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

@toolmark/judge-typesafe

v1.0.0

Published

Optional TypeSafe judge plugin for @toolmark/lint: Node-only probabilistic checks for description quality, tool overlap and consequential-hint coverage.

Readme

@toolmark/judge-typesafe

Optional toolmark lint judge backed by TypeSafe's Jev models: probabilistic checks for description quality, missing consequential/destructive hints, and tool pairs an agent could confuse for one another — on top of @toolmark/lint's built-in static rules.

pnpm add -D @toolmark/judge-typesafe

Install it as a devDependency: it's a lint-time judge, run by toolmark lint --judge, and is never bundled into an app. ESM, Node-only (its node export condition dynamic-imports the @typesafe-ai/sdk); Node ≥ 22.12.

It sends only the following to api.typesafe.ai:

  • the page's origin and path (for a --url page; query string, hash and any user:pass@ are stripped, since they can carry tokens or user data), or the manifest file's page name as-is;
  • tool names, titles and descriptions;
  • JSON Schema parameter paths and descriptions, with the Options: <value> = <label>; … list that DOM-synthesized select and radio descriptions end with removed.

Nothing else: never parameter values, option values or labels, default, enum, examples, const, or any other part of a schema.

Requires a TYPESAFE_API_KEY environment variable (or apiKey below). Without one, the judge prints a one-line warning to stderr, returns no findings, and makes no network calls or SDK client construction at all.

TYPESAFE_API_KEY=... pnpm toolmark lint --manifest tools.json --judge @toolmark/judge-typesafe

--judge loads and runs the named module's code (here, this package) as part of the lint run — only pass --judge specs you trust.

// or configure it programmatically and pass the Judge to `lint({ judges: [...] })`
import { typesafeJudge } from '@toolmark/judge-typesafe'

const judge = typesafeJudge({
  qualityThreshold: 1.5, // default
  hintThreshold: 0.85, // default
  overlapThreshold: 0.8, // default
  maxPairs: 200, // default
  strictHints: false, // default; `true` makes judge/consequential-hint an error
  pageTimeoutMs: 120_000, // default; overall bound per page
})

Findings: judge/description-quality, judge/consequential-hint, judge/overlap, judge/overlap-truncated (warn; all warn except judge/consequential-hint under strictHints), judge/unavailable (warn, on an SDK/network error — the judge never throws), and judge/timeout (warn, when a page's requests don't finish within pageTimeoutMs; each request is retried at most once, so a slow or unreachable API never stalls toolmark lint).

Docs: Lint guide · API reference.

License

MIT