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

agent-goal-runner

v0.2.2

Published

Run goal-driven local agent CLI loops from a browser UI.

Readme

agent-goal-runner

Run long agent coding tasks as a controlled series of fresh, goal-driven local CLI passes. This runs with your current codex or claude subscription, or with local models through the Pi harness.

agent-goal-runner is a local browser app for developers who use goal.md to steer agent work. It starts a localhost server, lets you select a local Git repository, shows the current goal file, and runs repeatable Codex, Claude, or Pi CLI passes with live logs, progress, summaries, optional verification commands, optional repair attempts after verification failures, and optional Git commits.

Agent Goal Runner UI

The app is local-first. It does not host your repository, proxy agent requests through a service, or manage provider authentication. Agent runs execute in the repository you select on your machine and use your local filesystem, local Git, and locally installed agent CLIs.

What It Does

  • Select a local Git repository and inspect or create its goal.md.
  • Run Codex, Claude, or Pi CLI passes from that repository with a consistent prompt and run count.
  • Stream logs, run status, progress, and final summaries into a local UI.
  • Re-read goal.md between passes and stop on completion, blocked state, errors, or user stop.
  • Run optional verification commands after successful passes, with a choice to stop or launch limited repair attempts when verification fails.
  • Optionally commit successful changes with generated commit messages.
  • Install the bundled goal-runner-framework skill globally or into the selected repository.
  • Manage local branches for the selected repository.

Requirements

  • Node.js 20 or newer
  • npm
  • Git available on PATH
  • Codex CLI installed and authenticated for Codex runs
  • Claude CLI installed and authenticated only if you use the Claude provider
  • Pi harness installed and available on PATH only if you use the Pi provider for local-model runs

Codex CLI authentication is handled by the Codex CLI itself. Sign in with Codex from a terminal before using Codex runs in this app. Claude and Pi setup are also handled by their local CLIs. For Pi runs, leave the Pi model empty to use the harness default or enter a local model name to pass as --model. agent-goal-runner does not store provider credentials or perform provider authentication.

Run With npx

Start the local app with:

npx agent-goal-runner

Then open:

http://127.0.0.1:4317

By default the server binds to 127.0.0.1 on port 4317. Override those values with HOST and PORT if needed:

HOST=127.0.0.1 PORT=4320 npx agent-goal-runner

On Windows PowerShell:

$env:PORT = "4320"
npx agent-goal-runner

Development From Source

Clone the repository, install dependencies, and start the development servers:

git clone https://github.com/Scott-Bauman/agent-goal-runner.git
cd agent-goal-runner
npm install
npm run dev

Development mode runs the Fastify backend and Vite frontend together. The backend listens on http://127.0.0.1:4317, and Vite prints the frontend URL, usually http://127.0.0.1:5173.

To run the built app from a clone:

npm run build
npm start

The built backend serves both the API and the built frontend from http://127.0.0.1:4317.

For source changes, validate the project with:

npm run typecheck
npm run lint
npm test
npm run build
npm pack --dry-run

Bundled Goal Skill

The package includes a bundled goal-runner-framework skill. The skill helps Agents create and maintain goal.md files in the structure expected by this app.

You can install it from the UI after selecting a repository, or from a source checkout:

npm run install:skill:global
npm run install:skill:repo -- "C:\path\to\target-repo"

Global install copies the skill to your user-level .agents/skills directory. Repo-local install copies it into the selected repository under .agents/skills, which is often the most reliable option when switching between repositories.

Documentation

License

MIT. See LICENSE.