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

# Onboarding

> Describe your assistant in plain language to generate a spec and training data.

The **Onboarding** flow is a conversational wizard that turns your natural‑language description into an assistant specification stored in Supabase.

## What Onboarding collects

Through a chat, Promptify asks about:

* **Who** the assistant is for (audience)
* **What** it should help with (scope)
* **How** it should respond (tone, formatting, constraints)
* **What it must not do** (guardrails)

Behind the scenes, each answer is sent to:

* `POST /api/onboarding/message` – to build up a session
* `POST /api/onboarding/reset` – when you restart the flow
* `GET /api/onboarding/history` – to restore prior conversations

Once the conversation reaches a complete spec, Promptify:

* Writes an `assistants` row in Supabase with:
  * `name`
  * `description`
  * `instructions` (system prompt)
  * `model` (base model to fine‑tune from)

## When you’re done

After finishing Onboarding:

1. The **Settings → Assistant Specification** page shows the generated spec.
2. The **Dashboard** becomes able to start fine‑tuning jobs.
3. The **Chat API** will default to your assistant’s base model until a fine‑tuned model is available.