← ClaudeAtlas

caching-strategylisted

Use when a ticket adds, changes, or debugs a cache — an in-memory memo, an HTTP cache header, a Redis layer, a CDN rule, a query-result cache — and the change must state what is cached, for how long, what invalidates it, and what happens when it is wrong. Invoke for "cache this", "it's showing stale data", "add Cache-Control", or when a profile shows the same read repeated.
tmj-90/gaffer · ★ 1 · AI & Automation · score 70
Install: claude install-skill tmj-90/gaffer
# Add or fix a cache A cache trades freshness for speed, and every cache goes stale in a specific way. Write the story before the code: key, scope, lifetime, invalidation, the race that can leave a stale value behind, and what happens when the cache is down. A cache without an invalidation plan is a bug on a delay. ## Steps 1. **Prove the need.** A measurement showing the same expensive read repeated (the `performance-profiling` skill). If the fix is "stop calling it twice", batch the query, or add an index (the `sql-query-performance` skill), do that instead. 2. **Choose the innermost layer that solves it.** Request-scoped memo (no invalidation problem); process memory (fast, per instance, diverges across instances); a shared store such as Redis (consistent across instances, one more dependency); HTTP caching (RFC 9111: `Cache-Control`, `ETag`/`If-None-Match` → `304`, CDN). Call `search_lore` for the repo's cache helper, key prefixing and serializer; use it. 3. **Design the key.** Every input that changes the output is in the key: identifiers, tenant, the caller's permission scope, locale, query parameters, and a schema or code version. Missing tenant = data leak; missing version = old shapes after a deploy. HTTP: `Vary` on every request header the response depends on. 4. **Set the TTL from the data.** Ask how stale this can be before someone is harmed; that is the TTL. Add ±10–20 % jitter so keys written together do not expire together. T