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

gsab-js

v0.2.0

Published

Google Sheets as a Backend — JavaScript/TypeScript client (CRUD, server-side query, realtime watch) for the browser and Node.

Readme

gsab — Google Sheets as a Backend (JavaScript / TypeScript)

A lightweight client that treats a Google Spreadsheet like a database: CRUD, server-side queries, and realtime watch(), in the browser and Node. Companion to the Python library (gsab on PyPI).

  • Docs: https://gsab.ajmalaksar.com
  • Status: early (0.1.0) — no-auth public reads and authenticated CRUD in Node work today; browser (GIS) auth is next.

No-auth public read (zero setup)

Reading a public ("anyone with the link") sheet needs no OAuth and no API key — it runs in a plain browser page or in Node:

import { connect } from "gsab-js";

const db = connect("https://docs.google.com/spreadsheets/d/<ID>/edit").sheet();

const rows = await db.read();                          // every row
const pro  = await db.read({ plan: "pro" });           // client-side filter
const top  = await db.query("SELECT A, D ORDER BY D DESC LIMIT 10"); // server-side gviz

for await (const change of db.watch({ interval: 2000 })) {
  console.log(change); // { added, updated, removed }
}

Reactive cache

createCache(db) keeps one in-memory snapshot of a sheet, polls + diffs, and dispatches granular events — one poller feeds many listeners (cheaper than each view re-reading). Works over a public sheet with no auth:

import { connect, createCache } from "gsab-js";

const db = connect("https://docs.google.com/spreadsheets/d/<ID>/edit").sheet();
const cache = createCache(db, { key: "id", interval: 2000 });

cache.on("insert", (row) => console.log("added", row));
cache.on("update", (row, prev) => console.log("changed", prev, "→", row));
cache.on("delete", (row) => console.log("removed", row));
cache.on("sync",   (rows) => {});  // after EVERY successful poll, changed or not
cache.on("error",  (err) => {});   // a poll failed (polling continues)

await cache.start();     // resolves once the initial snapshot is loaded ("ready")
cache.all();             // current rows;  cache.get(1);  cache.size;  cache.running
cache.stop();            // stop polling (the snapshot stays readable)

Experimental — polling (~1–2s), not push (same envelope as watch()).

React bindings (useSheet)

gsab-js/react turns a sheet into live component state — rows, loading, and error, re-rendered on every change. No providers, no config:

import { connect } from "gsab-js";
import { useSheet } from "gsab-js/react";

const db = connect("https://docs.google.com/spreadsheets/d/<ID>/edit").sheet();

function Users() {
  const { rows, loading, error } = useSheet(db, { key: "id" });
  if (loading) return <p>Loading…</p>;
  if (error) return <p>Sheet unreachable — showing the last known rows.</p>;
  return <ul>{rows.map((r) => <li key={String(r.id)}>{String(r.name)}</li>)}</ul>;
}

(In TypeScript, type the rows with useSheet<User>(db, ...) — rows are Record<string, unknown> otherwise.)

Passing a manager gives the component its own poller (started on mount, stopped on unmount). To share one poller across many components, create the cache yourself and pass it in — a never-started cache is started for you (and left running); one you stopped stays stopped, and the hook never stops a cache it didn't create:

const cache = createCache(db, { key: "id" });  // app level (or via context)
const { rows } = useSheet(cache);              // any number of components

useSheet also returns refresh() (re-read now) and the underlying cache (escape hatch for get() / on()). Works with any auth tier — public read-only sheets or an authed manager. React ≥18 is an optional peer dependency; the gsab and gsab-js/node entry points never load it. Experimental, like the cache it sits on.

Authenticated CRUD (Node)

Writes (and reads of private sheets) need a Google sign-in. In Node, loopbackAuth() reuses the same OAuth client the Python gsab CLI installed — sign in once in the browser; the token is cached after:

import { connect } from "gsab-js";
import { loopbackAuth } from "gsab-js/node";

const schema = {
  name: "users",
  fields: {
    id: { type: "integer", primaryKey: true },
    name: { type: "string", required: true },
    plan: { type: "string", default: "free" },
  },
};

const db = connect({ auth: await loopbackAuth() }).sheet(schema);

const id = await db.createSheet("My App DB");    // creates the spreadsheet, returns its id
await db.insert({ id: 1, name: "Ada", plan: "pro" });
await db.bulkInsert([{ id: 2, name: "Linus" }]);
await db.upsert({ id: 1, plan: "team" });        // insert-or-update on the primary key
await db.update({ id: 2 }, { plan: "team" });
await db.delete({ id: 2 });
const url = await db.share("reader");            // public link (reader | commenter | writer)

gsab-js/node is a separate entry point, so importing gsab in the browser never pulls in the Node auth dependencies. Constraints match the Python library: no transactions, and unique/primaryKey are enforced read-check-write (concurrent inserts of the same new key can still race).

Deploying (Vercel / serverless / CI)

A server has no browser to sign in with, so it uses a long-lived refresh token instead. One command on your own machine (it reuses — or triggers — the same loopback sign-in), one env var on the host:

npx gsab-js env
# GSAB_CREDENTIALS=…        ← set this one value on Vercel / Netlify / CI (it's a secret)

And the server code is one line different:

import { connect } from "gsab-js";
import { refreshTokenAuth } from "gsab-js/node";

const db = connect({ spreadsheetId, auth: refreshTokenAuth() }).sheet(schema);

Debugging a deploy? npx gsab-js doctor (in that environment) says what's configured and does a real token refresh. Prefer separate variables? npx gsab-js env --split prints the GSAB_CLIENT_ID / GSAB_CLIENT_SECRET / GSAB_REFRESH_TOKEN trio, which refreshTokenAuth() also accepts.

If it leaks: the credential is scoped to drive.file — it can only touch sheets gsab created, never the rest of your Drive — and you can revoke it any time at myaccount.google.com → Security → Third-party access. Keep it in your host's secret store; never commit it. The site's live demo (gsab.ajmalaksar.com/demo) runs exactly this recipe: public no-auth reads in the browser, refreshTokenAuth() writes in Next.js route handlers.

Concurrent editing: update() writes only the changed cells, so two clients editing different fields of the same row at the same time don't clobber each other — only a true same-cell edit is last-write-wins (Sheets has no conditional writes). Combined with watch() (which sees writes from your app, other clients, and people editing the Sheet directly), a table stays live-collaborative.

Teach your coding agent

A ready-made agent skill ships with the package (skills/gsab-js/) — quickstarts, the deploy recipe, and the rules of thumb (rate limits, per-cell writes, error types). Install it for Claude Code (or any skills-aware agent):

npx degit ajmalaksar25/gsab-js/skills/gsab-js ~/.claude/skills/gsab-js
# or, from a project that already has gsab-js installed:
cp -r node_modules/gsab-js/skills/gsab-js ~/.claude/skills/

The docs are also agent-friendly directly: every page has a raw-Markdown twin (https://gsab.ajmalaksar.com/docs/javascript.md) and the whole set is indexed at /llms.txt.

Roadmap

  • Now: read / query / watch over a public sheet (no auth) · authenticated CRUD in Node (createSheet / insert / bulkInsert / update / delete / upsert / share) via loopback OAuth · a reactive cache (createCache — snapshot + delta dispatch) · React bindings (useSheet over the cache).
  • Next: browser auth via Google Identity Services (a Web OAuth client), then a hosted sign-in page for near-zero-setup browser auth.