executeWorkflow(workflow);

3. Context-Aware Tool Selection

// Example code: Select tools based on context
const userContext = {
    message: "Help me generate a quarterly report",
    history: ["Open Excel", "Search for data"]
};

const selectedTools = await mcp.selectToolsForContext(userContext);

4. Extension Plugin Support

// Example code: Register and use a new tool
const customTool = () => {
    return 'Custom tool execution completed!';
};

mcp.registerPlugin('customTool', customTool);

const result = await mcp.execute_plugin('customTool');

💻 Usage Examples

Basic Usage

The above code examples in the "Features" section demonstrate the basic usage of Smart MCP, including multithreading task processing, workflow definition and execution, context-aware tool selection, and plugin registration and usage.

Advanced Usage

The following examples further show the application scenarios of Smart MCP:

// Example code: Context-aware tool selection in a more complex scenario
const complexUserContext = {
    message: "I need to analyze the sales data of this quarter and generate a report",
    history: ["Query sales data from the database", "Filter out abnormal data"],
    preferences: ["Use Excel for data visualization"]
};

const advancedSelectedTools = await mcp.selectToolsForContext(complexUserContext);

📚 Documentation

🔌 Available Tool Integrations

AI Tools

• Gemini API: Google's generative AI model, supporting text generation, conversation, etc.
• Step-by-step Reasoning: A step-by-step reasoning tool.
• Web Search: Web search functionality.

Development Tools

• GitHub: Repository management, issue tracking, and PR creation.
• File System: File operations and code editing.
• Database: PostgreSQL integration (planned).

Memory Tools

• Knowledge Graph: Store and retrieve structured information.
• Vector Storage: Semi-structured search and retrieval.

📊 Context-Aware Tool Selection

The context-aware selector recommends the most relevant tools based on the user's message history:

const userContext = {
    message: "Optimize code performance",
    history: ["Run tests", "Check logs"]
};

const selectedTools = await mcp.selectRelevantTools(userContext);

🌐 Workflow System

Define and execute complex task flows:

const workflow = {
    id: "Data analysis pipeline",
    description: "Data processing and analysis pipeline",
    steps: [
        { 
            id: "Data acquisition", 
            type: "fetcher",
            parameters: { source: "Database" }
        },
        { 
            id: "Data cleaning", 
            type: "cleaner",
            parameters: {} 
        },
        // Other steps...
    ]
};

const result = await mcp.executeWorkflow(workflow);

📝 API Documentation

1. Core Interfaces

• mcp.registerTool(name, func): Register a new tool.
• mcp.execute_tool(name, params?): Execute a specified tool.
• mcp.selectRelevantTools(context): Select relevant tools based on the context.
• mcp.executeWorkflow(workflow): Execute a defined workflow.

2. Extension Interfaces

• mcp.registerPlugin(name, func): Register a plugin.
• mcp.usePlugin(name): Use a specified plugin.
• mcp.listPlugins(): List all available plugins.

🔧 Technical Details

Smart MCP uses advanced algorithms and data structures to achieve efficient task processing and tool selection. It leverages multithreading technology to improve the execution efficiency of tasks and uses context-aware algorithms to select the most appropriate tools based on user needs and historical information. The workflow system allows users to define and execute complex task flows, improving the automation and efficiency of work.

🚀 Development and Contribution

We welcome community developers to participate in the project development and jointly improve its functionality. We look forward to your contributions!

1. Code Repository

git clone https://github.com/SmartMCP/smart-mcp.git
cd smart-mcp
npm install

2. Submit Issues

Submit encountered issues or suggestions in GitHub Issues.

3. Submit Contributions

Fork the repository, clone it for local development, and submit a Pull Request after completion.

📄 License

This project is licensed under the MIT License. For specific terms, please refer to the LICENSE file.