Scenario: You need to add a library to your awesome project that does something. You don't know what the library is, so what do you do? Search the cheeseshop, of course. And when you find it in the cheeseshop?
It'd be nice if we could have a bookmarklet that installs from the great big crane project (I have mine running at localhost:8001). When you click the bookmarklet, it could prompt for a project name, and then scrape the current cheeseshop page (or maybe just the url) to find the name of the dependency we want to install. Then it updates either the requirements.txt (for a pip project) or the buildout's zc.recipe.eggs recipe to include that dependency.