Recent posts
Open bot!
🎉 Introducing 𝙾𝚙𝚎𝚗 𝙱𝚘𝚝 An open source Grok Bot that works with ANY agent harness, designed for real companies. It includes: - AI Coworkers - Generative UI - Computer use (remote/local) - Agent-human handoffs - Full data recording, owned by you Repo → https://t.co/ssje0KRts5 We're using this internally at @CopilotKit and it's changing the way we work forever. Powered by CopilotKit and AG-UI. More info below 👇
very cool launch we did a webinar with jeff on "wiki" style memory and it's clear he'd thought about this problem a lot (webinar here: https://www.youtube.com/watch?v=Lsut4TCfygw)
I’ve been looking forward to today for 3 years Today we’re announcing Foundation - Chroma’s solution to memory Our research preview of this technology builds self-improving memory from your agent sessions. Try it out at https://trychroma.com/foundation
lots of observability & evals platforms not a lot of platforms that help close the loop and have an agent that suggests fixes, adds evals itself, etc
@hwchase17 BTW your NYC ads worked 😂 just learned that's why we chose LangSmith
top tier webinar tmrw! @Vtrivedy10 (@LangChain) @willcb (@PrimeIntellect) @AEllisBloor (@baseten) and I will will be jamming on how to automate more parts of the agent improvement loop (eval and environment engineering in particular) Come join us: https://events.langchain.com/webinar/Towards-Automating-Eval-and-Environment-Engineering/
+1 to this part of this is due to coding agent standards - agents.md and skills are just markdown files/directories!
agents are starting to look less like apps and more like directories instructions. skills. tools. memory. identity. channels. schedules. evals. the harness + runtime become infra underneath it all. really like where this abstraction is going
🎓New YouTube playlist: Managed Deep Agents Gives an overview of Managed Deep Agents, and then each video dives deep into core concepts. Launching with six videos! 1⃣ Intro: https://t.co/DD1J0vOr2D 2⃣ Conceptual Overview: https://t.co/CDs6ysKAax 3⃣ Quickstart: https://t.co/otB4xrCfNh 4⃣ Instructions and Context Hub: https://t.co/251flZZBZr 5⃣ Skills: https://t.co/rQtfQ8Jq28 6⃣ Tools: https://t.co/g1Fw8UEnj1 Playlist link: https://t.co/FRAWhAOwvB
new onboarding for managed deep agents
mda 0.5.3 is a thing of beauty
I love working with Sydney and team and you will too! Come join us
we're hiring open source devs @LangChain looking for people who are building at the frontier of agents and are excited to quickly develop ownership apply: https://www.langchain.com/careers?ashby_jid=74e5f9f4-e44a-4594-ba26-abdf71bf287d#explore-jobs
Slack is the best ux
@hwchase17 if you dont have ur agents working overtime in slack/teams , ngmi
channels are how you interact with your managed deepagents eg slack pretty diagram 👇
anatomy of an agent channel. https://langch.in/mda
🚀Today we launched LangSmith Tuned Evaluators, starting with Perceived Error. Tuned Evaluators run on production traces to catch undesirable agent behavior and attach feedback that you can use in your agent improvement processes. In our benchmark, our tuned model beat frontier models at 82% lower cost. https://t.co/2df6i35SsY
totally agree! here's how we architected deepagents to enable this deepagents runs connected to a "backend". this backend needs to expose filesystem like operations, but it does not have to be a filesystem. it could be a database, object storage, or a real filesystem - it just has to expose read/write/edit etc operations this backend could also be what we call a "sandbox". if a sandbox, it needs to expose an "execute" command which lets it execute code this backend is SEPARATE from where the agent loop runs. this allows us to "separate the brains from the hands" (https://t.co/Pi3XljhB67) deepagents is built on top of langgraph, which means we can easily deploy it with MCP, a2a, and other standard endpoints we use this architecture to power many different types of experiences first, we can create a classic TUI like coding experience. we do this by giving deepagents a "sandbox" that is running locally in the same directory; deloying deepagents locally behind a light weight server; and then connecting to it with the TUI acting like a frontend. see dcode for an example of this https://t.co/wj48PbCuSx second, we can create a cloud coding experience. we can do this by running deepagents on LangSmith deployments for a production scale deployment, and connecting to a sandbox running on modal, daytona, e2b that is running elsewhere. we can then build a frontend to connect to langsmith deployments and let users interract with it there, and also expose it in slack to let users interract with it there. note: both slack and web ui connect to the same backend, so you can switch between them seamlessly. code: https://t.co/Pdevl2PRrv of course - deepagents can be used to create agents that are NOT coding agents. a lot of agents still need to write and execute code, so this architecture is still very useful. but for some the code execution is overkill, and thats where you can swap to a "fake" backend, and still let it have the ability to interract with files (good for context engineering!) without having to spin up a full sandbox. for a really easy way to create these types of agents - see managed deepagents: https://t.co/NAXiKqZbi1
I love agentic coding harnesses, but they shouldn't be primarily terminal-based. The terminal is great for quick and precise commands, but information density is extremely low and UI affordances are minimal. Maybe provision of TUIs is worthwhile for occasional use (when establishing a tunnel is too annoying, or something), but it feels very strange for this to be the default modality. It took a long time for dynamic language REPLs to break out of the terminal (Jupyter notebooks and similar); I hope we don't have to wait as long for the harnesses.
