https://t.co/QYN9Z7crCw

TL;DR · AI 摘要
LangChain推出新工具LangSmith Engine和Sandboxes,解决生产环境中的代理失败问题,提升开发效率与安全性。
核心要点
- LangSmith Engine可自动分析生产日志,识别问题并提出修复建议。
- Sandboxes提供安全的代码执行环境,支持代理代码的规模化运行。
- LangChain Labs启动持续学习研究计划,与多家公司合作推进代理技术发展。
结构提纲
按章节快速跳转。
思维导图
用一张图看清主题之间的关系。
查看大纲文本(无障碍 / 无 JS 友好)
- LangChain Newsletter
金句 / Highlights
值得收藏与分享的关键句。
LangSmith Engine通过监控生产日志,将重复性失败聚类为命名问题并诊断根本原因。
Sandboxes提供安全、可扩展的环境,支持代理代码执行并与Deep Agents SDK集成。
LangChain Labs与Harvey、NVIDIA等公司合作,推进代理技术的持续学习研究。
Article

Fixing agent failures in production | LangChain Newsletter
Fresh off Interrupt 2026, we're bringing you major LangSmith launches (Engine is here, and Sandboxes are now GA), plus a quick roundup of everything else we shipped. You can also catch up on
, now available on demand.
LangSmith Engine: Accelerate the entire agent development lifecycle
Improve agents without spending hours buried in traces. LangSmith Engine watches your production traces, clusters recurring failures into named issues, diagnoses root causes, and proposes fixes for review.
Sandboxes GA: Secure code execution environments for agents
LangSmith Sandboxes are Generally Available: secure, scalable environments built for agent code execution, and integrated with the Deep Agents SDK and the LangSmith platform.
Everything else we shipped at Interrupt 2026
Other updates focused on two areas: production-grade agent infrastructure and stronger observability and governance.
Two days, 1,000+ builders, and what's actually working with agents in production. Watch all 23 talks from teams at Cisco Customer Experience, LinkedIn, Rippling, and more. Now available on demand.
See our favorite use cases, best practices, and customer stories covering the most top-of-mind topics in the AI space.
Introducing LangChain Labs: Advancing continual learning for agents
At Interrupt we launched LangChain Labs, a new effort focused on continual learning to advance open, applied research for every agent. We're excited to partner with Harvey, NVIDIA, Prime Intellect, Fireworks, and Baseten on this initiative.
On the latest episode of Max Agency, Harrison sits down with Cogent Co-Founder & CTO Geng Sng to explore the world of autonomous cyber defense and what it takes to get enterprises to trust an agent within their production environment. Watch on
or listen on
or
.
ICYMI, watch our
with Alex Shevchenko, Head of Applied Research at Ramp.
🗽 June 2 // New York City: Inside the Agent EngineA technical fireside chat with Harrison and Traversal CEO Anish Agarwal discussing how to build, test, and monitor agents in production.
🗽 June 3 // New York City: Interrupt ResumeIn terrupt is coming to New York! Expect a half-day session including a keynote from Harrison, fireside chat with Pendo, and a hands-on Deep Agents workshop.
⚙ June 10 // Pittsburgh: Improving Production AgentsHe ar best practices on how teams turn production traces into insights, issues, and better agent experiences from teams at LangChain team and Credit Genie.
💻 June 11 // Virtual: How to Shorten the Path from Agent Issue to PR with LangSmith EngineSe e how Engine automates the whole agent development loop, from production failure to fix to eval.
💻 June 24 // Virtual: The Agent Development Lifecycle 101Jo in Harrison Chase for a walkthrough of the Agent Development Lifecycle and why it matters for teams moving agents into production.
🏙 June 29 // Chicago: LangChain X AWS WorkshopA hands-on session on deep agents, including how to integrate them with AWS services to ensure they are scalable, reliable, and enterprise-ready.
How can you follow along with the Lang Latest? Check out the
,
, and
for more product and content updates.
For any additional questions or feedback, email us at hello@langchain.dev.