Does the Next Run Improve? Community Weighs In on Chase's Knowledge Factory
hwchase17 · x · 2026-11-22
Responding to Harrison Chase's "knowledge factory" article, the community crystallizes a practical test for enterprise agents: after you correct one, does the next run improve — or does someone have to explain the same thing again?
The concrete shape: agents flag gaps, a curator checks evidence, and reviewed knowledge becomes available for future work. The hard engineering calls are which corrections become shared rules vs. task-local, and how to test whether an update actually improves outcomes. The post also links an open-source LangChain 101 guide covering agents, state, memory, durable execution, tracing, and evals across LangChain 1.5, LangGraph 1.2, and Deep Agents 0.7.
More from coding & agent
- Stop Automating Everything: Why Humans Belong Inside the Agent Loop — dotey · 2026-11-22
- Grok Bot agents organized a 5,000-person summit in 6-7 months, vs the usual 18 — XFreeze · 2026-11-22
- One command adds a music-video generator plugin to Claude Code — AIandDesign · 2026-11-22
- Dev Gets Antigravity Account Banned After Using Its Provider Inside t3 Code — jeremyphoward · 2026-11-22
- Knowledge factory concept extended to computer use agents — hwchase17 · 2026-11-22
- Teacher asks if a 6GB RTX 3060 can run a local Claude math-grading agent — beekersavant · 2026-11-22