Automated crawlers had no dedicated trap route in the docs site. Add a hidden site-wide footer link to a noindex page whose existing Marlin pageview identifies requests. Co-Authored-By: Codex <codex@openai.com>
