Claude Managed Agents 新增 dreaming
New in Claude Managed Agents: dreaming, outcomes, and multiagent orchestrationMay 19, 2026
做 Agent 的同学必看，Managed Agents 新增的 dreaming 和 outcomes 能显著提升任务成功率，赶紧去申请 dreaming 预览并试试多智能体编排。
New in Claude Managed Agents: dreaming, outcomes, and multiagent orchestration
Claude 托管代理新功能：梦境、成果与多代理编排
- Category
- Product announcements
- Product
- Claude Platform
- Date
- May 19, 2026
- Reading time
- 5
- min
- Share
- Copy link
- https://claude.com/blog/new-in-claude-managed-agents
- 分类
- 产品公告
- 产品
- Claude 平台
- 日期
- 2026年5月19日
- 阅读时间
- 5
- 分钟
- 分享
- 复制链接
- https://claude.com/blog/new-in-claude-managed-agents
Today we're launching dreaming in Claude Managed Agents as a research preview. Dreaming extends memory by reviewing past sessions to find patterns and help agents self-improve. We're also making outcomes, multiagent orchestration, and webhooks available to developers building with Managed Agents. Together, these updates make agents more capable at handling complex tasks with minimal steering.
今天，我们在Claude托管代理中推出梦境功能，作为研究预览。梦境通过回顾过去的会话，寻找模式并帮助代理自我改进，从而扩展记忆。我们还向使用托管代理的开发者提供成果、多代理编排和Webhooks功能。这些更新共同使代理在处理复杂任务时更具能力，且只需极少的引导。
Build self-improving agents with dreaming
利用梦境构建自我改进的代理
Dreaming is a scheduled process in Claude Managed Agents that reviews agent sessions and memory stores, extracts patterns, and curates memories so agents improve over time. You decide how much control you want: dreaming can update memory automatically, or you can review changes before they land.
梦境是Claude托管代理中的一个计划进程，它审查代理会话和记忆存储，提取模式，并整理记忆，使代理随时间不断改进。您可以决定想要的控制程度：梦境可以自动更新记忆，或者您可以在更改生效前进行审查。
Dreaming surfaces patterns that a single agent can’t see on its own, including recurring mistakes, workflows that agents converge on, and preferences shared across a team. It also restructures memory so it stays high-signal as it evolves. This is especially useful for long-running work and multiagent orchestration.
梦境能揭示单个代理无法自行发现的模式，包括反复出现的错误、代理趋同的工作流程，以及团队共享的偏好。它还会重组记忆，使其在演变过程中保持高信号。这对于长期运行的工作和多代理编排尤其有用。
Together, memory and dreaming form a robust memory system for self-improving agents. Memory lets each agent capture what it learns as it works. Dreaming refines that memory between sessions, pulling shared learnings across agents and keeping it up-to-date.
记忆和梦境共同构成了一个强大的记忆系统，用于自我改进的代理。记忆让每个代理在工作时捕捉所学内容。梦境在会话之间精炼这些记忆，提取跨代理的共享学习成果，并保持其最新状态。
Dreaming is available in Managed Agents on the Claude Platform; developers can request access here.
梦境功能现已在Claude平台的托管代理中提供；开发者可在此处申请访问权限。
Outcomes: define the quality bar for agent work
成果：定义代理工作的质量基准
With outcomes, you write a rubric describing what success looks like and the agent works toward it. A separate grader evaluates the output against your criteria in its own context window, so it isn't influenced by the agent's reasoning. When something isn't right, the grader pinpoints what needs to change and the agent takes another pass.
通过成果，您可以编写一个评分标准，描述成功的样子，代理则朝着这个目标努力。一个独立的评分器会在自己的上下文窗口中根据您的标准评估输出，因此不会受到代理推理的影响。当某些内容不正确时，评分器会指出需要更改的地方，代理会再次尝试。
Agents do their best work when they know what "good" looks like. For example, a structural framework, a presentation standard, or a set of requirements that need to be met. With outcomes, agents can check their work against that bar and self-correct until the output is good enough, without a human needing to review each attempt.
当代理知道“好”的标准时，它们会发挥最佳表现。例如，一个结构框架、一个演示标准，或一组需要满足的要求。通过成果，代理可以对照这一基准检查自己的工作，并自我纠正，直到输出足够好，而无需人工审查每次尝试。
Outcomes is particularly useful for tasks that require attention to detail and exhaustive coverage. It also works for subjective quality, like whether copy matches a brand voice or a design follows visual guidelines. In testing, outcomes improved task success by up to 10 points over a standard prompting loop, with the largest gains on the hardest problems. Outcomes also improved file generation quality, with +8.4% task success on docx and +10.1% on pptx in our internal benchmarks.
