mcp-builder

Guide MCP server design in TypeScript or Python with tool schemas and transports.

190|27|Updated Apr 22, 2026
One-click install
npx skills add https://github.com/linxuan-sys/opencode-skills-chinese --skill mcp-builder-linxuan-sys
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-builder
Source: https://github.com/linxuan-sys/opencode-skills-chinese/tree/main/mcp-builder
Command: npx skills add https://github.com/linxuan-sys/opencode-skills-chinese --skill mcp-builder-linxuan-sys

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill helps you design and implement high-quality MCP (Model Context Protocol) servers so an LLM can reliably complete real-world tasks by calling well-defined tools that integrate with external APIs and services.

Core Features & Use Cases

  • End-to-end MCP server workflow: Plan, implement, review, test, and evaluate MCP servers using a structured, stage-based approach.
  • Tool and schema design guidance: Create input/output schemas (Zod for TypeScript, Pydantic for Python), craft precise tool descriptions, and add actionable error messages.
  • Transport-aware implementation strategy: Select the right transport (stdio for local, streamable HTTP for remote) and follow practical best practices for naming, pagination, and response formats.
  • Evaluation plan templates: Produce stable, read-only QA evaluations to validate that the LLM can solve complex questions using only your MCP tools.

Quick Start

Ask an AI to create an MCP server that integrates your external service by listing the MCP specification essentials, proposing tool interfaces with validated schemas, and drafting an evaluation XML with 10 stable read-only questions for quality testing.

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 with schema validation in TypeScript or Python?

To build an MCP server, you define tool interfaces using Zod for TypeScript or Pydantic for Python, ensuring strict schema validation. This process includes crafting precise tool descriptions, adding actionable error messages, and selecting the appropriate transport layer for communication.

When should I use stdio versus streamable HTTP transport for an MCP server?

Use stdio transport for local MCP server deployments and streamable HTTP for remote access. Transport selection depends on your deployment environment, with both options supporting structured tool use, pagination, and standardized response formats for LLM interactions.

What is the best way to evaluate multi-step tool usage in MCP servers?

The best way to evaluate multi-step tool usage is by creating stable, read-only XML evaluations. This involves drafting specific QA questions to validate that the LLM can successfully solve complex problems using only the defined MCP tools.

Does this MCP server design approach support pagination and error handling?

Yes, this MCP server design approach explicitly supports pagination and error handling. It guides you in adding actionable error messages and following practical best practices for naming, pagination, and response formats to ensure reliable external API integration.

How do I define tool interfaces for an LLM to complete real-world tasks?

You define tool interfaces by creating input and output schemas with Zod or Pydantic, ensuring the LLM can reliably complete real-world tasks. This requires following MCP best practices for tool design, including precise descriptions and validated schemas.