Skip to content
Kitbay

redis-semantic-cache

MITUpdated on Nov 5, 2026175GitHub stars

Redis's skill for LangCache, a semantic cache for LLM responses offered on Redis Cloud.

What the skill teaches

Redis's skill for LangCache, a semantic cache for LLM responses offered on Redis Cloud. It walks the agent through a cache-aside flow: search the cache for a similar prompt, return the stored answer on a hit, otherwise call the model and save the result, using the LangCache SDK or REST API. It also covers tuning the similarity threshold to trade hit rate against wrong answers, using a different cache for each kind of task, and narrowing lookups with attributes of your own. LangCache is a Redis Cloud service, so it needs a Redis Cloud account and API key.

Capabilities

  • Add a semantic cache in front of LLM calls
  • Tune LangCache similarity thresholds
  • Separate caches per task type
  • Filter cached entries with custom attributes