LLMs-full.txt Generator

llms-full.txt is the long companion to llms.txt. The short file is an index of links. The full file repeats those links and includes the page markdown so an assistant can read the content without opening each URL. Both files come from the same form.

The full file starts with the project title and the summary, then one section per page. Each section shows the title, the URL as a markdown link, and the markdown you wrote. A page with no title is skipped. A page that is included needs an http or https URL. The short file lists the same pages under one Pages heading.

Project name is one of the values LLMs-full.txt Generator puts on screen. The title of both files. It is required. It is not read from a live site.

Read Page title on its own before you mix it with the other rows. The heading for that page inside the full file. A page with no title is skipped. The title is what you type, not the HTML title of the URL.

Page URL answers a narrower question than the headline number. An absolute http or https link. A page that is included needs one. The URL is written into the file. It is not fetched.

Treat Markdown body as a label with a specific job. The content of that section. It is the text you type. It is not a scrape of the page.

The Short index line is worth a full stop. llms.txt lists the same pages. They sit under one Pages heading. The short file does not repeat the long bodies.

Update a page in the form and both files change together. The body is the markdown you type, not a download of the URL. Empty pages do not become sections. Publish the two files yourself. This page does not upload them.

If you remember one sequence from LLMs-full.txt Generator, remember the fields in the order they change a decision. Project name matters because The title of both files. In practice, It is required. The mistake to avoid is this: It is not read from a live site. Page title matters because The heading for that page inside the full file. In practice, A page with no title is skipped. The mistake to avoid is this: The title is what you type, not the HTML title of the URL. Page URL matters because An absolute http or https link. In practice, A page that is included needs one. The mistake to avoid is this: The URL is written into the file. It is not fetched. Markdown body matters because The content of that section. In practice, It is the text you type. The mistake to avoid is this: It is not a scrape of the page. Short index matters because llms.txt lists the same pages. In practice, They sit under one Pages heading. The mistake to avoid is this: The short file does not repeat the long bodies. After that, the checks are simple. Every kept page has a title and an absolute URL. The body is text you wrote, not a scrape. A title-less row was expected to be skipped. Both files were downloaded if you need both. You will publish the files yourself.

The worked example for LLMs-full.txt Generator, read as one scene, is this. Project name Example, one page titled Guide at https://example.com/guide with a short markdown body. llms-full.txt contains the title, the link, and that body. The short llms.txt lists Guide under Pages. The guide URL was not downloaded. Add a second page with a URL but no title. That page is skipped in both files. Add the title and it appears in both.

Keep LLMs-full.txt Generator as this step only. When the job moves on, the LLMs.txt Generator is the next page: LLMs.txt Generator writes an llms.txt markdown file from a project name, a summary, optional notes, and grouped links. The LLMs.txt Auditor covers a different piece of the same work: LLMs.txt Auditor scores the structure of llms.txt markdown you paste and shows the math behind the score.