Microsoft Learn MCP

Connect to the Microsoft Learn MCP (Model Context Protocol) Server to interact with Microsoft documentation through structured tools.

Endpoint

https://learn.microsoft.com/api/mcp

This is a remote MCP server using streamable HTTP. It provides three main tools:

• microsoft_docs_search — Search Microsoft documentation
• microsoft_docs_fetch — Fetch specific documentation pages
• microsoft_code_sample_search — Search for official code samples

Setup

Add to mcporter config

mcporter config add --name mslearn --url https://learn.microsoft.com/api/mcp --type http

Or manually add to ~/.config/mcporter/config.json:

{
  "servers": {
    "mslearn": {
      "type": "http",
      "url": "https://learn.microsoft.com/api/mcp"
    }
  }
}

Verify connection

mcporter list mslearn --schema

Usage

Search documentation

mcporter call mslearn.microsoft_docs_search query="Azure Functions triggers"

Fetch a specific doc page

mcporter call mslearn.microsoft_docs_fetch url="https://learn.microsoft.com/en-us/azure/azure-functions/functions-triggers-bindings"

Search code samples

mcporter call mslearn.microsoft_code_sample_search query="Python blob storage" language="python"

Tool Reference

Tool schemas are dynamic. Always check current schema with:

mcporter list mslearn --schema

Common patterns:

Tool	Typical parameters
microsoft_docs_search	query (required), locale (optional, e.g., "en-us")
microsoft_docs_fetch	url (required, full Learn URL)
microsoft_code_sample_search	query (required), language (optional), product (optional)

Output formats

Default output is human-readable. Use --output json for structured data:

mcporter call mslearn.microsoft_docs_search query="Entra ID" --output json

Notes

• No authentication required for the Learn MCP Server
• The server interface may change dynamically; always call list to get current tools
• For complex queries, prefer search over fetch; let the tool find relevant pages
• Locale defaults to en-us if not specified