LLM Access

This documentation site exposes a few AI-friendly text endpoints so you can feed mcMMO docs into ChatGPT, Claude, Cursor, your own scripts, or any other tool that works better with plain text than with full HTML pages.

These endpoints are for reading docs content, not for chatting with an AI on this site. There is no built-in Ask AI box here, so using these endpoints does not create model-token charges by itself.

Which endpoint should I use?

llms.txt

Use /llms.txt when you want an AI to discover the structure of the docs before choosing which pages to read.

Example:

https://mcmmo.hyronic.dev/llms.txt

This file is a lightweight map of the site. It is the best starting point for:

• AI tools deciding which docs pages matter for a question
• Search/indexing pipelines
• Prompt workflows where you want a small overview first

llms-full.txt

Use /llms-full.txt when you want to give an AI the full site in one pass.

Example:

https://mcmmo.hyronic.dev/llms-full.txt

This file contains the processed text of every docs page, joined into one large plaintext document. It is useful when:

• you want maximum context in a single upload
• your AI tool accepts a text file or pasted corpus
• you are building your own retrieval or summarization workflow

Single-page Markdown with .md

Every docs page can be requested as Markdown by adding .md to the end of the page URL.

Examples:

This is the best option when:

• you only need one page
• your AI tool follows links page by page
• you want language-specific content

The Markdown export includes the page title and the processed body text, without the normal site chrome around it.

Using the Accept header

If you are writing a bot, script, or crawler, you can request a normal docs URL and ask for Markdown through the HTTP Accept header.

Example:

curl -H "Accept: text/markdown" https://mcmmo.hyronic.dev/docs/admin/permissions

Instead of returning the regular HTML page, the site will serve the Markdown version of that docs page.

This is useful when:

• your client already knows the normal docs URL
• you do not want to manually append .md
• you want cleaner automation for crawlers and agents

Typical workflows

Give an AI one exact page

1. Open the docs page you want.
2. Add .md to the end of the URL.
3. Paste that Markdown into your AI tool, or let the tool fetch it directly.

Example:

/docs/admin/permissions.md

Let an AI explore the site

1. Start with /llms.txt.
2. Let the AI identify the relevant docs pages.
3. Fetch only those pages as .md.

This keeps prompts smaller and usually gives better answers than dumping the whole site every time.

Give an AI the whole docs set

1. Fetch /llms-full.txt.
2. Upload or paste it into the AI tool.
3. Ask questions against that full text.

This is the simplest workflow, but not always the most efficient one.

What this feature does not do

This site's LLM support is export-only.

It does not currently provide:

• an embedded chat assistant
• site-side model inference
• provider-managed search/chat billing
• guaranteed fact-checking beyond the docs content itself

If you use these files with an external AI tool, the quality of the answers still depends on:

• the quality of the model
• how much context you give it
• whether you feed it the correct page or language

Tips for best results

• Use page-level .md exports for precise questions.
• Use /llms.txt first when the AI needs to discover where an answer lives.
• Use /llms-full.txt only when broad cross-page context really matters.
• For configuration, commands, and permissions, link the AI to the exact reference page rather than a broad section landing page.

See also

• Permissions
• Localization
• Glossary