Guides
Short, sourced guides to the questions people ask before and after choosing an MCP server or skill. Each is written from the MCP specification and the AI clients' own documentation.
The basics
- What is an MCP server?An MCP server is a program that gives AI apps like Claude, ChatGPT or Cursor tools and data from another system through the open Model Context Protocol.4 min read · 6 sources
- What is a Claude skill?A Claude skill is a folder with a SKILL.md file of instructions, plus optional scripts and reference files, that Claude loads only when a task needs it.5 min read · 7 sources
Compare
- MCP vs skills: how MCP servers and agent skills differMCP servers connect an AI agent to outside apps and data; skills teach it a procedure. How they differ in format, cost and risk, and how to combine them.5 min read · 7 sources
- MCP vs API: what's the difference?An API is how one service talks to programs; MCP is a shared protocol that lets AI apps find and use tools from many services in one way. See how they fit.5 min read · 6 sources
How to
- How to connect MCP servers to Claude, ChatGPT and CursorHow to add MCP servers to Claude Code, Claude Desktop, ChatGPT, Cursor, VS Code and Codex: config file locations, commands, and local vs remote servers.4 min read · 10 sources
- How to build an MCP serverBuild an MCP server step by step: pick an official SDK, define tools, resources and prompts, choose stdio or Streamable HTTP, and test with MCP Inspector.5 min read · 8 sources
- How to create custom skills for Claude CodeCreate a custom skill for Claude Code: write SKILL.md, add scripts and reference files, test it, and use the skill-creator skill to measure and refine it.5 min read · 6 sources