Key Takeaways核心要点
- Create custom interfaces without starting from scratch. Use templates and coding-agent guidance to build on LangSmith data, then publish and share inside LangSmith.无需从零开始，即可创建自定义界面。利用模板和编码智能体（coding-agent）的指导，基于 LangSmith 数据进行构建，然后直接在 LangSmith 内部发布和共享。
- Review agent behavior your way. Build focused UIs for feedback, experiments, traces, and the behaviors that matter most to your application.以您自己的方式审查智能体行为。为反馈、实验、追踪记录以及对您应用至关重要的行为构建专属 UI。
- Create Custom Apps with a prompt. Use LangSmith Chat to describe what data they want and how to display it, then turn that into a working app.通过提示词创建自定义应用。使用 LangSmith Chat 描述您需要的数据及展示方式，即可将其转化为一个可用的应用程序。
Today, we’re launching Custom Apps in LangSmith, which lets you build, publish, and run custom UIs on top of LangSmith data and run them directly inside LangSmith.今天，我们正式在 LangSmith 中推出“自定义应用”（Custom Apps）功能，让您能够在 LangSmith 数据之上构建、发布和运行自定义 UI，并直接在 LangSmith 内部使用它们。
Thousands of AI engineering teams use LangSmith to debug, evaluate, and improve their agents. LangSmith provides default UIs for those workflows based on the most common user requirements we've seen, including prebuilt dashboards for latency, cost, and error rates, and a comparison view that highlights regressions across experiments. But we've seen that teams often need to interact with data in ways that align with their own quality criteria and review processes.成千上万的 AI 工程团队使用 LangSmith 来调试、评估和改进他们的智能体。LangSmith 根据我们观察到的最常见用户需求，为这些工作流提供了默认 UI，包括用于监控延迟、成本和错误率的预置仪表板，以及突出显示跨实验回归情况的对比视图。但我们发现，团队往往需要以符合自身质量标准和审查流程的方式与数据进行交互。
Some teams already solve this by building custom frontends around LangSmith data. They use coding agents, LangSmith APIs, and internal tools to create custom interfaces of their data. But these applications also require ongoing maintenance just like any other piece of software, which includes hosting, authentication, permissions, and sharing.一些团队已经通过围绕 LangSmith 数据构建自定义前端来解决这个问题。他们利用编码智能体、LangSmith API 和内部工具来创建数据的自定义界面。但与任何其他软件一样，这些应用也需要持续的维护，包括托管、身份验证、权限管理和共享。
Now, Custom Apps lets you build the interface you want with your LangSmith data, publish it into your workspace, and skip the hosting, auth, and permissions work.现在，通过自定义应用，您可以利用 LangSmith 数据构建所需的界面，将其发布到您的工作区中，并省去托管、身份验证和权限配置等繁琐工作。
How Custom Apps work自定义应用的工作原理
With Custom Apps, you can start with a chat prompt or build in code. On Plus and Enterprise plans, tell LangSmith Chat what data you want to see and how you want it visualized, and Chat turns that into a working app. If you'd rather build it yourself, start with provided templates and build against the LangSmith API with your coding agent.使用自定义应用，您可以通过聊天提示词开始，也可以通过代码构建。在 Plus 和 Enterprise 计划中，只需告诉 LangSmith Chat 您想查看哪些数据以及希望如何可视化，Chat 就会将其转化为一个可用的应用。如果您更愿意亲自动手，可以从提供的模板开始，并利用您的编码智能体针对 LangSmith API 进行开发。
Once published, the app becomes a shared interface inside the workspace. Instead of exporting data or rebuilding the same static view ad hoc, you can reuse the app for recurring workflows and share it with teammates in the same place they already review traces, experiments, annotations, and evaluation results.应用发布后，它将成为工作区内的一个共享界面。您无需导出数据或反复重建相同的静态视图，即可在重复的工作流中复用该应用，并与团队成员共享，而他们无需离开审查追踪记录、实验、标注和评估结果的界面即可直接使用。
Custom Apps are especially useful when you need to repeat the same review process. Here are a few common ways to use them.当您需要重复进行相同的审查流程时，自定义应用尤为有用。以下是一些常见的应用场景。
Give annotators the right context为标注人员提供正确的上下文
Human feedback is one of the most important inputs for improving AI systems. Automated evals can help you find patterns, but you still need people to review outputs, judge behavior, and explain what good looks like for your application.人工反馈是改进 AI 系统最重要的输入之一。自动化评估可以帮助您发现模式，但您仍然需要人工来审查输出、判断行为，并解释对于您的应用而言，什么是好的结果。
That review often depends on who is doing the work. An engineer may need the full trace, including tool calls, metadata, and intermediate steps. A subject matter expert (SME) may only need the user request, the model response, relevant context, and a clear rubric.这种审查往往取决于执行工作的人员。工程师可能需要完整的追踪记录，包括工具调用、元数据和中间步骤；而领域专家（SME）可能只需要用户请求、模型响应、相关上下文以及明确的评估准则。
With Custom Apps, you can create annotation interfaces that match the reviewer’s workflows. You can give reviewers the context they need, hide the details they don’t, and guide them through the specific feedback you want to collect.借助自定义应用，您可以创建符合审查人员工作流的标注界面。您可以为审查人员提供他们所需的上下文，隐藏不必要的细节，并引导他们提供您想要收集的具体反馈。
Our human-in-the-loop reviews used to live across spreadsheets and other separate tools. Reviewers switched between platforms, and our team spent manual effort prepping files and standardizing formats. So we built a Custom App for structured LLM evaluation and annotation. It runs inside LangSmith and connects through the API, so we can build review queues and dashboards that match how our teams work. A reviewer picks a queue, evaluates and edits the output, and adds the result to a dataset. This gives us a repeatable framework for human-in-the-loop review, simpler tracking, and evaluation that's easier to scale across teams.
