Building effective agents
构建有效的代理

Over the past year, we've worked with dozens of teams building large language model (LLM) agents across industries. Consistently, the most successful implementations weren't using complex frameworks or specialized libraries. Instead, they were building with simple, composable patterns.
在过去的一年里，我们与数十个团队合作，构建了跨行业的大型语言模型（LLM）代理。始终如一，最成功的实施并没有使用复杂的框架或专业的库。相反，他们使用简单、可组合的模式进行构建。

In this post, we share what we’ve learned from working with our customers and building agents ourselves, and give practical advice for developers on building effective agents.
在这篇文章中，我们分享了与客户合作和自己构建代理所学到的经验，并为开发者提供了构建有效代理的实用建议。

What are agents?  什么是代理？

"Agent" can be defined in several ways. Some customers define agents as fully autonomous systems that operate independently over extended periods, using various tools to accomplish complex tasks. Others use the term to describe more prescriptive implementations that follow predefined workflows. At Anthropic, we categorize all these variations as agentic systems, but draw an important architectural distinction between workflows and agents:
“Agent”可以通过多种方式定义。一些客户将代理定义为在较长时间内独立操作的完全自主系统，使用各种工具来完成复杂任务。其他人则用这个术语来描述遵循预定义工作流程的更具规范性的实现。在 Anthropic，我们将所有这些变体归类为代理系统，但在工作流程和代理之间做出一个重要的架构区分：

• Workflows are systems where LLMs and tools are orchestrated through predefined code paths.
  工作流是通过预定义代码路径协调LLMs和工具的系统。
• Agents, on the other hand, are systems where LLMs dynamically direct their own processes and tool usage, maintaining control over how they accomplish tasks.
  另一方面，代理是系统，其中LLMs动态地指导自己的过程和工具使用，保持对完成任务方式的控制。

Below, we will explore both types of agentic systems in detail. In Appendix 1 (“Agents in Practice”), we describe two domains where customers have found particular value in using these kinds of systems.
下面，我们将详细探讨这两种代理系统。在附录 1（“实践中的代理”）中，我们描述了客户在使用这些系统时发现特别有价值的两个领域。

When (and when not) to use agents
何时（以及何时不）使用代理

When building applications with LLMs, we recommend finding the simplest solution possible, and only increasing complexity when needed. This might mean not building agentic systems at all. Agentic systems often trade latency and cost for better task performance, and you should consider when this tradeoff makes sense.
在使用LLMs构建应用程序时，我们建议寻找尽可能简单的解决方案，仅在需要时增加复杂性。这可能意味着根本不构建自主系统。自主系统通常以延迟和成本为代价来提高任务性能，因此您应该考虑这种权衡何时是合理的。

When more complexity is warranted, workflows offer predictability and consistency for well-defined tasks, whereas agents are the better option when flexibility and model-driven decision-making are needed at scale. For many applications, however, optimizing single LLM calls with retrieval and in-context examples is usually enough.
当需要更多复杂性时，工作流程为明确定义的任务提供可预测性和一致性，而当需要灵活性和模型驱动的决策时，代理是更好的选择。然而，对于许多应用程序，优化单个 LLM 调用与检索和上下文示例通常就足够了。

When and how to use frameworks
何时以及如何使用框架

There are many frameworks that make agentic systems easier to implement, including:
有许多框架使得代理系统更容易实现，包括：

• LangGraph from LangChain;
  LangGraph 来自 LangChain;
• Amazon Bedrock's AI Agent framework;
  亚马逊 Bedrock 的 AI 代理框架
• Rivet, a drag and drop GUI LLM workflow builder; and
  Rivet，一个拖放式图形用户界面LLM工作流构建器；和
• Vellum, another GUI tool for building and testing complex workflows.
  Vellum，另一个用于构建和测试复杂工作流的 GUI 工具。

