LangChain Launches Managed Deep Agents: Build Production Agents Like a Folder
hwchase17 · x · 2026-08-14
LangChain has officially launched Managed Deep Agents (MDA), a streamlined way to build and deploy production agents. Developers can define agent behavior by simply uploading a directory of files (e.g., instructions, skills, tools, MCP connectors).
MDA handles the underlying 'Deep Agents harness' (including planning, tool calling, filesystem management, and subagent delegation) and the managed runtime (LangSmith Agent Server hosts and operates the agent across restarts). Developers can add Slack integration or daily runs simply by adding a file, with zero server configuration required.
More from coding & agent
- AI Engineering is Becoming Alchemy: A Reflection on Cognitive Extension — zakelfassi · 2026-08-14
- Claude Code Tip: Select DOM Elements Visually to Prompt Frontend Changes — EricBuess · 2026-08-14
- X Open-Sources 'For You' Feed Algorithm for Full Transparency — tetsuoai · 2026-08-14
- Open-Sourced YOLO Training Template Integrates Kaggle Datasets and Auto-Labeling — tom_doerr · 2026-08-14
- Arcee AI Open-Sources NAC: An Agent Harness for Long-Running Complex Tasks — code_star · 2026-08-14
- Teaching AI Agents to 'Dream': Building Long-Term Memory via Reflection — rseroter · 2026-08-14