managing-api-cache

name	managing-api-cache
description	Implement intelligent API response caching with Redis, Memcached, and CDN integration. Use when optimizing API performance with caching. Trigger with phrases like "add caching", "optimize API performance", or "implement cache layer".
allowed-tools	Read, Write, Edit, Grep, Glob, Bash(api:cache-*)
version	1.25.0
author	Jeremy Longshore <[e-mail]>
license	MIT
tags	["api","redis","performance"]
compatibility	Designed for Claude Code, also compatible with Codex and OpenClaw

Managing API Cache

Overview

Implement intelligent API response caching using Redis, Memcached, or in-memory stores with cache key generation, TTL management, cache invalidation strategies, and HTTP cache headers. Support read-through, write-through, and cache-aside patterns with tag-based invalidation for related resources and stale-while-revalidate behavior.

Prerequisites

• Redis 6+ or Memcached for distributed caching (required for multi-instance deployments)
• Cache client library: ioredis (Node.js), redis-py (Python), or Lettuce (Java)
• CDN with cache control support for edge caching (CloudFront, Cloudflare, Fastly) -- optional
• Monitoring for cache hit/miss ratios and eviction rates
• Understanding of data freshness requirements per endpoint

Instructions

1. Analyze endpoint characteristics using Read and Grep to classify endpoints by cacheability: fully cacheable (static data), conditionally cacheable (user-specific data), and never cacheable (mutations, real-time data).
2. Implement cache key generation middleware that creates deterministic keys from method, path, query parameters (sorted), and relevant headers (Accept, Authorization hash for user-specific caches).
3. Build a cache-aside middleware that checks the cache before executing the handler, returning cached responses with X-Cache: HIT header, or executing the handler, caching the result, and returning with X-Cache: MISS.
4. Configure TTL per endpoint category: long TTL (1 hour) for reference data, medium TTL (5 minutes) for frequently changing lists, short TTL (30 seconds) for near-real-time data.
5. Implement cache invalidation on mutations: when a POST/PUT/DELETE modifies a resource, invalidate all cached responses containing that resource using tag-based invalidation.
6. Add Cache-Control, ETag, and Last-Modified response headers for HTTP-level caching, enabling CDN and browser cache participation.
7. Implement stale-while-revalidate behavior: serve stale cached responses immediately while asynchronously refreshing the cache in the background, reducing perceived latency.
8. Add cache warming for critical endpoints: pre-populate cache entries on application startup or schedule for frequently accessed resources.
9. Write tests verifying cache hits, misses, invalidation correctness, TTL expiration, and stale-while-revalidate behavior.

See ${CLAUDE_SKILL_DIR}/references/implementation.md for the full implementation guide.

Output

• ${CLAUDE_SKILL_DIR}/src/middleware/cache.js - Cache-aside middleware with hit/miss tracking