Challenge 3: CLI or Client Library? — Possible Solution
====================================================================

(a) Quickly checking what's currently cached during debugging —
REDIS-CLI. This is a one-off, human-driven, exploratory task — exactly
what an interactive shell is for. Writing and running a small script
with a client library just to check one or two keys would be far more
setup than the task warrants.

(b) A running Express application reading a cached value on every
request — A CLIENT LIBRARY (e.g. node-redis). This is code that needs
to run repeatedly, programmatically, as part of the application's
normal request-handling logic — redis-cli is an interactive human
shell, not something an application process shells out to on every
request. The client library gives the application a direct,
programmatic connection to Redis, exactly the same relationship
mongodb1-8 established between application code and the MongoDB
driver.

The underlying distinction: redis-cli is for a HUMAN exploring or
debugging by hand; a client library is for CODE that needs to talk to
Redis as part of its own running logic.