The team over at @RunloopAI is awesome, great demo deployment pattern for LangChain’s Deep Agents The goal of our DeepAgents library is to give builders a full-open agent harness where everything is configurable, hackable, and observable. You can also use our default harness to get started in asap with create_deep_agent() I’m excited to build a future of fully open agent harnesses where builders can get to use frontier level intelligence (Opus 4.5, gpt-5.2, and Gemini 3, etc) but also get full control and observability into the details of prompts, tools, context management, etc. also hmu because i’m personally here to help anyone building Deep Agents, DMs are open 🚀
Loading thread detail
Fetching the original tweets from X for a clean reading view.
Hang tight—this usually only takes a few seconds.