skingflow

SkinFlow Multi-Agent Framework

🚀 Flexible flow engine for intelligent multi-agent applications - Supports complex task decomposition, intelligent planning, memory management, and tool integration

📖 Table of Contents

• Quick Start
• Core Features
• Architecture Overview
• Documentation
• Examples
• Language Support

🚀 Quick Start

5-Minute Quick Experience

# 1. Clone the repository
git clone https://github.com/skingko/skingflow.git
cd skingflow

# 2. Install dependencies
npm install

# 3. Configure environment variables
cp .env.example .env
# Edit .env file to set your LLM API key and database connection

# 4. Run the example
node examples/quick-start/index.js

Simple Usage Example

import { createMultiAgentFramework } from './lib/multi-agent/index.js';

// Create framework instance
const framework = await createMultiAgentFramework({
  llm: {
    provider: 'http',
    baseUrl: 'https://api.openai.com/v1/chat/completions',
    apiKey: 'your-api-key',
    model: 'gpt-4'
  },
  memory: {
    storage: {
      type: 'postgres',
      config: {
        host: 'localhost',
        database: 'skingflow',
        user: 'postgres',
        password: 'your-password'
      }
    }
  }
});

// Process request
const result = await framework.processRequest(
  "Create a simple web application",
  { userId: 'user123' }
);

console.log(result);

✨ Core Features

🧠 Intelligent Multi-Agent System

• Planning Agent: Automatically decomposes complex tasks and creates execution plans
• Professional Sub-Agents: Specialized in research, programming, data analysis, content creation
• Intelligent Coordination: Automatically selects the most suitable agents for specific tasks
• Context Isolation: Ensures secure collaboration between agents

💾 Advanced Memory System (mem0-based architecture)

• Short-term Memory: Session context and temporary information management
• Long-term Memory: Persistent knowledge storage and historical records
• User Preferences: Personalized settings and habit learning
• Semantic Search: Vector-based intelligent memory retrieval

🛠️ Unified Tool System

• YAML/XML Tool Definitions: Declarative tool configuration, easy to extend
• Virtual File System: Secure file operation environment
• MCP Protocol Support: Standardized tool integration
• Custom Tools: Flexible tool development and integration mechanism

🔄 Stream Processing Engine

• Asynchronous Stream Processing: High-performance concurrent execution
• Real-time Response: Supports streaming output and real-time feedback
• Workflow Orchestration: Intelligent management of complex workflows

🛡️ Enterprise-Grade Reliability

• Degradation Mechanism: Multi-layer error recovery strategies
• Circuit Breaker: Automatic fault isolation and recovery
• Health Monitoring: Real-time system status tracking
• Detailed Logging: Complete debugging and audit information

🏗️ Architecture Overview

┌─────────────────────────────────────────────────────────────┐
│                    SkinFlow Framework                        │
├─────────────────────────────────────────────────────────────┤
│  Multi-Agent System                                         │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐         │
│  │ Planning    │  │ Sub-Agents  │  │ Coordination│         │
│  │ Agent       │  │ Manager     │  │ System      │         │
│  └─────────────┘  └─────────────┘  └─────────────┘         │
├─────────────────────────────────────────────────────────────┤
│  Core Services                                              │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐         │
│  │ LLM         │  │ Memory      │  │ Tool        │         │
│  │ Abstraction │  │ System      │  │ Registry    │         │
│  └─────────────┘  └─────────────┘  └─────────────┘         │
├─────────────────────────────────────────────────────────────┤
│  Infrastructure                                             │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐         │
│  │ Fallback    │  │ Virtual     │  │ Stream      │         │
│  │ Manager     │  │ FileSystem  │  │ Engine      │         │
│  └─────────────┘  └─────────────┘  └─────────────┘         │
└─────────────────────────────────────────────────────────────┘

📚 Documentation

• Installation Guide - Detailed installation and configuration steps
• Basic Usage Tutorial - Step-by-step usage tutorial
• Advanced Configuration - Advanced features and custom configuration
• Tool System - Tool development and integration guide
• Memory System - Detailed memory management explanation
• Agent System - Multi-agent collaboration mechanism
• API Reference - Complete API documentation
• Best Practices - Production environment recommendations
• Troubleshooting - Common problem solutions

🎯 Examples

• Quick Start - Simplest usage example
• Intelligent Assistant - Complete intelligent assistant application
• Content Creation - Automated content generation
• Data Analysis - Intelligent data processing
• Web App Generator - Automated web development

🌐 Language Support

This project supports multiple languages:

• 🇺🇸 English - README.md
• 🇨🇳 中文 - README.zh.md
• 🇪🇸 Español - README.es.md
• 🇫🇷 Français - README.fr.md
• 🇩🇪 Deutsch - README.de.md

📖 Online Documentation: skingflow-docs.pages.dev

🚀 Production Ready

The SkinFlow framework is fully tested with the following production features:

• ✅ High Availability: Complete error handling and degradation mechanisms
• ✅ High Performance: Asynchronous stream processing and intelligent caching
• ✅ Scalable: Modular architecture, easy to extend
• ✅ Monitorable: Detailed logs and statistics
• ✅ Security: Virtual file system and permission control

📊 Benchmark

Metric	Performance
Simple request response time	< 2 seconds
Complex task processing time	< 30 seconds
Concurrent processing capability	100+ requests/minute
Memory usage	< 512MB
Success rate	> 95%

🤝 Contributing

We welcome community contributions! Please check the Contributing Guide to learn how to participate in project development.

📞 Support

• Documentation: Complete Documentation
• Examples: Example Code
• Issue Feedback: GitHub Issues
• Discussions: GitHub Discussions

📄 License

This project is open source under the MIT License. See LICENSE file for details.

🎉 Start using SkinFlow to build your intelligent applications!