mcp

Create MCP servers with FastMCP and Node/TypeScript SDKs.

2|Updated May 20, 2026
One-click install
npx skills add https://github.com/Dianshu-Liao/SkilLGuard --skill mcp-dianshu-liao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp
Source: https://github.com/Dianshu-Liao/SkilLGuard/tree/main/data/skills/mcp
Command: npx skills add https://github.com/Dianshu-Liao/SkilLGuard --skill mcp-dianshu-liao

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Building an MCP (Model Context Protocol) server often fails due to unclear tool design, weak schemas, missing error handling, and insufficient testing, which prevents agents from reliably using tools to complete real tasks.

Core Features & Use Cases

  • Tool-centric MCP design: Convert API capabilities into workflow-oriented tools that fit how agents actually execute tasks.
  • Protocol-aligned implementation guidance: Covers FastMCP (Python) and MCP SDK (Node/TypeScript), including schema design patterns and validation expectations.
  • LLM-friendly reliability: Emphasizes actionable, educational error messages; structured inputs/outputs; truncation and pagination considerations; and quality checklists.
  • Security and robustness practices: Focuses on best practices to reduce prompt-injection risk and improve safe integration habits.
  • Evaluation creation: Provides guidance for building read-only evaluation suites to verify tool usefulness and answer correctness.

Quick Start

Use the mcp skill to generate an MCP server plan for a new integration when a user asks you to create an MCP server or MCP tools for a specific external API.

Frequently Asked Questions about mcp

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

FAQPage Schema
How do I design MCP server tools that agents can reliably use for real workflows?

MCP server tools require workflow-oriented design, strict schema validation, and actionable error handling to ensure agents can reliably execute tasks. This involves converting API capabilities into structured interfaces that fit how agents operate.

What is the best way to validate inputs and handle errors when building an MCP server?

Input validation and error handling in an MCP server require strict schema definitions and actionable, educational error messages. This approach ensures agents receive clear feedback to self-correct during tool execution.

Does this MCP server design approach work with both Python FastMCP and Node TypeScript SDK?

MCP server design applies to both Python FastMCP and Node/TypeScript MCP SDK development scenarios. It provides protocol-aligned implementation guidance including schema design patterns and validation expectations for both environments.

How do I reduce prompt injection risk when integrating external APIs into MCP tools?

Reducing prompt injection risk in MCP tools involves following security best practices and robustness guidelines during tool design. This includes structured inputs and outputs to constrain agent interactions safely.

How do I create an evaluation suite to test MCP tool usefulness and answer correctness?

Creating an MCP evaluation suite involves building read-only tests to verify tool usefulness and answer correctness. This provides guidance for planning evaluations that validate agent task completion.

When should I implement pagination or truncation strategies in MCP server tools?

Pagination and truncation strategies in MCP server tools are needed when handling large data outputs from external APIs. These considerations prevent overwhelming the model context window during agent execution.