@gravityai-dev/aws-bedrock

AWS Bedrock integration for Gravity workflow system, providing access to Claude AI models and embedding services.

Features

• BedrockClaude: Chat completion with Claude models (Claude 3.5 Sonnet, Claude 3 Haiku, etc.)
• BedrockEmbedding: Text embeddings using Amazon Titan embedding models
• BedrockEmbeddingService: Batch embedding operations for multiple texts

Installation

npm install @gravityai-dev/aws-bedrock

Available Nodes

BedrockClaude

Chat completion using AWS Bedrock Claude models.

Inputs:

• prompt (string): The user message or prompt
• systemPrompt (string, optional): System instructions for the model
• maxTokens (number, optional): Maximum tokens to generate (default: 4096)
• temperature (number, optional): Sampling temperature (0.0-1.0, default: 0.7)

Outputs:

• response (string): Generated text response from Claude
• usage (object): Token usage statistics

Supported Models:

• anthropic.claude-3-5-sonnet-20241022-v2:0
• anthropic.claude-3-haiku-20240307-v1:0
• anthropic.claude-3-sonnet-20240229-v1:0

BedrockEmbedding

Generate embeddings for text using Amazon Titan models.

Inputs:

• text (string): Text to embed
• model (string, optional): Embedding model to use

Outputs:

• embedding (number[]): Vector embedding of the input text
• dimensions (number): Dimensionality of the embedding vector

BedrockEmbeddingService

Batch embedding service for processing multiple texts efficiently.

Inputs:

• texts (string[]): Array of texts to embed
• model (string, optional): Embedding model to use

Outputs:

• embeddings (number[][]): Array of embedding vectors
• count (number): Number of embeddings generated

Configuration

AWS Credentials

This package requires AWS credentials with Bedrock access. Configure using the AWS credential node:

• AWS Access Key ID
• AWS Secret Access Key
• AWS Region (e.g., us-east-1)

Required Permissions

Your AWS credentials need the following permissions:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": ["bedrock:InvokeModel"],
      "Resource": [
        "arn:aws:bedrock:*::foundation-model/anthropic.claude*",
        "arn:aws:bedrock:*::foundation-model/amazon.titan-embed*"
      ]
    }
  ]
}

Usage Examples

Basic Chat with Claude

// In your workflow, connect:
// 1. Input node with prompt text
// 2. BedrockClaude node with AWS credentials
// 3. Output node to display response

// The BedrockClaude node will automatically handle:
// - AWS authentication
// - Model invocation
// - Response parsing

Generating Embeddings

// Connect BedrockEmbedding node to:
// 1. Text input (string)
// 2. AWS credentials
// 3. Vector storage or similarity comparison nodes

Batch Embeddings

// Use BedrockEmbeddingService for multiple texts:
// 1. Array of texts input
// 2. AWS credentials
// 3. Batch processing output

Model Information

Claude Models

• Claude 3.5 Sonnet: Most capable model, best for complex reasoning
• Claude 3 Haiku: Fastest model, good for simple tasks
• Claude 3 Sonnet: Balanced performance and speed

Embedding Models

• Amazon Titan Text Embeddings: High-quality text embeddings
• Dimensions: 1536 (standard output size)

Development

Building

npm run build

Project Structure

src/
├── BedrockClaude/          # Claude chat completion node
├── BedrockEmbedding/       # Single text embedding node
├── BedrockEmbeddingService/# Batch embedding service
├── credentials/            # AWS credential definitions
├── shared/                 # Shared utilities and client
└── index.ts               # Plugin registration

Error Handling

Common issues and solutions:

• Authentication errors: Verify AWS credentials and region
• Model access denied: Ensure Bedrock model access is enabled in AWS console
• Rate limiting: Implement retry logic or reduce request frequency
• Invalid model: Check model name matches AWS Bedrock available models

License

MIT

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

Support

For issues and questions:

• GitHub Issues: https://github.com/gravityai-dev/bedrock/issues
• Documentation: See the main Gravity documentation for workflow integration