@launchdarkly/ai-node

Node.js convenience wrapper for the LaunchDarkly AI SDK.

This package re-exports the full public API of @launchdarkly/ai-server and carries @launchdarkly/node-server-sdk as a hard dependency. For standard Node.js applications, installing this single package is all that is needed — no peer dependency wiring required.

When to use this package

| Runtime | Install | |---|---| | Node.js (server, Lambda, containers) | @launchdarkly/ai-node ← this package | | Edge runtimes (Vercel, Cloudflare, etc.) | @launchdarkly/ai-server directly, then call initClient(preInitializedClient) |

Installation

npm install @launchdarkly/ai-node

Add whichever handler package(s) your application uses:

npm install @launchdarkly/ai-openai-messages
# or
npm install @launchdarkly/ai-claude-agents
# etc.

Optionally, install @launchdarkly/ai-otel to enable trace export to the LaunchDarkly Observability dashboard:

npm install @launchdarkly/ai-otel

initClient() detects the OTel packages at runtime and configures tracing automatically. If they are absent, a single console.warn is emitted and all AI calls continue normally with no-op spans.

Usage

Import from @launchdarkly/ai-node exactly as you would from @launchdarkly/ai-server — the API surface is identical.

import 'dotenv/config';
import { config, shutdown } from '@launchdarkly/ai-node';
import { createOpenAIHandler } from '@launchdarkly/ai-openai-messages';

const caller = config({
  key: 'my-ai-config-flag',
  handler: createOpenAIHandler(),
});

const result = await caller.invoke(
  'What is feature flagging?',
  { kind: 'user', key: 'user-123' },
);

console.log(result.response);
await shutdown();

@launchdarkly/node-server-sdk is auto-discovered by initClient() via dynamic import. The LD_SDK_KEY environment variable is read automatically — no explicit initClient() call is required unless you need programmatic options or early startup initialization.

Environment variables

| Variable | Required | Description | |---|---|---| | LD_SDK_KEY | Yes | LaunchDarkly server-side SDK key | | LD_BASE_URI | No | Override the LaunchDarkly polling base URI | | LD_STREAM_URI | No | Override the streaming URI | | LD_EVENTS_URI | No | Override the events URI | | LD_SERVICE_NAME | No | OTel service.name resource attribute (default: nodejs-sdk) | | LD_ENVIRONMENT | No | deployment.environment attribute attached to telemetry | | OTEL_EXPORTER_OTLP_ENDPOINT | No | OTLP endpoint override (default: LaunchDarkly Observability backend) |

inspectConfig(key, context)

Reads an AI Config flag variation without invoking any AI provider. Re-exported from @launchdarkly/ai-server — see the full reference there.

import { inspectConfig } from '@launchdarkly/ai-node';

const result = await inspectConfig('my-ai-config-flag', { kind: 'user', key: 'user-123' });
if (result.enabled) {
  console.log(result.config?.model?.name);
}

Never throws. Returns { enabled: boolean, config: AiConfigRep | null, meta: VariationMeta | null }.

Full API reference

See @launchdarkly/ai-server — all exports are re-exported unchanged from this package.