> ## Documentation Index
> Fetch the complete documentation index at: https://docs.zangtable.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Concepts

> The data model behind projects, tables, tokens, records, and exports.

## Project

A project is an isolated SQLite database addressed by a project slug. In API paths, the project slug acts as the Project ID. Slugs are lowercase letters, numbers, hyphens, and underscores: `^[a-z0-9_-]+$`.

## Project API Token

Project API Tokens are long-lived server-side credentials for one project. Raw tokens start with `pat_` and have full database power inside that project, so keep them on your backend/server only.

## Table

A table is a database table you create inside a project. Use tables to organize one kind of data, like recipes, customers, orders, posts, or app settings.

## Column

Column names must match `^[A-Za-z][A-Za-z0-9_]*$`. Supported column types are `INTEGER`, `REAL`, `TEXT`, `BLOB`, and `ANY`. If you create a column without passing `column_type`, ZangTable uses `TEXT`.

## Record

A record is one row in one table, sent and returned as a JSON object. Record values should be simple JSON values like strings, numbers, booleans, or `null`. Arrays and nested objects are rejected.

## Export

Exports are generated table, database, or schema files. Table exports can be CSV or JSON, database exports can be ZIP/CSV or JSON, and schema exports can be JSON or SQL.

## Operation grouping

API Endpoint requests may include `X-Operation-Key` and `X-Operation-Run-Id` header fields so developer-defined related requests appear grouped together in API journey tooling within the ZangTable admin panel.