These frameworks make it easy to get started by simplifying standard low-level tasks like calling LLMs, defining and parsing tools, and chaining calls together. However, they often create extra layers of abstraction that can obscure the underlying prompts ​​and responses, making them harder to debug. They can also make it tempting to add complexity when a simpler setup would suffice.
这些框架通过简化标准的低级任务，如调用 LLMs、定义和解析工具以及将调用链在一起，使得入门变得简单。然而，它们往往会创建额外的抽象层，这可能会掩盖底层的提示和响应，使调试变得更加困难。它们还可能使人倾向于增加复杂性，而实际上更简单的设置就足够了。

We suggest that developers start by using LLM APIs directly: many patterns can be implemented in a few lines of code. If you do use a framework, ensure you understand the underlying code. Incorrect assumptions about what's under the hood are a common source of customer error.
我们建议开发者直接使用 LLM APIs：许多模式可以用几行代码实现。如果你使用框架，请确保你理解底层代码。对底层实现的错误假设是客户错误的常见来源。

See our cookbook for some sample implementations.
请查看我们的食谱以获取一些示例实现。

Building blocks, workflows, and agents
构建模块、工作流程和代理

In this section, we’ll explore the common patterns for agentic systems we’ve seen in production. We'll start with our foundational building block—the augmented LLM—and progressively increase complexity, from simple compositional workflows to autonomous agents.
在本节中，我们将探讨我们在生产中看到的代理系统的常见模式。我们将从我们的基础构建块——增强的 LLM 开始，并逐步增加复杂性，从简单的组合工作流到自主代理。

Building block: The augmented LLM
构建模块：增强的 LLM

The basic building block of agentic systems is an LLM enhanced with augmentations such as retrieval, tools, and memory. Our current models can actively use these capabilities—generating their own search queries, selecting appropriate tools, and determining what information to retain.
代理系统的基本构建块是一个LLM，增强了检索、工具和记忆等扩展功能。我们当前的模型可以主动使用这些能力——生成自己的搜索查询，选择合适的工具，并确定保留哪些信息。

We recommend focusing on two key aspects of the implementation: tailoring these capabilities to your specific use case and ensuring they provide an easy, well-documented interface for your LLM. While there are many ways to implement these augmentations, one approach is through our recently released Model Context Protocol, which allows developers to integrate with a growing ecosystem of third-party tools with a simple client implementation.
我们建议专注于实施的两个关键方面：将这些功能量身定制以满足您的特定用例，并确保它们为您的 LLM 提供一个简单、文档齐全的接口。虽然有许多方法可以实现这些增强功能，但一种方法是通过我们最近发布的 Model Context Protocol，它允许开发者通过简单的客户端实现与不断增长的第三方工具生态系统集成。

For the remainder of this post, we'll assume each LLM call has access to these augmented capabilities.
在本帖的其余部分，我们将假设每个 LLM 调用都可以访问这些增强功能。

Workflow: Prompt chaining
工作流程：提示链

Prompt chaining decomposes a task into a sequence of steps, where each LLM call processes the output of the previous one. You can add programmatic checks (see "gate” in the diagram below) on any intermediate steps to ensure that the process is still on track.
提示链将任务分解为一系列步骤，其中每个 LLM 调用处理前一个步骤的输出。您可以在任何中间步骤上添加程序检查（请参见下图中的“门”），以确保过程仍在正轨上。

When to use this workflow: This workflow is ideal for situations where the task can be easily and cleanly decomposed into fixed subtasks. The main goal is to trade off latency for higher accuracy, by making each LLM call an easier task.
何时使用此工作流程：此工作流程非常适合将任务轻松且清晰地分解为固定子任务的情况。主要目标是通过将每个 LLM 调用变成更简单的任务来权衡延迟与更高的准确性。

Examples where prompt chaining is useful:
提示链的有用示例：

• Generating Marketing copy, then translating it into a different language.
  生成营销文案，然后将其翻译成另一种语言。
• Writing an outline of a document, checking that the outline meets certain criteria, then writing the document based on the outline.
  撰写文档大纲，检查大纲是否符合某些标准，然后根据大纲撰写文档。

