N8N Tools - Agno Reasoning Agent icon

N8N Tools - Agno Reasoning Agent

Level 3 AI Agent with advanced reasoning - Chain of thought, memory systems, and logic processing powered by Agno framework
Join our community

Overview

The Agno Reasoning Agent node is an advanced AI agent designed for complex problem-solving using step-by-step reasoning, memory systems, and logical verification. It leverages native AI models from providers like OpenAI and Anthropic (Claude) to perform deep analysis, multi-step reasoning, and reflection on input queries or messages.

This node is beneficial in scenarios where users need:

  • Detailed reasoning and explanation of complex problems.
  • Persistent conversational memory to maintain context over multiple interactions.
  • Logical consistency checks and self-reflection to improve answer quality.
  • Integration of various reasoning tools and external knowledge sources for enhanced insights.

Practical examples:

  • Analyzing business strategies by breaking down problems into smaller parts and exploring alternative solutions.
  • Conducting research assistance with memory of previous queries and logical verification of conclusions.
  • Automating decision-making processes that require multi-step reasoning and evidence tracking.

Properties

Name Meaning
Reasoning Agent Recommendation Informational notice about the node's capabilities and configuration.
Model Provider Choose the AI model provider: OpenAI or Anthropic (Claude).
Model Select the specific AI model based on the chosen provider:
- For OpenAI: GPT-4o, GPT-4o Mini, GPT-4 Turbo, GPT-3.5 Turbo
- For Anthropic: Claude 3.5 Sonnet, Claude 3.5 Haiku, Claude 3 Opus
Instructions System instructions guiding the AI assistant’s behavior, emphasizing reasoning and memory usage.
Message Input message or complex query for the agent to analyze and reason about.
Enable Memory Toggle to enable built-in conversation memory and context persistence.
Memory Configuration Settings for conversation memory including:
- Session ID (unique identifier)
- Memory Type (Conversation Buffer, Summary Memory, Window Memory, Token Buffer)
- Max Messages or Tokens limits
- Persistent Storage options (Memory only, N8N Tools DB, Supabase, PostgreSQL, MongoDB, File System)
- External database connection details if applicable
- Enable working memory during reasoning
- Memory retention duration (Session, Persistent, Temporary)
- Context window size in tokens
- Cache previous conclusions for faster processing
Reasoning Configuration Advanced reasoning options such as:
- Reasoning Method (Chain of Thought, Tree of Thought, Reflection, Multi-Step)
- Max reasoning steps
- Enable self-reflection
- Logic verification
Enable Tools Enable or disable Agno native tools integration (Reasoning Tools auto-enabled).
Agno Native Tools Select which native tools to use, including:
- Reasoning Tools
- Knowledge Tools
- Web Search Tools
- Data Analysis Tools
- Finance Tools
- Email Tools
- HTTP Request Tools
- File System Tools
- Database Tools
- Shell Tools
Tool Configurations Specific configurations for enabled tools, e.g., for Reasoning Tools:
- Problem decomposition
- Logical consistency check
- Evidence tracking
- Chain of thought depth (shallow to ultra deep)
- Contradiction detection
- Number of alternative solutions to explore
Advanced Options Additional fine-tuning parameters:
- Temperature (controls randomness)
- Max tokens for response
- Output format (Text, Markdown, JSON)
- Reasoning depth (Shallow, Medium, Deep, Expert)
Report Issue Notice with links to support form and GitHub issues page for reporting bugs or getting help.

Output

The node outputs a JSON object containing:

  • response: The AI-generated reasoning result or answer.
  • agent_type: Fixed value "reasoningAgent".
  • agent_level: Fixed value 3, indicating advanced reasoning level.
  • execution_time_ms: Time taken for the reasoning process in milliseconds.
  • reasoning_steps: Array detailing each reasoning step performed.
  • tools_used: List of tools utilized during reasoning.
  • memory_operations: Details of memory reads/writes during the session.
  • llm_calls: Number of language model calls made.
  • llm_tokens_used: Token usage statistics for the LLM calls.
  • cache_hits: Number of times cached results were used.
  • session_id: Unique identifier for the reasoning session.
  • agno_version: Version of the underlying Agno framework.
  • model_used: The AI model selected.
  • model_provider: The AI model provider selected.
  • logic_verified: Boolean indicating if logic verification passed.
  • reasoning_depth: Depth level of reasoning applied.
  • output_format: Format of the output (text, markdown, json).
  • input_data: Original input data provided to the node.

If an error occurs, the output JSON contains an error field with the error message and a success flag set to false.

The node does not output binary data.

Dependencies

  • Requires an API key credential for the N8N Tools API to authenticate requests.
  • Connects to the Agno backend service at https://n8ntools-agno-production.up.railway.app/api/v1/agent for processing.
  • Optional external databases (Supabase, PostgreSQL, MongoDB) can be configured for persistent memory storage.
  • Uses native Agno tools for ultra-fast reasoning and tool integrations.

Troubleshooting

  • Missing Credentials: If the API key credential is not provided or invalid, the node will throw an error indicating credentials are required.
  • API Errors: Errors returned from the Agno API will appear in the output under the error field. Common causes include network issues, invalid parameters, or exceeding rate limits.
  • Memory Configuration Issues: Incorrect or incomplete external database connection URLs or API keys may cause failures in persistent memory storage.
  • Timeouts: Complex reasoning tasks might hit timeout limits; increasing the timeout or reducing max tokens may help.
  • Logic Verification Failures: If logic verification is enabled, inconsistent reasoning steps may cause errors or warnings.
  • Tool Selection Conflicts: Disabling the mandatory "Reasoning Tools" option may lead to unexpected behavior since it is auto-enabled internally.

To resolve most issues:

  • Ensure valid API credentials are configured.
  • Verify all required parameters and external database settings.
  • Adjust advanced options like temperature, max tokens, and reasoning depth according to task complexity.
  • Use the provided support links to report bugs or seek help.

Links and References

Discussion