Q: Tools for Reproducible Computing?

Thank you to all of the panelists for your introductions and initial comments.  In my first posting, I suggested that we need better tools, better habits, better repositories, and higher expectations in order to achieve reproducibility.  That seems like a useful way to structure the next few questions, so I will start with tools:

What currently-available tools do you recommend for enabling reproducible scientific computing?  Is there a tool that we ought to have, but do not?

P.S. I am using "reproducibility" as an easy shorthand for re-usability, re-creation, verification, and related tasks that have already seen some discussion.  Please interpret the question broadly.