n8n-nodes-rahyana

Reliable AI Chatbot for Rahyana.ir - Simple, memory-enabled AI agent for n8n workflows.

🚀 Features

Core Capabilities

• ✅ Simple Chat - Single input field for user messages
• ✅ Reliable Memory - Proven conversation memory that actually works
• ✅ Tool Support - OpenAI-compatible tools integration
• ✅ Streaming - Real-time streaming responses
• ✅ Debug Output - See exactly what messages are sent to the AI

Ultra-Simple Design

• 🎯 One Input Field - Just ={{$json.chatInput}}
• 🧠 Automatic Memory - Built-in conversation tracking
• 🛠️ Optional Tools - Connect tools input if needed
• 📤 Clean Output - { "content": "AI response", "sessionId": "..." }
• ⚡ Works Immediately - Chat Trigger → Rahyana → Response

📦 Installation

Via npm

npm i n8n-nodes-rahyana

In n8n GUI

1. Go to Settings → Community Nodes
2. Click Add and enter: n8n-nodes-rahyana
3. Click Install

🔑 Credentials Setup

Create credentials named Rahyana API with:

• API Key (required) - Your Rahyana.ir API key
• Base URL (default: https://rahyana.ir/api/v1) - API base URL

🔧 Complete Overhaul - Version 0.6.0

What Was Wrong:

• ❌ Memory was not working - AI couldn't remember conversations
• ❌ Complex input structure with confusing message arrays
• ❌ Unreliable session management
• ❌ Memory being saved in wrong order
• ❌ No debugging to see what was happening

What Was Fixed:

• ✅ Memory Flow: Load → Add user → Send to API → Save assistant response
• ✅ Simple Input: Single chatInput field instead of complex arrays
• ✅ Reliable Sessions: Proper sessionId handling and continuity
• ✅ Debug Output: See exactly what messages are sent to the AI
• ✅ Error Handling: Graceful fallbacks when memory fails

🎯 Simple Usage

Basic Chatbot (No Setup Required)

Chat Trigger → Rahyana → Respond to Webhook

Configuration:

• Model: google/gemini-2.0-flash-lite-001
• Chat Input: ={{$json.chatInput}}
• System Prompt: (optional) Add if needed
• Memory: true (default)
• Session ID: ={{ $json.sessionId }} (default)

Debug Mode: Enable Debug Mode in the node settings to see detailed memory information:

{
  "content": "AI response",
  "sessionId": "session_123",
  "debug": {
    "messagesSent": 5,
    "lastMessage": "what was the number?",
    "memoryKey": "rahyana_memory_session_123",
    "fullConversation": [
      { "role": "user", "content": "hi" },
      { "role": "assistant", "content": "Hi there! How can I help..." },
      { "role": "user", "content": "remember this number..." },
      { "role": "assistant", "content": "Okay, I remember the number..." },
      { "role": "user", "content": "what was the number?" }
    ]
  }
}

Check the n8n console logs for detailed memory debugging information.

Advanced Chatbot with Tools

Chat Trigger → Rahyana (with Tools) → Respond to Webhook

Configuration:

• Tools: Add your tool definitions
• Tool Choice: auto (default)

External Memory Integration

Chat Trigger → Load Memory (Redis) → Rahyana (External Memory) → Save Memory (Redis) → Respond

Configuration:

• Internal Memory: false
• Memory Input: Connect to your memory source

📝 Simple Input Configuration

The node now uses a much simpler input format:

Chat Input:

• Single Field: Just ={{$json.chatInput}}
• Automatic Role: User messages are automatically added to conversation
• Memory Integration: Previous messages are automatically included from memory

System Prompt:

• Optional: Add system instructions if needed
• Automatic: Added to the beginning of the conversation

How Memory Works:

