Discord Agent for Structured Metadata Generation

• Dual-Language Metadata Generation: Creates structured metadata in both English (en) and Chinese (zh).
• AI-Powered Content Creation: Utilizes a Langchain AI Agent (AI Agent node) with a configurable system message to guide content generation.
• Contextual Conversation Memory: Implements Window Buffer Memory to maintain conversation history for more coherent responses.
• Flexible Triggering: Supports triggering via the When Executed by Another Workflow node for programmatic execution and When chat message received node for real-time chat interactions.
• Discord Integration: Employs discordTool nodes (Discord, Discord1, Discord2) to post messages to specific Discord channels ('ai-tools' and 'free-guides') within a given guild.
• OpenAI Model Integration: Connects to an OpenAI Chat Model (OpenAI Chat Model node) using GPT-4o-mini for AI processing.
• Character Limit Enforcement: The system message in the AI Agent node specifies a maximum character limit of 1800 for Discord messages.
• Structured Tool Usage: The workflow is set up to allow the AI agent to select and use specific Discord tools for posting messages.

Workflow Operation

This workflow can be triggered in two primary ways:

1. Workflow Execution Trigger (When Executed by Another Workflow):

   • To use this trigger, another n8n workflow must execute this workflow.
   • The input data should contain a Task parameter. The AI agent will use this as the primary input for generating content.
   • Connection Example: When Executed by Another Workflow -> AI Agent

2. Chat Message Trigger (When chat message received):

   • This trigger listens for incoming chat messages via a webhook.
   • When a message is received, it's passed to the AI Agent node.
   • Connection Example: When chat message received -> AI Agent

AI Agent Configuration (AI Agent node):

• text parameter: This is populated with ={{ $json.Task }}{{ $json.chatInput }}. This means the agent will process both a Task variable (if triggered by another workflow) and chatInput (if triggered by the chat trigger).
• options.systemMessage: This defines the AI's persona and instructions. It's crucial for controlling the output.
  • It instructs the AI to manage Discord, use channel IDs, format output stylishly, and enforce a maximum of 1800 characters per message.
  • It specifically mentions the ability to use tools for 'free guides channel' and 'ai-tools channel'.
• promptType: Set to define, suggesting the AI is defining its response based on the provided prompt and tools.

Language Generation (English & Chinese):

The AI Agent node is designed to generate output that can be used for both English and Chinese content. The system message implies it should handle this. To ensure specific language output:

• The input prompt (either from Task or chatInput) should ideally indicate the desired language or context.
• The AI model (OpenAI Chat Model) and its training data are responsible for generating multilingual responses. Explicit prompts for language conversion might be needed if the AI doesn't infer it automatically.

Discord Message Posting:

• Discord node: Used to post messages to the ai-tools channel (channelId: 1352242462520901632).
• Discord2 node: Used to post messages to the free-guides channel (channelId: 1352242462520901632).
• content parameter: For both these nodes, the content is set to ={{ /*n8n-auto-generated-fromAI-override*/ $fromAI('Message', ``, 'string') }}. This syntax indicates that the AI agent ($fromAI) is expected to generate the actual message content. The AI's system message guides it to use these tools.
• guildId: Both nodes are configured for the YungCEO SOCIETY💰 guild (1236784625196601386).

Memory Management (Window Buffer Memory node):

• sessionKey: Set to ={{ $json.Task }}. This means the conversation memory is keyed by the Task input, allowing for session-specific context.
• This node connects to the AI Agent to provide and store conversation history.

Sticky Notes:

• The workflow includes several Sticky Note nodes (Sticky Note, Sticky Note2, Sticky Note3, Sticky Note4) which provide documentation and guidance on workflow setup, operation modes, and troubleshooting. These are visual aids within the n8n canvas.

Credential Configuration:

• Ensure you have configured the following credentials in n8n:
  • Discord Bot API: For the discordTool nodes (e.g., 'Motion Assistant').
  • OpenAI API: For the OpenAI Chat Model node (e.g., 'OpenAi account').

Important Notes:

• The n8n-auto-generated-fromAI-override syntax suggests that the AI agent is intended to directly output content that is then passed to the Discord tools. The AI must be prompted to generate messages suitable for these specific channels.
• The Discord1 node is configured with operation: 'getAll', but its output is not explicitly connected to further processing within this snippet. Its purpose might be for fetching data rather than sending messages.
• To generate content in specific languages (English or Chinese), the prompt provided to the AI agent should either specify the language or the AI should be capable of inferring it from the context or user input.