We have a recent regression in one of our Mozmill tests which report to brasstack: could not validate element XPath: /html/body[@id='gsr']/div[@id='cnt']/div[@id='res']/div/p[1]/b with text oau45rtdgsh34nfr That test is running a feeling lucky search against Google and doesn't expect to get back a search result. But Google already knows about this search term because it scans all our test results: http://www.google.de/search?hl=en&source=hp&q=oau45rtdgsh34nfr&btnG=Google+Search&aq=f&oq= I feel that we should exclude search robots from scanning our Mozmill results on brasstacks. Clint, can we do that?

I have added a meta tag to the head section of the delivered content: <meta name="robots" content="noindex, nofollow"> Lets see if that helps.

Bustage fix landed as: http://hg.mozilla.org/qa/mozmill-tests/rev/06d676d1cae9 http://hg.mozilla.org/qa/mozmill-tests/rev/5e5cba320e6d I have also updated all the Litmus tests. Clint, please make sure we have a robots.txt or the meta line served for the tcm part on brasstacks.