Claude Code just Built me an AI Agent Team (Claude Code + Skills + MCP)

章节 1：Claude Code 核心价值与团队架构理念

📝 本节摘要：

本节主要阐述了 Claude Code 相比传统 Claude Projects 的核心优势。作者指出，Claude Code 打破了以往 AI 代理之间“孤岛式”运作的局限，允许代理共享工作区与上下文，使 Claude 能够作为“团队领导”自动协调任务。此外，本节还介绍了构建 AI 代理团队的基础架构，即每个代理需通过单一 Markdown 文件定义其角色、知识库及工具技能（MCP），并建议为代理分配互不重叠的职责以减少冲突。

[原文] [Speaker]: I use Claude more than any other ai, but even I dismiss Claudee as just for developers and I was wrong after using myself.

[译文] [Speaker]: 我使用 Claude 的频率比任何其他 AI 都高，但即便是我，也曾认为 Claude 只是给开发者用的，但在亲自使用后我发现自己错了。

[原文] [Speaker]: I am completely sold, and this is too powerful to ignore if you want to use AI to build system and do work for you end to end, not just chatting and copy pasting.

[译文] [Speaker]: 我完全被折服了，如果你想利用 AI 构建系统并为你完成端到端的工作，而不仅仅是聊天和复制粘贴，那么它的强大功能是不容忽视的。

[原文] [Speaker]: So in this video I'll share how to get started with Claude code, even if you don't use it for coding and most importantly, how to use it to build an AI work team and system that you can actually use.

[译文] [Speaker]: 因此，在这个视频中，我将分享如何开始使用 Claude Code，即使你不将其用于编程，最重要的是，如何利用它来构建一个你真正能投入使用的 AI 工作团队和系统。

[原文] [Speaker]: Let's go.

[译文] [Speaker]: 我们开始吧。

[原文] [Speaker]: So why Claude Code is even needed if Claude is already so powerful.

[译文] [Speaker]: 既然 Claude 已经如此强大，为什么还需要 Claude Code 呢？

[原文] [Speaker]: So most people use Claude projects with custom instructions to turn Claude into AI agents.

[译文] [Speaker]: 大多数人使用带有自定义指令的 Claude Projects 将 Claude 转化为 AI 代理。

[原文] [Speaker]: But they are siloed.

[译文] [Speaker]: 但它们是孤立的。

[原文] [Speaker]: You end up copy and pasting between them.

[译文] [Speaker]: 你最终不得不在它们之间复制粘贴内容。

[原文] [Speaker]: You are the coordinator.

[译文] [Speaker]: 你充当了协调员的角色。

[原文] [Speaker]: But with Claude code it's different.

[译文] [Speaker]: 但有了 Claude Code，情况就不同了。

[原文] [Speaker]: So these Claude agents share the same workspace, access to the same context file.

[译文] [Speaker]: 这些 Claude 代理共享同一个工作区，访问同一个上下文文件。

[原文] [Speaker]: They can talk to each other and hands off work automatically each with their own token context window.

[译文] [Speaker]: 它们可以互相交流并自动交接工作，且每个代理都有自己的 token 上下文窗口。

[原文] [Speaker]: And Claude now becomes the team lead.

[译文] [Speaker]: 而 Claude 现在变成了团队领导。

[原文] [Speaker]: You give the direction they execute together and that is what we are building today.

[译文] [Speaker]: 你给出方向，它们协同执行，这就是我们今天要构建的内容。

[原文] [Speaker]: So each of these Claude Code agents you create needs at least three core components all defined in one single markdown file.

[译文] [Speaker]: 因此，你创建的每个 Claude Code 代理都需要至少三个核心组件，所有这些都定义在一个 markdown 文件中。

[原文] [Speaker]: Their role and responsibilities, their knowledge like workflow details, reusable Claude agent skills, the MCP tools that they have access to.

[译文] [Speaker]: 它们的角色和职责，它们的知识（如工作流细节），可复用的 Claude 代理技能，以及它们可以访问的 MCP 工具。

[原文] [Speaker]: For this demo, we'll be building a five agent work team.

[译文] [Speaker]: 对于这个演示，我们将构建一个由五名代理组成的工作团队。

[原文] [Speaker]: And Claude, our main agent, will act as our team lead to coordinate between these agent.

[译文] [Speaker]: 我们的主要代理 Claude 将作为团队领导来协调这些代理。

[原文] [Speaker]: A bonus tip is when designing your AI teammate, give them specific and non-overlapping roles to minimize the potential conflicts.

[译文] [Speaker]: 一个额外的建议是，在设计你的 AI 队友时，给它们分配具体且不重叠的角色，以最大限度地减少潜在的冲突。

https://www.youtube.com/watch?v=oTa2RccMkJU