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

@getskillmd/mcp

v0.1.7

Published

Model Context Protocol server for skillmd — generate, fetch, and browse skill.md files from any IDE

Downloads

716

Readme

@getskillmd/mcp

Model Context Protocol server for skillmd. Paste a URL into Claude Code, Cursor, Windsurf, Antigravity, or any MCP-aware tool and get back a skill.md — plus design tokens, fonts, brand assets, and screenshots.

The server gives your agent:

  • Markdown skill files describing a product, API, or library
  • A typed design-token contract (tokens.json + globals.css variables + Tailwind preset)
  • Real brand screenshots (above-the-fold and full-page) as inline images
  • Real brand assets (favicon, OG image, Twitter card, logo) as inline images
  • A verify_design round-trip that scores how closely the UI you built matches the source

Tools

| Tool | What it does | | ------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------- | | generate_skill(url, mode?, waitForCompletion?) | Generate a fresh skill.md from any public URL. Modes: design (default), api, library, generic. Polls to completion by default. | | get_skill(slug, includeScreenshots?) | Fetch a saved bundle: markdown, token contract (tokens.json + globals.css vars + Tailwind preset), and brand screenshots as images. | | get_skill_screenshot(slug, fold?) | Fetch one screenshot — above (hero / above-the-fold) or full (entire scrolled page). | | get_skill_asset(slug, kind) | Fetch a brand asset — favicon, og (Open Graph image), twitter (Twitter card), or logo. Drops the real mark into your UI. | | verify_design(slug, url) | Compare a publicly reachable preview URL against the skill's brand. Returns a fidelity score, mismatches, and concrete fixes. | | list_skills(cursor?, limit?) | List skills owned or bookmarked by the authenticated user. Paginated (1–50 per page, default 20). | | find_inspiration(cursor?, limit?) | Browse inspiration entries (Dribbble, Pinterest, Figma, image uploads, merges) saved by the user. |

Resources

| URI | Type | What you get | | ------------------------- | --------------- | -------------------------------------------------------------------------------------------------- | | skill://<slug> | text/markdown | The latest skill.md for a saved skill | | skillshot://<slug> | image/png | The brand screenshot for a saved skill, as an image you can pin alongside your context |

Clients with a resource picker (Claude Desktop, Claude Code, Cursor) can browse and pin these next to other context.

Authentication

Pick whichever fits — all three work:

  1. Browser sign-in (recommended). Run npx @getskillmd/cli login once. It opens your browser, you approve, and a token is cached at ~/.config/getskillmd/credentials.json. The npx @getskillmd/mcp server picks it up automatically — no API key, no env var.
  2. API key. Create one at https://getskillmd.com/account and set GETSKILLMD_API_KEY. Best for CI and headless setups.
  3. OAuth (hosted endpoint). Connect a client to https://getskillmd.com/api/mcp — clients that support OAuth (Claude Desktop, Cursor, …) open a browser to authorize. No key to copy or store.

Install

Claude Code / Claude Desktop

Add this to your MCP config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, or via claude mcp add in Claude Code).

After npx @getskillmd/cli login, no key is needed:

{
  "mcpServers": {
    "skillmd": {
      "command": "npx",
      "args": ["-y", "@getskillmd/mcp"]
    }
  }
}

Or pin an API key explicitly:

{
  "mcpServers": {
    "skillmd": {
      "command": "npx",
      "args": ["-y", "@getskillmd/mcp"],
      "env": {
        "GETSKILLMD_API_KEY": "gskmd_pk_..."
      }
    }
  }
}

Cursor / Windsurf / Antigravity

Same JSON snippet under the editor's MCP settings panel.

Any other MCP client

npx -y @getskillmd/mcp

Speaks streamable HTTP and stdio MCP transports.

Hosted endpoint

Skip the npx install and connect to https://getskillmd.com/api/mcp from any client that supports the streamable HTTP MCP transport.

Clients that support OAuth just need the URL — they'll open a browser to authorize:

{
  "mcpServers": {
    "skillmd": {
      "type": "http",
      "url": "https://getskillmd.com/api/mcp"
    }
  }
}

For clients without OAuth support, pass an API key (or a skillmd login token) instead:

{
  "mcpServers": {
    "skillmd": {
      "type": "http",
      "url": "https://getskillmd.com/api/mcp",
      "headers": {
        "Authorization": "Bearer gskmd_pk_..."
      }
    }
  }
}

Environment

| Var | Required | Purpose | | ---------------------- | -------- | ------------------------------------------------------------------------------------ | | GETSKILLMD_API_KEY | no* | API key from https://getskillmd.com/account | | GETSKILLMD_TOKEN | no* | A token to use directly instead of the cached skillmd login credentials | | GETSKILLMD_API_URL | no | Override the API base — defaults to https://getskillmd.com/api/v1 |

* Credentials are resolved in order: GETSKILLMD_API_KEYGETSKILLMD_TOKEN → the skillmd login browser session. Set at least one, or just run npx @getskillmd/cli login.

Programmatic use

The package also exports a factory so you can embed the server in your own MCP runtime:

import { buildGetSkillMdServer } from "@getskillmd/mcp";

const server = buildGetSkillMdServer({
  apiKey: process.env.GETSKILLMD_API_KEY!,
  // baseUrl: "https://getskillmd.com/api/v1",
  // name: "skillmd",
  // version: "0.1.0",
});

For multi-tenant setups where the API key depends on the caller, use the resolver form:

import { buildGetSkillMdServerFromResolver, createApiClient } from "@getskillmd/mcp";

const server = buildGetSkillMdServerFromResolver({
  apiResolver: ({ authInfo }) =>
    createApiClient({ apiKey: authInfo?.token ?? process.env.GETSKILLMD_API_KEY! }),
});

License

MIT