Enforces Redis key naming, TTL, cache-aside, locking, and rate-limiting rules.
npx skills add https://github.com/balajirags/aifsd-kit --skill cache-redis-balajirags
Please help me install this Agent Skill. Skill: cache-redis Source: https://github.com/balajirags/aifsd-kit/tree/main/docs/skills/cache-redis Command: npx skills add https://github.com/balajirags/aifsd-kit --skill cache-redis-balajirags
<app>:<env>:<entity>:<id> key format and requires an expiry on every cache key. - Concurrency-safe patterns: Covers cache-aside reads, distributed locks with TTL plus unique owner ID and atomic Lua-based release, and atomic rate limiting via Lua scripts or pipelined transactions. - Durability and eviction guidance: Explains when Pub/Sub is acceptable versus Kafka/Streams, and how to choose maxmemory-policy per use case. - Use Case: While implementing a Spring service, a developer applies the cache-aside pattern with @Cacheable/@CacheEvict and uses a Redisson lock with tryLock and a TTL to guard a reservation workflow. ## Quick Start Review my Redis caching and locking code against the cache-redis rules and flag any violations.High-intent search queries and answers about installing and using this skill.
A distributed lock must carry a TTL so a crashed holder cannot block it forever, plus a unique lock ID so you only release locks you own. Release via an atomic check-and-delete Lua script, never a plain DEL, which could delete another owner's lock.
Cache-aside reads the cache first, falls back to the database on a miss, then writes the result back with a TTL. In Spring, use @Cacheable for reads and @CacheEvict for writes, as shown in the skill's examples.
Redis Pub/Sub is not durable: a disconnected subscriber loses messages permanently. Use it only for best-effort signals like cache invalidation broadcasts, and use Kafka or Redis Streams when delivery must be guaranteed, such as order events.
Reading then writing a counter as two round trips races under concurrent requests, allowing requests to interleave and exceed the limit. Use INCR, SET with NX, or an atomic Lua script or pipelined transaction instead.
Use allkeys-lru or allkeys-lfu for pure caches, volatile-lru when only some keys carry a TTL, and noeviction for session or lock data where silent key loss would be a correctness bug rather than a cache miss.
vecearImplement reusable Redis patterns for caching, sessions, rate limiting, pub/sub, and distributed locks.
omarluqConfigure samber/hot in-memory caches with TTL, eviction, and singleflight.
lgzarturoConfigure Spring Boot caching with Redis, Caffeine, and Ehcache providers.
yonatangrossProvides .NET implementation of distributed systems patterns including locking, circuit breaking, retries, bulkheads, and rate limiting with Redis and PostgreSQL.