The assistant often tries to fetch details from a github file for me, but then just gets stuck for quite long waiting for a timeout and gets nothing back. Sometimes it figures out it can get the contents from the raw link, but often it just goes ahead and tries an another file or an alternative source
The assistant should know to use the raw files instead of fetching the blocked pages, or have some redirection done by the harness
