Same brief, same model
Each brief is given word for word to the same model, the same model, for every row in the matrix.
SkillArena compares UI skills for AI coding agents. Every page was generated by the same model from the same product brief. The only variable is the skill added to the system prompt, and the first row is a no-skill control group.
One row per skill, starting with the control group. One column per brief. Open a skill to see all of its runs on the full results site.
Loading the latest results…
Swipe sideways to see every brief →
Skills promise better UI from coding agents. SkillArena tests the promise by holding everything else still, so the differences you see come from the skill alone.
Each brief is given word for word to the same model, the same model, for every row in the matrix.
The one variable is the skill added to the system prompt. The first row has no skill at all and serves as the control group.
Every skill and brief pair is generated three times, so a single lucky or unlucky run doesn't decide the verdict. Every link opens the full results on skillarena.dev.
New skills, new briefs and new models are added over time. Leave your email and we'll send updates when fresh results land.