---
title: Add your site
description: Choose the public pages Rinhelp should learn from and keep its index current.
sidebar:
  order: 1
---

Rinhelp indexes public pages. It does not bypass authentication, read private app screens, or treat a hidden URL as a secure source.

## Choose the source URL

Use the public origin that contains the pages you want to answer from. A homepage, docs root, or help-center root is a good starting point. Rinhelp discovers the site's sitemap when available and follows same-host links from the source.

The product site is one example: its marketing pages and `/docs` pages are part of the same public corpus. The console, API, inbox, and account routes are not documentation sources.

## Write for retrieval

Give each page one clear job. Use descriptive headings, state the answer directly, and keep important limits close to the claim they qualify. A page that is easy for a person to scan is usually easier for retrieval to use accurately.

## Refresh after changes

Run indexing again after changing pricing, policies, product behavior, or documentation. The index stores page-level content and citations, so stale pages can produce stale but well-cited answers.

## Keep private routes out

Only submit a source whose pages are intended for the public. Authentication and robots rules remain security boundaries; they are not substitutes for keeping private data off a public origin.