| name | evernote-reference-architecture |
| description | Reference architecture for Evernote integrations.
Use when designing system architecture, planning integrations,
or building scalable Evernote applications.
Trigger with phrases like "evernote architecture", "design evernote system",
"evernote integration pattern", "evernote scale".
|
| allowed-tools | Read, Write, Edit, Grep |
| version | 1.0.0 |
| license | MIT |
| author | Jeremy Longshore <jeremy@intentsolutions.io> |
| compatible-with | claude-code, codex, openclaw |
| tags | ["saas","evernote","scaling"] |
Evernote Reference Architecture
Overview
Production-ready architecture patterns for building scalable, maintainable Evernote integrations. Covers service layer design, caching strategy, sync architecture, and deployment topology.
Prerequisites
- Understanding of microservices or modular monolith architecture
- Cloud platform familiarity (AWS, GCP, or Azure)
- Knowledge of message queues and caching
Instructions
Architecture Layers
Client Layer [Web App / Mobile / CLI]
|
API Layer [Express/Fastify REST API]
|
Service Layer [NoteService | SearchService | SyncService]
|
Integration [EvernoteClient (rate-limited, instrumented)]
|
Infrastructure [Redis Cache | PostgreSQL | Message Queue]
Service Layer Design
Separate concerns into focused services:
- NoteService: CRUD operations, ENML formatting, tag management
- SearchService: Query building, pagination, result enrichment
- SyncService: Webhook handling, incremental sync, conflict resolution
- AuthService: OAuth flow, token storage, refresh logic
class ServiceRegistry {
constructor(noteStore, cache, db) {
this.notes = new NoteService(noteStore);
this.search = new SearchService(noteStore, cache);
this.sync = new SyncService(noteStore, db);
}
}
Caching Strategy
Cache at two levels: in-memory LRU for hot data (note metadata, user info) and Redis for shared state (notebook lists, tag lists, sync checkpoints). Invalidate on webhook notification.
Sync Architecture
Use webhooks as the primary change notification channel. Fall back to polling when webhooks are unavailable. Process changes through a message queue for reliability and retry. Store sync state (USN) in the database for crash recovery.
