Create Skill Test skill: what it does and how to install it

What it does. Guides the creation of eval.yaml specs, from choosing distinct stimuli and sizing for statistical power to arranging fixtures, writing graders and outcome-focused rubrics, and validating the result. It covers Vally's schema and checks for overfitting, weak evidence and fixture defects.

When to use it. For creating or reviewing evaluations and their fixtures in dotnet/skills. It does not diagnose existing evaluation failures or create and edit skill instructions.