The new rules of context engineering for Claude 5 models Claude 5 模型上下文工程的新规则
I’ve written previously about how to best and work with them iteratively to discover what you want to build.我之前写过关于如何为最新一代 Claude 5 模型编写最佳提示词，以及如何与它们迭代协作以探索你想要构建的内容。
But when you send a message to Claude, the prompt is only a small part of the context it gets. Much of your context is assembled from your system prompt, Skills, CLAUDE.md files, memory, and other sources. We call this , and it makes a big impact on the results you generate when using Claude Code or in building your own agents.但当你向 Claude 发送消息时，提示词只是它所获取上下文的一小部分。你的大部分上下文是由系统提示词、技能 (Skills)、CLAUDE.md 文件、记忆以及其他来源组成的。我们称之为上下文工程，它对你使用 Claude Code 或构建自己的智能体时生成的输出结果有着巨大的影响。
Unlike a prompt, context is used generally across many requests, so it cannot be as specific. How do you build these general prompts and guidance for Claude, especially when you don’t know what a user’s prompt might be?与提示词不同，上下文通常在多次请求中通用，因此它无法做到那么具体。那么，你该如何为 Claude 构建这些通用的提示词和指南，尤其是在你无法预知用户会输入什么提示词的情况下呢？
This can be surprisingly difficult as Claude’s own capabilities evolve. Most recently, we noticed a large jump in the way we prompt the newest generation of Claude models. We removed over 80% of Claude Code’s system prompt for models like Claude Opus 5 and Claude Fable 5 with no measurable loss on our coding evaluations.随着 Claude 自身能力的演进，这可能会变得出人意料地困难。最近，我们注意到我们在提示最新一代 Claude 模型的方式上有了巨大的跨越。我们删除了 Claude Code 针对 Claude Opus 5 和 Claude Fable 5 等模型系统提示词中超过 80% 的内容，且在我们的编码评估中没有出现可感知的性能损失。
Here’s what we’ve learned about prompting this new class of models, and how you can utilize it to update your context engineering. We’ve put these best practices in `claude doctor`, use the command /doctor in Claude Code to rightsize your skills, and CLAUDE.md files.以下是我们对这一新类模型提示方式的见解，以及你可以如何利用它来更新你的上下文工程。我们已将这些最佳实践放入 `claude doctor` 中，请在 Claude Code 中使用 /doctor 命令来优化你的技能和 CLAUDE.md 文件。
Unhobbling Claude为 Claude “松绑”
Overall, we found that we were over-constraining Claude Code, both through our system prompt and in our CLAUDE.md files and skills.总的来说，我们发现我们过去通过系统提示词、CLAUDE.md 文件和技能对 Claude Code 的限制太多了。
For example, when we read transcripts of our own internal usage of Claude Code, we see several conflicting messages in a single request like “leave documentation as appropriate,” or “DO NOT add comments” as our system prompt, skills, and user requests clash with each other.例如，当我们阅读内部使用 Claude Code 的记录时，会发现在同一个请求中存在多条相互冲突的信息，比如系统提示词要求“根据需要保留文档”，而技能和用户请求却要求“不要添加注释”，这些指令彼此冲突。
Generally, Claude can interpret the user’s intent to get to the right answer, but Claude must think more carefully about these overlapping and conflicting messages before deciding what to do.通常情况下，Claude 可以解读用户的意图并得出正确答案，但 Claude 必须在决定怎么做之前，更仔细地权衡这些重叠且冲突的信息。
And while these constraints were once needed to avoid worst case scenarios, we have since found we can delete many of them and let the model use surrounding context and judgement instead.虽然这些约束曾经是为了避免最坏的情况而必需的，但我们现在发现可以删除其中许多约束，让模型利用周围的上下文和判断力来处理。
