npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

jevkit-drift

v0.1.1

Published

Detect when a new TypeSafe Jev model version flips decisions your code depends on. Replays a golden set and diffs the answers.

Readme

jevkit-drift

TypeSafe's docs warn that jev-latest moves when a new version ships, and that confidence thresholds tuned against one version do not automatically hold on the next. jevkit-drift answers the question that warning implies: for the requests you actually care about, what changed?

Unofficial and unaffiliated with TypeSafe.

npm install jevkit-drift

Two kinds of change

A flip is a changed decision. Your code takes a different branch. This is what breaks things.

A shift is movement in the probability distribution with the same decision still on top. Harmless on its own, but it is what walks an answer toward a threshold, so a large shift is an early warning before anything has flipped.

Both are reported separately, because conflating them is how a drift report ends up either too noisy to read or too quiet to help.

Use

import { readRecords, writeRecords } from "jevkit-core";
import { compareSets, replay } from "jevkit-drift";

const baseline = readRecords("golden.jevl");
const candidate = await replay(baseline, (s, q) =>
  client.systemOne(s, q, { model: "jev-1.14.0" }),
);
writeRecords("candidate.jevl", candidate);

const report = compareSets(baseline, candidate);
console.log(report.summary());

for (const [requestId, delta] of report.flips) {
  console.log(requestId, delta.describe());
}

CLI

npx jevkit-drift golden.jevl candidate.jevl
npx jevkit-drift golden.jevl candidate.jevl --max-flips 3
npx jevkit-drift golden.jevl candidate.jevl --max-shift 0.15
npx jevkit-drift golden.jevl candidate.jevl --format json

Exit codes: 0 within tolerance, 1 drifted, 2 bad usage.

How records are paired

On request_id, the digest of state plus questions with the model deliberately excluded. That is the whole reason the record format keeps two digests: the full id changes when the model changes, so it cannot pair a baseline with its replay, while request_id can.

Change the state or a question and the records stop pairing, which is correct. They are no longer the same request, and comparing them would be meaningless. The CLI says so rather than reporting zero drift.

License

MIT