Home

DataVault

A framework-agnostic TypeScript data service that acts as a smart cache and observer hub. Any frontend or backend consumer can ask for a piece of data by key. The service fetches it, caches it, and pushes updates to every registered observer automatically.

What it does

1. A consumer requests data by key
2. If the data is cached and still valid, it is returned immediately
3. If not, the service looks up the registered API definition for that key, makes the upstream call, maps the response, stores it, and notifies observers
4. On every future change, all observers are notified automatically

Deployment modes

Mode	Description
Library Mode	Import directly into any frontend or Node app — runs in-process
Microservice Mode	Run as a standalone HTTP + WebSocket server — any client connects over the network

The core logic is identical in both modes. Only the transport layer differs.

Quick navigation

Topic	Page
System overview and diagrams	Architecture
Getting up and running fast	Getting-Started
Full DataVault public API	API-Reference
Defining data sources	Definitions
Cache internals	Cache-Design
Storage backends	Storage-Adapters
REST, WebSocket, and polling	Transport-Types
Field mapping and transforms	Data-Mapping
HTTP and WebSocket protocol	Microservice-Mode
Design patterns used	Design-Patterns
Security features	Security
Running the test suite	Testing

At a glance

import { DataVault } from '@banditintinc/datavault';

const ds = new DataVault({ storage: 'indexeddb' });

ds.registerDefinition({
  key: 'users',
  url: 'https://api.example.com/users',
  type: 'rest',
  method: 'GET',
  cacheTTL: 60_000,
  mapping: { name: 'profile.displayName', email: 'contact.email' }
});

ds.get('users', {
  id: 'UserList',
  onUpdate: (data) => renderTable(data)
});

Tech stack

• TypeScript 5.4
• Node 20
• Express 4 (microservice HTTP layer)
• ws (microservice WebSocket layer)
• Jest + ts-jest (tests)