jevals: agent evaluations that double as runtime guardrails

byebaybaySep 22

Jevals turns agent checks into typed questions so the same evaluation definitions run on saved traces and as live guardrails using Jev-style decisions.

Open on Reddit

Preview of jevals: agent evaluations that double as runtime guardrails

Launch: Jev
Use case: Agents
Format: Build
Source: reddit.com

byebaybay

See all

Gregor Zunic

@gregpr07 · Sep 17

NandhaKishorM

@NandhaKishorM · Sep 18

Open-source repo in the Jev ecosystem (31,378 stars): Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a singl.

browser-use

@browser-use · Sep 16

jaredpalmer

@jaredpalmer · Sep 17

Open on Reddit

jevals: agent evaluations that double as runtime guardrails

More built with Jev