{
  "ingredients": "eggs, rice, spinach"
}

Generate Recipes from Ingredients with Ollama AI Chef Agent

This n8n workflow leverages an AI Agent powered by Ollama to act as a "Chef Agent", generating recipes based on provided ingredients. It simplifies the process of getting creative recipe ideas directly through a chat interface or an HTTP request.

What it does

This workflow automates the following steps:

1. Listens for a Chat Message or HTTP Request: The workflow can be triggered in two ways:
   • Chat Trigger: It listens for incoming chat messages, making it suitable for conversational interfaces.
   • Webhook: It can also be triggered by an HTTP POST request, allowing integration with other systems or custom applications.
2. Initializes AI Agent: An AI Agent is set up to process the incoming request.
3. Configures Ollama Chat Model: The AI Agent uses the Ollama Chat Model as its large language model (LLM) to understand the request and generate responses.
4. Manages Conversation History: A "Simple Memory" (Buffer Window Memory) is used to maintain context during the conversation, allowing the AI Agent to remember previous turns in a chat.
5. Utilizes a "Think" Tool: The AI Agent is equipped with a "Think" tool, which likely helps it to process and reason about the request before generating a recipe.
6. Generates Recipe: Based on the input (ingredients) and the AI Agent's configuration, it generates a recipe.
7. Responds to Webhook: If triggered by a Webhook, the generated recipe is sent back as the response to the HTTP request.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n instance: A running n8n instance.
• Ollama Installation: Ollama needs to be installed and running, accessible by your n8n instance, to provide the AI model.
• Ollama Chat Model: A compatible chat model (e.g., llama2, mistral, etc.) must be pulled and running in Ollama.
• Langchain Nodes: Ensure the @n8n/n8n-nodes-langchain package is installed in your n8n instance.

Setup/Usage

1. Import the Workflow: Download the provided JSON and import it into your n8n instance.
2. Configure Ollama Chat Model:
   • Open the "Ollama Chat Model" node.
   • Ensure the "Model" field is set to the name of the model you have running in Ollama (e.g., llama2).
   • Verify the "Base URL" points to your Ollama server (default is http://localhost:11434).
3. Configure Webhook (if using HTTP trigger):
   • Activate the workflow.
   • Copy the "Webhook URL" from the "Webhook" trigger node.
   • You can now send POST requests to this URL with your ingredients in the request body.
4. Configure Chat Trigger (if using Chat interface):
   • The "When chat message received" node is pre-configured. Ensure your chat integration (e.g., Slack, Telegram, etc.) is properly set up to send messages to this n8n instance if you intend to use this trigger.
5. Test the Workflow:
   • Via Webhook: Send a POST request to the Webhook URL with a JSON body containing your ingredients, for example: {"ingredients": "chicken, broccoli, rice"}.
   • Via Chat: Send a chat message (e.g., "What can I make with chicken, broccoli, and rice?") to your configured chat integration.

The workflow will then process the request and return a generated recipe.