epic-caching

What problem does it solve?

Caching is a core performance technique, but without a structured approach it can introduce complexity, stale data, and unpredictable behavior. This Skill guides teams on implementing robust, layered caching across the Epic Stack, balancing speed, correctness, and maintainability.

Core Features & Use Cases

• Caching philosophy: Clear guidelines for when to cache, which data to cache, how to invalidate, and how to reason about performance benefits.
• Two cache backends: Persistent SQLite cache (with LiteFS replication) and a short-lived in-memory LRU cache for deduplication and fast access.
• Use cases: Caching expensive Prisma queries, external API responses, and derived data with support for stale-while-revalidate and server timing integration.
• Patterns: TTL, SWR, cache invalidation after mutations, and structured cache keys.

Quick Start

Install the caching utilities in your Epic Stack project, choose an appropriate cache backend, and wrap expensive data-fetching calls with cachified. Define a cache key, configure TTL and optional stale-while-revalidate, and wire server timing to monitor cache performance.