@affectively/edgework-sdk

Welcome! This SDK brings AI inference directly to the browser — fast, private, and offline-capable.

Install

npm install @affectively/edgework-sdk

What's Inside

• Storage — Local-first data with D1/Dash sync
• Inference — WebGPU-powered AI that runs on-device
• RLHF — Collect feedback to improve models over time
• React — Ready-to-use components for feedback UI

Quick Start

import { EdgeworkProvider, useInference } from '@affectively/edgework-sdk/react';

function App() {
  return (
    <EdgeworkProvider>
      <MyComponent />
    </EdgeworkProvider>
  );
}

Quality Autopilot

Run one command to lint common production routes/assets (/terms, /privacy, /sitemap.xml, /robots.txt, /.well-known/*, favicons, manifests) and optionally smoke-test deployed process URLs.

# From repo root
bun run audit:aeon-legal-routes --no-strict

# From packages/edgework-sdk
bun run quality -- --source=workspace --smoke --smoke-timeout=10000

Programmatic usage:

import { runQualityAutopilot } from '@affectively/edgework-sdk/deploy';

const report = await runQualityAutopilot({
  source: 'workspace',
  smoke: true,
});

Learn More

• Documentation
• GitHub

License

Apache-2.0

Made with care by the AFFECTIVELY team.

Last Updated: 2026-01-31

Sub-Directories

• Generated
• Rust
• Scripts
• Src