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

@git-stunts/plumbing

v3.3.1

Published

Robust async, stream-first Git plumbing for Node, Bun, and Deno.

Readme

@git-stunts/plumbing

npm version License CI

The Industrial Foundation for Git-Based Subsystems.

@git-stunts/plumbing is a low-level, environment-agnostic Git plumbing library designed for the modern JavaScript ecosystem. Built on Hexagonal Architecture and Domain-Driven Design (DDD) principles, it provides a secure, streaming-first, and type-safe interface for Git operations across Node.js, Bun, and Deno.

🪠 Why Plumbing?

Most Git libraries are either high-level wrappers that leak implementation details or low-level bindings that are runtime-locked. plumbing treats Git as a formal subsystem, isolating your application from the complexities of shell execution, runtime differences, and security risks.

  • Environment Agnostic: One API for Node, Bun, and Deno.
  • Async-First Architecture: Modernized for v3.0, every initialization and execution path is asynchronous and non-blocking.
  • Streaming by Default: Memory-efficient data handling using unified streams for all runtimes.
  • Defensive Engineering: Built-in OOM protection, argument sanitization, and lock-contention retries.
  • Mathematical Integrity: Powered by Zod validation to ensure every SHA, Ref, and Object is structurally sound before it touches the disk.

🛡️ Safety: Docker Mandate

To protect your host system from accidental corruption and ensure reproducible execution, execution on the host is strictly prohibited. This project uses @git-stunts/docker-guard to enforce isolation.

# Run the industrial test suite
docker-compose run --rm node-test

📦 Installation

npm install @git-stunts/plumbing

🛠️ Quick Start

Async Initialization

Modern Git plumbing requires asynchronous initialization to safely validate the environment and working directory.

import GitPlumbing from '@git-stunts/plumbing';

// Safely open a repository (Async factory)
const git = await GitPlumbing.createRepository({ cwd: './my-assets' });

// Execute a command with full telemetry and OOM protection
const status = await git.execute({ args: ['rev-parse', '--is-inside-work-tree'] });
console.log(`Ready: ${status}`);

Atomic Commit Flow

Create blobs, trees, and commits programmatically without the fragility of manual shell scripts.

import GitPlumbing, { GitSha } from '@git-stunts/plumbing';

const git = await GitPlumbing.createRepository({ cwd: './my-repo' });

const commitSha = await git.createCommitFromFiles({
  branch: 'refs/heads/main',
  message: 'Feat: immutable asset storage',
  author: { name: 'James Ross', email: '[email protected]' },
  committer: { name: 'James Ross', email: '[email protected]' },
  parents: [GitSha.from(await git.revParse({ revision: 'HEAD' }))],
  files: [
    { path: 'manifest.json', content: JSON.stringify({ version: '1.0' }) },
    { path: 'data.bin', content: new Uint8Array([0xde, 0xad, 0xbe, 0xef]) },
  ],
});

Non-Mutating Prune Inspection

Stream loose unreachable objects that Git considers older than a canonical UTC cutoff. This API always invokes Git with --dry-run --no-progress; unrestricted git prune remains prohibited by the command sanitizer.

const plumbing = await GitPlumbing.createDefault({ cwd: './my-repo' });
const stream = await plumbing.inspectPrunableObjects({
  expiresBefore: '2026-07-01T00:00:00.000Z',
});

for await (const chunk of stream) {
  // Parse Git's `<object-id> <type>` records incrementally.
}
await stream.finished;

Persistent Git Protocol Sessions

Repeated object operations can share one stock Git process. The typed session wrappers handle protocol framing while preserving explicit lifecycle ownership:

const plumbing = await GitPlumbing.createDefault({ cwd: './my-repo' });
const objects = await plumbing.openCatFileSession();

try {
  const metadata = await objects.infoMany([firstOid, secondOid]);
  const values = await objects.readMany([firstOid, secondOid], {
    maxBytes: 8 * 1024 * 1024,
  });
  for (const { oid, type, content } of values) {
    consume(oid, type, content);
  }
} finally {
  await objects.close();
}

maxBytes is a cumulative content budget for readMany(). A response that exceeds the remaining budget is drained without being retained, the operation fails with OBJECT_BUFFER_LIMIT_EXCEEDED, and the session remains aligned for later requests. The default budget is 10 MiB.

Additional typed protocols are available through openMktreeSession() and openFastImportSession(). A single tree supplied to write() remains iterable or async iterable and is streamed without constructing a second whole-tree buffer. writeMany() assembles one bounded framed write for up to 256 independent trees, while writeBlobs() assembles one bounded protocol write for up to 256 blobs under a caller-selectable content budget capped at 64 MiB. infoMany() accepts up to 1,000 object names under a 64 KiB command budget. All batch results preserve input order.

openUpdateRefSession() reuses one git update-ref --stdin process across explicit update-ref transactions. Its update() method performs an unconditional update when expectedOldOid is undefined; null means create-only and an OID means compare-and-swap. It validates Git's ordered start/prepare/commit acknowledgements. A rejected or malformed transaction poisons the session. noDeref is explicit; on the minimum supported Git, consumers must retain any separate symbolic-ref safety check they require.

Sessions have no implicit timeout. Their owner must call close() for orderly completion or terminate()/abort() when abandoning work. Plumbing owns the process and Git protocol mechanics only; callers such as @git-stunts/git-cas own pooling, chunking, caching, retention, and eviction policy.

The Docker-only microbenchmark compares one-at-a-time operations with bounded batch windows or a persistent ref session against fresh SHA-256 repositories. It records elapsed time, Git process topology, stdin write count, and exact object-identity witnesses:

docker compose run --build --rm node-test npm run benchmark:protocol-sessions

📖 Deep Dives

📄 License

Apache-2.0