This article clarifies the distinctions between MCP, skills, and agents in coding environments. It explains how skills function as reusable prompts for tasks, while MCP provides tools that can enhance functionality. The author critiques common misconceptions and highlights the practical benefits of each approach.

This article introduces the Skills Leaderboard, a platform for discovering and installing reusable capabilities for AI agents. Users can enhance their agents' functionalities with a simple command. The focus is on procedural knowledge that can be integrated easily.

The article discusses a study comparing two methods for teaching AI coding agents about Next.js: using skills and embedding documentation in an agents.md file. The results showed that the embedded documentation approach achieved a 100% pass rate, while the skill-based method struggled, highlighting the effectiveness of providing direct access to relevant information.

The Codex app transforms how developers interact with AI agents, enabling them to manage multiple tasks and collaborate effectively. It offers new skills that extend beyond code generation, allowing Codex to perform a variety of tasks on a computer. Developers can customize their interactions with Codex, choosing between a direct or conversational style.