Use the GetSkillary MCP server with your AI agent

GetSkillary runs a public Model Context Protocol (MCP) server that gives AI agents direct access to the GetSkillary Agent Skills Library. Instead of copying instructions from a web page, your agent can search the library, read a skill's details, fetch download links and get installation guidance as part of a conversation. This tutorial explains what the server provides, how to connect a client to it, and how to use it safely.

What the server provides

The server indexes the GetSkillary Agent Skills Library, a collection of 232 reusable agent skills. A skill is a packaged set of instructions, and sometimes scripts or reference files, that teaches an agent how to carry out a specific kind of task. The library also groups skills into solutions, which are sets of skills that work together for a broader goal.

The tools it exposes

When a client connects, the server advertises eight tools. Your agent decides when to call them based on your request, but knowing what each one does helps you ask better questions.

Skills

Solutions

Step 1: Check that your client supports remote MCP servers

Many AI clients now support MCP, including Claude, Cursor and Codex. Support for remote servers over Streamable HTTP varies by client, version and plan, so confirm in your client's documentation that it can connect to a remote MCP server by URL. Some clients only support local servers launched as a command; those will need a newer version or a bridging tool.

</div>
<p class="tool-card__desc">AI assistant from Anthropic for writing, analysis, research and software development.</p>
<div class="tool-card__foot">
  <span class="price-tag">Freemium</span>
  <a class="btn btn--secondary btn--sm" href="/go/claude?p=embed-use-getskillary-mcp" rel="nofollow noopener" target="_blank" data-out="claude">Visit website<svg class="i" width="14" height="14" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.8" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg></a>
</div>

Step 2: Add the server URL

Every client has its own place to manage MCP servers, usually in a settings area for connectors, integrations or tools, or in a configuration file. Menu names and file locations change between releases, so follow your client's current documentation for where to add a server. What you need to provide is the same everywhere:

• Name: something recognisable, such as getskillary
• URL: https://mcp.getskillary.com/mcp
• Authentication: none required for the public server

Clients that use a JSON configuration file commonly accept a format like the following. Treat it as the common shape rather than an exact specification, since some clients use different key names or require a transport type field:

{
  "mcpServers": {
    "getskillary": {
      "url": "https://mcp.getskillary.com/mcp"
    }
  }
}

After saving, restart or reload the client if it does not pick up the change automatically. Most clients show a list of connected servers and their tools; check that the eight GetSkillary tools appear.

</div>
<p class="tool-card__desc">AI-first code editor built on VS Code with agentic editing across your codebase.</p>
<div class="tool-card__foot">
  <span class="price-tag">Freemium</span>
  <a class="btn btn--secondary btn--sm" href="/go/cursor?p=embed-use-getskillary-mcp" rel="nofollow noopener" target="_blank" data-out="cursor">Visit website<svg class="i" width="14" height="14" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.8" stroke-linecap="round" stroke-linejoin="round" aria-hidden="true"><path d="M14 4h6v6M20 4l-9 9M18 14v5a1 1 0 0 1-1 1H5a1 1 0 0 1-1-1V7a1 1 0 0 1 1-1h5"/></svg></a>
</div>

Step 3: Try a few requests

Once connected, describe what you want in plain language. The agent chooses which tools to call. Some example prompts:

• "Search GetSkillary for skills that help write release notes from Git history."
• "Show me the details of the top result and explain what files it includes."
• "Find a GetSkillary solution for setting up a data analysis workflow and recommend which skills I need for CSV cleaning."
• "Give me the install plan for that solution in this project."

Many clients ask for confirmation before each tool call. That is a useful checkpoint: read what the agent intends to do before approving.

Step 4: Review skills before running them

A skill can include instructions that change how your agent behaves and, in some cases, scripts that run on your machine. Treat any downloaded skill the way you would treat code from a public repository:

1. Read the skill's main instruction file in full before installing it.
2. Open any bundled scripts and check what they execute, which files they touch and whether they make network requests.
3. Install into a project or test environment first, not into a workspace with sensitive data or credentials.
4. Keep your client's permission prompts enabled so that file writes and commands require approval.
5. Remove skills you no longer use.

The GetSkillary server provides discovery and installation guidance. It does not run skills on your behalf, and responsibility for what executes locally stays with you and your client's permission settings.

Troubleshooting

Summary

Adding https://mcp.getskillary.com/mcp as a remote MCP server gives your agent searchable access to the GetSkillary Agent Skills Library and its solutions. Setup takes a few minutes in any client that supports remote servers. Review each skill before it runs locally, and see the Claude and Cursor profiles for more on compatible clients.