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

@b3-business/cherry

v0.3.2

Published

A tree-shakeable, minimal API client factory. Import only the routes you need — nothing more.

Readme

cherry

Cherry-pick your API routes.

A tree-shakeable, minimal API client factory. Import only the routes you need — nothing more.

npm JSR GitHub


Latest Changelog - 0.3.2

  • Fix: Ship TypeScript source files in npm package to fix module resolution

See CHANGELOG.md for full history.


What is Cherry?

Cherry is a lightweight API client library that separates route definitions from the client runtime. Routes are plain objects with validation schemas — import only what you use, bundle only what you import.

import { createCherryClient } from "@b3b/cherry";
import { listZones, getZone } from "./routes/cloudflare";

const cf = createCherryClient({
  baseUrl: "https://api.cloudflare.com/client/v4",
  headers: () => ({ Authorization: `Bearer ${process.env.CF_TOKEN}` }),
  routes: { listZones, getZone },
});

// Fully typed, fully tree-shakeable
const zones = await cf.listZones({ account_id: "abc" });

Why Cherry?

The Problem

Official API clients (e.g., Cloudflare, AWS) bundle everything:

  • Every endpoint, even ones you'll never use
  • Massive web shims for Node.js compatibility
  • Complex class hierarchies that defeat tree-shaking

The result? A simple "list DNS records" call pulls in megabytes of unused code, bloating serverless deployments and slowing cold starts.

# Real-world bundle size comparison (hypothetical)
cloudflare-sdk:     2.4 MB (bundled)
cherry + 3 routes:   12 KB (bundled)

The Solution

Cherry inverts the architecture:

| Traditional SDK | Cherry | |-----------------|--------| | Monolithic client class | Minimal client factory (~50 lines) | | All endpoints registered | Routes are plain imports | | Tree-shaking impossible | Only imported routes are bundled | | Runtime schema validation optional | Validation built-in (Valibot) |

Routes are data, not code. They're plain objects describing endpoints — completely decoupled from the client that executes them.


Installation

# npm
npm install @b3b/cherry valibot

# pnpm
pnpm add @b3b/cherry valibot

# bun
bun add @b3b/cherry valibot

# jsr (Deno)
deno add jsr:@b3b/cherry

Quick Start

1. Define a Route

import * as v from "valibot";
import { route, path, param } from "@b3b/cherry";

export const listZones = route({
  method: "GET",
  path: path`/zones`,
  queryParams: v.object({
    account_id: v.string(),
    page: v.optional(v.number()),
  }),
  response: v.object({
    result: v.array(v.object({ id: v.string(), name: v.string() })),
  }),
});

2. Create a Client

import { createCherryClient } from "@b3b/cherry";
import { listZones, getZone, createDnsRecord } from "./routes/cloudflare";

const cf = createCherryClient({
  baseUrl: "https://api.cloudflare.com/client/v4",
  headers: () => ({ Authorization: `Bearer ${process.env.CF_TOKEN}` }),
  routes: { listZones, getZone, createDnsRecord },
});

3. Call Your API

// Named method — discoverable, autocomplete-friendly
const zones = await cf.listZones({ account_id: "abc" });

// Generic call — works with any route, even ones not in `routes`
const zones = await cf.call(listZones, { account_id: "abc" });

Features

Path Parameters

Use path tagged template with param() markers for dynamic URL segments:

import { route, path, param } from "@b3b/cherry";

export const getUser = route({
  method: "GET",
  path: path`/users/${param("id")}`,
  pathParams: v.object({ id: v.number() }),
  response: UserSchema,
});

// Multiple params
export const getComment = route({
  method: "GET",
  path: path`/posts/${param("postId")}/comments/${param("commentId")}`,
  pathParams: v.object({ postId: v.number(), commentId: v.number() }),
  response: CommentSchema,
});

