What problem does it solve? Building MCP (Model Context Protocol) servers that LLMs can actually use effectively requires careful tool design, schema definition, error handling, and evaluation — this Skill provides a structured four-phase workflow so your server exposes well-designed tools instead of raw API wrappers. ## Core Features & Use Cases - Four-phase workflow: Research and planning, implementation, review and testing, and evaluation, covering the full MCP server lifecycle. - Dual-language guidance: Supports Python (FastMCP) and Node/TypeScript (MCP SDK) with input schemas via Zod or Pydantic, output schemas, and tool annotations like readOnlyHint and destructiveHint. - Evaluation harness: Generates roughly 10 independent, read-only, multi-tool evaluation questions with verifiable answers in XML format. - Use Case: You want to expose your company's internal REST API to Claude. Use this Skill to design action-oriented tools with consistent naming, pagination, actionable errors, and then validate the server with the MCP Inspector. ## Quick Start Ask the AI to help you build an MCP server in TypeScript that wraps the GitHub API with well-named tools and structured output schemas.