– Roxanne Wang, Chime®“我们过去的人机协作审查流程散落在电子表格和其他独立的工具中。审查人员需要在不同平台间切换，我们的团队也花费了大量人工来准备文件和标准化格式。因此，我们为结构化的 LLM 评估和标注构建了一个自定义应用。它在 LangSmith 内部运行并通过 API 连接，这样我们就可以构建符合团队工作方式的审查队列和仪表板。审查人员选择一个队列，评估并编辑输出，然后将结果添加到数据集中。这为我们提供了可重复的人机协作审查框架，简化了追踪流程，并使评估更容易在团队间扩展。” – Roxanne Wang, Chime®
Customize how you compare experiments自定义您的实验对比方式
Experiments help you assess whether changes to your agent achieved the intended outcome. Before shipping a change, you usually need to review individual results, segments, or failure patterns that matter to your application.实验有助于评估对智能体的更改是否达到了预期效果。在发布更改之前，您通常需要审查对您的应用至关重要的单个结果、细分数据或失败模式。
That might mean comparing outputs across prompt versions, checking whether a model swap introduced regressions, or slicing results by customer segment, failure category, model provider, or internal evaluation rubric.这可能意味着对比不同提示词版本的输出，检查模型替换是否引入了回归，或者按客户细分、失败类别、模型提供商或内部评估准则对结果进行切片分析。
With Custom Apps, you can create experiment review interfaces around those decisions. That means you can highlight the slices that matter most, show representative examples, compare outputs side by side, or package results into the format you use during release reviews.借助自定义应用，您可以围绕这些决策创建实验审查界面。这意味着您可以突出显示最重要的切片，展示具有代表性的示例，并排对比输出，或者将结果打包成您在发布审查时使用的格式。
This is especially useful when experiment review is a recurring process. Instead of rebuilding the same static chart or custom analysis each time, you can create a shared app that stays connected to LangSmith data and reuse it for every review cycle.当实验审查是一个重复性流程时，这一点尤为有用。您无需每次都重新构建相同的静态图表或自定义分析，而是可以创建一个与 LangSmith 数据保持连接的共享应用，并在每个审查周期中复用它。
Make trace review more focused使追踪记录审查更具针对性
Your trace review process typically follows the type of agent you're running. For a customer support agent, you might focus on tone and whether the agent resolved the issue, which you can only judge by reading the exchange. On the other hand, a research agent may be judged on its sources and the number of tool calls required.您的追踪记录审查流程通常取决于您运行的智能体类型。对于客户支持智能体，您可能关注语气以及智能体是否解决了问题，这只能通过阅读对话来判断。另一方面，研究型智能体则可能根据其来源以及所需的工具调用次数来评估。
When that review happens on a schedule, Custom Apps let you create a more focused UI around the questions you ask most often, such as where behavior changed, which step caused a failure, or whether the application followed the expected process.当此类审查按计划进行时，自定义应用让您可以围绕最常询问的问题构建更具针对性的 UI，例如行为在何处发生变化、哪一步导致了失败，或应用是否遵循了预期的流程。
Use the LangSmith open source design system使用 LangSmith 开源设计系统
Custom Apps use Macaw, LangSmith’s design system, by default. This gives your apps the same visual language as the rest of LangSmith, including consistent components and iconography.自定义应用默认使用 LangSmith 的设计系统 Macaw。这使您的应用与 LangSmith 的其他部分具有相同的视觉语言，包括统一的组件和图标。
You also have the freedom to customize your look. If you want an app to match your company’s brand instead, you can add any customization in chat or with your coding agent (ie. "Make this look like old school Macintosh software”).您也可以自由定制外观。如果您希望应用符合公司的品牌形象，可以在聊天中或通过编码智能体添加任何自定义设置（例如：“让它看起来像老式 Macintosh 软件”）。
Along with today’s launch, we’re making Macaw open source. We’ll continue expanding it and welcome feedback from the community as it evolves.伴随今天的发布，我们将 Macaw 开源。我们将继续扩展它，并欢迎社区在它不断演进的过程中提供反馈。
Availability and pricing可用性与定价
Custom Apps are available for Plus and Enterprise plans:自定义应用适用于 Plus 和 Enterprise 计划：
- Plus plans include one Custom App per organization and the chat experience.Plus 计划包括每个组织一个自定义应用以及聊天体验。
- Enterprise plans include unlimited Custom Apps with the chat experience.Enterprise 计划包括无限数量的自定义应用以及聊天体验。
For general usage and billing information, see our pricing page.有关一般使用和计费信息，请参阅我们的定价页面。
Get started开始使用
You can try Custom Apps today by logging in or signing up for LangSmith, and visit the docs for more detail.您今天就可以通过登录或注册 LangSmith 来试用自定义应用，并访问文档了解更多详情。
