🔗 View on GitHub: anthropics/skills/mcp-builder
🚀 Quick Installation
Guide for creating high-quality MCP (Model Context Protocol) servers.
Claude Code Plugin
/plugin marketplace add anthropics/skills
/plugin install example-skills@anthropic-agent-skillsWhat It Does
Create MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use this skill when:
- 🔌 Building MCP servers to integrate external APIs or services
- 🛠️ Working with Python (FastMCP) or Node/TypeScript (MCP SDK)
- 📐 Designing tool schemas and APIs for LLM consumption
Four Key Principles
1. API Coverage vs. Workflow Tools
Balance comprehensive API endpoint coverage with specialized workflow tools. When uncertain, prioritize comprehensive API coverage.
2. Tool Naming and Discoverability
Use clear, descriptive tool names with consistent prefixes (e.g., github_create_issue, github_list_repos) and action-oriented naming.
3. Context Management
Design tools that return focused, relevant data with concise descriptions and pagination support.
4. Actionable Error Messages
Error messages should guide agents toward solutions with specific suggestions and next steps.
Recommended Stack
| Component | Recommendation |
|---|---|
| Language | TypeScript (high-quality SDK, good AI code generation) |
| Transport | Streamable HTTP for remote, stdio for local |
| Schema | Zod (TypeScript) or Pydantic (Python) |
Development Workflow
- Understand the API — Review service documentation, endpoints, auth
- Tool Selection — Prioritize comprehensive API coverage
- Project Setup — Use language-specific guides for structure
- Implementation — Create tools with proper schemas, error handling, pagination
- Testing — Use MCP Inspector:
npx @modelcontextprotocol/inspector - Evaluation — Create comprehensive evals to test effectiveness
Tool Annotations
readOnlyHint— true/falsedestructiveHint— true/falseidempotentHint— true/falseopenWorldHint— true/false