How it works
Large language models (LLMs) are great at reading markdown because, unlike HTML, there’s a lot less noise for them to parse. To make our documentation more accessible to AI, we’ve provided markdown-accessible versions of our documentation so you can reference it with your own AI tools. You can access plain-text markdown versions of our docs in a few ways:- Our llms.txt file: This is an emerging standard and acts as a sort of LLM-friendly sitemap. It provides links to our markdown files so your LLM can access markdown versions of our pages without having to download them.
- Download individual markdown pages: You can download or reference individual pages to add to your own knowledge base or use them as a reference for your LLM.
Use our LLM sitemap
You can access or download our LLM-friendly documentation files to help your LLM access, and provide answers based on, markdown versions of our documentation. We provide two versions of our documentation sitemap based on the emerging llms.txt standard:- llms.txt: A concise version with links to all our documentation pages
- llms-full.txt: A comprehensive version with full content of all documentation pages included inline
- Access our concise llms.txt file
- Access our comprehensive llms-full.txt file
- Learn more about the proposed llms.txt standard
Download plain text documentation
You can download or copy our documentation as plain text markdown files for use with your LLM of choice—you can even create your own GPTs and tools! All of our documentation is written in MDX (Markdown with JSX), which is easily parseable by LLMs. You can access the source markdown for any page by adding.md to the end of most URLs to get the raw markdown version.
For example, you can check out the markdown for this page at https://www.ona.com/docs/ona/agents/use-docs-with-ai.md.