Using LLMs with the Pismo Developers Portal
The Pismo Developers Portal includes an llms.txt file to make it easier for large language models (LLMs) to parse and understand the documentation. The file provides a machine-friendly Markdown-based index of all documentation pages on the portal, and is kept in sync with site updates.
What is an llms.txt file?
An llms.txt file is an emerging AI convention designed to help LLMs understand a website's content, reducing the need for scraping HTML pages.
Instead of crawling the entire website, LLMs can parse a single file to understand the structure of the documentation and follow explicit links to source content in Markdown. This approach ensures LLMs are working with the most up-to-date first-party information.
To view the llms.txt file, do the following:
- Go to the Pismo Developers Portal.
- Scroll down to the page footer.
- Select Read llms.txt.
Why Pismo provides an llms.txt file
The llms.txt file is designed to make Pismo's developer documentation more accessible and usable for LLM-based tools by:
- Reducing the need for web scraping—Websites contain complex HTML structures and boilerplate content, and without an explicit index, LLMs waste tokens parsing irrelevant HTML markup. The llms.txt file provides a singular, structured entry point to the portal.
- Providing an authoritative content index—The llms.txt file lists every documentation page published on the portal, as well as the most-recent changelog entries. This ensures LLMs work from the same single authoritative source as human users.
- Helping LLMs understand the site structure—The portal is designed for for human users, not LLMs. Navigation menus, breadcrumbs, and UI-driven grouping aren't always reliable signals for automated tools. The llms.txt file makes the structure explicit by clearly specifying the grouping of documentation pages, based on guides, APIs, events, and the changelog.
- Improving accuracy—The platform automatically updates the llms.txt file any time Pismo makes a change to the documentation. This means LLMs don't have to rely on old or cached data, reducing the risk of AI hallucinations.
Access documentation in Markdown
The llms.txt file provides direct access to a Markdown version of every documentation page on the portal. Because the content is formatted in Markdown, it preserves the same information as HTML, but without any layout or styling. This makes it easier for LLMs to parse the content accurately and for you to manually copy and paste page content into LLM tools.
Markdown-formatted content offers the following advantages:
- Clear page and content structure defined with Markdown notation, reducing ambiguity in the page layout
- Improved token efficiency, allowing LLMs to process more meaningful information within the same context window
- Reduced risk of extracting incomplete content, as the content doesn't rely on client-side rendering or custom JavaScript functions to expose page sections
Updated about 3 hours ago