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

@unitflow/devtools

v0.3.1

Published

Unitflow runtime inspector bridge and MCP server for agent-driven debugging.

Readme

@unitflow/devtools

Runtime devtools for agents. An MCP server exposes your running unitflow app — live model instances, store values, and a causal event log ("this emit ran this handler which wrote this store") — to Claude Code or any MCP client.

Everything is automatic: ports are named from the model contract (task-model(42).inputs.rename), no manual { name } needed.

Setup

1. Install (dev dependency of your app):

pnpm add -D @unitflow/devtools

2. One line in your app entry:

import { devtools } from "@unitflow/devtools";

if (import.meta.env.DEV) devtools(runtime);

Production builds tree-shake this away; a running app without the MCP server just retries quietly in the background.

3. Register the MCP server with Claude Code (once per project):

claude mcp add unitflow -- npx -y --package=@unitflow/devtools unitflow-mcp

or in .mcp.json:

{
  "mcpServers": {
    "unitflow": {
      "command": "npx",
      "args": ["-y", "--package=@unitflow/devtools", "unitflow-mcp"]
    }
  }
}

That's it. Start your dev server, open the app, and ask the agent.

Tools

| Tool | Returns | | --- | --- | | list_instances | Live model instances: id, key, lease count | | get_stores | Store values by name filter — derived (combined) stores included, evaluated on demand | | event_log | Writes, emits, and instance lifecycle, oldest first, with causal links | | trace | The full causal chain of one event: ancestors to the root, plus descendants |

A typical answer to "what happened when I typed in the search box":

#5      write  app/search.ui.setQuery.target   "relay"
#6  ←#5 write  app/search.outputs.results      { _tag: "Success", value: [...] }

Options

devtools(runtime, {
  app: "my-app",          // how the app introduces itself
  url: "ws://localhost:4477",
  capacity: 2000,          // inspector ring buffer
});

The hub port is 4477 by default; override with UNITFLOW_MCP_PORT on the server and url on the app side.

How it works

Debug.attach() (from @unitflow/core) installs an inspector on the registry — when detached, the hot paths pay a single property check. The bridge streams its ring buffer to the unitflow-mcp process over a local WebSocket; the MCP side is built on effect/unstable/ai/McpServer. Causality comes from the synchronous dispatch windows of the core, so chains are exact for synchronous cascades; work after an async suspension records no cause.