结果（Outcomes）特别适用于需要注重细节和全面覆盖的任务。它也适用于主观质量，比如文案是否符合品牌语调或设计是否遵循视觉指南。在测试中，结果（Outcomes）相比标准提示循环，将任务成功率提高了最多10个百分点，其中在最具挑战性的问题上提升最大。结果（Outcomes）还提升了文件生成质量，在我们的内部基准测试中，docx任务成功率提高了8.4%，pptx提高了10.1%。
You can also now define an outcome, let the agent run, and get notified by a webhook when it's done.
您现在还可以定义结果，让代理运行，并在完成时通过webhook接收通知。
Multiagent orchestration: Handle complex tasks with multiple agents
多代理编排：使用多个代理处理复杂任务
When there is too much work for a single agent to do well, multiagent orchestration lets a lead agent break the job into pieces and delegate each one to a specialist with its own model, prompt, and tools. For example, a lead agent can run an investigation while subagents fan out through deploy history, error logs, metrics, and support tickets.
当单个代理无法出色完成过多工作时，多代理编排让主导代理将任务分解，并将每个部分委托给拥有自己模型、提示和工具的专业代理。例如，主导代理可以开展调查，而子代理则分散处理部署历史、错误日志、指标和支持工单。
These specialists work in parallel on a shared filesystem and contribute to the lead agent's overall context. The lead agent can check back in with other agents mid-workflow because events are persistent and every agent remembers what it's done. You can also trace every step in the Claude Console: which agent did what, in what order, and why, giving you full visibility into how your task was delegated and executed.
这些专业代理在共享文件系统上并行工作，并为主导代理的整体上下文做出贡献。由于事件是持久的，每个代理都记得自己做了什么，主导代理可以在工作流程中随时与其他代理进行核对。您还可以在Claude控制台中追踪每一步：哪个代理做了什么、顺序如何、为什么这样做，从而全面了解任务的委派和执行情况。
What teams are building
团队正在构建的内容
Teams are using dreaming, outcomes, and multiagent orchestration to ship agents that verify their own work, learn across sessions, and parallelize complex jobs:
团队正在利用梦想（dreaming）、结果（outcomes）和多代理编排来构建能够自我验证工作、跨会话学习以及并行处理复杂任务的代理：
- Harvey uses Managed Agents to coordinate complex legal work like long-form drafting and document creation. With dreaming, their agents remember what they learned between sessions, including filetype workarounds and tool-specific patterns. Completion rates went up ~6x in their tests.
- Netflix's platform team built an analysis agent that processes logs from hundreds of builds across different sources. With changes that affect thousands of applications, what matters is finding the issues that recur across many of them. Multiagent orchestration lets the agent analyze batches in parallel and surface only the patterns worth acting on.
- Spiral by Every is using multiagent orchestration and outcomes to power the writing agent behind their new API and CLI. The lead agent runs on Haiku: it fields incoming requests, poses quick follow-up questions when needed, then delegates the drafting to subagents running on Opus. When a user asks for multiple drafts, the subagents run in parallel. Writing quality is Spiral's core value, so they use outcomes to enforce it. Each draft is scored against a rubric of Every's editorial principles and the user's voice, both pulled from memory. Only drafts that clear the bar are returned.
- Wisedocs built a document quality check agent on Managed Agents, using outcomes to grade each review against their internal guidelines. Reviews now run 50% faster, while staying aligned with their team's standards.
- Harvey使用托管代理来协调复杂的法律工作，如长篇起草和文档创建。借助梦想（dreaming），他们的代理能记住会话间学到的内容，包括文件类型变通方法和工具特定模式。在他们的测试中，完成率提高了约6倍。
- Netflix的平台团队构建了一个分析代理，用于处理来自不同来源的数百个构建日志。对于影响数千个应用程序的变更，关键在于找出许多应用程序中反复出现的问题。多代理编排让代理能够并行分析批次，并仅呈现值得采取行动的模式。
- Every 公司的 Spiral 正在利用多智能体编排和结果来驱动其新 API 和 CLI 背后的写作代理。主代理运行在 Haiku 上：它处理传入的请求，在需要时提出快速的后续问题，然后将草稿任务委派给运行在 Opus 上的子代理。当用户请求多个草稿时，子代理会并行运行。写作质量是 Spiral 的核心价值，因此他们使用结果来强制执行。每个草稿都会根据 Every 的编辑原则和用户声音的评分标准进行评分，这两者都从记忆中提取。只有达到标准的草稿才会被返回。
- Wisedocs 在 Managed Agents 上构建了一个文档质量检查代理，使用结果根据其内部指南对每次审查进行评分。现在审查速度提高了 50%，同时保持与其团队标准的一致性。
No items found.
