GuideIntermediate

Claude Code & MCP: Building Custom Integrations

Extend Claude Code's capabilities by building your own MCP servers. The Model Context Protocol is the "USB-C of AI": a standard protocol that turns any API, database, or service into a tool Claude Code can use natively. In this guide you build MCP servers with the official TypeScript and Python SDKs (`MCPServer` from the Python SDK 2.x), connect them over stdio and Streamable HTTP, create MCP Apps with interactive UI, and finish with a production-ready server connected to real data. It covers the two startup forms that coexist today: the classic handshake and the stateless 2026-07-28 revision. SDK and protocol data as of October 2026.

lessons
42
modules
8
available in
English · Spanish
certificate
Yes
access
Free
Outcomes

What you'll be able to do

  • What MCP is and why it's the "USB-C of AI": the standard protocol for integrations
  • The Host-Client-Server architecture and how data flows between Claude Code and your server
  • The 3 primitives: Resources (data), Tools (functions), Prompts (templates)
  • Build MCP servers in TypeScript with `@modelcontextprotocol/sdk` 1.x and `registerTool`, and move them to v2 (`@modelcontextprotocol/server`)
  • Build MCP servers in Python with the official `mcp` 2.x SDK and `MCPServer`
  • Choose a transport: stdio for local, Streamable HTTP for remote
  • MCP Apps: tools that return interactive UI (dashboards, forms)
  • Testing and debugging with the MCP Inspector and test suites
  • Configure MCP servers in Claude Code: local, project (`.mcp.json`), and user scopes, variables, and permissions
  • Project: a complete MCP server connected to a real API/database
Before you start

What you need to bring

Requirements and materials

  • Having completed [Claude Code Foundations](https://nieva.team/en/courses/claude-code-foundations-guide) (using Claude Code and existing MCP servers)
  • Recommended: [Testing with Claude Code](https://nieva.team/en/courses/testing-with-claude-code-guide), because modules 5 and 7 use pytest
  • Basic experience with TypeScript or Python
  • Familiarity with REST APIs
  • Comfortable with the terminal/CLI
  • Claude Code installed and configured
Content

The syllabus, module by module

Open any of them to see its lessons.

Common questions

What people usually ask

Start whenever you like

Reviews

What students say

These reviews are from enrolled students who completed at least 50% of the course. We moderate reviews only on content grounds (spam, offensive language, personal data), never for being critical or negative.

No approved reviews yet.

Be the first to share your experience!