CAIN-42 CAIN Studio
25CAIN-42 Top 25 › AGENT-CI › FAQ

FAQ

Why no AI judge?

Because a release gate has to give the same verdict twice. Every rule here is arithmetic on the runs you send, and a second, independent implementation must agree before anything is signed.

Which agent frameworks?

Any whose runs you can export as steps with a tool, arguments, cost and dependencies.

Can the team under test still lie in its traces?

Yes: traces are self-reported. The gate checks them against the baseline (a dropped path, missing costs, a lowered risk tier are flagged), not against reality.

FIRME AI CORPORATION · CAIN-42 is a Byzantine trust fabric. Numbers marked "our own check" were run by us, not by an independent party. Nothing on these pages is a certification.