Additionally, Claude Code now has many more tools. Claude used to rely on CLAUDE.md as a source of memory, information, and guidance. Now we have memory, artifacts, and skills, which Claude can use to create new ways of loading and sharing context across sessions.此外，现在的 Claude Code 拥有了更多工具。过去，Claude 依靠 CLAUDE.md 作为记忆、信息和指导的来源。现在我们有了记忆、工件 (Artifacts) 和技能，Claude 可以利用它们创造出在会话间加载和共享上下文的新方式。
Then and now过去与现在
There were a number of previous context engineering best practices that had become myths. Including:此前有许多上下文工程的最佳实践已经变成了误区，包括：
Then: Give Claude rules
Now: Let Claude use judgement
When we first rolled out Claude Code, we needed to be sure that Claude avoided worst case scenarios, such as deleting files. This meant we would give particularly strong guidance that might not always be true, For example, in the system prompt we used to say:过去：给 Claude 制定规则
现在：让 Claude 运用判断力
当我们首次推出 Claude Code 时，我们需要确保 Claude 避免最坏的情况，例如删除文件。这意味着我们会给出特别强硬的指导，但这些指导并不总是适用。例如，在系统提示词中，我们曾写道：
In code: default to writing no comments. Never write multi-paragraph docstrings or multi-line comment blocks — one short line max. Don't create planning, decision, or analysis documents unless the user asks for them — work from conversation context, not intermediate files.在代码中：默认不写注释。永远不要写多段式的文档字符串或多行注释块——最多写一行简短的。除非用户要求，否则不要创建规划、决策或分析文档——基于对话上下文工作，而不是基于中间文件。
But for a certain subset of prompts, this guidance would be wrong. In the case of documentation, the user may have their own preferences, or specific parts of very complex code might need multi-line comment blocks.但对于某些特定的提示词，这种指导是错误的。在文档编写方面，用户可能有自己的偏好，或者非常复杂的代码的特定部分可能确实需要多行注释块。
Still, without these guardrails for older models, the comments Claude wrote would be incorrect in many cases and we had to accept this tradeoff. But newer models have better judgement and can handle these decisions well without explicit rules.尽管如此，在旧模型中，如果没有这些护栏，Claude 编写的注释在许多情况下会出错，我们不得不接受这种权衡。但较新的模型拥有更好的判断力，无需明确的规则也能很好地处理这些决策。
In the new system prompt we say: Write code that reads like the surrounding code: match its comment density, naming, and idiom.在新的系统提示词中，我们写道：编写的代码应与周围代码风格一致：匹配其注释密度、命名习惯和惯用法。
Then: Give Claude examples
Now: Design interfaces
The number one rule for tool usage was to give Claude examples on how to use them. With our newest models, we’ve found that giving examples actually constrains them to a certain exploration space.过去：给 Claude 提供示例
现在：设计接口
工具使用的第一准则曾是为 Claude 提供使用示例。但对于我们最新的模型，我们发现提供示例反而会将它们限制在特定的探索空间内。
Instead of using examples, think more about the design of your tools, scripts and files- what parameters does Claude have and how can they be more expressive?与其使用示例，不如多考虑工具、脚本和文件的设计——Claude 有哪些参数？它们如何才能更具表现力？
For example, in the Todo tool example, just listing status as an enumeration between pending, in_progress, and completed, hints to Claude about how to use it. The instruction on keeping one item in_progress helps define our requested behavior.例如，在 Todo 工具示例中，仅将状态列为 pending（待办）、in_progress（进行中）和 completed（已完成）的枚举，就能向 Claude 暗示如何使用它。关于保持一项处于 in_progress 的指令有助于定义我们所要求的行为。
Then: Put it all upfront
Now: Use progressive disclosure过去：把所有内容都放在前面
现在：使用渐进式披露
Because Claude Code was focused on coding, our system prompt included detailed information on how to do code review and verification. These were not always needed, but when they were, it was crucial information.由于 Claude Code 专注于编码，我们的系统提示词包含了关于如何进行代码审查和验证的详细信息。这些信息并不总是必需的，但在需要时，它们至关重要。
