> For the complete documentation index, see [llms.txt](https://docs-v4.nftx.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs-v4.nftx.io/agentic-development/llm-ready-docs.md).

# LLM-Ready Docs

You don't need MCP to feed these docs to a model. Every page is published in machine-friendly formats that any LLM or script can pull directly.

## Every page as markdown

Add `.md` to any page URL to get the raw markdown instead of the rendered HTML — smaller, cleaner, and cheaper for a model to read:

```
https://gitbook.flaunch.gg/nftx/how-it-works/floor-tokens.md
```

## llms.txt and llms-full.txt

GitBook keeps two index files at the site root, always in sync with the docs:

| File                                                             | What it is                                                                                                             |
| ---------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------- |
| [`llms.txt`](https://gitbook.flaunch.gg/nftx/llms.txt)           | A compact index — every page and its markdown URL — so an agent can discover what exists and fetch only what it needs. |
| [`llms-full.txt`](https://gitbook.flaunch.gg/nftx/llms-full.txt) | The entire documentation in a single file, ready to drop into a model's context as one blob.                           |

Use `llms.txt` when you want the agent to navigate; use `llms-full.txt` when you want the whole thing in context at once.

## Ask a page directly

Any markdown page can also answer questions inline. Append an `ask` parameter (and an optional `goal`) to a `.md` URL, and GitBook returns an answer assembled from the relevant pages, with source links and suggested follow-ups:

```
https://gitbook.flaunch.gg/nftx/contracts/listings.md?ask=How is listing tax refunded?
```

This is handy from a script or a single tool call when you want a targeted answer without wiring up a full MCP client.

## Ask in the docs: GitBook Assistant

These docs also ship with **GitBook Assistant**, an in-page AI you can simply talk to. Open it by:

* pressing <kbd>⌘</kbd>/<kbd>Ctrl</kbd> + <kbd>I</kbd>,
* clicking the Assistant button next to the search bar, or
* typing a question into the **Ask or search…** bar and choosing **Ask…**.

It uses agentic retrieval — taking into account the page you're on and what you've already read — to answer from the documentation. It's the fastest way to get an answer without leaving the page.

{% hint style="info" %}
All of these read the same published content the site is built from, so they stay current automatically as the docs change.
{% endhint %}