MCP Lwh

🚀 MCP Service Tools

This is a collection of MCP (Model Context Protocol) service tools for invoking various AI service APIs.

🚀 Quick Start

This MCP service toolset is designed to simplify the process of interacting with multiple AI service APIs. It provides a unified interface for different tasks across various AI platforms.

✨ Features

Hugging Face MCP Service

• Supports tasks such as text generation, classification, question - answering, mask filling, summarization, and translation.
• Supports image classification, image segmentation, image - to - text conversion, and zero - shot image classification.
• Offers speech recognition functionality.

Dify MCP Service

• Provides multiple AI application interfaces. Specific parameters and return values may vary depending on the configuration.
• Refer to the relevant documentation for specific features.

📦 Installation

1. Install Dependencies

Use the Node.js package manager npm to install:

npm install mcp-service-tools

2. Configure Environment Variables

Set the corresponding API keys:

export HUGGING_FACE_KEY='your-api-key'
export DIFY_API_KEY='your-api-key'

💻 Usage Examples

Basic Usage - Hugging Face Tools

const { hfTool } = require('mcp-service-tools');

// Text Generation
await hfTool.textGeneration('gpt2', 'Input text', { max_new_tokens: 50 });

// Text Classification
await hfTool.textClassification('distilbert-base-uncased-finetuned-sst-2-english', 'This is a piece of text');

// Question Answering
await hfTool.questionAnswering('deepset/roberta-base-squad2', {
  question: 'Question?',
  context: 'Context containing the answer'
});

// Fill Mask
await hfTool.fillMask('bert-base-uncased', 'Paris is the [MASK] of France.');

// Summarization
await hfTool.summarization('facebook/bart-large-cnn', 'Long text content');

// Translation
await hfTool.translation('t5-base', 'Text to be translated');

// Image Classification
await hfTool.imageClassification('google/vit-base-patch16-224', imageBlob);

// Image Segmentation
await hfTool.imageSegmentation('facebook/detr-resnet-50-panoptic', imageBlob);

// Image to Text
await hfTool.imageToText('nlpconnect/vit-gpt2-image-captioning', imageBlob);

// Zero - Shot Image Classification
await hfTool.zeroShotImageClassification(
  'openai/clip-vit-base-patch32',
  imageBlob,
  ['Cat', 'Dog', 'Bird']
);

// Zero - Shot Text Classification
await hfTool.zeroShotClassification(
  'facebook/bart-large-mnli',
  'This is a piece of text',
  ['Category 1', 'Category 2', 'Category 3'],
  false
);

// Automatic Speech Recognition
await hfTool.automaticSpeechRecognition('facebook/wav2vec2-base-960h', audioBlob);

Advanced Usage - Dify Tools

const { difyTool } = require('mcp-service-tools');

// Example call (refer to the official documentation for specific parameters)
await difyTool.someFunction('parameter1', 'parameter2');

📄 License

This project is licensed under the MIT license.

⚠️ Important Note

You need to obtain the corresponding API keys before use. Please note the API call limits and costs. The specific parameters and return values of the Dify API may vary depending on the application configuration.