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

@ikenga/skill-groundwork

v0.8.3

Published

Reusable research → design → plan → orchestrate → act skill for Claude Code.

Readme

@ikenga/skill-groundwork

Groundwork: research → design → plan → orchestrate for Claude Code, with plan folders that don't rot.

A Claude Code skill that scaffolds and maintains a reusable research → design → plan → orchestrate → act folder for any non-trivial work — software features, marketing campaigns, org changes.

It drops a domain-agnostic spine (00-README · 01-plan · 02/03 research · 04-discussion newest-first · 05-tracking · 09-orchestration) plus stateless action-skills that augment the docs in place without clobbering hand-written prose. Profile-driven: software, general, content, and design-system. Each plan also gets self-contained HTML views — a board (artifact/board.html), a file explorer (fully offline) with search (artifact/explorer.html), and a cross-plan plans index (<plans-dir>/_index.html).

Install

# Global install (recommended — available across all projects)
npx skills add ikenga-hq/groundwork -g

# Project install (committed with your repo)
npx skills add ikenga-hq/groundwork

npx skills add resolves skills/groundwork/SKILL.md and symlinks it into ~/.claude/skills/groundwork/.

Source of truth & publish flow

There is one editable copy. As of the 2026-08-21 unification, the dev source moved into this package — ./skills/groundwork/ is the source. Edit it directly.

| Copy | Path | Role | |---|---|---| | Source (this package) | ikenga-pkgs/packages/skills/groundwork/skills/groundwork/ | ADR-009 home, Changesets-versioned. Edit here. | | Workspace dogfood | <ikenga-workspace>/.claude/skills/groundwork/ | A symlink into the path above — not a copy. | | Published mirror | ikenga-hq/groundwork | The npx skills add install surface. Built FROM this package. |

Why the change: the previous layout kept the dev source in ikenga-workspace, which is private, and copied it into this public package via sync-from-dev.mjs. Contributors could read the published skill but not the file anyone actually edited, and every change needed a sync hop that could silently drift. The script is gone; there is nothing to drift.

Files no longer carry a GENERATED banner, because they are no longer generated. PORTABILITY.md was emitted by the old sync script and is now hand-maintained source.

  • pnpm test — the deterministic hard gate: skills/groundwork/scripts/test_groundwork_state.py (state machine, profiles, ids, designs) and test_fence_integrity.py (fixture plan: writes touch only fenced regions, identical re-runs are byte- and mtime-exact no-ops, and deliberately broken writers are caught). Python 3 stdlib only. Runs in ikenga-pkgs CI (ci.yml, on PRs that touch this package and nightly).
  • pnpm build:mirror — runs pnpm test first and refuses to build if it fails (no skip flag; the mirror push is manual, so this is where the publish depends on the tests), then emits the standalone ikenga-hq/groundwork mirror tree (package.json + README + install.sh + skills/groundwork/) to ./dist-mirror for review before pushing. Note the mirror repo's own README.md, CONTRIBUTING.md, .github/ and assets/ are hand-maintained there and are not emitted — a blanket sync would delete them.

How it compares

groundwork is one option among several planning skills. Superpowers and etr/groundwork are different tools; the points below describe only what this repo implements and make no claims about their internals.

  • Plan state lives in a .groundwork.json anchor plus fenced generated regions in the markdown docs; re-runs write only inside fences, and test_fence_integrity.py checks this.
  • Stable IDs (WP-NN, G-NN, D-NN) link plan, tracking, orchestration and the HTML views.
  • Profiles (software, general, content, design-system, film) change vocabulary and optional docs without changing the spine.
  • Produces standalone HTML views (board, explorer, plans index) per plan, plus an optional orchestrate --emit-workflow multi-agent kickoff.
  • State logic is stdlib-only Python with a deterministic test suite (npm test).

Commands this skill runs

For users who want to allowlist rather than blanket-approve Bash. Derived from SKILL.md and actions/*.md; <skill> is the installed skills/groundwork/ directory.

python3 <skill>/scripts/groundwork_state.py <subcommand> --plan <plan>
    scaffold, spine-gate, resolve-profile, next-id, write-region, stamp-research,
    status-data, board-data, explorer-data, living-spec-data, plans-index-data,
    register-subplan, register-id, register-design, register-design-impl,
    design-data, design-lock, design-migrate,
    issue-sync-data, register-issue, register-milestone, apply-issue-sync
python3 <skill>/scripts/generate_explorer.py --plan <plan> | --all-under <plans-dir> [--missing-only]
python3 <skill>/scripts/generate_plans_index.py --plans-dir <plans-dir>
python3 <skill>/scripts/watch.py --plan <plan> | --plans-dir <plans-dir>
node    <skill>/scripts/build-offline-artifacts.mjs     # maintainers only

issue-sync additionally runs gh repo view, gh issue create|close|reopen|list, gh pr list, gh api repos/{owner}/{repo}/milestones and glab repo view, glab issue create|list, glab api projects/:id/milestones with your credentials. It asks for confirmation (target repo, visibility, exact text) before any write.

Security notes

  • Web research is untrusted data. The research action writes fetched web content into 02-research-external.md, and plan docs later become briefs for build agents. Fetched text is kept in the delimited research section, cited, and must be treated as data, never instructions, when briefing agents. Review 02 before orchestrating.
  • issue-sync writes to your Git forge with your credentials and publishes plan text to the repo. It confirms the target repo, its visibility and the exact text before any write.
  • CDN scripts. board.html and the living spec (artifact/index.html) load React and @babel/standalone from cdn.jsdelivr.net, pinned to exact versions with SRI (integrity) hashes. They need network on first open; only the explorer and plans index are fully offline.
  • Explorer iframes. Gallery thumbnails use sandbox="allow-scripts" (opaque origin, no access to the explorer). The full viewer keeps allow-same-origin because previewed plan artifacts (the board, living spec) use localStorage and fetch sibling plan files; the explorer only talks to them via postMessage (theme push). Plan folders from untrusted clones are untrusted input: previewed HTML runs scripts, so do not open them in the viewer.

Portability

The skill's docs and the standalone board carry a few this-workspace-internal references (plans/studio / plans/groundwork) that are illustrative in a target project. See skills/groundwork/PORTABILITY.md for the full disclosure.

License

Apache-2.0. Copyright © 2026 Royalti.io.