Optional Path Parameters

Use optional() for optional URL segments:

import { route, path, optional } from "@b3b/cherry";

export const getApiResource = route({
  method: "GET",
  path: path`/api${optional("version")}/resource`,
  pathParams: v.object({ version: v.optional(v.string()) }),
  response: ResourceSchema,
});

Query Parameters

Define query string parameters with their own schema:

export const listUsers = route({
  method: "GET",
  path: path`/users`,
  queryParams: v.object({
    page: v.optional(v.number()),
    limit: v.optional(v.number()),
    search: v.optional(v.string()),
  }),
  response: v.array(UserSchema),
});

// Usage: client.listUsers({ page: 1, limit: 10, search: "john" })
// → GET /users?page=1&limit=10&search=john

Query Array Formats

Configure how arrays are serialized in query strings:

export const filterItems = route({
  method: "GET",
  path: path`/items`,
  queryParams: v.object({ tags: v.array(v.string()) }),
  queryParamOptions: { arrayFormat: "comma" }, // or "repeat" | "brackets" | "json"
  response: v.array(ItemSchema),
});

// arrayFormat examples for tags=["a","b"]:
// "repeat"   → ?tags=a&tags=b        (default)
// "comma"    → ?tags=a,b
// "brackets" → ?tags[]=a&tags[]=b
// "json"     → ?tags=["a","b"]

Body Parameters

Define request body schema for POST/PUT/PATCH:

export const createPost = route({
  method: "POST",
  path: path`/posts`,
  bodyParams: v.object({
    title: v.string(),
    body: v.string(),
    userId: v.number(),
  }),
  response: PostSchema,
});

Namespaced Routes

Group related routes into nested objects for better organization:

// routes.ts
export const posts = {
  list: route({ method: "GET", path: path`/posts`, response: v.array(PostSchema) }),
  get: route({ method: "GET", path: path`/posts/${param("id")}`, pathParams: v.object({ id: v.number() }), response: PostSchema }),
  create: route({ method: "POST", path: path`/posts`, bodyParams: PostInputSchema, response: PostSchema }),
};

export const users = {
  list: route({ method: "GET", path: path`/users`, response: v.array(UserSchema) }),
  get: route({ method: "GET", path: path`/users/${param("id")}`, pathParams: v.object({ id: v.number() }), response: UserSchema }),
};

// client.ts
const api = createCherryClient({
  baseUrl: "https://api.example.com",
  routes: { posts, users },
});

// Usage with namespacing
await api.posts.list({});
await api.posts.get({ id: 1 });
await api.users.get({ id: 42 });

Dynamic Headers

Provide headers dynamically (supports async for token refresh):

const client = createCherryClient({
  baseUrl: "https://api.example.com",
  headers: async () => ({
    Authorization: `Bearer ${await getAccessToken()}`,
    "X-Request-ID": crypto.randomUUID(),
  }),
  routes: { /* ... */ },
});

Generic call() Method

Call any route without registering it in the client:

import { createCherryClient } from "@b3b/cherry";
import { getUser, listPosts } from "./routes";

const client = createCherryClient({ baseUrl: "https://api.example.com" });

// Works with any route - useful for one-off calls or dynamic route selection
const user = await client.call(getUser, { id: 1 });
const posts = await client.call(listPosts, {});

Custom Fetcher

Replace the underlying fetch logic for logging, retries, auth refresh, etc.:

createCherryClient({
  baseUrl: "...",
  fetcher: async (req) => {
    console.log(`→ ${req.init.method} ${req.url}`);
    const res = await fetch(req.url, req.init);
    if (!res.ok) throw new Error(`HTTP ${res.status}`);
    return res;
  },
});

Composable Middleware

Composition is userland — no magic middleware system:

import type { Fetcher } from "@b3b/cherry";

const withRetry = (fetcher: Fetcher, attempts = 3): Fetcher =>
  async (req) => {
    for (let i = 0; i < attempts; i++) {
      try { return await fetcher(req); }
      catch (e) { if (i === attempts - 1) throw e; }
    }
    throw new Error("unreachable");
  };

const withLogging = (fetcher: Fetcher): Fetcher =>
  async (req) => {
    console.log(`→ ${req.init.method} ${req.url}`);
    return fetcher(req);
  };

const baseFetcher: Fetcher = (req) => fetch(req.url, req.init);

createCherryClient({
  baseUrl: "...",
  fetcher: withLogging(withRetry(baseFetcher)),
});