Workflow: Routing  工作流程：路由

Routing classifies an input and directs it to a specialized followup task. This workflow allows for separation of concerns, and building more specialized prompts. Without this workflow, optimizing for one kind of input can hurt performance on other inputs.
路由将输入分类并将其引导到专门的后续任务。此工作流程允许关注点的分离，并构建更专业的提示。如果没有这个工作流程，针对一种输入进行优化可能会影响其他输入的性能。

When to use this workflow: Routing works well for complex tasks where there are distinct categories that are better handled separately, and where classification can be handled accurately, either by an LLM or a more traditional classification model/algorithm.
何时使用此工作流程：路由适用于复杂任务，其中存在可以更好地单独处理的不同类别，并且分类可以通过LLM或更传统的分类模型/算法准确处理。

Examples where routing is useful:
路由有用的示例：

• Directing different types of customer service queries (general questions, refund requests, technical support) into different downstream processes, prompts, and tools.
  将不同类型的客户服务查询（一般问题、退款请求、技术支持）引导到不同的下游流程、提示和工具中。
• Routing easy/common questions to smaller models like Claude 3.5 Haiku and hard/unusual questions to more capable models like Claude 3.5 Sonnet to optimize cost and speed.
  将简单/常见问题路由到较小的模型，如 Claude 3.5 Haiku，将困难/不寻常的问题路由到更强大的模型，如 Claude 3.5 Sonnet，以优化成本和速度。

Workflow: Parallelization
工作流程：并行化

LLMs can sometimes work simultaneously on a task and have their outputs aggregated programmatically. This workflow, parallelization, manifests in two key variations:
LLMs 有时可以同时在一个任务上工作，并且可以通过程序化方式聚合它们的输出。这个工作流程的并行化表现为两个关键变体：

• Sectioning: Breaking a task into independent subtasks run in parallel.
  分段：将一个任务分解为独立的子任务并行运行。
• Voting: Running the same task multiple times to get diverse outputs.
  投票：多次运行相同任务以获得多样化的输出。

When to use this workflow: Parallelization is effective when the divided subtasks can be parallelized for speed, or when multiple perspectives or attempts are needed for higher confidence results. For complex tasks with multiple considerations, LLMs generally perform better when each consideration is handled by a separate LLM call, allowing focused attention on each specific aspect.
何时使用此工作流程：当划分的子任务可以并行化以提高速度，或者当需要多个视角或尝试以获得更高的置信度结果时，平行化是有效的。对于具有多个考虑因素的复杂任务，LLMs 通常在每个考虑因素由单独的 LLM 调用处理时表现更好，从而允许对每个特定方面进行集中关注。

Examples where parallelization is useful:
并行化有用的示例：

• Sectioning:  分段：
  • Implementing guardrails where one model instance processes user queries while another screens them for inappropriate content or requests. This tends to perform better than having the same LLM call handle both guardrails and the core response.
    实施保护措施，其中一个模型实例处理用户查询，而另一个则筛选不当内容或请求。这通常比让同一个 LLM 调用同时处理保护措施和核心响应表现更好。
  • Automating evals for evaluating LLM performance, where each LLM call evaluates a different aspect of the model’s performance on a given prompt.
    自动化评估LLM性能的评估，其中每个LLM调用评估模型在给定提示上的不同性能方面。
• Voting:
  • Reviewing a piece of code for vulnerabilities, where several different prompts review and flag the code if they find a problem.
    审查一段代码以查找漏洞，其中多个不同的提示会审查并标记代码，如果发现问题。
  • Evaluating whether a given piece of content is inappropriate, with multiple prompts evaluating different aspects or requiring different vote thresholds to balance false positives and negatives.
    评估给定内容是否不当，使用多个提示评估不同方面或要求不同的投票阈值，以平衡假阳性和假阴性。

