Platform Hierarchy

Revisium organizes data in a hierarchy inspired by GitHub:

Organization (= username)
  └── Project
        └── Branch (master, staging, dev)
              └── Revision (immutable snapshot) / Draft (WIP)
                    └── Table (JSON Schema)
                          └── Row (id + data)

Analogy with GitHub

Revisium	GitHub	Description
Organization	GitHub Org / User	Top-level namespace
Project	Repository	Independent data space with tables, branches, and endpoints
Branch	Branch	Isolated development line
Revision	Commit	Immutable snapshot of all tables and rows
Draft	Working directory	Mutable state where changes are made
Table	Folder	Collection of records sharing the same schema
Row	File in folder	A data record with a unique id, conforming to the table's schema

Organization

The top-level namespace. Your username is your organization — they are the same thing. Each organization owns one or more projects and has members with roles (admin, editor, viewer).

Limitation

Currently, you cannot create separate organizations. Your username acts as your single organization. Multi-organization support is planned.

Project

A project is an independent data space. It has:

• A root branch (usually master)
• Its own set of tables and schemas
• API endpoints bound to specific revisions

Projects are isolated — schemas and data don't leak between projects.

Branch

Each branch has its own revision history and a current draft. Changes on one branch don't affect others. See Versioning & Branches for details.

Revision and Draft

A revision is an immutable, point-in-time snapshot of every table and row in a branch. Once committed, a revision cannot be changed.

A draft is the mutable working state. All modifications happen in the draft. When ready, the draft is committed to produce a new revision.

Draft (mutable) → commit → Revision (immutable) = new HEAD

See Versioning & Branches for the full commit workflow, diff, and rollback.

Table

A table is defined by a JSON Schema and contains rows. Tables are created, modified, and deleted in a draft revision. The schema enforces structure on every write.

Row

A row is a data record identified by a unique string id within a table. The row data must conform to the table's JSON Schema.

System fields (createdAt, updatedAt, hash, etc.) are stored separately from user data and can be optionally included in the schema via $ref. See System Fields.