February 2026: LangChain Newsletter

February gave us only twenty-eight days, but we made them count. From new Agent Builder capabilities to production monitoring insights, here's your monthly roundup of updates from the LangChain team.

What’s new for LangChain?

LangSmith

🔨 Agent Builder now has a central chat with all your tools, one-click conversation-to-agent, file uploads, and a unified tool registry.

⚙️ Configure which parts of a trace’s inputs and outputs appear directly in the tracing table.

👀 Schedule reports to run daily, weekly, or on a custom cadence with Insights Agent.

📌 Pin any experiment as your baseline and instantly see performance deltas across every run without manual side-by-side comparison.

Open Source

🚀 deepagents v0.4 allows you to run your agents in fully isolate sandboxes.

Interrupt 2026

Interrupt 2026 tickets are live! Join us May 13-14 at The Midway in San Francisco for two days of can't-miss keynotes, hands-on workshops, and real-world production stories from teams building agents at scale. Hear from Andrew Ng, Harrison Chase, and AI teams at Clay, Rippling, and more.

Speak the Lang

Production monitoring for agents Traditional monitoring tools weren't built for agents. Their infinite input space, non-deterministic outputs, and multi-step reasoning chains require a different approach entirely. We broke down what production observability for agents actually looks like.

Improving Deep Agents with harness engineering Our coding agent jumped from Top 30 to Top 5 on Terminal Bench 2.0, without touching the model. The secret was harness engineering: self-verification loops, loop detection middleware, and using LangSmith traces as a feedback signal to systematically debug failure modes at scale.

Tracing in LangSmith is as easy as copy/paste Get started in seconds with Claude Agent SDK, OpenAI, LangChain, Vercel AI SDK, and 20+ other frameworks.

LangChain Academy course: Building reliable agents Learn how to take an agent from first run to production-ready system through iterative cycles of improvement using LangSmith.

Customer story highlight

monday.com Service sped up their evals feedback loop nearly 9x by making evals a day 0 requirement using LangSmith.

Upcoming Events

• Chicago (March 3): Community Meetup: Detecting Errors Early with LangSmith.
• San Francisco (March 4): Deep Agents Meetup.
• London (March 6): Agents & Knowledge Graphs Hackathon.
• San Francisco (March 18): Contrary SF Tech Talk.
• San Francisco (March 19): It Worked on My Laptop - Agents in Production.
• Stockholm (March 22): Lovable x LangChain Proptech/Contech Hackathon.
• New York (March 24): It Worked on My Laptop - Agents in Production.
• Amsterdam (March 26): The Rise of Autonomous Systems.
• Mexico City (March 26): Agents in Production.
• San Francisco (March 27): LangSmith for Startups Office Hours.
• New York (March 29): LangSmith for Startups: Sunday Hacking Brunch.
• Miami (April 2): Community Meetup: Evals 102.