Review agent changes before release
Versalist tests an agent change before it ships. Run the skill you use today and the revision you want to ship on the same cases, read the scores side by side, and record the decision.
Why we build it
A good demo can't tell you whether an agent change survives its hardest cases. Versalist gives teams versioned skills, fixed cases and side-by-side results, so a person can see what changed and decide. We don't train model weights or deploy for you. Try the six-case example on your own machine.
Who runs it
Brij Singh builds Versalist at Social Protocol Labs LLC, a Delaware company. Terms · Privacy