Blockchain Data Processing Agent with Strands Agents on AWS

This project demonstrates how to build an intelligent blockchain data analysis agent using the Strands Agents framework with the Model Context Protocol (MCP) to integrate with the awslabs.aws-dataprocessing-mcp-server.

🚀 Features

• Intelligent AWS Data Processing: Leverage AI to help with AWS data processing tasks
• MCP Integration: Seamlessly connect to awslabs.aws-dataprocessing-mcp-server tools
• Structured Output: Get organized analysis of data processing tasks
• Interactive Chat: Natural language interface for AWS operations
• Direct Tool Access: Call AWS tools directly when needed
• Conversation Management: Smart context management for long conversations

🚀 Quick Start: Amazon Bedrock AgentCore Deployment

Pre-requisites

• Python 3.10+ (use uv or pyenv to manage Python versions)
• pip (have a version of pip installed that's compatible with your Python version)
• AWS CLI (if developing locally, use aws configure to set up credentials)
• Amazon Bedrock model access (use the AWS Console to enable the required foundation models, such as Claude Sonnet 4)

1. Install Dependencies

cd agentcore-deployment
pip install -r requirements.txt

2. Deploy to AgentCore

# Automated deployment with starter toolkit
python3 deploy_blockchain_agent.py --agent-name "my_blockchain_data_agent" --skip-local-test

# Or deploy manually (see Manual Deployment section)

3. Test Deployment

# Test the deployed agent
agentcore invoke --agent "my_blockchain_data_agent" '{"prompt": "How many Bitcoin blocks have been mined today?"}'