The serverless WorkPaper example now has runnable persistence adapters for three common storage shapes:

• Postgres JSONB
• Redis or string key-value storage
• object storage for S3/R2-style buckets

The example keeps the same request handler boundary: load a WorkPaper JSON document, apply a verified cell edit, read the calculated output, and save the updated document back through the adapter.

Links:

• example README: https://github.com/proompteng/bilig/tree/main/examples/serverless-workpaper-api#persistence-adapters
• adapter source: https://github.com/proompteng/bilig/blob/main/examples/serverless-workpaper-api/persistence-adapters.ts
• service docs: https://proompteng.github.io/bilig/serverless-workpaper-api-route.html
• starter storage issues: https://github.com/proompteng/bilig/issues?q=is%3Aissue%20state%3Aopen%20label%3Afirst-timers-only%20storage

What I want feedback on: which storage layer should have the most production-shaped recipe next?

Candidates already queued for first-time contributors include Drizzle, Kysely, Upstash Redis, Neon Postgres, Cloudflare R2, and AWS S3. A useful reply would name the framework or storage driver, the expected deployment target, and the minimum verification you would want before trusting formula-backed workbook JSON in a service.