Vyriy

@vyriy/services

Part of Vyriy - a calm architecture toolkit for TypeScript, React, SSR, SSG, APIs, and cloud-ready apps.

Small AWS service SDK for Vyriy projects.

Purpose

This package wraps common AWS SDK operations behind calm, reusable helpers. It is not meant to hide AWS. It keeps the AWS command inputs available where they are useful, while removing repeated client setup and common boilerplate from app, script, service, and deployment code.

The package is AWS-first and serverless-oriented. Current helpers cover:

  • cloudfront: create invalidations and optionally wait for completion
  • dynamodb: create tables and read/write/query/scan document items
  • ecs: start one Fargate task with environment overrides
  • lambda: invoke Lambda functions with JSON string payloads
  • logs: wait for marker strings in CloudWatch Logs
  • s3: upload, download, and check object existence
  • sns: publish messages to topics
  • ssm: read one or many Parameter Store values

Install

With npm:

npm install @vyriy/services

With Yarn:

yarn add @vyriy/services

LocalStack

Some clients are prepared for local AWS-compatible development:

  • dynamodb uses LOCALSTACK_HOST and LOCALSTACK_PORT as its endpoint when @vyriy/env reports the local stage.
  • s3 uses the same LocalStack endpoint in local stage and enables forcePathStyle.
  • ecs uses the configured region in local stage, but does not point to a LocalStack endpoint.
  • cloudfront, lambda, logs, sns, and ssm currently use their normal AWS SDK client configuration.

Service clients are created lazily inside each helper call. Importing @vyriy/services or a service subpath does not construct AWS SDK clients.

LocalStack defaults come from @vyriy/env:

  • LOCALSTACK_HOST: defaults to localhost
  • LOCALSTACK_PORT: defaults to 4566

Usage

Import from the package root:

import * as services from '@vyriy/services';

await services.s3.upload('assets-bucket', 'pages/index.html', '<html>...</html>', 'text/html');
await services.lambda.invoke('my-function', JSON.stringify({ ok: true }));

Or import a service subpath directly:

import { getParameter } from '@vyriy/services/ssm';
import { upload } from '@vyriy/services/s3';

const apiUrl = await getParameter('/app/api-url');

await upload('assets-bucket', 'config/api-url.txt', apiUrl, 'text/plain');

Wait for a Lambda log marker and clean up the source S3 object:

import { logs, s3 } from '@vyriy/services';

await s3.upload('email-bucket', 'messages/test.eml', rawEmail, 'message/rfc822; charset=utf-8');

try {
  await logs.waitForMarker('/aws/lambda/process-email', marker);
} finally {
  await s3.remove('email-bucket', 'messages/test.eml');
}

Client factories are also exported from service subpaths when direct AWS SDK access is needed. The service path gives the generic createClient name its context:

import { createClient } from '@vyriy/services/s3';

const client = createClient({ region: 'eu-central-1' });

Examples

Invalidate CloudFront after publishing static files:

import { invalidate } from '@vyriy/services/cloudfront';

await invalidate('DISTRIBUTION_ID', ['/index.html', '/assets/*'], true);

Run a Fargate task:

import { runTask } from '@vyriy/services/ecs';

await runTask('regenerate-pages', [{ name: 'CONTENT_ID', value: 'home' }]);

Read application settings from SSM:

import { getParameters } from '@vyriy/services/ssm';

const settings = await getParameters(['/app/api-url', '/app/cdn-url']);