npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@cookiemonsterdev/kafka-studio

v1.1.0

Published

A local web UI for inspecting and interacting with an Apache Kafka cluster, built on @cookiemonsterdev/kafka-core.

Readme

@cookiemonsterdev/kafka-studio

A local web UI for inspecting and interacting with an Apache Kafka cluster — browse and manage topics, tail and produce messages, inspect consumer groups, and watch a live view of traffic moving through the cluster. Launched from the command line and served on localhost, in the spirit of Prisma Studio. Built on top of @cookiemonsterdev/kafka-core and @cookiemonsterdev/kafka-config, and lives in the kafka monorepo.

Status: early — the CLI, HTTP server, and web shell run end to end. The studio can browse, create and configure topics; produce messages (single sends and rate-limited bursts); browse and tail live messages; inspect consumer groups (members, per-partition lag, offset reset, deletion) and share groups; render a live topology board of cluster activity; and browse read-only ACL, client quota, and transaction state. A command palette (⌘K/Ctrl+K) jumps between pages. Every session is authenticated (see Security), and --read-only is enforced by the server, not just hidden in the UI. Launchable standalone (kafka-studio, this package's own bin) or as kafka studio from @cookiemonsterdev/kafka-cli, which resolves this package at runtime and prints an install hint if it isn't present. Not published to npm; install and usage instructions will follow.

Contents

Local development

From the workspace root:

pnpm --filter @cookiemonsterdev/kafka-studio build
node packages/studio/dist/bin.js

build produces the server bundle (dist/) and the browser SPA (dist/web/). Run node packages/studio/dist/bin.js --help for the flag list (port, host, browser, read-only mode).

pnpm --filter @cookiemonsterdev/kafka-studio dev watches and rebuilds the server bundle only. To serve src/web with Vite's own dev server instead of a static build, set KAFKA_STUDIO_DEV=1 before starting the server from source — or run the one-liner below, which does both:

pnpm studio:dev

Builds @cookiemonsterdev/kafka-config, @cookiemonsterdev/kafka-core, and the studio itself, then starts the server with KAFKA_STUDIO_DEV=1 set, so src/web is served through Vite's dev middleware (live reload) instead of the static dist/web build. Re-run it after a server-side source change; src/web changes hot-reload on their own.

Local Kafka with Docker

The studio has no built-in cluster — it needs a real (or locally hosted) broker to point at. docker-compose.dev.yml in this package brings up a single-node KRaft broker on localhost:9092, PLAINTEXT only, for exactly this:

cd packages/studio
docker compose -f docker-compose.dev.yml up -d

Wait for it to report healthy (docker compose -f docker-compose.dev.yml ps), then point the studio at it with KAFKA_BROKERS — the studio (via @cookiemonsterdev/kafka-core's fromEnv) reads this the same way the CLI does, so no config file is required for a quick local check:

KAFKA_BROKERS=localhost:9092 node packages/studio/dist/bin.js

To use a named cli.profiles connection instead (SASL/SSL, a remote cluster, …), point KAFKA_CONFIG at a kafka.config.* file the same way @cookiemonsterdev/kafka-cli does — see @cookiemonsterdev/kafka-config.

Tear the broker down (and drop its data) with:

docker compose -f docker-compose.dev.yml down -v

This compose file is for manual, local use only — it is not part of pnpm test or pnpm test:integration for this package. pnpm test:integration instead brings up its own broker from @cookiemonsterdev/kafka-core's test/assets/ (see Tests below), reusing those compose files rather than shipping a second copy of them.

Security

A local server that can speak to a possibly-production Kafka cluster is a real attack surface, so none of this is optional:

  • Localhost by default. The server binds 127.0.0.1 unless you pass an explicit --host. Doing so prints a loud warning to stderr — it means this machine's network can now reach a server that can read and mutate the connected cluster.
  • A session token per process. On startup the server generates a random token and opens the browser to a URL carrying it in the hash (#token=…), never in a query string or header the server itself would log. The page reads it once, keeps it in memory only (not sessionStorage or any other Web Storage, which stays readable by an XSS payload for as long as the tab is open) for the tab's lifetime, and strips it from the visible URL. Every /api/* request after that carries it as x-kafka-studio-token (or, for the SSE streams EventSource can't attach headers to, a token query param instead). A request with a missing or wrong token gets 401. A hard reload starts a fresh session — reopen the studio from the terminal's own printed URL.
  • An Origin/Host allowlist. The standard defense against DNS rebinding against a local dev server: a request naming a Host (or, when present, Origin) other than the address the server was actually told to bind is rejected with 403, before the token is even checked.
  • --read-only is enforced by the server, not by hiding buttons in the UI — a mutating request is rejected with 403 regardless of what the client sent.

Design system

The UI is dark-only — there is no light palette, no .dark class, and no theme toggle. Every token lives in the single :root block of src/web/styles/theme.css, and scripts/check-theme-drift.mjs keeps that block byte-identical to packages/docs/src/styles/global.css. Change one, change both, then run pnpm theme:check from the workspace root.

The accent (--primary) is the brand mark's own green, and --chart-1--chart-5 are a categorical set — src/web/lib/topic-accent.ts hashes a topic name onto one of them so a topic keeps the same colour everywhere it appears.

Two colour literals are duplicated by hand because Vite does not process them through the module graph: the theme-color meta and the splash screen in src/web/index.html. Both mirror --background; keep them in sync when that token changes.

Loading and failure states go through the shared components in src/web/components/ui/skeleton, spinner, empty-state, error-state, query-boundary and toast. A failed read gets a skeleton then an error panel with a working retry; a failed mutation always raises a toast via the QueryClient's MutationCache, so it survives the dialog that started it closing.

Tests

pnpm --filter @cookiemonsterdev/kafka-studio test
pnpm --filter @cookiemonsterdev/kafka-studio test:integration

Unit tests live beside source as src/**/*.test.ts and never start Docker; test/suites/tarball.test.ts and test/suites/bundle-budget.test.ts are also unit tests in this sense — both build and pack the package, neither needs a broker. test/suites/studio-e2e.test.ts is the one broker-backed integration suite, walking the studio's HTTP API (topic create/describe/configure, produce, a bounded message read, consumer group listing) against a real cluster brought up by test/helpers/global-setup.ts — the same compose files @cookiemonsterdev/kafka-core's own integration suite uses, not a second copy of them. KAFKA_EXTERNAL=1 (point KAFKA_BROKERS at an already-running broker) and DO_NOT_STOP=1 work the same way here as they do for core and the CLI.

Contributing

See CONTRIBUTING.md at the workspace root.

License

MIT