Workflow: Orchestrator-workers
工作流：Orchestrator-workers

In the orchestrator-workers workflow, a central LLM dynamically breaks down tasks, delegates them to worker LLMs, and synthesizes their results.
在协调器-工作者工作流程中，一个中央LLM动态地分解任务，将其委派给工作者LLMs，并综合它们的结果。

When to use this workflow: This workflow is well-suited for complex tasks where you can’t predict the subtasks needed (in coding, for example, the number of files that need to be changed and the nature of the change in each file likely depend on the task). Whereas it’s topographically similar, the key difference from parallelization is its flexibility—subtasks aren't pre-defined, but determined by the orchestrator based on the specific input.
何时使用此工作流程：此工作流程非常适合复杂任务，在这些任务中，您无法预测所需的子任务（例如，在编码中，需要更改的文件数量和每个文件中更改的性质可能取决于任务）。尽管在拓扑上相似，但与并行化的关键区别在于其灵活性——子任务不是预定义的，而是由协调者根据特定输入确定的。

Example where orchestrator-workers is useful:
orchestrator-workers 有用的示例：

• Coding products that make complex changes to multiple files each time.
  编码产品，每次对多个文件进行复杂更改。
• Search tasks that involve gathering and analyzing information from multiple sources for possible relevant information.
  搜索涉及从多个来源收集和分析信息以寻找可能相关信息的任务。

Workflow: Evaluator-optimizer
工作流程：评估者-优化器

In the evaluator-optimizer workflow, one LLM call generates a response while another provides evaluation and feedback in a loop.
在评估-优化工作流程中，一个LLM调用生成响应，而另一个在循环中提供评估和反馈。

When to use this workflow: This workflow is particularly effective when we have clear evaluation criteria, and when iterative refinement provides measurable value. The two signs of good fit are, first, that LLM responses can be demonstrably improved when a human articulates their feedback; and second, that the LLM can provide such feedback. This is analogous to the iterative writing process a human writer might go through when producing a polished document.
何时使用此工作流程：当我们有明确的评估标准，并且迭代改进提供可衡量的价值时，此工作流程特别有效。良好适配的两个标志是，首先，当人类表达他们的反馈时，LLM 的响应可以明显改善；其次，LLM 可以提供这样的反馈。这类似于人类作者在制作精 polished 文档时可能经历的迭代写作过程。

Examples where evaluator-optimizer is useful:
评估器-优化器有用的示例：

• Literary translation where there are nuances that the translator LLM might not capture initially, but where an evaluator LLM can provide useful critiques.
  文学翻译中，翻译者LLM可能最初无法捕捉到的细微差别，但评估者LLM可以提供有用的批评。
• Complex search tasks that require multiple rounds of searching and analysis to gather comprehensive information, where the evaluator decides whether further searches are warranted.
  复杂的搜索任务需要多轮搜索和分析以收集全面的信息，评估者决定是否需要进一步的搜索。

Agents

Agents are emerging in production as LLMs mature in key capabilities—understanding complex inputs, engaging in reasoning and planning, using tools reliably, and recovering from errors. Agents begin their work with either a command from, or interactive discussion with, the human user. Once the task is clear, agents plan and operate independently, potentially returning to the human for further information or judgement. During execution, it's crucial for the agents to gain “ground truth” from the environment at each step (such as tool call results or code execution) to assess its progress. Agents can then pause for human feedback at checkpoints or when encountering blockers. The task often terminates upon completion, but it’s also common to include stopping conditions (such as a maximum number of iterations) to maintain control.
随着LLMs在关键能力上成熟，代理在生产中逐渐出现——理解复杂输入、进行推理和规划、可靠地使用工具以及从错误中恢复。代理的工作始于来自人类用户的命令或互动讨论。一旦任务明确，代理便独立规划和操作，可能会返回人类以获取更多信息或判断。在执行过程中，代理在每一步都至关重要的是从环境中获取“真实情况”（例如工具调用结果或代码执行）以评估其进展。代理可以在检查点暂停以获取人类反馈或在遇到障碍时暂停。任务通常在完成时终止，但也常常包括停止条件（例如最大迭代次数）以保持控制。

