n8n-nodes-devana

Connect your n8n workflows to Devana AI's powerful RAG and Agentic AI capabilities. Build intelligent automation with access to your knowledge base, documents, and AI agents.

✨ Features

• 🤖 Dynamic Agent Selection - Dropdown list with searchable agent picker
• 📊 Rich Metadata Output - Separate outputs for response and metadata (tokens, sources, scores)
• 🎯 JSON Mode Support - Get structured JSON responses automatically parsed
• 🔄 Dual Outputs - Response on first output, metadata & sources on second output
• 🌍 Multi-language Support - 10+ languages including EN, FR, ES, DE, IT, PT, ZH, AR, HI, ID
• 📁 File Attachments - Upload and attach files to conversations
• 💬 Conversation Management - Continue existing conversations with context
• 🎨 Custom Parameters - Stop sequences, custom metadata, identity override, hidden mode

🚀 Installation

Via n8n Community Nodes (Recommended)

1. Open your n8n instance
2. Go to Settings → Community Nodes
3. Click Install
4. Enter n8n-nodes-devana
5. Click Install

Via npm

npm install n8n-nodes-devana

Then restart n8n.

🔑 Credentials Setup

1. In n8n, go to Credentials → New
2. Search for "Devana API"
3. Fill in:
   • API Key: Your Devana AI API key
   • Base URL: https://api.devana.ai (default)

Getting an API Key

1. Log in to Devana AI
2. Navigate to your agent settings
3. Generate an API key in the "API Access" section

📖 Resources

Chat

Chat Completion

Send messages to your Devana AI agents and get intelligent responses powered by RAG.

Required Parameters:

• Agent ID (Model): Select from your agents list or enter manually
• Messages: At least one message (role + content)

Optional Parameters:

• Temperature (0-1): Control randomness. Higher = more creative
• Max Tokens: Maximum tokens to generate
• Top P (0-1): Nucleus sampling for diversity control
• Stream: Enable streaming responses
• Conversation ID: Continue an existing conversation
• Language: Response language (en, fr, es, de, it, pt, zh, ar, hi, id)
• Response Format:
  • text: Plain text response (default)
  • json_object: Structured JSON output (auto-parsed)
• Files: Comma-separated file IDs to attach
• Frequency Penalty (-2 to 2): Reduce repetition
• Presence Penalty (-2 to 2): Encourage new topics
• Stop Sequences: Comma-separated stop sequences
• Metadata: Custom JSON metadata
• Identity: Custom identity override
• Hidden Mode: No tracking mode

Outputs:

1. Response: The AI message content
2. Metadata & Sources: Usage stats (tokens), sources with RAG scores, finish reason

Agent

Manage your Devana AI agents programmatically.

Operations:

• List: Get all your agents
• Get: Retrieve a specific agent by ID
• Create: Create a new agent
• Update: Update an existing agent
• Delete: Delete an agent

Conversation

Manage conversation history.

Operations:

• Get Messages: Retrieve all messages from a conversation
• Delete: Delete a conversation

File

Upload files to your Devana knowledge base.

Operations:

• Upload: Upload a file from binary data

💡 Example Workflows

Simple Chat

┌─────────────┐      ┌──────────────┐      ┌─────────────┐
│   Webhook   │─────▶│    Devana    │─────▶│  Response   │
│   Trigger   │      │  Chat (GPT)  │      │   Format    │
└─────────────┘      └──────────────┘      └─────────────┘

Advanced RAG Workflow

┌─────────────┐      ┌──────────────┐      ┌──────────────┐      ┌──────────────┐
│   Manual    │─────▶│    Devana    │─────▶│  Filter by   │─────▶│   Send to    │
│   Trigger   │      │  (RAG Mode)  │      │  RAG Score   │      │    Slack     │
└─────────────┘      └──────────────┘      └──────────────┘      └──────────────┘
                              │
                              │ (Metadata Output)
                              ▼
                      ┌──────────────┐
                      │  Log Tokens  │
                      │   & Sources  │
                      └──────────────┘

JSON Mode Example

Set Response Format to "JSON Object" and prompt:

"Return a JSON object with fields: summary, sentiment (positive/negative/neutral), and key_points (array)"

The node will automatically parse the JSON response into an object you can use in subsequent nodes.

📊 Output Structure

Response Output (Output 1)

{
  "message": "The AI response text or parsed JSON object"
}

Metadata Output (Output 2)

{
  "usage": {
    "prompt_tokens": 150,
    "completion_tokens": 85,
    "total_tokens": 235
  },
  "finish_reason": "stop",
  "model": "agent_id",
  "response_format": "text",
  "sources": [
    {
      "content": "Source content...",
      "score": 0.95,
      "metadata": {...}
    }
  ],
  "score": 0.92
}

🎯 Use Cases

Customer Support Automation

Email → Devana AI (RAG on knowledge base) → Auto-reply or escalate to human

Content Generation

Trigger → Devana AI (JSON mode) → Parse structured data → Save to database

Document Analysis

Upload file → Devana AI (with file attachment) → Extract insights → Send report

Multi-language Support

Detect language → Devana AI (with lang parameter) → Translate & respond

🔧 Advanced Features

Token Usage Tracking

Use the Metadata output to track token consumption:

// In a Code node after Devana
const tokens = $input.item.json.usage.total_tokens;
if (tokens > 1000) {
  // Alert or log high usage
}

RAG Score Filtering

Filter responses based on source confidence:

// In a Filter node on Metadata output
const score = $json.score;
return score > 0.8; // Only high-confidence responses

Conversation Threading

Build multi-turn conversations:

// Store conversation_id from first response
// Pass it back in subsequent Devana nodes

🛠️ Development

Build Commands

npm run build     # Compile TypeScript
npm run dev       # Watch mode for development
npm run format    # Format code with Prettier
npm run lint      # Check code with ESLint
npm run lintfix   # Auto-fix ESLint errors

Project Structure

n8n-nodes-devana/
├── credentials/
│   └── DevanaApi.credentials.ts    # API credentials configuration
├── nodes/
│   └── Devana/
│       ├── Devana.node.ts          # Main node implementation
│       └── devana.svg              # Node icon
├── dist/                            # Compiled JavaScript (auto-generated)
├── package.json                     # Package configuration
├── tsconfig.json                    # TypeScript config
└── README.md                        # This file

📚 Resources

• 🌐 Website: devana.ai
• 📖 Documentation: docs.devana.ai
• 💬 Support: support@devana.ai
• 🐛 Issues: GitHub Issues
• 📦 npm: n8n-nodes-devana

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

MIT © Devana AI

---