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

grepai-skills

v0.2.0

Published

Connect AI agents to GREP for research, domain expertise, apps, slide decks, spreadsheets, and multi-step workflows.

Readme

GREP Skills

Connect your agent to GREP for research, deliverables, reusable experts, and multi-step workflows. The package is now named grepai-skills (formerly grep-research-skills).

Install

npx grepai-skills

Requires Node.js 18+. The installer copies the package to ~/.grepai-skills/, installs Claude Code skills, and detects OpenClaw. Cowork uses the ZIP below; local Claude Code installation does not install a Cowork connector or skill.

Existing installations can run the new installer to repoint GREP-owned skill links. Credentials remain in ~/.grep/session.json; unrelated skill links and directories are preserved. The old npm package remains a separate published package; changing this repository does not update it.

Until the renamed npm package is published, install this checkout with node bin/install.js.

Claude Code marketplace

/plugin marketplace add parcha-ai/grepai-skills
/plugin install grepai-skills@grepai-skills

Claude app and Cowork

  1. Download the skill ZIP from GitHub releases, or build it from source.
  2. Open Customize → Skills → + → Create skill → Upload a skill and upload the ZIP.
  3. Enable the skill. For team-wide provisioning, an owner uses the organization's Skills settings.

This is a single-skill ZIP, not a plugin archive. Its research/SKILL.md routes to bundled workflow references, scripts, and resources. See Claude's upload instructions.

npm ci
npm run build:cowork

The ZIP is written under dist/ using the package version. Allow api.grep.ai, api.descope.com, and grep.ai in environments with restricted network access. Preview deployments additionally need their configured host. Adding the MCP connector is a separate step.

Git checkout

git clone https://github.com/parcha-ai/grepai-skills.git
cd grepai-skills
./setup

Connect native MCP tools

For Claude Code:

claude mcp add --transport http --scope project grep https://api.grep.ai/api/v2/mcp

Then use /mcp to authenticate through the browser. In Claude app/Cowork, add the same URL as a custom connector. For other agents, use their Streamable HTTP and OAuth setup.

The current server covers research, continuation, cancellation, files, attachments, usage, expert configuration and builds, server-side skills, context files, versions, and sharing. Discover actual tools and schemas; feature gates and scopes vary. Batch tools exist in Parcha main but were not advertised by production when checked on 2026-09-06.

The MCP reference explains request fields and capabilities. The setup follows Claude Code's MCP documentation.

Authentication

| Surface | Authentication | |---|---| | Native MCP | Browser OAuth for https://api.grep.ai/api/v2/mcp | | REST v2 CLI | parcha-* API key; the server also supports properly scoped REST OAuth | | Legacy REST v1 CLI | Descope email-OTP session, refreshed automatically |

The MCP verifier rejects REST API keys and ordinary OTP session JWTs. Signing into the CLI does not sign into MCP. The scripts select v1 for OTP sessions and v2 for API-key sessions.

Run /grep-login for signup, email login, or API-key setup. New users complete website onboarding and return to the agent. Existing connected MCP users can work without a local CLI session.

Workflows

| Skill | Purpose | |---|---| | research | Sourced research, default medium effort | | quick-research | Focused lookup, low effort | | ultra-research | Explicitly exhaustive investigation, high effort | | grep-plan | Research with project context before planning | | grep-skill-creator | Research, draft, test, and improve a local agent skill | | grep-domain-expert | Discover/use experts and manage custom experts when requested | | grep-build-app | Build an app or requested deliverable | | grep-build-slidedeck | Research-backed presentation | | grep-build-spreadsheet | Research-backed table or spreadsheet | | grep-with-context | Research using supplied files | | grep-continue | Follow up on an existing job | | grep-research-workflow | Chain research and deliverables; use batches when available | | grep-mcp | Connect and verify native MCP | | grep-login | CLI authentication and onboarding | | grep-status | Account and job status | | grep-upgrade | Inspect current plans and upgrade options |

Installed skill names remain stable. The Claude/Cowork ZIP exposes one research router. Native GREP tools are preferred when connected; scripts provide the fallback.

Research and builds run in the background. Preserve job IDs and resume polling after timeouts instead of submitting duplicates. Present findings, citations, the research link, and actual artifact links. Timing, prices, and expert catalogs are discovered rather than guaranteed by the package.

CLI

node scripts/auth.js login [email protected]
node scripts/auth.js status
node scripts/grep-api.js research "Compare these vendors" --effort=medium
node scripts/grep-api.js status <slug>
node scripts/grep-api.js result <slug> --no-wait
node scripts/grep-api.js files <slug>
node scripts/grep-api.js continue <slug> "Add a migration checklist"
node scripts/grep-api.js experts
node scripts/grep-api.js discovery

run and waiting result commands poll in-process; agents should launch them through their background mechanism. v2 continuation, timeline, and attachments require v2 credentials. Describe deliverables in the question/context and select a discovered expert; v2 does not accept the old output_type field.

Configuration: GREP_API_BASE defaults to https://api.grep.ai; GREP_UI_BASE overrides generated links; GREP_API_BASE_PATH overrides automatic v1/v2 routing.

Development

  • scripts/: authentication, REST API, billing, update check.
  • skills/: individual workflows.
  • resources/: shared product/MCP guidance and examples.
  • dist/cowork/research/: consolidated skill source.
  • bin/: installer and ZIP builder.
  • .claude-plugin/: plugin and marketplace manifests.
  • .well-known/skill-manifest.json: package discovery.
  • test/: isolated automated verification.

Product contracts are checked against Parcha main edb6304f34; production discovery can lag main. Test and build before publishing. Renaming the GitHub repository does not publish an npm package or create a release.

MIT — Parcha Labs, Inc.