@LangChain: How do agents remember? The longer an agent runs, the more it takes in (instructions, tool results, files, and drafts),…
Summary
LangChain 发布新博客，介绍其 Deep Agents 框架如何通过将上下文移出活跃窗口（而非删除）来解决智能体长期运行时的记忆与推理准确度下降问题。
View Cached Full Text
Cached at: 10/14/26, 10:04 PM
How do agents remember?
The longer an agent runs, the more it takes in (instructions, tool results, files, and drafts), which results in decreased accuracy and reasoning.
Read how LangChain’s Deep Agents harness addresses this by moving context out of the active window without deleting it ⤵️
Jess Weng@jess_weng18·4h: New blog dropped 👀
https://langchain.com/blog/context-engineering-deep-agents-remember…
Similar Articles
@GitHub_Daily: AI for long-running complex tasks: as context grows, models tend to 'forget' and output quality drops sharply. The LangChain team has open-sourced a tutorial: Deep Agents from Scratch, which deconstructs core design patterns of mainstream agents from scratch, explained thoroughly...
The LangChain team has open-sourced the tutorial 'Deep Agents from Scratch', which deconstructs the core design patterns of mainstream agents from scratch, covering task planning, context offloading to a file system, and sub-agent isolation. It includes 5 progressive notebooks, allowing you to build a complete deep research agent hands-on.
@LangChain: We simplified agent authentication, memory, and channels, and brought in web search as a pre-built tool for Managed Dee…
LangChain introduced Managed Deep Agents at the Interrupt NYC conference, bundling the agent harness with infrastructure to simplify authentication, memory, and channel integrations, while adding a pre-built web search tool, Harbor evaluation integration, and Parallel tool integration — allowing teams to focus on business logic.
@LangChain: Managed Deep Agents keeps the project shape you already know: ↳ AGENTS.md, skills/, subagents/, + tools.json Context Hu…
LangChain introduces Managed Deep Agents, maintaining a familiar project layout with AGENTS.md, skills/, subagents/, and tools.json, along with Context Hub for persistent context management across sessions.
@Huahuazo: I want to understand how to build an Agent from scratch, not just call APIs—I've looked through many resources, and they're either too abstract or just hand you a finished product to tweak parameters, leaving the 'why design this way' part basically blank. LangChain's official Agents From Scratch...
LangChain's official Agents From Scratch tutorial provides a detailed guide on how to build an AI agent from scratch with human-in-the-loop and memory capabilities, covering four parts: basic setup, evaluation system, human-in-the-loop (HITL), and long-term memory.
@LangChain: What separates a simple agent from one that can reliably work through complex, long-running tasks? It comes down to the…
LangChain announces virtual workshops on building deep agents, focusing on harnesses for reliable complex task management in AI systems.