Claude Skills MCP Connector
A remote MCP server that exposes a library of Claude-style SKILL.md packages, enabling Claude to search, retrieve skills, and read supporting files via MCP tools.
README
Claude Skills MCP Connector
A remote MCP server for Claude that exposes a local library of Claude-style SKILL.md
packages stored in this repository.
Included skill sources
This project was assembled from the three uploaded ZIP files:
taste-skill-main.zipskills-main.zipimpeccable-main.zip
The actual skill folders are under skills/. Supporting files inside each skill are
preserved so a model can load references/templates/recipes on demand.
MCP tools
list_skills
Returns all discovered skills, their descriptions, and their available files.
search_skills
Searches skill names/descriptions.
Example:
{"query":"portfolio design"}
get_skill
Loads the full SKILL.md for one skill.
Example:
{"name":"impeccable"}
read_skill_file
Loads a supporting file inside a skill without allowing path traversal.
Example:
{"name":"impeccable","file":"reference/layout.md"}
Deploy to Vercel
Requirements:
- Node.js 20+
- A Vercel account
- A GitHub repository
Install:
npm install
Check TypeScript:
npm run typecheck
Run locally:
npx vercel dev
The MCP endpoint is:
http://localhost:3000/mcp
After deployment it will be:
https://YOUR-PROJECT.vercel.app/mcp
Connect to Claude
Add the deployed /mcp endpoint as a remote MCP/custom connector in the Claude
interface available to your account.
How Claude should use it
Recommended flow:
search_skillsfor the task.get_skillfor the best matching skill.read_skill_fileonly when the skill references additional files.- Apply the returned instructions to the user's task.
Important
This connector does not modify Claude's built-in Skill system. It exposes the uploaded skill packages as MCP-accessible content. Claude reads the returned instructions and uses them in the conversation.
Updating skills
Replace/add folders under skills/, commit, and push to GitHub. Vercel will deploy
the new version automatically if Git integration is enabled.
Security
The connector is read-only. It does not execute scripts from the skill packages.
read_skill_file is restricted to paths inside the selected skill directory.
Vercel configuration
This project intentionally does not set a functions.runtime value in vercel.json.
Vercel automatically detects the Node.js runtime for api/mcp.ts.
The Node.js major is pinned to 24.x in package.json. Vercel supports Node.js 22+,
and Node.js 20 is scheduled for deprecation on October 1, 2026.
If an older deployment is cached, redeploy after committing the updated
vercel.json and package.json.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.