@hypertune/openfeature-server-provider

Hypertune OpenFeature Server Provider

This package contains a provider for using Hypertune with the OpenFeature JS Server SDK. If you want to use Hypertune with OpenFeature in the client then checkout the Hypertune OpenFeature Web Provider.

Hypertune overview

Hypertune is the most flexible platform for feature flags, A/B testing, analytics and app configuration. Built with full end-to-end type-safety, Git-style version control and local, synchronous, in-memory flag evaluation. Optimized for TypeScript, React and Next.js.

Quick start

1. Installation

npm install @openfeature/server-sdk
npm install @hypertune/openfeature-server-provider

2. Usage

import { OpenFeature } from "@openfeature/server-sdk";
import { HypertuneProvider } from "@hypertune/openfeature-server-provider";

// Register Hypertune feature flag provider
await OpenFeature.setProviderAndWait(
  new HypertuneProvider({
    token: "YOUR_HYPERTUNE_TOKEN",
  })
);

// Create a new client
const client = OpenFeature.getClient();

// Define a context that matches schema of your Hypertune project
const context = {
  environment: "test",
  user: {
    id: "user_123",
    name: "Test User",
    email: "[e-mail]",
  },
};

// Set global context or pass it to each flag evaluation
await OpenFeature.setContext(context);

// Evaluate your feature flag
const exampleFlag = await client.getBooleanValue("exampleFlag", false);

if (exampleFlag) {
  console.log("exampleFlag is enabled");
}

OpenFeature Specific Considerations

Accessing nested flags

To access nested flags you can use a dot separated path to your flag e.g. backend.exampleFlag. Values in a list can only be accessed by first evaluating the list flag as an object using the getObjectValue method.

Publishing events

To publish an event you can evaluate an event trigger flag using the getBooleanValue method.

Flag-specific arguments

Flag-specific arguments are not supported for individual flags with this OpenFeature provider. Trying to evaluate a flag with arguments will result in an error and the default/fallback value will be used. The only way to pass inputs to Hypertune is via the top level context.