Railway-Oriented Error Handling

All client methods return ResultAsync<T, CherryError> from neverthrow:

const result = await client.posts.get({ id: 1 });

// Pattern 1: Check and unwrap
if (result.isOk()) {
  console.log(result.value); // typed as Post
} else {
  console.error(result.error); // typed as CherryError
}

// Pattern 2: Map/chain operations
const title = await client.posts.get({ id: 1 })
  .map(post => post.title)
  .unwrapOr("Unknown");

// Pattern 3: Match both cases
result.match(
  (post) => console.log(`Got: ${post.title}`),
  (error) => console.error(`Failed: ${error.message}`),
);

Typed Error Hierarchy

All errors extend CherryError with type and retryable properties:

import {
  CherryError,
  HttpError,        // HTTP 4xx/5xx (retryable for 5xx and 429)
  ValidationError,  // Valibot schema validation failed
  NetworkError,     // fetch() threw (always retryable)
  SerializationError, // JSON serialization failed
  UnknownCherryError, // Catch-all
  isCherryError,
} from "@b3b/cherry";

const result = await client.posts.get({ id: 1 });

if (result.isErr()) {
  const err = result.error;
  
  if (err instanceof HttpError) {
    console.log(err.status, err.statusText, err.body);
    if (err.retryable) { /* retry logic */ }
  }
  
  if (err instanceof ValidationError) {
    console.log(err.target); // "request" or "response"
    console.log(err.issues); // Valibot issues array
  }
}

Type Inference Utilities

Extract input/output types from route definitions:

import type { InferRouteInput, InferRouteOutput } from "@b3b/cherry";
import { getUser } from "./routes";

type GetUserInput = InferRouteInput<typeof getUser>;
// { id: number }

type GetUserOutput = InferRouteOutput<typeof getUser>;
// { id: number; name: string; email: string; ... }

Route Validation at Definition Time

Routes validate configuration immediately — catch mistakes during development:

// ❌ Throws: "Route has path params [id] but no pathParams schema"
const bad = route({
  method: "GET",
  path: path`/users/${param("id")}`,
  response: UserSchema,
  // Missing pathParams!
});

// ❌ Throws: "Path param ':userId' not found in pathParams schema"
const mismatch = route({
  method: "GET",
  path: path`/users/${param("userId")}`,
  pathParams: v.object({ id: v.number() }), // Wrong key!
  response: UserSchema,
});

Design Principles

  1. Tree-shakeable by default — Routes are plain imports, not registered in a global client
  2. Minimal runtime — Client is ~50 lines, no dependencies beyond Valibot
  3. User owns composition — No built-in middleware, just a replaceable fetcher
  4. Type-safe end-to-end — Params validated in, response validated out
  5. No magic — Everything is explicit and inspectable

Stack

  • Runtime: Bun
  • Validation: Valibot
  • Language: TypeScript (strict)
  • Formatting: Prettier (100 char width)
  • Bundling: tsdown

Dev Info

Publishing

Publishing to npm and JSR uses trusted OIDC publishing via GitHub Actions - no tokens required. Manual npm publish or jsr publish will not work locally.

To release a new version:

  1. Update version in package.json and jsr.json
  2. Update CHANGELOG.md and README (Latest Changelog section)
  3. Commit, tag, and push:
    git add -A && git commit -m "release: X.Y.Z" && git tag X.Y.Z && git push && git push --tags

The GitHub Actions workflow will automatically publish to both npm and JSR.

Dry-run Scripts

npm run dry-npm   # Verify npm package contents
npm run dry-jsr   # Verify JSR package contents

License

MIT