Build with AI

Point Claude, ChatGPT, or Cursor at the VUZ llms.txt and have it wire up the API for you.

The whole VUZ API is published in the llms.txt format — a single, LLM-optimised text file. Give its URL to Claude, ChatGPT, Cursor, or any coding assistant and ask it to build your integration; the model reads the full reference and writes the code.

The two files

Compact indexhttps://developers.vuz.co.il/llms.txt

A structured table of contents — every public endpoint and guide, linked. Best when the assistant can follow links, or as a lightweight primer.

Full referencehttps://developers.vuz.co.il/llms-full.txt

The index plus the complete text of every guide, inlined. Paste this (or its URL) when you want the model to have everything in one shot.

How to use it

Copy a URL above

Use the Full reference for most tasks — it contains the whole API in one file.

Paste it into your AI assistant

Claude, ChatGPT, Cursor, Windsurf — anything that can read a URL or a pasted document.

Describe what you want to build

Be specific about the outcome and mention the Sandbox so it tests safely first.

Example prompt

Read https://developers.vuz.co.il/llms-full.txt — that is the full VUZ API reference.

Write me a Node.js script that issues a tax invoice receipt (חשבונית מס קבלה) whenever an
order is marked paid in my system. Requirements:
- Use the Sandbox first: base URL https://sandbox-api.vuz.co.il/api/v1 with a vuz_test_ key
  in the X-Api-Key header.
- Create the client inline if it doesn't exist yet.
- Handle Israeli VAT correctly and print the returned document number + PDF URL.
- Then show me exactly what to change to go live.

The assistant only ever sees the public API surface (every endpoint reachable with an API key). It never sees your data — you run the code it writes with your own Sandbox key.

Prefer a starting point?

Quickstart — your first call by hand, in 5 minutes.
Authentication — how API keys and scopes work.
API Reference — browse every endpoint yourself.