---
title: "Building llms-full.txt"
description: "How to author and maintain the LLM-friendly URL index."
type: "page"
canonical: "https://audigeo.ai/docs/integrations/markdown-mirrors/llms-full-txt"
---

# Building `llms-full.txt`

`llms.txt` is a short root-level descriptor; `llms-full.txt` is the comprehensive index of every page available as `.md`.

## Minimum format

```
# My Site — LLM-friendly URL index

> Every page below is available as Markdown for AI crawlers.

## Marketing
- https://example.com/index.md
- https://example.com/pricing.md

## Blog
- https://example.com/blog/post-1.md
- https://example.com/blog/post-2.md
```

## Build it from your sitemap

If you already have a sitemap, transform it:

```bash
curl -s https://example.com/sitemap.xml \
  | grep -oE "<loc>[^<]+</loc>" \
  | sed -E 's/<\/?loc>//g' \
  | awk '{print "- " $0 ".md"}'
```

## Validation

Use AudiGEO — its `markdown_mirror` analyzer fetches your `llms-full.txt`, checks that each declared URL serves valid markdown, and reports inconsistencies.

## Companion files

- `/llms.txt` — short pointer to `/llms-full.txt`
- `/robots.txt` — allow `*.md` and welcome AI crawlers explicitly

See the [README](README.md) for the full adoption checklist.