---
title: "Point your AI to LLM-friendly Maventa documentation"
canonical: https://documentation.maventa.com/changelogs/product-changelog/2026-05-06-point-your-ai-to-llm-friendly-maventa-documentation/
---

Maventa's documentation publishes an LLM-friendly markdown copy of every page alongside the HTML. If you use an AI coding assistant when working on Maventa integrations (Claude, Cursor, GitHub Copilot, ChatGPT, or anything else that can fetch a URL), you can point it at the docs directly and get answers grounded in the current source rather than the assistant's training data.

## What is available

Three things, all on `documentation.maventa.com`:

- [**`/llms.txt`**](https://documentation.maventa.com/llms.txt): a hierarchical index of every public page, following the [`llms.txt`](https://llmstxt.org/) convention. Most AI tools recognise this URL pattern automatically.
- [**`/llms-full.txt`**](https://documentation.maventa.com/llms-full.txt): every documentation page concatenated into one file, around 1.1 MB of plain markdown. Useful when you want to load all the docs into context in one go.
- **`<url>.md` for every page**: for example, the [Getting Started guide](https://documentation.maventa.com/integration-guide/getting-started/) is also available as [plain markdown](https://documentation.maventa.com/integration-guide/getting-started.md). Useful for fetching one specific page on demand.

A typical workflow: add [`/llms.txt`](https://documentation.maventa.com/llms.txt) to your assistant's documentation sources or paste it at the start of a session, then ask integration questions as normal. The assistant fetches the relevant pages on demand and can quote them back.

## REST API endpoints in the bundle

Every REST API page now ships filtered markdown summaries of its endpoints — request parameters, body schemas, and response shapes — derived from the live OpenAPI specs. Ask your assistant something like *"what is the request body for `POST /v1/invoices`?"* and the answer comes from the same source the Swagger UI renders.

The summaries are snapshots refreshed periodically. The canonical machine-readable spec URLs are documented on each REST API page if you need them directly.

## What is not included

Changelog entries are kept out of the bundled `/llms-full.txt`. Individual changelog posts are historical records, and an entry from a year ago may describe a flow that has since changed. Concatenated alongside current docs, that context is easy to lose. Each entry remains fetchable individually as `<url>.md` under `/changelogs/`.

The Peppol participant lookup tool is interactive and renders as a placeholder in its `.md` form. The HTML version on the live site remains the primary way to use it.