Lingmengcan

Lingmengcan AI Platform English | 中文

Lingmengcan is an end-to-end AI application development platform powered by large language models. It provides comprehensive solutions including knowledge base management, intelligent conversations, workflow orchestration, and AI image generation. Built with modern microservice architecture, supporting fully local deployment to ensure enterprise data security.

✨ Key Features

• 🤖 Multi-Model Support: Compatible with OpenAI, Ollama, DeepSeek and other LLMs
• 💬 Intelligent Chat: Multi-turn conversations, context memory, streaming output
• 📚 Knowledge Base RAG: Document upload, vectorization, intelligent retrieval enhancement
• 🎨 AI Image Generation: Integrated Stable Diffusion for text-to-image generation
• 🔄 Workflow Engine: Drag-and-drop visual designer with rich node types and real-time debugging
• 🧩 Plugin Marketplace: Extensible plugin system supporting custom node development
• 👥 Permission Management: Complete RBAC permission system
• 🔒 Private Deployment: Fully local operation without external dependencies

🏗️ Architecture

Frontend Layer (web/)

Vue 3 + TypeScript + Vite
├── UI Framework: TDesign + Tailwind CSS
├── State Management: Pinia
├── Routing: Vue Router
└── HTTP Client: Axios

Service Layer (service/)

NestJS + TypeScript
├── Database: TypeORM + MySQL
├── Authentication: JWT + Passport
├── AI Integration: LangChain + OpenAI API
├── File Storage: Local + Cloud Storage
└── Vector Database: ChromaDB

Core Modules

• User Management: Registration, login, permission control
• Chat System: Multi-turn conversations, history, streaming output
• Knowledge Base: Document parsing, vectorization, similarity search
• Model Management: Multi-model configuration, load balancing, monitoring
• Workflow Engine: Visual designer, node orchestration, real-time debugging
• Plugin Marketplace: Plugin management, dynamic loading, custom extensions
• Drawing System: Stable Diffusion integration, parameter adjustment

🚀 Quick Start

Requirements

| Component | Version | Description | |-----------|---------|-------------| | Node.js | 18+ | Frontend & Backend runtime | | Python | 3.10+ | AI model environment | | MySQL | 8.0+ | Primary database | | pnpm | Latest | Package manager |

1️⃣ Clone Repository

git clone https://github.com/lingmengcan/lingmengcan.git
cd lingmengcan

2️⃣ Database Setup

# Create database
mysql -u root -p -e "CREATE DATABASE lingmengcan_ai CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;"

# Import schema
mysql -u root -p lingmengcan_ai  **⚠️ Important Notice**: The following screenshots are from early versions. Current version has significant UI and feature updates - screenshots need to be updated urgently!

### 💬 Intelligent Chat System
**New Features**:
- ✅ Multi-turn conversations with context understanding
- ✅ Knowledge base RAG enhancement
- ✅ Streaming output with typewriter effect
- ✅ Real-time model switching and comparison
- ✅ Conversation history management
- ✅ Export conversation records

  
  

### 🎨 AI Image Generation Studio
**New Features**:
- ✅ Text-to-Image generation
- ✅ Image style transfer and editing
- ✅ Advanced parameter fine-tuning
- ✅ Batch generation and management
- ✅ ControlNet precise control
- ✅ Image version history management

  

### 🤖 Model Library

  

### ⚙️ System Management Center
Complete enterprise-level admin system

| 👥 User Permission Management | 🔐 Role Permission System |
|-------------------------------|---------------------------|
|  |  |
| • User information management• Permission assignment• Login logs | • RBAC permission model• Fine-grained control• Permission inheritance |

| 📋 Menu Route Management |
|--------------------------|
|  |
| • Dynamic menu configuration• Route permissions• Menu icons |

### 📚 Knowledge Base Management (New Feature)
- 📄 Multi-format document upload (PDF, Word, Markdown)
- 🔍 Intelligent document parsing and chunking
- 🧠 Vectorization storage and retrieval
- 💡 Knowledge base Q&A with citations
- 📊 Usage statistics
- 🔄 Version management and rollback

### 🔄 Workflow Engine (New Feature)
Visual drag-and-drop workflow designer powered by Vue Flow:

**Designer Features**:
- 🎨 Drag-and-drop canvas with node connections
- 🔍 Canvas zoom, auto-layout, minimap navigation
- ↩️ Undo/Redo history support
- 💾 Auto-save (30s interval)
- 📤 Import/Export workflows (JSON format)

