Amazon Bedrock Agentcore Samples

Active
GitHub Jupyter Notebook Apache-2.0

Description

Amazon Bedrock Agentcore samples that accelerate AI agents into production with scale, reliability, and security for real-world deployment.

Key Features

  • Framework and model agnostic - Support for Strands Agents, CrewAI, LangGraph, LlamaIndex and any LLM
  • Secure runtime - Secure serverless runtime for deploying agents and tools at scale
  • Gateway API - Convert APIs, Lambda functions, and services into MCP-compatible tools
  • Managed memory infrastructure - Hosted memory service for personalized agent experiences
  • Observability - Trace, debug, and monitor agent performance with OpenTelemetry
  • Fine-grained access control - Cedar policies for detailed permission management

Use Cases

💡 Quick-start Amazon Bedrock AgentCore with CLI to create and deploy production-ready agents
💡 Use Gateway to convert existing AWS APIs and Lambda functions into MCP-compatible tools for agents
💡 Integrate enterprise identity providers (Okta, Entra, Cognito) for secure agent identity management
💡 Automate AgentCore infrastructure deployment with CloudFormation/CDK/Terraform templates

Tags

aws agent-runtime production enterprise python

Categories

🏗️ Sandbox & Execution ⚡ Agent Tools

Quick Start

Prerequisites: AWS account (aws configure), Node.js 20+, uv (Python) or Node.js (TypeScript). Install CLI: npm install -g @aws/agentcore, run agentcore create to scaffold a project. Ensure Claude 4.0 model access is enabled in Amazon Bedrock console with BedrockAgentCoreFullAccess policy.
Visit GitHub

Related Projects