N8N Tools - Agno Agent

⚡ Ultra-fast AI Agent powered by the Agno framework - Multi-agent systems with ~3μs performance, advanced reasoning, memory, and multi-modal capabilities.

🚀 Features

🎯 Agent Levels (Progressive Complexity)

• ⚡ Level 1: Basic Agent - Tools + instructions with ultra-fast execution
• 📚 Level 2: Knowledge Agent - Knowledge base + storage capabilities
• 🧮 Level 3: Reasoning Agent - Memory + advanced reasoning capabilities
• 👥 Level 4: Team Agent - Collaborative multi-agent team systems
• 🚀 Level 5: Workflow Agent - State management + automation workflows

⚡ Performance & Capabilities

• 🏆 Ultra-fast Performance: ~3μs agent instantiation (powered by Agno)
• 🌐 23+ Model Providers: Anthropic, OpenAI, Google, Groq, Perplexity, Ollama, and more
• 🎬 Multi-Modal: Native support for text, images, audio, and video
• 🧠 Advanced Reasoning: Built-in reasoning tools and logic capabilities
• 💾 Smart Memory: Conversation context and persistent long-term memory
• 🛠️ Native Tools: 10+ ultra-fast tool categories optimized for performance
• 👥 Team Collaboration: Multi-agent systems with different collaboration strategies
• 🔄 Real-time Streaming: Live response streaming capabilities
• 🛡️ Fallback Protection: Automatic fallback to secondary models

📦 Installation

1. Install the package:

   npm install n8n-nodes-n8ntools-agno-agent

2. Add to your N8N installation or restart N8N to load the node.

3. Configure your N8N Tools API credentials in N8N.

🔧 Configuration

🔌 Required Inputs

• 🤖 AI Language Model (Required): Connect any N8N LLM node (Claude, OpenAI, Gemini, etc.)
• 📊 Main Data (Standard): Regular workflow data flow

🔑 Required Credentials

• N8N Tools API: Get your API key from n8ntools.io

💡 Simplified Architecture

The Agno Agent has 2 clean inputs:

1. 🤖 AI Language Model (Required)

   • Connect Claude, OpenAI, Gemini, Groq, or any LLM node
   • Automatically inherits model configuration (temperature, max tokens, etc.)
   • Supports 23+ model providers for maximum flexibility

2. 📊 Main Data (Standard)

   • Regular data flow from previous nodes
   • Used as context and input for agent processing

🛠️ Native Tools (Built-in)

Unlike traditional agents, Agno uses native ultra-fast tools selected via dropdown:

• 🧮 Reasoning Tools - Advanced logic and chain-of-thought processing
• 🔍 Web Search - Search engines and web information retrieval
• 📊 Data Analysis - Statistical analysis and data processing
• 💰 Finance Tools - Stock prices and market analysis (YFinance)
• 📧 Email Tools - Send emails and manage communications
• 🌐 HTTP Request - Make API calls and HTTP requests
• 📁 File System - Cloud storage operations (S3, R2, MinIO, GCS)
• 🗃️ Database - SQL queries and database operations
• 🧠 Knowledge - Knowledge base search and retrieval
• 🔄 Shell Tools - Execute shell commands and scripts

🎯 Agent Types

⚡ Basic Agent (Level 1)

Simple agent with tools and instructions - Ultra-fast execution for straightforward tasks.

📚 Knowledge Agent (Level 2)

Agent with knowledge base integration and storage capabilities for information retrieval.

🧮 Reasoning Agent (Level 3)

Advanced agent with memory and reasoning capabilities for complex problem-solving.

