Jiabin Tang

AutoAgent & Auto-Deep-Research v0.2.0 (formerly known as MetaChain) is coming! We fix some bugs and add more easy-to-use commands like `auto main` or `auto deep-research`. Code: github.com/HKUDS/AutoAgent, github.com/HKUDS/Auto-Dee…. A wild example running successfully only once:

MetaChain has become Auto Agent 👇 A name that better reflects its 2 main features: a fully automated & zero-code framework. And to demonstrate the value of the framework, the team offers its own DR version = Auto Deep Research. Bravo to the team 👏

🚀 2025 is shaping up to be the Year of AI Agents! Meet ✨AutoAgent✨ - our new framework that lets you Create Your Own AI Assistant to kickstart your year. 👩‍💻 For Developers: ✨AutoAgent✨ (github.com/HKUDS/AutoAgent) is a Fully-Automated & Zero-Code LLM Agent framework. Anyone…

Come and try it !

arxiv.org/pdf/2502.05957… Jiabin Tang Tianyu Fan Chao Huang The University of Hong Kong {jiabintang77, tianyufan0504, chaohuang75}@gmail.com ‡ Source Code: github.com/HKUDS/MetaChain

🚀 Revolutionizing LLM Agent Creation: No Code Needed! Today we are lanching our Fully-Automated & Zero-Code LLM Agent Framework-🔥MetaChain🔥, enabling Anyone Can Build Their Own Agent through Natural Language Alone. Forget the complexity of LangChain & AutoGen - ✨MetaChain✨…

I made SWEBench-verified-mini SWEBench-verified-mini is a subset of SWEBench-verified that uses 50 instead of 500 datapoints, requires 5GB instead of 130GB of storage, and has approximately the same distribution of performance, test pass rates, and difficulty as the original…

Agent will re-define the game in 2025😎