📚 Tutorials Overview
Learn by doing - each tutorial is copy-paste code for your own project explained step by step: what it sets up, why each call matters, what output to expect, and what to try next.
| Tutorial | What you learn |
|---|---|
| Offline Memory - No Keys, No Network | Batched + local-fs + keyword/local embedder; the fastest first memory |
| Durable Memory on Filecoin | Batched + lh-ipfs-filecoin: batching, flush, rebuild, cloud backup |
| Batched Memory on Walrus | Batched + lh-ipfs-walrus: quota-aware batching, blob IDs, gateway reads |
| Memory on S3 / R2 / MinIO | Batched + s3: content-addressed keys, hard-delete forget(), local MinIO dev |
| Encrypted Memory with Memwal | Relayer engine: SEAL encryption, analyze(), verify(), pinning |
| Cross-Device Backup & Restore | Snapshots + pointer service (backupIndex / restoreIndex, raw HTTP reference inside) |
| BYOK Keys | Save / check / delete your Lighthouse key so uploads bill your plan |
Prerequisites for all tutorials: Node 18+, a fresh project with the @lighthouse-ai/* packages installed from npm, and npx tsx available (each page lists its exact npm install line). Durable tutorials add a Lighthouse API key and/or memwal credentials - each page tells you exactly where to get them.
Load keys with dotenv (import 'dotenv/config' at the top of your script) from a project-root .env - explicit shell export still wins.