The GenAIID AgentCore Starter Pack (GASP) is a starter project repository that enables users (delivery scientists and engineers) to quickly deploy a secured, web-accessible React frontend connected to an AgentCore backend. Its purpose is to accelerate building full stack applications on AgentCore from weeks to days by handling the undifferentiated heavy lifting of infrastructure setup and to enable vibe-coding style development on top. The only central dependency of GASP is AgentCore. It is agnostic to agent SDK (Strands, LangGraph, etc) and to coding assistant platforms (Q, Kiro, Cline, Claude Code, etc).
GASP is designed with security and vibe-codability as primary tenets. Best practices and knowledge from experts are codified in documentation in this repository rather than in code. By including this documentation in an AI coding assistant's context, or by instructing the AI coding assistant to leverage best practices and code snippets found in the documentation, delivery scientists and developers can quickly vibe-build AgentCore applications for any use case. AI coding assistants can be used to fully customize the frontend and the cdk infrastructure, enabling scientists to focus the areas where their knowledge is most impactful: the actual prompt engineering and GenAI implementation details.
With GASP as a starting point and development framework, delivery scientists and engineers will accelerate their development process and deliver production quality AgentCore code following architecture and security best practices without having to learn any frontend or infrastructure (cdk) code.
GASP comes deployable out-of-the-box with a fully functioning, full-stack application. This application represents starts as a basic multi-turn chat agent where the backend agent has access to some basic tools. Do not let this deter you, even if your use case is entirely different! If your application requires AgentCore, customizing GASP to any use case is extremely straightforward. That is the intended use of GASP!
The application is intentionally kept very, very simple to allow developers to easily build up whatever they want on top of the baseline. The tools shipped out of the box are implemented as lambda endpoints behind an AgentCore Gateway with authentication. One such tool is the "text analysis" tool which counts number of words and frequency of letters in a block of text. Try asking the agent to analyze a block of text and see what comes out.
If you are a delivery scientist or engineer who wants to use GASP to build a full stack application, this is the section for you.
GASP is designed to be forked and deployed out of the box with a security-approved baseline system working. Your task will be to customize it to create your own full stack application to to do (literally) anything on AgentCore.
Deploying the full stack out-of-the-box GASP baseline system is only a few cdk commands once you have forked the repo, namely:
cd infra-cdk
npm install
cdk bootstrap # Once ever
cdk deploy
cd ..
./scripts/deploy-frontend.shSee the deployment README for detailed instructions on how to deploy GASP into an AWS account.
What comes next? That's up to you, the developer. With your requirements in mind, open up your coding assistant, describe what you'd like to do, and begin. The steering docs in this repository help guide coding assistants with best practices, and encourage them to always refer to the documentation built-in to the repository to make sure you end up building something great.
The out-of-the-box architecture is shown above. Note that Amazon Cognito is used in four places:
- User-based login to the frontend web application on CloudFront
- Token-based authentication for the frontend to access AgentCore Runtime
- Token-based authentication for the agents in AgentCore Runtime to access AgentCore Gateway
- Token-based authentication when making API requests to API Gateway.
- Frontend: React with Next.js, TypeScript, Tailwind CSS, and shadcn components - infinitely flexible and ready for coding assistants
- Agent Providers: Multiple agent providers supported (Strands, LangGraph, etc.) running within AgentCore Runtime
- Authentication: AWS Cognito User Pool with OAuth support for easy swapping out Cognito
- Infrastructure: CDK deployment with Amplify Hosting for frontend and AgentCore backend
genaiid-agentcore-starter-pack/
├── frontend/ # Next.js React frontend application
│ ├── src/
│ │ ├── app/ # Next.js app router pages
│ │ ├── components/ # React components (shadcn/ui)
│ │ ├── hooks/ # Custom React hooks
│ │ ├── lib/ # Utility libraries
│ │ ├── services/ # API service layers
│ │ └── types/ # TypeScript type definitions
│ ├── public/ # Static assets and aws-exports.json
│ ├── components.json # shadcn/ui configuration
│ └── package.json
├── infra-cdk/ # CDK infrastructure code
│ ├── lib/ # CDK stack definitions
│ ├── bin/ # CDK app entry point
│ ├── lambdas/ # Lambda function code
│ └── config.yaml # Deployment configuration
├── patterns/ # Agent pattern implementations
│ └── strands-single-agent/ # Basic strands agent pattern
│ ├── basic_agent.py # Agent implementation
│ ├── requirements.txt # Agent dependencies
│ └── Dockerfile # Container configuration
├── scripts/ # Deployment and test scripts
│ ├── deploy-frontend.sh # Frontend deployment helper
│ ├── post-deploy.py # Configuration generation
│ └── test-*.py # Various test utilities
├── docs/ # Documentation
│ ├── DEPLOYMENT.md # Deployment guide
│ ├── AGENT_CONFIGURATION.md # Agent setup guide
│ ├── MEMORY_INTEGRATION.md # Memory integration guide
│ ├── GATEWAY.md # Gateway integration guide
│ └── STREAMING.md # Streaming implementation guide
├── gateway/ # Gateway utilities and tools
│ ├── tools/ # Gateway tool implementations
│ └── utils/ # Gateway utility functions
├── tests/ # Test suite
├── vibe-context/ # AI coding assistant context
└── README.md
Note: this asset represents a proof-of-value for the services included and is not intended as a production-ready solution. You must determine how the AWS Shared Responsibility applies to their specific use case and implement the needed controls to achieve their desired security outcomes. AWS offers a broad set of security tools and configurations to enable our customers.
Ultimately it is your responsibility as the developer of a full stack application to ensure all of its aspects are secure. We provide security best practices in repository documentation and provide a secure baseline but Amazon holds no responsibility for the security of applications built from this tool.
This project is licensed under the MIT-0 License.