mcp-builder

Define blueprints for building MCP servers with TypeScript and Python.

970|497|Updated Apr 3, 2024
One-click install
npx skills add https://github.com/Azure-Samples/AI-Gateway --skill mcp-builder-azure-samples
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-builder
Source: https://github.com/Azure-Samples/AI-Gateway/tree/main/.github/skills/mcp-builder
Command: npx skills add https://github.com/Azure-Samples/AI-Gateway --skill mcp-builder-azure-samples

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mcp, and includes scripts (resource) components.

What problem does it solve?

This Skill provides a comprehensive blueprint for designing, implementing, and validating MCP (Model Context Protocol) servers. It helps teams standardize tooling, schemas, transport, and security to accelerate development and ensure consistency across projects.

Core Features & Use Cases

  • Structured tool registration: guidance on naming, descriptions, input validation (Zod/Pydantic), and tool annotations to improve discoverability and safety.
  • SDK-driven patterns: recommended workflows for TypeScript and Python MCP servers, including common transport options (stdio and streamable HTTP), error handling, and pagination.
  • Quality and security practices: guidance on code quality, tests, error messages, and safe defaults to minimize risk.
  • Use Case: you’re building a library of MCP templates for internal teams to copy, adapt, and extend when creating new MCP servers.

Quick Start

Scaffold a minimal MCP server using the MCP SDKs, register a single read-only tool, and run under stdio to validate tooling and shape.

Frequently Asked Questions about mcp-builder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a scalable MCP server with standardized tooling?

Building scalable MCP servers requires standardizing tool registration, schemas, and transport. You can use SDK-driven patterns for TypeScript and Python to enforce consistent naming, structured outputs, and input validation with Zod or Pydantic.

What is the best way to register tools in an MCP server for discoverability?

The best way to register tools in an MCP server is to use structured registration patterns with clear descriptions, input validation, and tool annotations. This improves discoverability and safety for developers using your server templates.

Does the MCP SDK support both TypeScript and Python for server development?

Yes, the MCP SDK supports both TypeScript and Python for server development. It provides recommended workflows, common transport options like stdio and streamable HTTP, and specific validation libraries like Zod and Pydantic for each language.

How do I handle input validation and error handling in MCP servers?

Handle input validation in MCP servers by using Zod for TypeScript and Pydantic for Python to enforce schemas. The blueprint guides you on quality practices, safe defaults, and structured error messages to minimize risk and improve reusability.

Can I use streamable HTTP as a transport option for MCP servers?

Yes, you can use streamable HTTP as a transport option for MCP servers. The blueprint includes recommended workflows for TypeScript and Python that cover both stdio and streamable HTTP, alongside guidance on pagination and error handling.