**Rich Node Types**:
| Node Type | Description |
|-----------|-------------|
| **LLM Node** | LLM API calls with model selection, temperature/top_p/max_tokens, system/user prompts |
| **Condition Node** | IF/ELIF/ELSE branches with AND/OR logic combinations |
| **HTTP Node** | External API calls (GET/POST/PUT/DELETE), headers, auth, timeout/retry |
| **Loop Node** | Three loop modes: for (count), while (condition), foreach (iterate) |
| **Parallel Node** | Parallel branch execution with strategies: wait all/any/race mode |
| **Database Node** | Database CRUD operations with multi-datasource support |
| **Transform Node** | Data format transformation and processing |

**Real-time Debugging**:
- 🐛 Integrated debug panel
- 📊 Stream/non-stream execution modes
- 📝 Execution logs viewer
- 🖼️ Text and image input support

### 🧩 Plugin Marketplace (New Feature)
Extensible plugin system for custom workflow nodes:

- 📦 **Plugin Categories**: AI plugins, Rule plugins, Logic plugins, HTTP plugins, Custom plugins
- 🔌 **Dynamic Loading**: Auto-load enabled plugins in workflow designer
- ⚙️ **JSON Schema Config**: Define node configuration forms via JSON Schema
- 📋 **Version Control**: Plugin versioning and author tracking
- 🎯 **Search & Filter**: Quick search and category filtering

## 📁 Project Structure

lingmengcan/ ├── 📁 web/ # Frontend (Vue 3 + TypeScript) │ ├── src/ │ │ ├── api/ # API layer │ │ ├── components/ # Reusable components │ │ ├── views/ # Page views │ │ ├── store/ # State management (Pinia) │ │ ├── router/ # Route configuration │ │ └── utils/ # Utility functions │ └── package.json ├── 📁 service/ # Backend (NestJS + TypeScript) │ ├── src/ │ │ ├── controllers/ # Controller layer │ │ ├── services/ # Business logic layer │ │ ├── entities/ # Data models │ │ ├── modules/ # Feature modules │ │ ├── dtos/ # Data transfer objects │ │ └── utils/ # Utility classes │ └── package.json ├── 📁 doc/ # Documentation │ ├── lingmengcan-ai.sql # Database schema │ └── *.md # Documentation files ├── 📁 images/ # Screenshots └── README.md

## 🔧 Configuration

### Backend Config (`service/config.development.yaml`)

```yaml
# Database configuration
database:
  host: localhost
  port: 3306
  username: root
  password: your_password
  database: lingmengcan_ai

# AI model configuration
llm:
  openai:
    apiKey: your_openai_key
    baseURL: https://api.openai.com/v1
  
  ollama:
    baseURL: http://localhost:11434

# Stable Diffusion configuration
stablediffusion:
  apiUrl: http://localhost:7860

Frontend Config (web/.env.development)

# API base URL
VITE_API_BASE_URL=http://localhost:3000

# Application title
VITE_APP_TITLE=Lingmengcan AI Platform

🤝 Contributing

We welcome all forms of contributions!

1. 🍴 Fork the repository
2. 🌿 Create feature branch (git checkout -b feature/AmazingFeature)
3. 💾 Commit changes (git commit -m 'Add some AmazingFeature')
4. 📤 Push to branch (git push origin feature/AmazingFeature)
5. 🔀 Open Pull Request

Development Guidelines

• Code Style: ESLint + Prettier
• Commit Convention: Conventional Commits
• Test Coverage: Add tests for new features

📄 License

This project is licensed under the [MIT](LICENSE) License

🌟 Star History

[](https://star-history.com/#lingmengcan/lingmengcan&Date)

💬 Community

• 💬 QQ Group: 651535270
• 📧 Email: lingmengcan@example.com
• 🐛 Issues: GitHub Issues

🙏 Acknowledgments

Thanks to the following open source projects:

• Vue.js - Progressive JavaScript Framework
• NestJS - Enterprise Node.js Framework
• LangChain - AI Application Development Framework
• Stable Diffusion - AI Image Generation
• TDesign - Vue 3 Component Library

---

⭐ If this project helps you, please give us a Star! Made with ❤️ by Lingmengcan

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: lingmengcan
• Source: lingmengcan/lingmengcan
• License: MIT

Install and usage instructions live in the source repository linked above.