A toolkit for evaluating Jev probabilities on labeled data, selecting confidence thresholds and checking model drift.
Real data. Real impact.
Emerging
Developers
git clone https://github.com/abhixhek/jevcal.gitPer week
Excellent
One command unlocks capabilities that would take hours to build.
A toolkit for evaluating Jev probabilities on labeled data, selecting confidence thresholds and checking model drift.
Runs fixed questions and measures accuracy, calibration, coverage and escalation rates.
Connects threshold selection and model-change checks to reports and CI.
The README demo table uses a simulator, not measured Jev results; production thresholds need workload-specific validation. Not run here.
Jev is TypeSafe AI's System One decision model: it returns typed scores and choices (not free text), so agents can route, filter and gate at low latency and cost.
Source: abhixhek/jevcal — 10 stars, MIT. Listed by Awesome Jev.
MIT
4,600+ AI skills, agents & workflows. Install in 60 seconds. Part of the Torly.ai family.
© 2026 Torly.ai. All rights reserved.