Log inSubmit server
Skillmd logo

Skillmd

Searches, evaluates, lints, recommends, and safely installs SKILL.md Agent Skills through MCP.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Skillmd

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "skillmd": {
      "command": "npx",
      "args": [
        "-y",
        "skillmds"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

We couldn’t confirm it starts correctly

The install command started, but didn't respond the way we expected when we tried to talk to it.

npx -y skillmds

initialize succeeded but no response to tools/list.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked Today.

Overview

skillmd MCP server connects MCP-capable agents to the SkillMD registry of SKILL.md Agent Skills. It provides tools for registry search, skill inspection, linting, recommendations, trending results, saved-skill lookup, and validated installation. Install it locally with the skillmds npm package or connect to the hosted streamable HTTP endpoint. Reach for it when an agent needs to discover, assess, or add reusable skills during a conversation.

Use cases

  • Search the registry for a skill matching a development task
  • Inspect provenance and safety flags before adopting a skill
  • Install a validated skill into selected agent directories
  • Lint a local or registry SKILL.md during development

Key features

  • Registry search with ranked matches and install snippets
  • Skill retrieval with provenance, license, and capability flags
  • Validated installation without script execution
  • SKILL.md linting with diagnostics and quality scores
  • Trending, recommendations, and saved-skill lookup

Tools7

What Skillmd lets your agent do. Expand a tool for its parameters.

Verified live~1.1k tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 1d ago

Recent health check succeeded.

15/15 checks healthy over the last 38d

What it can do

From the 7 tools this server returned when we ran it; 7 declare their own safety hints.

Handles credentials · 1Makes network requests · 2Read-only tools · 7
Network at startup3 connections
Files written at startupNone
Install scriptsNone
npm provenanceSigned build
Package age116 days
Maintainers1

Observed 2026-10-24 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 15d ago via OSV.dev · skillmds (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars144Stargazers on the source repository.
npm downloads2.8kPackage downloads in the last 30 days.
Last commit2d agoMost recent push to the default branch.
Install checkInconclusiveStarted but did not complete the MCP handshake in our test.
Tools exposed7Callable tools this server registers over MCP.
Directory activity9 viewsConfig copies, upvotes, and views on AllMCPs.

How Skillmd works

What skillmd MCP server does

The skillmd MCP server exposes the SkillMD Agent Skills registry to MCP clients. Agent Skills are directories containing a SKILL.md file with YAML frontmatter and Markdown instructions that compatible agents load when needed.

The server covers the main registry workflow: find skills by task or keyword, inspect a specific skill, assess its format and capabilities, review popular or related entries, and install a selected skill. Registry records can include descriptions, categories, licenses, provenance, safety verdicts, capability flags, and installation snippets.

Installation creates a canonical copy under .agents/skills/<name> and links it into the selected agents’ skill directories. It validates the skill before writing it and does not execute scripts. Capability information such as network access, script execution, secret access, and documentation-only status is returned for review.

How it works

A local deployment runs through the skillmds npm package and communicates over MCP using standard input and output. The hosted deployment is available as a streamable HTTP MCP endpoint at https://api.skillmd.com/mcp.

Search accepts a task, name, or keyword and returns matching registry entries with the strongest match first. Retrieval uses an owner/name slug and returns the skill body along with source repository and pinned-commit provenance, license data, safety results, and capability flags.

Linting accepts either raw SKILL.md content or a registry slug. Results include diagnostics, a quality score from 0 to 100, and capability classifications. Recommendations can be based on a particular skill’s category or on skills trending during the previous 30 days. Trending results can be narrowed by category and time window.

Setup and configuration

Install the package globally with npm:

Terminal
npm i -g skillmds

A local MCP client can start the server with npx -y skillmds. For example, Claude Code can register it with:

Terminal
claude mcp add skillmd -- npx -y skillmds

Other MCP clients use an equivalent server entry with npx as the command and -y skillmds as its arguments. The hosted endpoint can be used instead when the client supports streamable HTTP.

The saved-skills tool requires a SkillMD personal access token in the SKILLMD_TOKEN environment variable. Other registry operations do not require an account according to the documented quick-start flow.

Tools and capabilities

  • skillmd_search finds registry skills by task, name, or keywords and returns descriptions, categories, safety verdicts, slugs, and install snippets.
  • skillmd_get retrieves a skill’s SKILL.md, provenance, license, safety verdict, and capability flags.
  • skillmd_install validates and installs a skill, writes its canonical copy, and links it into selected agent directories without running scripts.
  • skillmd_trending reports leaderboard entries ranked by installs, saves, and ratings, with optional category and time-window filters.
  • skillmd_recommend suggests related skills by category or returns recent trending skills.
  • skillmd_list_saved lists the authenticated user’s saved skills.
  • skillmd_lint validates supplied or registry-fetched SKILL.md content and reports diagnostics, capability flags, and a quality score.

Limitations and notes

The server manages Agent Skills rather than arbitrary agent plugins or application APIs. Installation writes files and links them into agent skill directories; it does not run scripts bundled with a skill. The capability flags are presented as information for the agent or developer, including docs_only, network_calls, executes_scripts, and reads_secrets.

The saved-skills operation depends on SKILLMD_TOKEN. A registry slug uses the owner/name format. The broader SkillMD project also supports CLI and GitHub Action workflows, but those are separate from the MCP tools described here.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Skillmd

skillmd is an MCP server that connects AI agents to the SkillMD registry of SKILL.md Agent Skills. Its main tools search and fetch skills, lint skill content, show trending and recommended entries, list saved skills, and validate and install skills into agent directories without executing scripts.

Install it through the `skillmds` npm package and configure your MCP client to run `npx -y skillmds`. Claude Code can use `claude mcp add skillmd -- npx -y skillmds`; clients that support streamable HTTP can instead connect to `https://api.skillmd.com/mcp`.

Yes, skillmd works with Cursor through its MCP server configuration. Add a server entry that runs `npx` with `-y skillmds`, or configure Cursor to use the hosted streamable HTTP endpoint.

No, skillmd validates a skill and writes or links its files without executing scripts. The installation response also reports capability flags, including whether the skill contains executable scripts, network calls, or secret access.

Set the `SKILLMD_TOKEN` environment variable to a SkillMD personal access token before using `skillmd_list_saved`. The token is required for saved-skill lookup, while the other documented registry operations can be used without an account.

Related MCP servers

All in Developer Tools All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewSkillmd AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/skillmd?style=directory)](https://allmcps.com/mcp/skillmd)
HTML Embed
<a href="https://allmcps.com/mcp/skillmd"><img src="https://allmcps.com/api/badge/skillmd?style=directory" alt="Skillmd on AllMCPs" /></a>