CareRelay MCP Server
Integrates with the Amazon Bedrock Converse API, using Amazon Nova 2 Lite to generate a calm, factual family brief from structured check-in and task data while limiting output length and forbidding diagnosis, prescriptions, or speculation.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@CareRelay MCP ServerDad missed his meds and feels tired. Check in, share with Maya, and get follow-up tasks."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
CareRelay π
Consent-first family coordination
CareRelay turns family check-ins into clear, actionable coordination plans while respecting each person's choice about what information can be shared.
Live Demo: https://carerelay-7wal.onrender.com GitHub Repository: https://github.com/vasu-repoo/CareRelay
π‘ The Problem
Family members often coordinate care through scattered messages, calls, and reminders. Important updates can get lost, responsibilities may be unclear, and private information can be shared without sufficient care.
Related MCP server: Hearth MCP Server
β¨ Our Solution
CareRelay provides a simple workflow for recording a check-in, choosing whether it can be shared, generating a family coordination brief, and organizing follow-up tasks.
Key Features
Consent-first sharing: A private check-in does not generate a shareable family brief.
Family coordination briefs: Summarize shared updates into a clear, readable message.
Actionable tasks: Organize follow-ups by task owner and due date.
Simple dashboard: Keep check-ins, family briefs, and tasks in one interface.
Privacy-conscious design: Keep private information out of shareable briefs when sharing has not been permitted.
AI integration: The application is configured to use Amazon Bedrock with Amazon Nova 2 Lite. Successful invocation in the deployed environment must be verified separately from local fallback behavior.
π οΈ Technology Stack
TypeScript
Node.js
HTML and CSS
Amazon Bedrock β Amazon Nova 2 Lite
Render β web application hosting
Git and GitHub
π How It Works
Record a family member's check-in.
Select the appropriate consent option.
If sharing is permitted, generate a family brief.
Review the follow-up plan and task owners.
Use the plan to coordinate the next steps.
Private check-ins should remain private and must not generate a shareable family brief.
π Run Locally
Prerequisites
Node.js and npm
Git
1. Clone the repository
git clone https://github.com/vasu-repoo/CareRelay.git
cd CareRelay2. Install dependencies
npm install3. Configure environment variables
Create a .env file in the project root and configure the values required by your application.
USE_BEDROCK=false
AWS_REGION=ap-southeast-2
BEDROCK_MODEL_ID=global.amazon.nova-2-lite-v1:0The example disables Bedrock so you can test the local fallback first.
To enable Bedrock, set USE_BEDROCK=true and configure AWS credentials securely in your local environment. Never commit .env files, access keys, or secret keys to GitHub.
4. Start the application
npm startUse the local address printed by the application, if applicable.
βοΈ Deployment
CareRelay is deployed on Render and connected to its GitHub repository.
For a deployed Bedrock integration, configure the required environment variables in Render's service settings. Verify AWS credentials, region, model access, and successful model invocation in the service logs.
Do not expose AWS credentials in source code, screenshots, public logs, or repository files.
π Privacy and Safety
Respect the user's sharing choice.
Do not create a shareable family brief for a private check-in.
Share only information needed for coordination.
Treat generated content as a coordination aid, not medical advice.
Use fictional data in public demonstrations and screenshots.
Important: A consent gate in the application is not a substitute for secure storage, access controls, and appropriate handling of sensitive information.
πΊοΈ Roadmap
Verify successful Amazon Bedrock invocation in the deployed environment.
Improve error handling and integration diagnostics.
Expand automated tests for consent and task workflows.
Improve accessibility and responsive behavior.
Prepare a concise end-to-end project demonstration.
π― Project Goal
CareRelay aims to make family coordination clearer, more actionable, and more respectful of consent.
Built for a developer hackathon project.
This server cannot be deployed
Maintenance
Related MCP Connectors
Authenticated MCP for a user's care circle: read CareEvents, list loved ones, inspect workspace.
Read-only A2Me family context tools for AI assistants (members, dates, activity, relationships)
OAuth scope approvals and consent receipts for remote MCP servers.
Community reads and consent-based member onboarding for local communities, over MCP.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables Alexa+ to deliver consent-aware family care handoffs, track shared observations and follow-ups, surface a prioritized next action, and keep private notes out of shared briefs.MIT
- AlicenseNot gradedqualityCmaintenanceEnables an Alexa+ family concierge to personalize household evening plans, suggest timelines, create reminders, and log gentle elder/accessibility care check-ins through MCP tools, resources, and prompts. It supports real local MCP invocation over Streamable HTTP with virtual Bedrock-style AWS demo behavior.2 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables families to care for an aging parent by voice, handling medication check-ins, appointments, daily summaries, missed-dose escalation, and parent-controlled sharing. Writes are tiered and confirmed by the user through MCP elicitation, and each session exposes only the tools of the caller's role.2 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables consented, daily Alexa check-ins that score speech and memory answers against a personal baseline, flag sustained drift to family, produce clinician summaries, and export de-identified data for research. It also provides MCP tools, resources, prompts, and MCP Apps views for compatible agent hosts.3 npmMIT