Technical blog on how we built SmithDB (our database purpose built for agent traces)
@hwchase17
-
Letta always leader, sleep time compute comes from them
By
–
Letta has always been at the top of that, "sleep time compute" comes from them!
-
LangSmith Engine as Sleep Time Calculation
By
–
Memory for agents is often described as a 'sleep time calculation' or 'daydreaming'. This involves running a background process to analyze agent trajectories and update a
-
Three-step process for agent memory
By
–
Many agent memories follow a simple three-step process:
– execute the agent
– execute a process to examine those traces and analyze them
– update memory accordingly We show how to do this! -
Initial deployment: small step, essential repeatable improvement cycle
By
–
One thing we have observed quite consistently: deploying the first version is only a small part of the work. A key part of building reliable agents is having a repeatable lifecycle to improve them over time.
-
Auto-Harness: Self-Improving Harnesses
By
–
New paper on agents that shape their own harnesses to improve over time. No LangChain, but built on DeepAgents! Three key steps: 1/ Exploitation of weaknesses: identify the modes of
-
Difference between routing and model advising
By
–
I've been thinking a lot lately about model routing and related things. Current thoughts here, I'd like feedback: 1/ there is a difference between 'model routing' and 'model advising'. 'model routing' = routing to a single
-
Build your own Claude Code with Deep Agents
By
–
Great community article showing how to build a Claude Code-like agent using Deep Agents. Particularly relevant given how powerful GLM-5.2 appears to be! https://pub.towardsai.net/build-your-own-claude-code-using-langchin-a-deepdive-into-langchains-deep-agents-9ef98d98a69a …
-
Leve: filesystem-first durable agent framework built on LangGraph
By
–
Very cool work from @jit_infinity:
— Harrison Chase (@hwchase17) 20 juin 2026
🔥Leve: filesystem-first, durable agent framework built on LangGraph.
You describe an agent as a directory of files. Leve compiles that directory into an agent and runs it
Inspired by Vercel's Evehttps://t.co/cfWpii90Yn pic.twitter.com/S7FhJEirITVery cool work from @jit_infinity
: Leve: filesystem-first, durable agent framework built on LangGraph. You describe an agent as a directory of files. Leve compiles that directory into an agent and runs it Inspired by Vercel's Eve https://
github.com/prasanjit101/L
eve
… -
Complete course on agentic AI with LangChain and LangGraph
By
–
One of the best courses on agentic AIs I have seen. Nearly 10 hours of excellent content. Covers LangChain, LangGraph, RAG, deepagents, guardrails, and more. Are there any other good Lang* resources for people interested in learning?