n8n-nodes-roundtable/README.md

n8n node for orchestrating multi-agent LLM discussions

# n8n Roundtable Node

This community node for [n8n](https://n8n.io) allows you to orchestrate multi-agent LLM discussions using a single language model with different personas.

The Roundtable node lets you create simulated roundtable discussions between multiple AI "participants" with different roles, expertise, and perspectives. It uses a single LLM with different system prompts to create the illusion of multiple participants collaborating on a topic.

## How It Works

1. **Connect an LLM**: The node requires an LLM input (such as OpenAI, Anthropic, or any other n8n LLM node)
2. **Configure Participants**: Define multiple participants with unique roles and system prompts
3. **Set Discussion Topic**: Specify the topic and any necessary context
4. **Run Conversation**: The node orchestrates a multi-round discussion between the participants
5. **Process Output**: The conversation is summarized in your chosen format

## Installation

Follow the [installation guide](https://docs.n8n.io/integrations/community-nodes/installation/) in the n8n community nodes documentation.

```
npm install n8n-nodes-roundtable
```

## Node Configuration

### Basic Settings

- **Discussion Topic**: The main subject to be discussed
- **Context**: Additional background information
- **Speaking Points**: Specific aspects of the topic to address (one per line)
- **Number of Rounds**: How many discussion rounds to conduct
- **Output Format**: Choose from several output formats

### Host Configuration

- **Role**: The role of the discussion facilitator
- **System Prompt**: Custom system prompt for the host persona
- **Temperature**: Controls randomness in responses (0-1)

### Participants

Add multiple participants, each with:
- **Role**: The participant's role or expertise
- **System Prompt**: Custom system prompt for this persona
- **Temperature**: Controls randomness in responses (0-1)

## Input/Output

### Input
- Requires a connection to an LLM node (OpenAI, Anthropic, etc.)

### Output
- Structured conversation summary in the selected format
- Raw conversation log (optional)
- Metadata about the discussion

## Example Usage

### Basic Workflow

1. Start with an AI model node (e.g., OpenAI)
2. Connect it to the Roundtable node
3. Configure the discussion topic, participants, and settings
4. Execute the workflow to generate the multi-agent discussion

### Example Use Cases

- **Product Development**: Create discussions between product managers, engineers, designers, and marketers
- **Decision Making**: Simulate debates between stakeholders with different priorities
- **Content Creation**: Generate conversational content with multiple perspectives
- **Education**: Create example discussions for teaching complex topics

## Compatibility

- Works with any n8n-compatible LLM node
- Tested with OpenAI, Anthropic Claude, and custom LLM nodes
- Compatible with the latest n8n LangChain nodes

## Troubleshooting

- **If participants aren't responding properly**: Check your system prompts and ensure they clearly define each role
- **If conversations are repetitive**: Increase temperature settings and ensure varied system prompts
- **If you get rate limit errors**: Reduce the number of rounds or participants, or add more delay between API calls

## License

[MIT](LICENSE)