Description

n8n-nodes-gravity

Official n8n nodes for Gravity AI – Build AI workflows with ease.

Features

• 🤖 Multiple AI Models – Claude (via AWS Bedrock), OpenAI GPT-4, and more
• 📡 Real-time Communication – Seamless integration with Gravity's event-driven architecture
• 🔄 Streaming Support – Handle streaming responses from AI models
• 🎯 Type-Safe – Full TypeScript support with comprehensive type definitions
• 🔧 Easy Setup – Simple credential configuration

Installation

Community Nodes (Recommended)

1. In n8n, go to Settings > Community Nodes
2. Select Install
3. Enter https://github.com/gravityai-dev/n8n-nodes-gravity
4. Click Install

Manual Installation

npm install -g @gravityai-dev/n8n-nodes-gravity

Then restart n8n.

Quick Start

1. Configure Credentials

Add your Gravity API credentials:

• Go to Credentials > New
• Select Gravity API
• Enter your server URL and API key
• Save

2. Create Your First Workflow

1. Add a Gravity Input node to receive messages
2. Process with Gravity Claude or Gravity OpenAI
3. Send responses with Gravity Output

Available Nodes

🎯 Gravity Input

Entry point for Gravity messages in your workflows.

Outputs:

• chatId – Unique chat identifier
• conversationId – Conversation thread ID
• userId – User identifier
• message – Incoming message text
• metadata – Additional context

🤖 Gravity Claude

Process messages with Claude AI via AWS Bedrock.

Inputs:

• message – Text to process
• systemPrompt – System instructions
• model – Claude model version
• temperature – Response creativity (0-1)
• maxTokens – Maximum response length

Outputs:

• response – Claude's response
• usage – Token usage statistics

🧠 Gravity OpenAI

Process messages with OpenAI's GPT models.

Inputs:

• message – Text to process
• systemPrompt – System instructions
• model – GPT model version
• temperature – Response creativity (0-1)

Outputs:

• response – GPT's response
• usage – Token usage statistics

📤 Gravity Output

Send responses back to Gravity.

Message Types:

• text – Simple text response
• json – Structured data
• mdx – Rich markdown components
• image – Image responses
• tool – Tool/function outputs
• action – Suggested actions

📊 Gravity Update

Send progress updates during processing.

Inputs:

• message – Status message
• progress – Progress percentage (0-100)
• state – Current state

🔍 Gravity Embed

Generate embeddings for semantic search.

Inputs:

• text – Text to embed
• model – Embedding model

Outputs:

• embedding – Vector representation
• dimensions – Vector size

Example Workflows

Basic Chat Assistant

[Gravity Input] → [Gravity Claude] → [Gravity Output]

Multi-Model Comparison

[Gravity Input] → [Split]
                    ├→ [Gravity Claude] → [Merge]
                    └→ [Gravity OpenAI] → [Merge] → [Gravity Output]

Document Processing

[Gravity Input] → [Extract Text] → [Gravity Embed] → [Vector Store] → [Gravity Output]

AWS Bedrock Setup

To use Gravity Claude nodes:

1. Create AWS credentials in n8n
2. Ensure your AWS account has Bedrock access
3. Enable Claude models in your region

Best Practices

1. Error Handling: Always add error outputs to handle failures gracefully
2. Progress Updates: Use Gravity Update nodes for long-running processes
3. Metadata: Include relevant metadata in responses for better tracking
4. Streaming: Enable streaming for better user experience with long responses

Support

• 📧 Email: support@gravityai.dev
• 💬 Discord: Join our community
• 📚 Docs: Full documentation
• 🐛 Issues: GitHub Issues

License

MIT – See LICENSE file for details

Built with ❤️ by the Gravity AI Team