gave a talk "owning your intelligence" - ty @sequoia @sonyatweetybird for having me talked about harnesses and evals and the role they play in owning your intelligence TLDR: > agents = model + harness + context > model - own the weights using something like @FireworksAI_HQ > context - memory needs to be portable > harness - needs to be model agnostic. also needs to be good at bringing right context to llm. "right" context may depend on your use case, which is why an open/configurable harness helps > how to use middleware in langchain/deepagents to configure your harness > how to use langgraph to fully own your cognitive architecture > why evals/obs matters - some quotes from @satyanadella - “Create your private evals, because evals define what “good” looks like inside the organization” - “retain ownership of your organization’s memory, traces, feedbacks, decisions, and institutional context” - “you create your own continuous learning loop (i.e. hill climbing machine) that will allow your AI investments to compound the value of your firm” > how to use harbor for evals > tracing is important > evals + observability only matter so you can set up a data flywheel > data flywheel = run agent -> collect traces -> find interesting traces -> use those to improve > demo of langsmith engine which does exactly this! full video: https://t.co/k6li5hu6D9
some great new LangSmith docs on traces vs threads vs trajectories (new concept) observability data is no longer just for observability - its also for memory & learning having a really clear mental model of this data is incredibly helpful! https://docs.langchain.com/langsmith/observability-concepts
new langchain oss release!
What's new in LangChain? 🚀 🔌 Support for OpenAI's 3.0 SDK (using httpx2) ✨ Support for gemini-3.7-flash Plus a wave of core reliability fixes from external contributors: 🛠️ Tool calling & structured output: clearer errors, respect for pydantic aliases when validating tool inputs (https://t.co/27M0O5MAUm, https://t.co/Rd54vFPv7M, https://t.co/vLprs6D0Ik) 📊 Usage metadata fixes: token usage callback bugs, and cost metadata when streaming OpenRouter (https://t.co/57ronBXWt0, https://t.co/aVViSEobGj) 🧱 Content & prompt hardening: guard malformed Anthropic content blocks, and preserve non-str/non-dict items in prompt templates for content blocks (https://t.co/NL9xKsBhNv, https://t.co/Kjtvg4DmxC) Huge thanks to everyone who contributed! 🙏
Can use a lightweight classifier step to first decide if even worth running Then more expensive agent if that criteria is met
I’m afraid not everyone can afford to burn trillions of tokens
agents running in the background will be the future - lets work scale beyond people prompting them directly crons are one way to do this. first class support in managed deepagents
you can build agents that prompt themselves by giving them a schedule. with Managed Deep Agents, that only takes a few lines of code code snippet is from our shiny new MDA docs: https://langch.in/schedules
talked about how harnesses & evals help you own your intelligence how they fit in to the big picture: owning your intelligence means three things: - open agent system (harness is a big part of this!) - compounding loop (evals are a big part of this!) - governed runtime (harness also important here - we see managed harnesses growing rapidly)
An agent is three things: a harness, a model, and context. If you're serious about owning your intelligence, you probably want to own all three. @LangChain founder @hwchase17 joined us at our @sequoia Own Your Intelligence to talk about the piece that often gets the least attention: the harness. He offers a clear heuristic for when to build your own. The more out of distribution you are from what the models were trained on, the more you'll want to customize. And good technical content on how to actually measure performance with evals and langsmith. 00:00 Introduction 00:58 The three parts of an agent: harness, model, context 02:12 What a harness actually does 03:25 Customizing the core loop with middleware 04:41 Sandboxes, file systems, sub-agents, summarization 05:47 Cognitive architectures — and when you still need them 07:03 Build your own harness or use off the shelf? 08:24 In-distribution vs. out-of-distribution: the file-editing example 09:39 Why evals define what "good" means in an organization 11:04 Harbor: what an eval task actually looks like 12:11 Comparing harnesses and models on accuracy, latency, and cost 13:20 Why observability is underrated — it's usually the context 14:34 The data flywheel: traces → curation → experiments 15:42 Getting feedback through UX design and online evaluators 16:51 Demo: LangSmith Engine 19:23 Q&A: Running Engine on Engine, and "codex-ification" 20:44 Q&A: Will harnesses converge or diverge?
one of the strengths of managed deepagents is how easy it is to define all the pieces of your production agent we updated our docs to reflect this. each component is its own file. you can easy click through to see where it lives, what it looks like, etc https://docs.langchain.com/langsmith/python/managed-deep-agents-overview#example-agent
