MCP Server — AI Assistant Docs

Every page in this Documentation section is also available to your AI coding assistant directly — Claude Code, Cursor, VS Code with GitHub Copilot Chat, Claude Desktop, Windsurf, or any other tool that speaks the Model Context Protocol (MCP). Point it at one URL and it can look up setup, configuration, theming, and licensing details itself instead of you copy-pasting docs into chat, or it guessing (and sometimes getting it wrong).

What is MCP, in plain terms?

MCP (Model Context Protocol) is an open standard that lets an AI assistant call "tools" exposed by an external server, the same way it already calls tools like a file editor or a terminal. A tool is just a function with a name, a description, and typed inputs/outputs — the assistant decides when to call it based on what you ask. This server exposes three tools backed by the real, live content of this Documentation section (see Tools exposed below): one to list every doc page, one to fetch a specific page as markdown, and one to search across all of them. Nothing is installed on your machine — it's a remote server your assistant talks to over HTTP, the same way a browser talks to a website.

Quick start

The server lives at one URL. Every setup below is really just "tell your assistant about this URL":

Server URL: https://guidemyhair.com/mcp

Connect your assistant

Claude Code

Create a .mcp.json file in your project root (this is checked into git, so the whole team gets it automatically when they open the repo):

{
  "mcpServers": {
    "guidemyhair-docs": {
      "url": "https://guidemyhair.com/mcp"
    }
  }
}

Or add it without hand-editing JSON:

claude mcp add --transport http guidemyhair-docs https://guidemyhair.com/mcp

Cursor

Settings → Cursor Settings → MCP → "Add new MCP server", or create .cursor/mcp.json directly (same shape as Claude Code's config):

{
  "mcpServers": {
    "guidemyhair-docs": {
      "url": "https://guidemyhair.com/mcp"
    }
  }
}

VS Code (GitHub Copilot Chat)

Command Palette → "MCP: Add Server" → HTTP, or create .vscode/mcp.json directly. Note the shape is slightly different from the others — servers instead of mcpServers, and an explicit type field:

{
  "servers": {
    "guidemyhair-docs": {
      "url": "https://guidemyhair.com/mcp",
      "type": "http"
    }
  }
}

Claude Desktop

Settings → Connectors → "Add custom connector" → paste the server URL. This is the supported path for a remote HTTP server like this one in the desktop app — no JSON file to edit.

Windsurf

Windsurf Settings → Cascade → MCP Servers → "Add custom server", or edit ~/.codeium/windsurf/mcp_config.json directly (same mcpServers shape as Claude Code/Cursor).

{
  "mcpServers": {
    "guidemyhair-docs": {
      "url": "https://guidemyhair.com/mcp"
    }
  }
}

Any other MCP client

Any client that supports the Streamable HTTP transport just needs the URL above — there's nothing GuideMyHair-specific about the connection itself, and no authentication is required (it's read-only, public documentation).

Tools exposed

Example prompts

Once connected, you don't need to mention the tools by name — just ask your assistant normally and it'll reach for them when relevant:

• "How do I set up the granular shade catalog for TreatmentConsultant?"
• "What CSS custom properties can I override to theme the widget?"
• "Why is my widget showing a watermark — what license reasons trigger that?"
• "Search the GuideMyHair docs for anything about the header slot"

Troubleshooting

My assistant doesn't seem to be using it

Most tools list connected MCP servers somewhere in their settings/UI — confirm guidemyhair-docs (or whatever you named it) shows as connected, and that its tools (list_docs, get_doc, search_docs) are listed. Some assistants only enable newly added servers after a restart.

I'm getting a "Method not allowed" or connection error

Double-check the URL has no typo and doesn't have a trailing slash or extra path segment — it should be exactly https://guidemyhair.com/mcp.