Pinned
Stop manually triaging agent failures.
Let LangSmith Engine fix it.
00:00
LangChain
11K posts
LangChain
@LangChain
Joined November 2022
- LangChain reposted🗣️Voice traces in LangSmith just got an update! Now, LangSmith has an audio player that runs inline with spans. No new tabs necessary. See what was being played when a particular span was active, or see what span was active when a particular piece of audio was being played.
- You can turn an existing LangGraph agent into a fully functional voice agent with @pipecat_ai. This 17-minute walkthrough shows you exactly how to do it.
00:00🗣️Voice traces in LangSmith just got an update! Now, LangSmith has an audio player that runs inline with spans. No new tabs necessary. See what was being played when a particular span was active, or see what span was active when a particular piece of audio was being played.Watch the full tutorial: - Most teams building today have traces, what they don't have is a system for continuously improving their agents. The teams that do use an agent development lifecycle: build, test, deploy, and monitor. The problem with this lifecycle? It scales at the speed of their engineers.
00:00That's why we built LangSmith Engine, an agent built to improve your agent. Engine identifies issues from traces, clusters patterns, drafts fixes, and proposes evals to prevent regressions. Connect your tracing project + start improving your agent today: - Interpreters in Deep Agents: Small embedded runtimes where agents can write and execute code inside the agent loop. An explainer from @huntlovell on the @LangChain_OSS team langchain.com/blog/give-your…
- 🗣️Voice traces in LangSmith just got an update! Now, LangSmith has an audio player that runs inline with spans. No new tabs necessary. See what was being played when a particular span was active, or see what span was active when a particular piece of audio was being played.Get started today, and let us know what you think!
- The art of loop engineering, powered by Deep Agents!self-harness: harnesses that improve themselves spoiler alert, it's all loop engineering! 1. weakness mining - run agent and observe failures 2. propose improvements to the harness 3. confirm improvements are beneficial and don't cause other regressions and then you loop! the
- One thing we’ve seen pretty consistently: shipping the first version is only a small part of the work A key part of building reliable agents is having a repeatable lifecycle for improving them over time Here’s what that looks like: 1/ Build. Start with the agent, tools,
- LangSmith Fleet has two types of agents: General Purpose Chat and Specialized Agents. A breakdown on why this is a deliberate choice and when to use each.
- LangChain reposted
🧠Self-Harness: Harnesses that improve themselves New paper on agents shaping their own harnesses to improve over time. Not from LangChain, but builds on top of DeepAgents! Three key steps: 1/ Weakness mining: find failure modes from traces 2/ Harness proposal: suggest changes
- LangChain reposted
Speaking of "this is your opportunity to try GLM 5.2" 👀i was already using GLM-5.2 but WOW it's even better through baseten! give it a try with deepagents code (dcode), our open harness built to work with any model/provider
- When the EU AI Act goes into effect, compliance will become an ongoing measurement obligation. With LangSmith, you can turn tracing into compliance evidence. Run our customizable evaluators on production traffic, and score on the EU AI Act’s requirements, including bias,
Ready to see how you can meet the core technical requirements of the EU AI Act with LangSmith? - LangChain reposted
interesting point here: loops amplify behavior, making them a double edged sword but we know loops are the future, so how do we avoid amplifying bad patterns? you need an *engaged* human in the (stacked) loop(s) your agent needs to learn your taste! x.com/sydneyrunkle/s…Some more thoughts on looping in coding agents. lucumr.pocoo.org/2026/6/23/the-…
