Redis
Your agent now follows Redis Engineering's own guidance for data structures, the Query Engine (RQE), vector search with RedisVL, semantic caching with LangCache, and performance tuning.
What it installs
| Artifact | Path (in your project) | Purpose |
|---|---|---|
| Skill | .agents/skills/redis-development/ | Protocol the agent follows on Redis work — sourced from redis/agent-skills |
| Rule | .claude/rules/redis.md | Always-loaded pointer: triggers on Redis data structures, RQE queries, RedisVL, LangCache, or perf work |
How it works
Redis looks like a simple KV store until you hit production — wrong data structure choices (hashes vs. strings), missing TTLs, hot keys, unbounded lists, blocking commands on the main thread. The skill encodes Redis's own guidance: pick the right structure per access pattern, use RQE for secondary indexes instead of KEYS *, choose RedisVL for vector similarity, and use LangCache for LLM response caching.
The rule loads a pointer into every session, so the agent consults the skill before writing Redis client code or designing a keyspace — no explicit invocation needed.
Setup
No env vars or external accounts required. The skill is fetched from GitHub during harness-kit add via npx skills add.
Pairs well with
langchain/langgraph— LangCache fits naturally on top of LLM pipelinespostgresql— Redis as cache/queue in front of Postgres as source of truthcode-review-gates— enforce TTL and blocking-command checks on every commit