@rootbeer/zero NPM

GraphQL Zero

Mock your GraphQL API with generative AI fake data... zero config. Powered by LLMs that look at your schema and populate a "parallel universe" of fake data so you can prototype and test your frontend w/o implementing the backend or manually entering a bunch of data.

Use for your next prototyping session, product demo, or QA bug bash!

Example schemas are provided in the examples folder

npm install -g @rootbeer/zero
zero schema.graphql

Then swap out the URL in your frontend code with

http://localhost:9000

Note: We don't support persisted queries yet

To proxy to a real graphql server. Write your SDL file with extended types then use this command.

zero extended_schema.graphql --extend https://spacex-production.up.railway.app

Motivation

We live in modern times, why poke around in a 100 tabs trying to populate data for your app?

Zero is unlike anything you've seen before

It's zero config. No faker! No annotating your SDL! No directives!
It's static. static consistent data: it feels as if its real. you can play around with your app, not just stuck on one page with a chunk of lorem ipsum
:soon: It's dynamic. AI generated business logic, so you can query and MUTATE. Inspired by Backend GPT
It's incremental. Mock your entire API to completely separate from prod, or extend an existing API with proxying zero schema.graphql --extend https://existing-server.com/graphql

Credit: This repo is a hard fork of graphql-faker

Features

Releasing super early and unpolished. There's a 20% chance this works with your GraphQL schema. If it fails, try an easier schema :flushed: If it still fails, make an issue!

Optionally proxy an existing GraphQL API to extend it with faked data
Runs as local graphql server
queries (using heuristics for arguments)
persisted queries
field arguments for leaf fields (using heuristics)
custom scalars (starting with popular standards)
directives (starting with popular standards)
mutations (including file uploads and auth)
subscriptions
pagination

How it works

This CLI tool sends our server an introspection AST of your GraphQL SDL (including docstrings). Our server runs fancy AI algorithms to create a blob of mock data, which gets sent back to the CLI tool. From then on, all queries are executed locally on your machine. We mock at the GraphQL level so it's data source agnostic and client agnostic (Apollo, iOS, Java, etc).