I incepted my worktree by following some reddit's user advice
which had worked for me a few years ago when I started rfdupes:
https://www.reddit.com/r/openbsd/comments/1b7wl8h/

The trouble is very simple, for a start : how is `got' supposed to
know where to look for a repository to add the files to ?

The error is the following:

$ pwd
/var/www/got/public/saboua.xyz
$ got add /var/www/htdocs/saboua.xyz/doc/*.css
got: 'got add'needs a work tree in addition to a git repository
Work trees can be checked out from this Git repository with 'got 
checkout'.
The got(1) manual page contains more information.

On 2025-03-06 10:10, Stefan Sperling wrote:
> On Thu, Mar 06, 2025 at 05:41:24AM +0100, Sylvain wrote:
>> Prior to putting my website source code in a got repo for public 
>> display, I performed a little trick involving mainly mv commands in 
>> order to trim my website's folder of irrelevant content, which I 
>> restored afterwards.
>> 
>> I neglected to add the *.css files, located in /doc along with a bunch 
>> of other documents. Is there a command to commit those files, and only 
>> those files, to the repo ?
>> 
>> My ~/.history file has several got checkout -p doc [...] attempts that 
>> failed.
> 
> Hmm. 'got add' is indeed the command you should be using.
> I don't see anything wrong with your general approach.
> 
> Generally, adding files involves checking out a work tree (with or 
> without
> the -p option, it doesn't really matter), then putting new files into 
> this
> work tree and running 'got add' on them, i.e. passing the file paths 
> you
> want to add to 'got add'.
> 'got status' should then show the relevant files in 'A' status.
> Now run 'got commit' to commit the new files. This also accepts 
> filenames
> on the command line in case you want to limit the commit to specific 
> files
> while other unrelated changes are sitting in the same work tree.
> 
> Since you were asking for an equivalent to 'git add': Git's add command
> also performs staging functionality. Got has a separate command for 
> this
> called 'got stage'. You can optionally use that too, but it is not 
> necessary
> at all in this trivial use case.
> 
> So there is nothing super complicated about adding new files, and it 
> seems
> you were intending to perform the required steps correctly.
> 
> Except for some reason the checkout -p command ran into an error, so 
> you
> couldn't get a work tree. What was the error message?