Node-RED Make IoT Smart

🌐 Languages

An AI intelligent assistant extension designed specifically for Node-RED, making IoT development smarter and more efficient.

Overview

Node-RED Make IoT Smart is a comprehensive AI agent designed specifically for Node-RED development. It provides intelligent code assistance, automated flow optimization, and smart debugging capabilities to enhance your IoT development experience. The extension now supports six major scenarios: Learning, Solution, Integration, Development, Configuration, and Management.

Features

🤖 AI Assistant

• Intelligent Code Suggestions: Provides context-aware code recommendations for Node-RED flows.
• Smart Flow Analysis: Analyzes flows and provides optimization suggestions.
• Natural Language Interface: Interact with the Node-RED environment using natural language commands.
• Multi-language Support: Supports Chinese, English, Japanese, Korean, and other languages. Follows Node-RED language configuration changes.
• Multi-provider Support: Based on LangChain.js framework, supports OpenAI, Anthropic, Google, DeepSeek, and other AI models.
• Intelligent Memory Management: SQLite-based short-term and long-term memory system, supporting conversation history, user preferences, and flow template storage.
• Scenario-based Prompts: JSON-configured scenario-based prompt management with dynamic parameter injection support.
• MCP Tools Integration: Supports Model Context Protocol (MCP) tool calls to extend AI assistant capabilities.

🔧 Development Tools

• Real-time Code Analysis: Continuously analyzes Node-RED flows.
• Configuration Management: Centralized API configuration for different AI providers.
• Interactive Sidebar: Dedicated AI assistant panel integrated into the Node-RED editor.
• JSON Editor: Built-in configuration file editor with syntax highlighting.
• MCP Tool Integration: Supports Model Context Protocol (MCP) tool calls to extend AI assistant capabilities.
• LangChain Tool Management: Unified tool management framework supporting built-in tools and MCP tools.
• Scenario-based Support: Provides customized support for seven major scenarios:
  • Learning: Explains nodes and concepts, provides example flows.
  • Solution: Provides multiple IoT solutions with flow JSON and node installation guidance.
  • Integration: Assists with protocol integration (e.g., MQTT, Modbus) or software.
  • Development: Optimizes existing flows and function node code.
  • Configuration: Guides modification of Node-RED settings (e.g., settings.js).
  • Management: Supports remote access, Git integration, and batch deployment.

🚀 Upcoming Features

• Remote Debugging: AI-assisted remote debugging of Node-RED flows.
• Team Management: Collaborative development with team management features.
• Advanced Analytics: Deep insights into IoT system performance.
• Smart Deployment: AI-guided deployment strategies for IoT applications.

Installation

Install from npm

npm install @jhe.zheng/node-red-make-iot-smart

Install from Node-RED Palette Manager

1. Open the Node-RED editor.
2. Go to Menu → Manage Palette.
3. Search for @jhe.zheng/node-red-make-iot-smart.
4. Click Install.
5. After installation is complete, restart Node-RED.
6. After installation, you will see a new AI Assistant tab in the Node-RED sidebar.
7. Click the Configure button to set up your AI provider.
8. Choose from supported providers:
   • DeepSeek: Cost-effective choice with strong coding capabilities.
   • OpenAI: Industry-leading GPT models.
   • Anthropic: Advanced reasoning capabilities with Claude models.
9. Enter your API key and select the appropriate model.
10. After configuration is complete, you can start using the AI assistant. Note that after saving the configuration, Node-RED will automatically generate a configuration node, and Node-RED will prompt for flow changes - just click merge.
11. Start interacting with your AI assistant!

Quick Start

Type "Analyze current node"

Type "Create a sample flow"

Type "Health check"

Configuration

LangSmith Debug Configuration (Optional)

For better debugging and monitoring of LangChain execution processes, you can configure LangSmith support:

1. Copy the .env.example file to .env:

   cp .env.example .env

2. Edit the .env file and fill in your LangSmith configuration:

   LANGCHAIN_TRACING_V2=true
   LANGCHAIN_API_KEY=your_langsmith_api_key_here
   LANGCHAIN_PROJECT=your_project_name

3. Restart Node-RED to apply the configuration.

4. Visit LangSmith to view detailed execution traces and debugging information.

Note: LangSmith configuration is optional and will not affect basic functionality if not configured.

Usage

Basic Chat Interface

• Open the AI Assistant sidebar tab.
• Enter your questions or instructions in natural language.
• Get intelligent responses including code suggestions and explanations.

