Let's mimic git.git auto tag fetch behavior.

Related to https://github.com/libgit2/libgit2sharp/pull/927/files#r23450279

git fetch against a newly initialized repository with the following config

[core]
    repositoryformatversion = 0
    filemode = false
    bare = false
    logallrefupdates = true
    symlinks = false
    ignorecase = true
    hideDotFiles = dotGitOnly
[remote "origin"]
    url = https://github.com/libgit2/TestGitRepository
    fetch = +refs/heads/*:refs/remotes/origin/*

outputs

$ git fetch
remote: Counting objects: 68, done.
remote: Total 68 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (68/68), done.
From https://github.com/libgit2/TestGitRepository
 * [new branch]      first-merge -> origin/first-merge
 * [new branch]      master     -> origin/master
 * [new branch]      no-parent  -> origin/no-parent
remote: Counting objects: 1, done.
remote: Total 1 (delta 0), reused 1 (delta 0)
Unpacking objects: 100% (1/1), done.
 * [new tag]         annotated_tag -> annotated_tag
 * [new tag]         blob       -> blob
 * [new tag]         commit_tree -> commit_tree

git is noticing that the server did not automatically send the tags and is performing a second fetch. This second fetch is completely different from the first one

No source file or test is named in the issue. Start by locating the fetch implementation and its existing fetch tests, then compare the initial fetch with Git's follow-up tag fetch when the server does not send tags automatically. Done means matching the demonstrated two-fetch behavior and reporting the newly fetched tags.