👥 Team Agent (Level 4)

Collaborative multi-agent team system with different strategies:

• Sequential: Agents work in sequence
• Parallel: Agents work simultaneously
• Hierarchical: Manager-worker structure
• Democratic: Consensus-based decisions

🚀 Workflow Agent (Level 5)

Agentic workflow with state management and automation:

• Linear: Step-by-step execution
• Branching: Conditional paths
• Loop: Iterative processing
• State Machine: Complex state management

☁️ Cloud Storage Support

File System Tools support multiple cloud storage providers:

• Amazon S3 - Native AWS S3 storage
• Cloudflare R2 - Cost-effective S3-compatible storage
• MinIO - Self-hosted S3-compatible storage
• Google Cloud Storage - GCS with S3-compatible API

🎨 Examples

⚡ Simple Setup (Current Architecture)

{
  "workflow": "Ultra-fast Agno Agent",
  "nodes": {
    "llm": "Claude 3.5 Sonnet node → AI Language Model input",
    "agno_agent": {
      "agentType": "reasoningAgent",
      "instructions": "You are a research assistant with web search and reasoning capabilities.",
      "message": "Research quantum computing developments and analyze the data",
      "agnoNativeTools": ["reasoning", "webSearch", "dataAnalysis"],
      "enableMemory": true,
      "memoryKey": "quantum_research"
    }
  }
}

🏢 Enterprise File Processing

{
  "agentType": "basicAgent",
  "instructions": "Process documents from cloud storage and analyze content.",
  "message": "Extract insights from uploaded reports",
  "agnoNativeTools": ["fileSystem", "dataAnalysis", "reasoning"],
  "toolsConfig": {
    "fileSystemConfig": {
      "storageType": "s3",
      "s3BucketName": "company-documents",
      "s3Region": "us-east-1"
    }
  }
}

Team Agent Configuration

{
  "agentType": "teamAgent",
  "teamConfig": {
    "teamSize": 3,
    "collaborationStrategy": "sequential",
    "specialistRoles": ["researcher", "contentWriter", "designer"]
  }
}

Advanced Options

{
  "advancedOptions": {
    "temperature": 0.7,
    "maxTokens": 4000,
    "enableMemory": true,
    "memoryKey": "conversation_1",
    "enableStreaming": true,
    "outputFormat": "markdown",
    "fallbackModel": "gpt-4o-mini"
  }
}

🔗 Model Providers

Supported Providers

• Anthropic: Claude Sonnet, Haiku, Opus
• OpenAI: GPT-4, GPT-3.5, o1 models
• Google: Gemini Pro, Flash, Ultra
• Groq: Ultra-fast inference
• Perplexity: Search-powered responses
• Ollama: Local models (Llama, Mistral, etc.)
• 20+ Others: Including Azure OpenAI, AWS Bedrock, Hugging Face, etc.

📊 Response Format

{
  "response": "Agent's response text",
  "usage": {
    "promptTokens": 150,
    "completionTokens": 300,
    "totalTokens": 450
  },
  "model": "claude-3-5-sonnet-20241022",
  "agentType": "basicAgent", 
  "executionTime": 1250,
  "reasoning": ["Step 1: Analysis", "Step 2: Synthesis"],
  "toolsUsed": ["reasoning", "dataAnalysis"],
  "memoryKey": "conversation_1"
}

🚀 Performance

• Agent Instantiation: ~3 microseconds (Agno framework optimization)
• Memory Usage: ~6.5KiB per agent instance
• Concurrent Agents: Supports hundreds of simultaneous agents
• Response Time: Optimized for sub-second responses

🆚 vs N8N LangChain Agent

🔌 Architecture Comparison

N8N LangChain Agent (Complex):

• Inputs: LLM + Memory + Tools + Main Data (4 connections)
• Processing: LangChain framework (slower)
• Setup: Multiple node connections required

N8N Tools Agno Agent (Simplified):

• Inputs: LLM + Main Data (2 connections) ✨
• Processing: Agno framework (~3μs execution) ✨
• Setup: Simple connection, native tools built-in ✨
• Memory: Built-in + configurable persistence ✨
• Tools: Native ultra-fast tools via dropdown ✨

📚 Documentation

• Official Documentation
• Agno Framework
• API Reference
• Examples Repository

🤝 Support

• Community: Discord
• Documentation: docs.n8ntools.io
• Issues: GitHub Issues
• Email: support@n8ntools.io

📄 License

MIT License - see LICENSE file for details.

Powered by Agno Framework | Created by N8N Tools