Since then, Claude Code has gotten very competent at using progressive disclosure- loading the right context at the right times. For example, we moved verification and code review into their own skills that Claude Code could selectively call.从那时起，Claude Code 在使用渐进式披露方面变得非常熟练——即在正确的时间加载正确的上下文。例如，我们将验证和代码审查功能移到了它们各自的技能中，Claude Code 可以有选择地调用它们。
But progressive disclosure is not just for skills, we also use it for tools. Some of our tools are ‘deferred loading,’ which means the agent must search for their full definitions using ToolSearch before using them. This allows us to have more tools (such as our Task tools) that don’t take up context until they’re needed.但渐进式披露不仅适用于技能，我们也将其用于工具。我们的一些工具是“延迟加载”的，这意味着智能体必须在使用前使用 ToolSearch 搜索它们的完整定义。这使我们能够拥有更多的工具（例如我们的任务工具），而在需要之前它们不会占用上下文。
The same can be applied to your own CLAUDE.md and Skill.md files. A common myth is that you want to make these a central repository for every known practice that you might run into, because Claude would not find it otherwise. Instead, .同样的方法也适用于你自己的 CLAUDE.md 和 Skill.md 文件。一个常见的误区是，你认为应该将这些文件作为所有已知实践的中央存储库，否则 Claude 就找不到它们。相反，请考虑构建一个可以在适当时机加载的文件树。
Then: Repeat yourself
Now: Simple tool descriptions过去：重复自己
现在：简单的工具描述
Earlier Claude models could sometimes need repeated instructions or be more likely to listen to instructions at the end of their context window than at the start. This meant our system prompt would sometimes have references to tools in the main system prompt as well as instructions in the tool description.早期的 Claude 模型有时需要重复的指令，或者相比上下文窗口的开头，它们更容易听从结尾处的指令。这意味着我们的系统提示词有时会在主系统提示词中引用工具，同时又在工具描述中包含指令。
We found we could delete these repeat examples and put instructions on how to use tools in the tool descriptions rather than the system prompt.我们发现可以删除这些重复的示例，并将如何使用工具的指令放在工具描述中，而不是系统提示词中。
Then: Memory in CLAUDE.md files
Now: Auto-memory过去：CLAUDE.md 文件中的记忆
现在：自动记忆
We used to encourage users to save things to Claude’s memory, by using the # hotkey to write to their automatically. Instead, Claude now automatically saves memories that are relevant to the work and to you.我们过去鼓励用户通过使用 # 热键自动写入 CLAUDE.md 来将内容保存到 Claude 的记忆中。现在，Claude 会自动保存与工作和你相关的内容。
Then: Simple specs
Now: Rich references过去：简单的规范
现在：丰富的引用
In plan mode, Claude Code has heavily relied on markdown files with plans. Storing these files as plans helped Claude refer to them when needed. Another similar best practice was to store specs in the codebase for Claude to refer to while working across longer projects.在规划模式下，Claude Code 严重依赖带有计划的 Markdown 文件。将这些文件存储为计划有助于 Claude 在需要时进行参考。另一个类似的最佳实践是将规范存储在代码库中，以便 Claude 在进行长期项目时参考。
But we’ve found that Claude can handle increasingly more complicated references. Instead of simple markdown files, Claude can reference HTML artifacts created by our new artifacts feature.但我们发现 Claude 可以处理越来越复杂的引用。除了简单的 Markdown 文件，Claude 还可以引用我们新工件功能创建的 HTML 工件。
You may also give Claude references in the form of code. A spec may also be a detailed test suite, or a function in a different codebase that Claude might port.你也可以以代码的形式为 Claude 提供引用。规范也可以是一个详细的测试套件，或者是一个 Claude 可能需要移植的其他代码库中的函数。
