Caching Strategy Review for Applications
Design and verify application-level caches.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
ROLE: You are a backend architect reviewing caches. CONTEXT: Application: [APPLICATION_NAME]. Current caches (what, where, time to live, invalidation): [PASTE_CACHE_DETAILS]. Access patterns: [ACCESS_PATTERNS]. Consistency requirements: [CONSISTENCY_REQUIREMENTS]. Stack: [STACK]. STEPS: 1. Evaluate expected hit ratios, key design, stampede protection, invalidation correctness, stale-read risk, memory limits and eviction, serialization cost, and behavior during a cache outage. 2. Name the three places where caching would help most and two where current caching is dangerous. 3. For each recommendation choose the pattern (cache-aside, write-through, request coalescing) and show example code. 4. Explain how to measure hit rate and latency effect, and the failure test to run. OUTPUT FORMAT: Review table, recommendations with code, and measurement plan. QUALITY BAR: Correctness before speed; every cache needs a clear invalidation story. For every cache, require four facts: what is cached, who invalidates it and when, how stale it may be, and what happens if it is empty. Flag any cache where I could not answer those from the description, and include a stampede test. FOLLOW-UP: Ask me which data must never be stale.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.