InhaltDocs
GitHubSign in

Concepts

The whole of Inhalt is four ideas. Once they click, the Getting started walkthrough and the API reference read as just the operations on them.

Project

A project is your workspace and the unit of isolation. Each account owns one, and everything below (content types, entries, API keys) lives inside it. A project is addressed by its slug, which is how the Read API reaches your content from anywhere:

GET /api/read/<project>/...

Content type

A content type is the shape of one kind of content, like a post or an author. It has a key (post), a display name, and a set of typed fields. The fields are the contract: every entry of the type is validated against them, so a client always works inside real, typed fields and never freeform text.

Each field has a key, a type, and whether it is required. The types are:

  • string - a short text value (optional maxLength).
  • richtext - a longer body. Its format (html, markdown, or plaintext) declares what markup the string holds. See rich text.
  • reference - a link to another entry (to names the target type).
  • enum - one value from a fixed list of options.
  • list - many values of string, number, or boolean.

You define types with the schema tools (schema.create, schema.update). Editing a type's fields does not migrate existing entries; stored values are re-checked on the next write to that entry.

Entry

An entry is one piece of content of a given type, identified by its slug (unique within the type). Its values are the field values, validated against the type. Entries are what you create, edit, and read.

Draft and published

Every entry has two copies: a draft you edit and a published copy the public Read API serves.

  • Edits (create, patch) always land on the draft. Nothing is public yet.
  • publish copies the draft over the published copy, pushing it live in one step.
  • The Read API only ever returns the published copy. An entry that has never been published is a 404 there.

There is no version history. The model is drafts -> save -> publish.

Two API surfaces

Those concepts are reached through two surfaces, covered in the API reference:

  • The MCP API is the typed tool surface your AI client uses to read and edit content. Authenticated, and the way you make changes.
  • The Read API is the public, key-less HTTP API that serves your published content to a website or app.