Agents can handle sophisticated tasks, but their implementation is often straightforward. They are typically just LLMs using tools based on environmental feedback in a loop. It is therefore crucial to design toolsets and their documentation clearly and thoughtfully. We expand on best practices for tool development in Appendix 2 ("Prompt Engineering your Tools").
代理可以处理复杂的任务，但它们的实现通常很简单。它们通常只是 LLMs 使用基于环境反馈的工具循环。因此，清晰而周到地设计工具集及其文档至关重要。我们在附录 2（“为您的工具进行提示工程”）中扩展了工具开发的最佳实践。

When to use agents: Agents can be used for open-ended problems where it’s difficult or impossible to predict the required number of steps, and where you can’t hardcode a fixed path. The LLM will potentially operate for many turns, and you must have some level of trust in its decision-making. Agents' autonomy makes them ideal for scaling tasks in trusted environments.
何时使用代理：代理可以用于开放性问题，在这些问题中，很难或不可能预测所需的步骤数量，并且无法硬编码固定路径。LLM可能会运行多个回合，您必须对其决策能力有一定程度的信任。代理的自主性使它们在可信环境中非常适合扩展任务。

The autonomous nature of agents means higher costs, and the potential for compounding errors. We recommend extensive testing in sandboxed environments, along with the appropriate guardrails.
代理的自主性意味着更高的成本，以及潜在的累积错误。我们建议在沙盒环境中进行广泛测试，并设置适当的保护措施。

Examples where agents are useful:
代理有用的例子：

The following examples are from our own implementations:
以下示例来自我们自己的实现：

• A coding Agent to resolve SWE-bench tasks, which involve edits to many files based on a task description;
  一个编码代理，用于解决 SWE-bench 任务，该任务涉及根据任务描述对多个文件进行编辑；
• Our “computer use” reference implementation, where Claude uses a computer to accomplish tasks.
  我们的“计算机使用”参考实现，其中 Claude 使用计算机完成任务。

Combining and customizing these patterns
结合和定制这些模式

These building blocks aren't prescriptive. They're common patterns that developers can shape and combine to fit different use cases. The key to success, as with any LLM features, is measuring performance and iterating on implementations. To repeat: you should consider adding complexity only when it demonstrably improves outcomes.
这些构建模块并不是规定性的。它们是开发者可以塑造和组合以适应不同用例的常见模式。成功的关键，与任何 LLM 功能一样，是衡量性能并对实现进行迭代。重申一下：你应该考虑在明显改善结果时才增加复杂性。

Summary

Success in the LLM space isn't about building the most sophisticated system. It's about building the right system for your needs. Start with simple prompts, optimize them with comprehensive evaluation, and add multi-step agentic systems only when simpler solutions fall short.
在LLM领域取得成功并不是关于构建最复杂的系统，而是关于为您的需求构建合适的系统。从简单的提示开始，通过全面评估进行优化，仅在更简单的解决方案不足时才添加多步骤的自主系统。

When implementing agents, we try to follow three core principles:
在实施代理时，我们尽量遵循三个核心原则：

1. Maintain simplicity in your agent's design.
   保持您代理设计的简单性。
2. Prioritize transparency by explicitly showing the agent’s planning steps.
   通过明确展示代理的规划步骤来优先考虑透明度。
3. Carefully craft your agent-computer interface (ACI) through thorough tool documentation and testing.
   仔细通过全面的工具文档和测试来精心设计您的代理-计算机接口（ACI）。

Frameworks can help you get started quickly, but don't hesitate to reduce abstraction layers and build with basic components as you move to production. By following these principles, you can create agents that are not only powerful but also reliable, maintainable, and trusted by their users.
框架可以帮助你快速入门，但在进入生产阶段时，不要犹豫减少抽象层次，使用基本组件进行构建。遵循这些原则，你可以创建不仅强大而且可靠、可维护并受到用户信任的代理。

