> ## Documentation Index
> Fetch the complete documentation index at: https://docs.t3planet.de/en/latest/llms.txt
> Use this file to discover all available pages before exploring further.

# Usage

> Practical usage of T3AF for admins, editors, and stakeholders.

This guide focuses on practical operation for editors, administrators, and
non-technical stakeholders.

## For administrators

Daily responsibilities:

* Keep provider API keys valid in **T3AF → AI Providers**.
* Maintain the default provider and model selections.
* Monitor OpenAI usage trends (requires `openai_admin_api_key` when used).
* Keep credentials and access permissions under control.

### Admin checklist

1. Confirm a default provider is enabled in **AI Providers**.
2. Run **Test connection** on critical provider rows.
3. Test extension-dependent AI features in your connected modules.
4. Review usage statistics regularly (cost/rate-control).

## For editors

Editors usually do not configure providers directly. They interact with
features built by other extensions that depend on T3AF.

When AI features fail in a backend module:

* Retry once.
* Capture exact error text.
* Inform administrator with module/page context.

## For non-technical stakeholders

T3AF helps organizations by:

* Reducing duplicated AI integration work across extensions.
* Centralizing provider and model governance.
* Improving consistency of AI capabilities across teams.

## What to expect operationally

* Some providers have rate limits and temporary outages.
* Model behavior can differ between providers and versions.
* Statistics data may be cached and not always real-time.

## Known boundaries

* No standalone frontend plugin is provided by this extension.
* This package is a service layer; UI features come from dependent extensions.