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

# Glossary

> The terms that appear across Octavia's APIs, defined once.

Terms are defined once here rather than repeated on every endpoint page. Where a term
has a service-specific meaning, that is noted.

***

## Platform

**Team**
: The container for everything you build. Owns API keys, content, and usage. Your
first team is created at sign-up. See [Teams & API keys](/platform/teams-and-keys).

**Tenant**
: The server-side identity a request resolves to. It is determined by your API key and
attached by the gateway — you never send it. If you need to serve several customers
from one integration, give each one its own key rather than trying to pass a tenant.

**Service**
: A product you activate from the dashboard, such as AI CMS. Each has its own reference
and its own key scope.

**Dashboard**
: [dashboard.octaviatech.app](https://dashboard.octaviatech.app) — where you activate
services, manage teams and keys, and inspect usage.

**Plan**
: The entitlement level on a service. It decides how much you can do before the API
returns <code>426</code>. See [Plans](/api-reference/ai-cms/plans).

***

## Requests

**API key**
: The credential sent as <code>x-api-key</code> on every request. It is scoped to one
team and one service. See [Authentication](/api-reference/ai-cms/authentication).

**Base URL**
: The root every request is built on — for AI CMS,
<code>[https://api.octaviatech.app/cms](https://api.octaviatech.app/cms)</code>. See [Base URL](/api-reference/ai-cms/base-url).

**Envelope**
: The single JSON shape every endpoint returns:{" "}
`{ success, statusCode, message, data }`, with `pagination` on list endpoints. See
[Response format](/api-reference/ai-cms/response-fromat).

**Path parameter**
: A value substituted into the URL, e.g. `/articles/{id}`.

**Query parameter**
: A value that filters a list, sent after <code>?</code> in the URL.

**Body**
: The JSON payload of a `POST`, `PUT` or `PATCH`. Anything that creates or updates
data goes here.

***

## Data

**ObjectId**
: A 24-character hex string identifying a record, e.g.
<code>6810f2c3a1b2c3d4e5f60718</code>. Passed in the path or the body.

**Slug**
: A URL-facing unique string, e.g. <code>hello-world</code>. Some create endpoints
accept one explicitly; others derive it from the name.

**Multilingual field**
: A map from a language code to a string, e.g.
`{ "en": "Hello", "es": "Hola" }`. Articles can only be written in a language that has
been registered first.

**Soft delete**
: A delete that marks the record rather than removing it, so it can be restored or
kept out of public queries.

**Archive**
: Removing an article from the published site without deleting it.

***

## AI CMS

**Article**
: The central content record. It has a title, body, author, categories, and a
publication state.

**Draft / published**
: An article's visibility. Drafts are visible to authenticated callers with the right
role; published articles are public.

**Category / Subcategory**
: The two-level taxonomy an article is filed under. `category` and `subCategory` on an
article take **arrays** of IDs, even when you only have one.

**Tag**
: A free-form label. Optional, and nothing depends on it.

**Form**
: A schema of sections and fields that a visitor submits. Forms are defined per
language.

**Submission**
: One completed response to a form. Submissions can be related to an article.

**Repurpose**
: Turning one article into platform-specific copy for LinkedIn, Twitter/X or Telegram.

**Conversation**
: A multi-turn AI session that drafts an article from a prompt and your feedback.

***

## Limits and errors

**Rate limit**
: A transient throttle. Answered with <code>429</code>, and retryable.

**Plan quota**
: An entitlement ceiling on your plan. Answered with <code>426</code>, and **not**
retryable — it only clears when you upgrade, the period rolls over, or you delete
something.

**AI tokens**
: The metered AI credit pool. It is *consumed*, not a count of resources, and it is
checked with a projection: the API estimates the cost before running the call.

**`retry-after`**
: A response header telling you how long to wait. Available where the SDK surfaces
response headers.

***

<CardGroup cols={2}>
  <Card title="Status codes" icon="list" href="/api-reference/ai-cms/status-codes" arrow="true">
    Every code and its meaning.
  </Card>

  <Card title="Content model" icon="file-text" href="/api-reference/ai-cms/content-management/introduction" arrow="true">
    Every AI CMS resource and its fields.
  </Card>
</CardGroup>
