n8n AI Agent for Recursive Writing and Editing

This n8n workflow demonstrates a foundational setup for an AI agent capable of recursive writing and editing tasks. While the provided JSON defines the core components, it serves as a robust starting point for building sophisticated content generation and refinement processes using Large Language Models (LLMs).

What it does

This workflow establishes the basic architecture for an AI agent, including:

1. Receives Chat Messages: Initiates the workflow upon receiving a chat message, likely containing a prompt or instruction for the AI agent.
2. Initializes AI Agent: Sets up an AI Agent node, which is the core orchestrator for the AI's actions.
3. Configures OpenAI Chat Model: Defines the specific OpenAI Chat Model to be used by the AI Agent for generating and processing text.
4. Establishes Simple Memory: Integrates a basic memory buffer to allow the AI Agent to retain context from previous interactions within a window.
5. Prepares Structured Output Parser: Includes a structured output parser, indicating an intention to guide the AI to produce output in a specific, parseable format (e.g., JSON).
6. Conditional Logic: Incorporates an "If" node, suggesting the potential for branching logic based on conditions, which could be used for decision-making in recursive processes (e.g., "if content needs editing, then edit").
7. Data Manipulation: Utilizes "Edit Fields (Set)" and "Code" nodes for transforming and manipulating data within the workflow, essential for preparing prompts or processing AI responses.
8. Documentation: Includes a "Sticky Note" for internal documentation or notes within the workflow.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance (cloud or self-hosted).
• OpenAI API Key: Credentials for an OpenAI account with access to their chat models. This will need to be configured in the "OpenAI Chat Model" node.

Setup/Usage

1. Import the workflow:
   • Save the provided JSON content as a .json file.
   • In your n8n instance, click "Workflows" in the left sidebar.
   • Click "New" or the "+" icon, then select "Import from JSON".
   • Upload the saved JSON file.
2. Configure Credentials:
   • Locate the "OpenAI Chat Model" node.
   • Click on it and configure your OpenAI API credentials.
3. Activate the Workflow:
   • Once configured, activate the workflow by toggling the "Active" switch in the top right corner of the workflow editor.
4. Trigger the Workflow:
   • Send a chat message to the configured "Chat Trigger" to initiate the AI agent. The exact method for sending a chat message depends on how the "Chat Trigger" is exposed (e.g., through a specific chat platform integration).

This workflow provides a powerful foundation. To achieve full recursive writing and editing capabilities, you would typically expand upon this by:

• Adding more sophisticated logic in the "If" node to determine if content needs further iteration (e.g., checking for specific keywords, length, or quality metrics).
• Looping back AI responses to the agent for further refinement (e.g., taking an AI-generated draft, feeding it back with an "edit this" prompt).
• Integrating external tools (e.g., for spell-checking, grammar, or content analysis) as custom tools for the AI Agent.
• Defining specific "tools" for the AI Agent to use for various writing and editing tasks.