mcp-builder

Build an MCP server with custom tools and register it in VS Code.

7|2|Updated Dec 14, 2025
One-click install
npx skills add https://github.com/asafelobotomy/iCloud-Drive-Downloader --skill mcp-builder-asafelobotomy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-builder
Source: https://github.com/asafelobotomy/iCloud-Drive-Downloader/tree/main/.github/skills/mcp-builder
Command: npx skills add https://github.com/asafelobotomy/iCloud-Drive-Downloader --skill mcp-builder-asafelobotomy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Builds a Model Context Protocol (MCP) server end-to-end so you can expose custom tools and resources to an AI safely and consistently, without manual guesswork about the MCP lifecycle.

Core Features & Use Cases

  • Clarify server purpose: Converts a vague “need an integration” into a concrete set of tools, optional resources, and required credentials.
  • Choose the right transport: Selects between stdio (recommended), SSE, or Streamable HTTP based on where the server runs.
  • Scaffold, implement, and register tools: Generates a working server structure, implements tool input validation, and adds the server to .vscode/mcp.json for discovery in VS Code.
  • Test with MCP Inspector: Verifies server startup and tool behavior before you rely on it in production workflows.
  • Use case: Create an MCP server that provides tools like “searchTickets” and “getTicketDetails” backed by your internal Jira API, then register it so you can call those tools from your editor-driven agent.

Quick Start

Tell the AI to scaffold an MCP server named “my-integration” for stdio and include three tools with input schemas, then register it into .vscode/mcp.json and run MCP Inspector to verify the tools.

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 that exposes custom tools to an AI client?

To build an MCP server, you scaffold the server code, define validated tool input schemas and handlers, select a transport like stdio, and register the server in .vscode/mcp.json for AI client discovery.

What is the best way to choose an MCP transport for my server integration?

Choosing an MCP transport depends on your runtime environment: stdio is recommended for local integrations, while SSE or Streamable HTTP suit remote server integrations requiring validated tool behavior.

How do I register an MCP server in VS Code for editor discovery?

You register an MCP server in VS Code by adding the server configuration to the .vscode/mcp.json file, enabling the editor-driven agent to discover and call your custom tools directly.

Can I test MCP server tool behavior before relying on it in production?

Yes, you can test MCP server tool behavior using the MCP Inspector to verify server startup and validate tool execution before relying on the integration in production agent workflows.

Does scaffolding an MCP server include input validation for custom tools?

Yes, scaffolding an MCP server includes implementing tool input validation by defining structured tool schemas and handlers, ensuring deterministic tool behavior when an AI client invokes exposed resources.

When should I not use stdio transport for an MCP server?

You should not use stdio transport for an MCP server when deploying to remote environments, opting instead for SSE or Streamable HTTP transports to support remote AI client integrations.