Use cases
Progress, training, or activity journal
Use ThimbleDB for milestones, session notes, structured observations, and user-owned progress history.
Workload assumptions
- One user owns each journal.
- Entries are structured JSON, not large documents or media.
- New entries and corrections are occasional.
- The UI reads recent entries often.
- Aggregate analytics are optional and can be derived elsewhere.
Why it fits
User scopes keep journals physically separated. Trie layout supports direct reads and incremental entry growth. Snapshot collections work for the user profile, goals, or a small summary record.
Retained deletion supports recovery from an accidental removal. External identity linking allows the same user to keep one journal when changing login providers.
Suggested collections
| Collection | Starting layout | Notes |
|---|---|---|
profile | Snapshot | Goals and current summary |
sessions | Trie | One record per session |
milestones | Trie | Direct ID reads |
templates | Snapshot | Small reusable prompts or forms |
Keep leaderboards, population analytics, and cross-user comparisons in a separate derived system.
Minimal entry
await db.write("sessions", "session-2026-09-24", {
id: "session-2026-09-24",
occurredAt: "2026-09-24T18:00:00.000Z",
category: "practice",
observations: ["Improved consistency"],
nextActions: ["Repeat the same exercise"],
});Scaffold prompt
Build a private progress journal with ThimbleDB 1.x.
Store profile and templates as snapshots. Store sessions and milestones as
tries in the signed-in user's `user:<uuid>` scope. Use stable IDs, structured
observations, explicit timestamps, retained deletion, and restore.
Do not add cross-user scans, public leaderboards, full-text search, local
passwords, or large media in JSON records. Add tests for user isolation,
identity linking, recent-entry reads, corrections, delete/restore, account
disablement, and scope erasure.Validation checklist
- Journal data always uses the internal user scope.
- A linked identity resolves to the same journal.
- A different user cannot fetch the journal key or objects.
- Recent entries use cached reads after first load.
- Deleted entries are hidden immediately.
- User erasure covers every documented journal collection.
- Cross-user analytics use a separate pipeline.