Acknowledgements  致谢

Written by Erik Schluntz and Barry Zhang. This work draws upon our experiences building agents at Anthropic and the valuable insights shared by our customers, for which we're deeply grateful.
由 Erik Schluntz 和 Barry Zhang 撰写。此项工作基于我们在 Anthropic 构建代理的经验以及客户分享的宝贵见解，我们对此深表感谢。

Appendix 1: Agents in practice
附录 1：实践中的代理人

Our work with customers has revealed two particularly promising applications for AI agents that demonstrate the practical value of the patterns discussed above. Both applications illustrate how agents add the most value for tasks that require both conversation and action, have clear success criteria, enable feedback loops, and integrate meaningful human oversight.
我们与客户的合作揭示了两个特别有前景的人工智能代理应用，展示了上述模式的实际价值。这两个应用都说明了代理在需要对话和行动的任务中如何增加最大价值，具有明确的成功标准，能够启用反馈循环，并整合有意义的人类监督。

A. Customer support  A. 客户支持

Customer support combines familiar chatbot interfaces with enhanced capabilities through tool integration. This is a natural fit for more open-ended agents because:
客户支持将熟悉的聊天机器人界面与通过工具集成增强的功能相结合。这对于更开放式的代理来说是一个自然的选择，因为：

• Support interactions naturally follow a conversation flow while requiring access to external information and actions;
  支持交互自然地遵循对话流程，同时需要访问外部信息和操作；
• Tools can be integrated to pull customer data, order history, and knowledge base articles;
  工具可以集成以提取客户数据、订单历史和知识库文章；
• Actions such as issuing refunds or updating tickets can be handled programmatically; and
  诸如发放退款或更新票务等操作可以通过编程方式处理；并且
• Success can be clearly measured through user-defined resolutions.
  成功可以通过用户定义的分辨率清晰地衡量。

Several companies have demonstrated the viability of this approach through usage-based pricing models that charge only for successful resolutions, showing confidence in their agents' effectiveness.
几家公司通过基于使用的定价模型展示了这种方法的可行性，该模型仅对成功解决的案例收费，显示了对其代理人有效性的信心。

B. Coding agents  B. 编码代理

The software development space has shown remarkable potential for LLM features, with capabilities evolving from code completion to autonomous problem-solving. Agents are particularly effective because:
软件开发领域展现出LLM功能的显著潜力，能力从代码补全发展到自主问题解决。代理特别有效，因为：

• Code solutions are verifiable through automated tests;
  代码解决方案可以通过自动化测试进行验证；
• Agents can iterate on solutions using test results as feedback;
  代理可以根据测试结果作为反馈对解决方案进行迭代；
• The problem space is well-defined and structured; and
  问题空间定义明确且结构化；并且
• Output quality can be measured objectively.
  输出质量可以客观地衡量。

In our own implementation, agents can now solve real GitHub issues in the SWE-bench Verified benchmark based on the pull request description alone. However, whereas automated testing helps verify functionality, human review remains crucial for ensuring solutions align with broader system requirements.
在我们自己的实现中，代理现在可以仅根据拉取请求描述解决 SWE-bench Verified 基准中的真实 GitHub 问题。然而，尽管自动化测试有助于验证功能，但人工审查仍然对确保解决方案与更广泛的系统要求一致至关重要。

Appendix 2: Prompt engineering your tools
附录 2：为您的工具进行提示工程

No matter which agentic system you're building, tools will likely be an important part of your agent. Tools enable Claude to interact with external services and APIs by specifying their exact structure and definition in our API. When Claude responds, it will include a tool use block in the API response if it plans to invoke a tool. Tool definitions and specifications should be given just as much prompt engineering attention as your overall prompts. In this brief appendix, we describe how to prompt engineer your tools.
无论你正在构建哪个代理系统，工具很可能是你代理的重要组成部分。工具使 Claude 能够通过在我们的 API 中指定其确切结构和定义与外部服务和 API 进行交互。当 Claude 作出响应时，如果它计划调用工具，它将在 API 响应中包含一个工具使用块。工具的定义和规范应与整体提示一样受到提示工程的关注。在这个简短的附录中，我们描述了如何对你的工具进行提示工程。

