mcp-builder

Guides creation of MCP servers in TypeScript or Python with tool design, testing, and evaluation workflows.

1|Updated May 19, 2026
One-click install
npx skills add https://github.com/thomasmartinsen/talks --skill mcp-builder-thomasmartinsen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-builder
Source: https://github.com/thomasmartinsen/talks/tree/main/20260519%20-%20Agentic%20engineering/primitives/skills/mcp-builder
Command: npx skills add https://github.com/thomasmartinsen/talks --skill mcp-builder-thomasmartinsen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Building MCP (Model Context Protocol) servers that LLMs can actually use effectively requires careful tool design, protocol knowledge, and testing—this Skill provides a structured four-phase process to create high-quality MCP servers that integrate external APIs and services. ## Core Features & Use Cases - Four-Phase Workflow: Covers research and planning, implementation, review and testing, and evaluation creation for MCP server development. - Dual Language Support: Provides implementation guides for both TypeScript (MCP SDK with Zod schemas) and Python (FastMCP with Pydantic models). - Evaluation Framework: Generates 10 complex, verifiable evaluation questions in XML format to test whether LLMs can effectively use your server. - Use Case: You need to expose your company's internal REST API to LLM agents. Use this Skill to plan tool coverage, implement tools with proper schemas and annotations, test with MCP Inspector, and validate with realistic evaluation questions. ## Quick Start Help me build an MCP server in TypeScript that wraps the GitHub API with tools for issues and repositories.

Frequently Asked Questions about mcp-builder

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

FAQPage Schema
How do I build an MCP server in TypeScript?

Use the Model Context Protocol TypeScript SDK with Zod for input schemas and server.registerTool for tool registration. Set up streamable HTTP transport for remote servers or stdio for local servers, then test with MCP Inspector using npx @modelcontextprotocol/inspector.

How to create an MCP server in Python with FastMCP?

Use the Python SDK's FastMCP with Pydantic models for input validation and the @mcp.tool decorator for tool registration. Verify syntax with python -m py_compile and test the server using MCP Inspector.

TypeScript vs Python for MCP server development?

TypeScript is recommended for its high-quality SDK support, static typing, linting tools, and compatibility with execution environments like MCPB. Python with FastMCP is a solid alternative when your team or existing codebase is Python-based.

What transport should an MCP server use?

Use streamable HTTP with stateless JSON for remote servers since it scales and maintains more easily than stateful sessions. Use stdio transport for local servers that run on the same machine as the client.

How do I test whether my MCP server works well with LLMs?

Create 10 complex, read-only, verifiable evaluation questions in XML format with question and answer pairs. Each question should require multiple tool calls, have a single stable answer, and reflect realistic use cases.