Testing Document Import with Linux from Scratch

Alistair McLaurin - February 2nd 2026




A key use case for the Clouds and Light Platform is to convert existing company documents into training, run-books and how-to guides. This can be applied to everything from creating onboarding guides for new team members to DR, BCP and incident response manuals.

It is also a very powerful platform for developing system documentation using LLMs. If you have an application or system which lacks comprehensive documentation, using an LLM agent can be a powerful way to consolidate existing data and augment it with code analysis or cloud service configuration discovery.

Equally, once a system is documented in Markup it then becomes easier to build processes and tests to verify your data is up to date. You can then create a feedback loop for updating areas which may have changed.

By creating standardised document templates you can create a uniform way to bring together system information, quickly identify gaps and present the results back in a human readable format for further analysis and investigation.

To demonstrate the capabilities of the PDF import and Edit feature we wanted to start testing some longer documents, which might better reflect a company manual.

One good set of open source manuals are the Linux from Scratch series of books. If you want a good primer in IT, from how operating systems work to user management, networking, file system structure etc. then there are few better ways to learn than by building your own operating system.

The Linux from Scratch website has the guide in both PDF and HTML format. I've worked through it in the past and it is complex and sometimes painful but very rewarding when you get to the end and have a bootable OS you built yourself.

The book is a 366 page PDF and the Markup Conversion takes just under 2 minutes to convert this to Markup. It also seems to capture the text with pretty much 100% accuracy as well as identifying code blocks and inline code, list items and chapters and headings.

I then used the Markup Editor to edit the document for style, add hyperlinks and some additional call out boxes, fix the occasional error where text or a table spanned multiple pages etc. In total the editing of 360 pages took around 5 hours (or under 1 minute per page)

If you would like to see the results they are on the site at Demo - Linux from Scratch.

The editing was almost entirely done in the online editor, even working on the longest chapter, Chapter 8, both the editor and live preview functionality worked seamlessly.

The platform has the ability to import and convert documents while maintaining strict security guardrails and access control. It can be hosted on a dedicated environment in our SaaS platform or can be deployed in your own AWS account and use your own encryption keys for comprehensive security of your company data.

If you would like to book a meeting to discuss how this might be useful in your environment or see a more detailed demo demo please email alistair@cloudsandlight.com or use our feedback form.





Terms and Conditions Privacy Policy

Copyright © Clouds and Light Limited 2026