Rubrics are another form of references. Rubrics allow Claude to try and verify your taste in a particular field (e.g. what does a good API design look like) by using and spinning up verifier agents with those rubrics.准则 (Rubrics) 是另一种形式的引用。准则允许 Claude 通过使用动态工作流并启动带有这些准则的验证智能体，来尝试验证你在特定领域品味（例如，好的 API 设计是什么样的）。
Applying this to your context将其应用到你的上下文中
Pulling this all together, what does this look like when you assemble your context?综合以上所述，当你组装上下文时，它看起来应该是怎样的？
System Prompt
A system prompt is heavily tied to the product context. It tells Claude what product it’s operating in and what it’s doing. For Claude Code, you will likely never modify this, but if you are building your own agent harness, this is where you should spend a lot of time.系统提示词
系统提示词与产品上下文紧密相关。它告诉 Claude 它在什么产品中运行以及正在做什么。对于 Claude Code，你可能永远不需要修改它，但如果你正在构建自己的智能体框架，这就是你应该投入大量精力的地方。
CLAUDE.md
Keep your CLAUDE.md lightweight and briefly describe what your repo is for, but spend most of the tokens on gotchas inside of the codebase. For example, you may organize your code to keep types in one monolithic file and nowhere else. Avoid stating ‘the obvious’ things Claude should know by looking at your file system or your repo.CLAUDE.md
保持你的 CLAUDE.md 轻量化，简要描述你的仓库用途即可，但要把大部分 Token 用在代码库内部的“坑”上。例如，你可以组织代码将类型保留在一个单体文件中。避免陈述 Claude 通过查看你的文件系统或仓库就能知道的“显而易见”的事情。
Use progressive disclosure for more details, for example if you have several unique instructions on how to verify your work, create a verification skill and reference it from your CLAUDE.md.对于更多细节，请使用渐进式披露。例如，如果你有关于如何验证工作的多项独特指令，请创建一个验证技能，并在 CLAUDE.md 中引用它。
Skills
Think of skills as lightweight guides to let Claude find information when needed. Avoid making them overconstrained, except in highly important areas.技能 (Skills)
将技能视为轻量级的指南，让 Claude 在需要时找到信息。除非在非常重要的领域，否则避免过度限制它们。
For long skills, try and use progressive disclosure as much as possible- divide it into many files and split them out.对于较长的技能，请尽量使用渐进式披露——将其拆分为多个文件并分离开来。
It’s best when skills encode particular opinions, knowledge, or best practices that are particular to you, your team, or product.当技能编码了特定于你、你的团队或产品的观点、知识或最佳实践时，效果最好。
References
You can @ mention files to include them as references. References allow Claude to refer to in-depth information about the current plan.引用 (References)
你可以使用 @ 提及文件以将它们作为引用包含进来。引用允许 Claude 参考有关当前计划的深入信息。
This might be in specs files, mockups, or even entire codebases. Generally you should prefer files that are in code as it provides clear, high-fidelity instructions to Claude in a language it knows very well. For example, a HTML mockup of a design will generally produce better results than a description of the design or a screenshot.这可以是规范文件、模型图，甚至是整个代码库。通常你应该优先选择代码中的文件，因为它能以 Claude 非常熟悉的语言提供清晰、高保真的指令。例如，设计的 HTML 原型通常比设计的描述或截图能产生更好的结果。
Try simplifying尝试简化
Across your system prompt, skills, and CLAUDE.md files, you may need to simplify just like we did. We rolled out a new command called `claude doctor,` which will help you do this automatically as well. For more details on prompting more advanced models specifically, check out our .在你的系统提示词、技能和 CLAUDE.md 文件中，你可能需要像我们一样进行简化。我们推出了一个名为 `claude doctor` 的新命令，它也可以帮你自动完成此操作。有关专门提示更先进模型的更多详细信息，请查看我们的 Fable 实战指南。
想发布自己的文章？想发布自己的文章？
升级为 Premium升级为 Premium