the http interface is just an apache dir listing (a “find-links” repo) - pants/pex will parse out the html links and resolve against what’s available there.