Configure async Redis caching with get, set, and delete operations in FastAPI projects.
npx skills add https://github.com/abdulazeezoj/monovella-poc --skill redis-abdulazeezoj
Please help me install this Agent Skill. Skill: redis Source: https://github.com/abdulazeezoj/monovella-poc/tree/main/.agents/skills/redis Command: npx skills add https://github.com/abdulazeezoj/monovella-poc --skill redis-abdulazeezoj
redis_client and the exact conventions for using it without blocking the event loop or serving stale data. ## Core Features & Use Cases - Module-level async client: A single redis.asyncio.Redis instance in core/redis.py, built from REDIS_URL with decode_responses=True, safe to import anywhere. - Caching conventions: get/set with ex TTL, delete for invalidation, JSON serialization handled manually, and a <resource>:<qualifier> key naming pattern. - End-to-end guide: A before/after walkthrough that adds caching to the GET /items/ endpoint and invalidates the key in the create handler. - Use Case: You cache an expensive list endpoint, then add a write handler—this Skill ensures you also add the matching redis_client.delete(...) call so users never see stale data. ## Quick Start Ask the AI to add Redis caching with a 60-second TTL to the items list endpoint and invalidate the cache key on every write.High-intent search queries and answers about installing and using this skill.
Import the module-level redis_client, call await redis_client.get(key) before the real lookup, and on a miss store the result with await redis_client.set(key, json.dumps(payload), ex=60). The handler must become async def once it awaits the client.
Call await redis_client.delete(key) in every create, update, or delete handler that modifies the cached resource. Delete is a no-op if the key does not exist, so it is safe to call unconditionally before returning.
Use redis.asyncio inside async FastAPI handlers. The synchronous redis.Redis client blocks the event loop during each call, stalling every in-flight request, and low local traffic hides this until real concurrent load arrives.
The cached response was not invalidated when the underlying data changed, so it stays stale until its TTL expires. Add an explicit redis_client.delete call for the affected key in every write handler; a TTL is only a safety net.
No. redis.asyncio ships inside the standard redis package (redis>=5.2.0), so nothing extra is installed for async usage. This project uses the plain dependency without the hiredis extra.
MrJmpl3Load environment variables into validated pydantic-settings objects for Python services.
HZeroxiumImplement Redis-backed caching with TTLs, invalidation, and stampede protection for Java services.
renderincConfigure Render background workers for queue-based job processing.
hoangnh2412Invalidate local memory caches across ASP.NET Core instances via Redis pub/sub.