I would take a look at these threads, for more specifics on how Scrivener is best used in this context:
-
Scrivener and technical manuals: covers and discusses a broad range of tools, from LaTeX, to InDesign to Quarto and a few others. Scrivener is quite adept at working with all of these systems, so long as you start from the right place (i.e. not treating it like a word processor, but more like a Markdown writing platform).
-
Speaking more specifically on LaTeX—which can be approached much more directly as another plain-text syntax, rather than using Markdown—here are some thoughts on that debate, but it’s worth noting that’s a little dated, as these days we have an official non-fiction LaTeX template in the software, which fully demonstrates a number of the concepts spoken of in that post. For those where LaTeX is the one and only end-game, this can often be the best approach. Markdown is by necessity an abstraction, and to get around its limitations one needs to employ tactics to do so (more on that below, you absolutely can and make it beautiful in Scrivener).
The “General Non-Fiction” template is more… I would say for biographies, topical books and the like. You will on the whole find very little in Scrivener’s basic feature set that supports the needs of a technical writer, and this template is rooted entirely in that more basic approach: fonts and styles basically, not even real captions you can cross-ref to, much less easily build a list of tables/figures etc. from.
-
Definitely search the Markdown & LaTeX category for Quarto. This is an up-and-coming system that promises to be a one-stop Markdown-based publication system, specifically for those in the sciences and technical writing fields. Scrivener is quite suitable as a starting point for production there, as it can produce Pandoc-compliant output, which is what Quarto needs. There are already some fantastic project templates and workflows documented by the user base. While Pandoc is good on its own, if I needed a DOCX workflow I would very much consider Quarto over it, because it plugs some gaps about DOCX output I don’t care for (like captions that are much as limited as Scrivener’s styled text with baked in numbering).
-
Also search for Scrivomatic, by another forum member, which aims to streamline and fully automate Scrivener+Pandoc processing, with an emphasis on academic, sciences and technical needs.
I believe I explain the basics of what makes Scrivener tick as a syntax content generating platform in some of those links, but to put it more succinctly: what sets it apart from many other systems is that it lets you build your own “front end”, making the text easier to work with and edit around ultimately. For a bit more on what that means, at a practical level, here is a discussion that goes a bit into that concept, which shows a demonstration of how the user manual project for Scrivener employs styles and section layouts to generate syntax—meaning as the one working with that text, I don’t ever have to see it because the compiler inserts all of that for me. The “front end” for a menu label is some blue text in the text editor.
Some take that principle all the way to its end conclusion, and do not even see a bit of Markdown (or other syntax, again, Scrivener is flexible in what input you use, but I focus on that since Obsidian was mentioned, and it makes all kinds of sense to normalise the writing interface if you can). Some use the “Emphasis” style to insert asterisks around the marked text, for example. That is entirely a matter of taste with a program like this! You can go from 100% pure Markdown in the text editor (which makes for maximum cross-software compatibility when using external folder sync to use multiple tools on the same source, like Obsidian), to more like I do with a mix of whatever makes sense in the moment, to like I say, something that really looks a lot more like a rich text user’s project at first blush.
Hope that gives you some ideas at any rate! The big advantage of considering Markdown w/ Pandoc/Quarto in Scrivener is that your output options go from half a dozen or so with mediocre results, to this.
As a postscript, the Scrivener user manual project, while not a technical document per se, at least does employ a number of features of such a document. I referred to some of its techniques above, and this is where you can see them in action and in settings. Two caveats: (a) it still does not compile properly in the Windows version and (b) the source material is quite old (but that matters little for the technical implementation of it). You can compile it, but it won’t typeset because it produces malformed Markdown in some places, and a few other issues.