未找到项目。
PrevPrev
上一页上一页
0/5
0/5
NextNext
下一页下一页
eBook
电子书
Getting started
入门指南
Dreaming is available in research preview, outcomes, multiagent orchestration, and memory are available in public beta as part of Managed Agents. To get started with dreaming, request access here. Explore our documentation to learn more or visit the Claude Console to deploy your first agent.
Dreaming 目前处于研究预览阶段，结果、多智能体编排和记忆作为 Managed Agents 的一部分处于公开测试阶段。要开始使用 dreaming，请在此处请求访问。浏览我们的文档以了解更多信息，或访问 Claude Console 部署您的第一个代理。
FAQ
常见问题解答
No items found.
未找到项目。
Related posts
相关文章
Explore more product news and best practices for teams building with Claude.
探索更多产品新闻和针对使用 Claude 构建的团队的最佳实践。
Aug 28, 2026
2026年8月28日
Claude for Teachers, now available for U.S. K-12 schools and districts
Claude for Teachers 现已面向美国 K-12 学校和学区推出
Product announcements
产品公告
Claude for Teachers, now available for U.S. K-12 schools and districtsClaude for Teachers, now available for U.S. K-12 schools and districts
Claude for Teachers 现已面向美国 K-12 学校和学区推出Claude for Teachers 现已面向美国 K-12 学校和学区推出
Claude for Teachers, now available for U.S. K-12 schools and districtsClaude for Teachers, now available for U.S. K-12 schools and districts
Claude for Teachers 现已面向美国 K-12 学校和学区推出Claude for Teachers 现已面向美国 K-12 学校和学区推出
Aug 26, 2026
2026年8月26日
Claude gets its own browser in Cowork
Claude 在 Cowork 中拥有了自己的浏览器
Product announcements
产品公告
Claude gets its own browser in CoworkClaude gets its own browser in Cowork
Claude 在 Cowork 中拥有了自己的浏览器Claude 在 Cowork 中拥有了自己的浏览器
Claude gets its own browser in CoworkClaude gets its own browser in Cowork
Claude 在 Cowork 中拥有了自己的浏览器Claude 在 Cowork 中拥有了自己的浏览器
Aug 26, 2026
2026年8月26日
Claude in Chrome is generally available
Claude 在 Chrome 中正式可用
Product announcements
产品公告
Claude in Chrome is generally available Claude in Chrome is generally available
Claude 在 Chrome 中正式可用 Claude 在 Chrome 中正式可用
Claude in Chrome is generally available Claude in Chrome is generally available
Claude 在 Chrome 中正式可用 Claude 在 Chrome 中正式可用
Aug 11, 2026
2026年8月11日
Compliance API coverage extends to Claude Cowork and Claude Code
合规 API 覆盖范围扩展至 Claude Cowork 和 Claude Code
Enterprise AI
企业级 AI
Compliance API coverage extends to Claude Cowork and Claude CodeCompliance API coverage extends to Claude Cowork and Claude Code
合规 API 覆盖范围扩展至 Claude Cowork 和 Claude Code 合规 API 覆盖范围扩展至 Claude Cowork 和 Claude Code
Compliance API coverage extends to Claude Cowork and Claude CodeCompliance API coverage extends to Claude Cowork and Claude Code
合规 API 覆盖范围扩展至 Claude Cowork 和 Claude Code 合规 API 覆盖范围扩展至 Claude Cowork 和 Claude Code
Transform how your organization operates with Claude
利用 Claude 改变您组织的运营方式
See pricing
查看定价
See pricingSee pricing
查看定价 查看定价
Contact sales
联系销售
Contact salesContact sales
联系销售 联系销售
Get the developer newsletter
获取开发者通讯
Product updates, how-tos, community spotlights, and more. Delivered monthly to your inbox.
产品更新、操作指南、社区亮点等。每月发送至您的邮箱。
SubscribeSubscribe
订阅 订阅
Please provide your email address if you'd like to receive our monthly developer newsletter. You can unsubscribe at any time.
如果您希望接收我们的月度开发者通讯，请提供您的电子邮件地址。您可以随时退订。
Thank you! You’re subscribed.
谢谢！您已成功订阅。
Sorry, there was a problem with your submission, please try again later.
抱歉，您的提交出现问题，请稍后重试。
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力