Interactive AI Agent with Custom Chat Interface and Multi-Tool Capabilities - n8n Workflow

n8n users seeking advanced demonstrations of AI agent architecture.
Developers needing a customizable, conversational frontend for complex back-end automations.
Businesses aiming to deploy a specialized, low-code internal helper bot.
Users looking for ready-to-use n8n templates that integrate LangChain concepts and specialized n8n tools.

This comprehensive n8n workflow provides the complete infrastructure required to deploy a functional, multi-tool AI Agent. The agent is built using specialized n8n node types that adhere to LangChain principles, allowing the chosen Large Language Model (LLM, default Gemini) to intelligently decide when to use specific tools to fulfill complex, dynamic requests. Unlike basic conversational interfaces, this n8n template integrates short-term memory and a diverse toolbox, enabling the agent to perform tasks like fetching live RSS feeds, calculating financial details, generating secure passwords, and looking up Wikipedia summaries. This demonstrates how the n8n automation platform can orchestrate sophisticated AI interactions, offering significant value to users looking to build advanced conversational applications.

1. Input (n8n trigger): The process starts with the 'Example Chat Window' n8n trigger, which acts as the publicly accessible, custom-styled chat interface. User messages initiate the entire n8n workflow execution.


2. Context & Reasoning: The message is passed to the 'Your First AI Agent' n8n node. Before processing, the 'Simple Memory' node provides the conversation history (context window of 30 messages) to the LLM ('Gemini' or 'OpenAI').


3. Decision Making: The Agent n8n node uses the LLM to analyze the user's intent against the detailed system message instructions. It intelligently determines whether the request requires an external action (a tool) or a direct conversational response.


4. Tool Execution: If a tool is necessary (e.g., asking for a joke or a loan calculation), the Agent dynamically calls the specific tool n8n node (like getajoke or calculateloanpayment). The tool executes the specific automation logic.


5. Output: The Agent receives the tool's result, synthesizes a polite and coherent response using its conversational instructions, and returns the final output back to the user via the initial Example Chat Window n8n trigger, completing the cycle of the n8n workflow.

1. Import the n8n workflow: Copy the provided JSON data and paste it into your n8n canvas using the 'New' -> 'Import from JSON' function.


2. Configure LLM Credentials: Locate the 'Gemini' (or 'OpenAI') n8n node. Select or create your required API credential. If you prefer using OpenAI, disconnect the link from the Gemini node to the Agent and connect the OpenAI node instead, ensuring the OpenAI n8n node is enabled.


3. Review Agent Instructions: Navigate to the 'Your First AI Agent' n8n node and review the extensive 'System Message'. This text defines the agent's personality, goal, and tool usage rules, and should be customized to your application's needs.


4. Activate and Test: Save and activate the n8n workflow. Copy the webhook URL from the 'Example Chat Window' n8n trigger and open it in your browser to start chatting with your personalized AI Agent.

Example Chat Window (n8n trigger): The user-facing conversational interface. It uses the Chat Trigger n8n node type and is highly customized using CSS to provide a 'Glass & Glow' aesthetic. It serves as the primary n8n trigger for the agent.
Gemini (LLM): The Large Language Model selected for powering the agent's intelligence. Configured to use models/gemini-2.5-flash for low-latency reasoning.
Simple Memory: A Memory Buffer Window n8n node set to a Context Window Length of 30, which maintains conversational context for the agent.
Your First AI Agent: The central orchestration n8n node. It manages the flow between the n8n trigger, the LLM, the memory, and all available tools, relying heavily on its detailed System Message for decision-making.
getajoke (HTTP Request Tool): A general-purpose n8n node used as a tool to fetch external data (a joke API).
daysfromnow (Date/Time Tool): Used to perform date calculations, demonstrating the agent's ability to extract date parameters from conversation.
wikipedia (Wikipedia Tool): A specialized n8n node tool for searching and summarizing general knowledge.
createpassword (Crypto Tool): Generates secure strings, illustrating how the agent can call specialized n8n node functions for utility tasks.
calculateloanpayment (Code Tool): Executes custom JavaScript code to perform a complex loan amortization calculation, showcasing the agent's ability to handle high-precision numeric tasks.
n8nblogrssfeed (RSS Feed Read Tool): Retrieves recent posts from the n8n blog, demonstrating live data fetching as a tool.