Scenario Selection

• Select scenarios (Learning, Solution, Integration, Development, Configuration, Management) through the dropdown menu in the sidebar.
• AI customizes responses based on the selected scenario, providing relevant tools and flow JSON.

JSON/Code Handling

• Large JSON or code outputs are hidden behind View JSON/Code buttons to keep the UI clean.
• Edit flow JSON in the built-in editor with syntax highlighting and apply changes directly.

Supported Scenarios

Scenario Overview

Preset Prompt Examples

Smart Keyword Triggers

Dynamic Input Parameters

All scenarios support the following dynamic parameter injection:

• nodeRedVersion - Node-RED version information
• nodeVersion - Node.js version information
• currentTime - Current timestamp
• selectedFlow - Currently selected flow
• selectedNodes - Currently selected nodes
• lang - Current language setting
• mcpTools - Available MCP tools list

Each scenario also supports specific dynamic parameters:

• Learning Mode: userLevel (user skill level)
• Solution Mode: projectRequirements (project requirements)
• Integration Mode: integrationTargets (integration targets)
• Development Mode: developmentTask (development task)
• Configuration Mode: configurationNeeds (configuration needs)
• Management Mode: projectStatus (project status)

System Prompt Features

Each scenario is configured with specialized system prompts to ensure the AI assistant can:

1. Role Positioning: Clearly define its professional role in specific scenarios
2. Output Format: Provide structured response formats according to scenario needs
3. Tool Integration: Intelligently call appropriate MCP tools and Node-RED APIs
4. Context Awareness: Utilize dynamic parameters to provide personalized recommendations

Supported AI Providers

API Configuration

• API keys are stored locally and encrypted.
• Support for multiple provider configurations.
• Easy switching between different providers and models.
• Separate model settings for planning and execution phases.

Development

Project Structure

├── ai-sidebar.html          # Main sidebar interface
├── ai-sidebar-config.json   # UI configuration
├── make-iot-smart.html      # Node configuration template
├── make-iot-smart.js        # Backend node implementation
├── lib/
│   ├── langchain-manager.js # LangChain core manager
│   ├── memory-manager.js    # SQLite memory management
│   └── scenario-manager.js  # Scenario-based prompt management
├── config/
│   └── scenarios.json       # Scenario configuration file
├── data/
│   └── memory.db           # SQLite database file
└── package.json            # Package configuration

Technical Architecture

This project is built on the LangChain.js framework with a modular architecture design:

• LangChain Manager: Core AI model management, supporting multiple LLM providers
• Memory Manager: SQLite-based intelligent memory system, supporting short-term and long-term memory
• Scenario Manager: Scenario-based prompt management, supporting JSON configuration and dynamic parameters
• Tool Manager: Unified tool management framework, integrating MCP tools and built-in tools
• API Layer: RESTful API interface, supporting streaming chat and tool execution

Contributing

1. Fork the repository.
2. Create a feature branch.
3. Make changes and commit.
4. Submit a pull request.

Roadmap

Phase 1 (Completed)

• ✅ AI assistant integration
• ✅ Multi-provider support
• ✅ Interactive sidebar
• ✅ Configuration management
• ✅ Scenario-based support
• ✅ LangChain.js architecture migration
• ✅ SQLite memory management system
• ✅ MCP tool integration
• ✅ Unified tool management framework

Phase 2 (Coming Soon)

• 🔄 Remote debugging functionality
• 🔄 Team collaboration features
• 🔄 Advanced flow analysis
• 🔄 Smart deployment tools

Phase 3 (Future)

• 📋 Team management system
• 📋 Enterprise features
• 📋 Advanced security options
• 📋 Custom model training

System Requirements

• Node.js >= 18.0.0
• Node-RED >= 2.0.0

License

Licensed under the MIT License. See LICENSE file for details.

Support

AI development is more like art than technology. Mastering LLMs is not a simple task and requires deep understanding of AI models, data, and application scenarios. Each Q&A session may yield different results. Initial versions are often imperfect, but with the improvement of prompt engineering, they will gradually truly meet the daily needs of Node-RED users, whether IT or OT engineers. We welcome more interested people to join the project.

• Issue Reports: GitHub Issues
• Documentation: Wiki
• Discussions: GitHub Discussions

Author

Zheng He - Email: jhe.zheng@gmail.com - GitHub: @jimmyfreecoding - Website: https://www.makeiotsmart.com

Let AI-driven assistance make your IoT development smarter!