Documentation

Run Crowkis. Build on it. In that order.

Everything is here, but you only need a little of it to start. Pick the path that matches what you're doing, or jump straight to the section you need from the sidebar.

$ docker pull crowkis/crowkis:latest

then the quickstart, first cache hit in about five minutes

Start where you are

5 min

I just want it running

Pull the image, fire your first commands, watch a cache hit in the dashboard.

Start the quickstart →

ship it

I'm deploying to production

The hardened Docker setup, health checks, the auth boundary, and day-2 ops.

Docker deployment →

build

I'm wiring it into my app

Drop-in SDKs for Python and Node, or let an AI agent use the cache over MCP.

Browse the SDKs →

Getting started

From zero to a cache hit.

Quickstart →

Pull, run, first hit, about five minutes.

Docker deployment →

The hardened, production-shaped way to run it.

Reference

Every command, knob, and gate.

Commands →

The C* semantic family plus the Redis commands it speaks.

Configuration →

Every environment variable, explained, with defaults.

Security model →

Auth per surface, tenant isolation, the trust pipeline.

SDKs & integrations

Talk to Crowkis from your stack.

Python SDK →

Sync + async, get_or_compute, streaming, multimodal.

Node.js / TypeScript →

Typed client, retry/backoff, CachedOpenAI wrapper.

MCP for AI apps →

Let Claude Code and agents use the cache as a tool.

Stuck, or something missing?

The FAQ covers the common questions, and the inbox is open for the rest.

Read the FAQ Email us →