1. First Request: User sends "hi"

   • Load memory: [] (empty)
   • Add user message: [{role: "user", content: "hi"}]
   • Save to memory: [{role: "user", content: "hi"}]
   • Send to AI: [{role: "user", content: "hi"}]
   • AI responds: "Hello! How can I help you today?"
   • Save response: [{role: "user", content: "hi"}, {role: "assistant", content: "Hello!"}]

2. Second Request: User sends "remember 123"

   • Load memory: [{role: "user", content: "hi"}, {role: "assistant", content: "Hello!"}]
   • Add user message: [{role: "user", content: "hi"}, {role: "assistant", content: "Hello!"}, {role: "user", content: "remember 123"}]
   • Save to memory: [full conversation]
   • Send to AI: [full conversation with all messages]
   • AI sees everything and responds appropriately!

🧠 Memory Management

Internal Memory (Default)

• Automatic: Stores conversation history in workflow static data
• Session-based: Separate memory for different conversations
• Context Window: Configurable message limit (default: 20)
• Persistent: Survives workflow restarts
• Session ID: Uses {{ $json.sessionId }} from Chat Trigger by default
• Memory Continuity: Each response includes sessionId for next request

External Memory

• Flexible: Connect to any memory source (Redis, DB, etc.)
• Custom Logic: Implement your own memory management
• Scalable: Handle large-scale deployments

🛠️ Tool Integration

Basic Tools

{
  "name": "get_weather",
  "description": "Get current weather for a location",
  "parameters": {
    "type": "object",
    "properties": {
      "location": {
        "type": "string",
        "description": "City name"
      }
    },
    "required": ["location"]
  }
}

Tool Choice Options

• Auto: Let the model decide when to use tools
• None: Disable tool calling
• Required: Force the model to use a tool

🔄 Streaming Support

Basic Streaming

{
  "stream": true,
  "messages": [{"role": "user", "content": "Tell me a story"}]
}

Streaming Output

• Returns { streamed: true, response: ... }
• Handle streaming in subsequent nodes
• Perfect for real-time chat applications

📚 Example Workflows

1. Simple Chatbot

Chat Trigger → Rahyana → Respond to Webhook

Perfect for: Basic chatbot functionality

2. Chatbot with Tools

Chat Trigger → Rahyana (Tools) → Respond to Webhook

Perfect for: AI agents with capabilities

3. External Memory Chatbot

Chat Trigger → Load Memory → Rahyana → Save Memory → Respond

Perfect for: Production deployments

4. Streaming Chat

Webhook → Rahyana (Streaming) → Webhook (SSE Response)

Perfect for: Real-time applications

🔧 Configuration Options

Basic Settings

• Model: Rahyana model slug
• Messages: Message array with role/content
• System Prompt: Optional system instructions
• Temperature: Response randomness (0-2)
• Max Tokens: Response length limit

Memory Settings

• Internal Memory: Enable/disable built-in memory
• Context Window: Number of messages to keep
• Session ID: Custom session identifier

Advanced Settings

• Stream: Enable streaming responses
• Tools: Tool definitions
• Tool Choice: Tool calling behavior
• Return Raw: Return full API response

🛡️ Security Notes

• API Keys: Stored securely via n8n's encrypted credentials
• Input Validation: All inputs are validated before API calls
• Error Handling: Comprehensive error messages without exposing sensitive data
• Rate Limiting: Built-in retry logic with exponential backoff

📖 Model Selection

Browse available models at: https://rahyana.ir/models

Popular models:

• google/gemini-2.0-flash-lite-001 - Fast, general purpose
• google/gemini-2.5-flash-image-preview - Image generation
• openai/gpt-4o - Advanced reasoning
• anthropic/claude-3.5-sonnet - Creative writing

🤝 Support

• Documentation: https://rahyana.ir/docs
• Issues: https://github.com/rahyana-ai/n8n-nodes-rahyana/issues
• Community: https://community.n8n.io

📄 License

MIT License - see LICENSE file for details.

Built with ❤️ for the n8n community