There are often several ways to specify the same action. For instance, you can specify a file edit by writing a diff, or by rewriting the entire file. For structured output, you can return code inside markdown or inside JSON. In software engineering, differences like these are cosmetic and can be converted losslessly from one to the other. However, some formats are much more difficult for an LLM to write than others. Writing a diff requires knowing how many lines are changing in the chunk header before the new code is written. Writing code inside JSON (compared to markdown) requires extra escaping of newlines and quotes.
通常有几种方法可以指定相同的操作。例如，您可以通过编写差异来指定文件编辑，或者通过重写整个文件来指定。对于结构化输出，您可以在 markdown 或 JSON 中返回代码。在软件工程中，这些差异是表面的，可以无损地从一种格式转换为另一种格式。然而，有些格式对于LLM来说比其他格式更难编写。编写差异需要在写入新代码之前知道块头中有多少行正在更改。将代码写入 JSON（与 markdown 相比）需要额外转义换行符和引号。

Our suggestions for deciding on tool formats are the following:
我们对选择工具格式的建议如下：

• Give the model enough tokens to "think" before it writes itself into a corner.
  给模型足够的标记以“思考”，然后再让它自己陷入困境。
• Keep the format close to what the model has seen naturally occurring in text on the internet.
  保持格式接近模型在互联网上自然出现的文本。
• Make sure there's no formatting "overhead" such as having to keep an accurate count of thousands of lines of code, or string-escaping any code it writes.
  确保没有格式化的“开销”，例如必须准确计算成千上万行代码，或对其编写的任何代码进行字符串转义。

One rule of thumb is to think about how much effort goes into human-computer interfaces (HCI), and plan to invest just as much effort in creating good agent-computer interfaces (ACI). Here are some thoughts on how to do so:
一个经验法则是考虑人机界面（HCI）投入了多少精力，并计划在创建良好的代理-计算机界面（ACI）上投入同样多的精力。以下是一些实现这一目标的想法：

• Put yourself in the model's shoes. Is it obvious how to use this tool, based on the description and parameters, or would you need to think carefully about it? If so, then it’s probably also true for the model. A good tool definition often includes example usage, edge cases, input format requirements, and clear boundaries from other tools.
  把自己放在模型的角度上。根据描述和参数，使用这个工具是否显而易见，还是需要仔细思考？如果是这样，那么模型也可能是如此。一个好的工具定义通常包括示例用法、边界情况、输入格式要求以及与其他工具的明确界限。
• How can you change parameter names or descriptions to make things more obvious? Think of this as writing a great docstring for a junior developer on your team. This is especially important when using many similar tools.
  如何更改参数名称或描述以使其更明显？把这看作是为你团队中的初级开发者编写一个优秀的文档字符串。当使用许多相似的工具时，这一点尤其重要。
• Test how the model uses your tools: Run many example inputs in our workbench to see what mistakes the model makes, and iterate.
  测试模型如何使用您的工具：在我们的工作台上运行多个示例输入，以查看模型犯了哪些错误，并进行迭代。
• Poka-yoke your tools. Change the arguments so that it is harder to make mistakes.
  对你的工具进行防错设计。改变参数，使出错变得更加困难。

While building our agent for SWE-bench, we actually spent more time optimizing our tools than the overall prompt. For example, we found that the model would make mistakes with tools using relative filepaths after the agent had moved out of the root directory. To fix this, we changed the tool to always require absolute filepaths—and we found that the model used this method flawlessly.
在为 SWE-bench 构建我们的代理时，我们实际上花了更多时间来优化我们的工具，而不是整体提示。例如，我们发现模型在代理移出根目录后使用相对文件路径的工具时会出错。为了解决这个问题，我们将工具更改为始终要求绝对文件路径——我们发现模型使用这种方法毫无瑕疵。