Happy slice of mcp pizzaMCP.pizza
Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT and other tools is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AI Fire in da houseCheck it out free

simple-mcp-server-on-lambda

MCP.Pizza Chef: aal80

simple-mcp-server-on-lambda is a minimalistic MCP server prototype designed to run natively on AWS Lambda and Amazon API Gateway without additional bridging or custom transports. Leveraging the Streamable HTTP transport introduced in MCP v2025-03-26, it offers a straightforward serverless architecture ideal for quick deployment and experimentation within AWS environments. This server simplifies MCP hosting by eliminating complex infrastructure components, making it accessible for developers familiar with AWS CLI and Terraform.

Use This MCP server To

Deploy MCP serverless backend on AWS Lambda quickly Prototype MCP server implementations without extra infrastructure Host MCP servers using native AWS API Gateway endpoints Experiment with Streamable HTTP transport for MCP Integrate MCP server into AWS-based AI workflows Test MCP client-server interactions in serverless environments

README

DEPRECATED

Use this repo instead - https://github.com/aws-samples/sample-serverless-mcp-servers

DEPRECATED

A simple MCP Server running natively on AWS Lambda and Amazon API Gateway without any extra bridging components or custom transports. This is now possible thanks to the Streamable HTTP transport introduced in v2025-03-26. This is merely a prototype.

Architecture is as simple as it gets:

Prereqs

  • AWS CLI
  • Terraform

Instructions

Install dependencies:

cd src
npm install
cd ..

Bootstrap server and set env var with MCP Server endpoint:

cd terraform
terraform init
terraform plan
terraform apply
export SIMPLE_MCP_SERVER_ENDPOINT=$(terraform output --raw endpoint_url) 
cd ..

Note: It might take a few seconds for API Gateway endpoint to become operational.

Run client:

node src/client.js

Observe the response:

> node client.js
> listTools response:  { tools: [ { name: 'ping', inputSchema: [Object] } ] }
> callTool:ping response:  { content: [ { type: 'text', text: 'pong' } ] }

Learn about mcp

Intro

Protocol specification

simple-mcp-server-on-lambda FAQ

How do I deploy the simple-mcp-server-on-lambda?
Use Terraform to bootstrap the server and AWS CLI to manage deployment.
What prerequisites are needed to run this MCP server?
AWS CLI and Terraform are required to deploy and manage the server.
Does this server require custom transports or bridging components?
No, it uses the Streamable HTTP transport introduced in MCP v2025-03-26, eliminating extra components.
Is this MCP server suitable for production use?
No, it is a prototype intended for experimentation and learning.
How do I interact with the server once deployed?
Run the provided client script (node src/client.js) to communicate with the server endpoint.
Can this server be integrated with other LLM providers?
Yes, it supports standard MCP protocols compatible with providers like OpenAI, Claude, and Gemini.
How long does it take for the API Gateway endpoint to become operational?
It may take a few seconds after deployment for the endpoint to be ready.
Where can I find an updated or maintained version of this MCP server?
Refer to the recommended repo at https://github.com/aws-samples/sample-serverless-mcp-servers for current implementations.