@acumenlogs/sdk

Official Acumen Logs SDK for AI monitoring — trace LLM calls on the server and record user feedback in the browser.

Install

npm install @acumenlogs/sdk

Server (Node.js)

Set your monitor report URL:

ACUMEN_LOGS_MONITOR_URL=https://app.acumenlogs.com/api/monitor/{uuid}/report

import AcumenLogs from '@acumenlogs/sdk';

const traceId = await AcumenLogs.beginTrace('meal_planner', {
  session_id: sessionId,
  user_id: userId,
});

let response, error;
try {
  response = await openai.chat.completions.create({ model: 'gpt-4o', messages });
} catch (err) {
  error = err;
} finally {
  AcumenLogs.logSpan(traceId, 'draft_plan', response, error, { kind: 'llm' });
}

AcumenLogs.recordFeedback(traceId, 'accepted', { source: 'explicit' });

Browser (feedback only)

import { createAcumenLogsClient } from '@acumenlogs/sdk/browser';

const acumen = createAcumenLogsClient({
  monitorUrl: import.meta.env.VITE_ACUMEN_LOGS_MONITOR_URL,
});

acumen.trackAccept(traceId);
acumen.trackRegenerate(traceId);

React

import { useAcumenFeedback } from '@acumenlogs/sdk/browser/react';

const { accept, reject, regenerate } = useAcumenFeedback(acumen, traceId);

Documentation

• AI Monitoring Quickstart
• SDK reference

License

MIT