Using these docs with AI tools

This documentation is written in markdown and published as markdown. If you want to hand a page to ChatGPT, Claude, or your own agent, you never have to scrape rendered HTML.

Copy one page

At the top of every page:

• Copy page — puts that page’s markdown on your clipboard.
• View as markdown — opens the raw file.
• Open in ChatGPT / Open in Claude — opens a new chat with the page’s markdown URL and a prompt already written.

The .md twin

Any page URL with .md on the end returns the markdown source:

https://docs.santara.ai/channels/airbnb/      → the page
https://docs.santara.ai/channels/airbnb.md    → the markdown
https://docs.santara.ai/id/channels/airbnb.md → the Indonesian markdown

The home page’s twin is /index.md. Every page carries a <link rel="alternate" type="text/markdown"> pointing at its own, so a crawler does not have to guess the convention.

The whole manual in one file

File	What it is
/llms.txt	The map: every page, its URL, one line of description
/llms-full.txt	The entire documentation as a single markdown file
/id/llms.txt, /id/llms-full.txt	The same, in Bahasa Indonesia
/vi/llms.txt, /vi/llms-full.txt	The same, in Tiếng Việt

A locale falls back to English for pages it has not translated yet, so a non-English llms.txt is always a complete map of the product rather than a partial one.

Why bother

A rendered docs page is a few hundred kilobytes of HTML wrapped around a few kilobytes of prose. An answer engine fetching the markdown gets the same words for a fraction of the bytes, and quotes them correctly, because there is no navigation, no sidebar and no cookie banner in the way.

Asking about your workspace

These docs describe the product. For questions about your own data — your bookings, your revenue, your listings — use Ask Santara inside the product. It has your workspace; a chatbot with this URL does not.